V -?-'"" /^ '-^ >f < "'T.s^^ aC>' »# "^o •"•■/: ^^. ^„^^^ ',#% • ,^^ °^ "\.^^ . <5. •\,^^ r \ \.^^" '^ >' -», %. '•" 0" .^"r % %<^ :^M^:''\<^ :. ■^^''J^^S ^^^'^4^^^% ^^^'^m^'S> ^^ ,^ ^ -„^/ .^Q '^---^^^ /^^&'^° ^--^'^^ ^°^:i;^'^% cp^^^i;;^".^ ^^i^^^'-^t. <^° \.d^ = ^<3f> ^•»,%:'°'^-'>\^-«.%*'°'^^ v^\r°' "-. rt '«-.K^^ <^^. ^k.J' \..^^/^ THE PROBLEM OF KNOWLEDGE •Tl h^)(i^ THE MACMILLAN COMPANY NEW YORK • BOSTON • CHICAGO • DALLAS ATLANTA • SAN FRANCISCO MACMILLAN & CO., Limited LONDON • BOMBAY • CALCUTTA MELBOURNE THE MACMILLAN CO. OF CANADA, Ltd. TORONTO THE PROBLEM OF KNOWLEDGE BY DOUGLAS CLYDE MACINTOSH, Ph.D. ASSISTANT PROFESSOR OF SYSTEMATIC THEOLOGY IN YALE UNIVERSITY THE MACMILLAN COMPANY 1915 All rights reserved^ Copyright, 1915, By the MACMILLAN COMPANY. Set up and clectrotypcd. Published October, 1915. J. 8. Cashing Co. — Berwick & Smith Co. Norwood, Mass., U.S.A. NOV I 1915 ©CI.A41435e4 "VvD ( « MY FIRST COMPANION JOHN EVERETT MACINTOSH WITH THE GRATITUDE AND AFFECTION OF A YOUNGER BROTHER >^' ACKNOWLEDGMENT i If I were to undertake to make complete acknowledgment of my indebtedness to others for stimulus and direction in connec- tion with the study of the problems discussed in this volume, I should want to begin by referring to my first teacher of philoso- phy, Professor James Ten Broeke, of McMaster University, Toronto. But where to end I should scarcely know. I shall therefore mention only those whom I have to thank for one direct service or another in connection with the actual production of the present volume. Through the kindness of my former teachers, Professors J. E. Angell and A. W. Moore, of the University of Chicago ; of my colleagues, Professors E. W. Hopkins, A. K. Eogers, and C. A. Bennett, of Yale University; of my former colleagues in Yale University, Professor W. E. Hocking (now of Harvard Univer- sity) and Dr. H. T. Costello (now of Columbia University) ; and, finally, of Professor E. B. Perry, of Harvard University, and Professor D. S. Miller, of the G-eneral Theological Seminary, New York, various portions of the volume have been read in the manuscript, and as a result not a few helpful criticisms and suggestions have been received. I am particularly indebted to Dr. Costello in connection with the references to the logisticians in the last chapter. It should not be concluded, however, that here, or elsewhere in the book, any one except the author is to be held responsible for any of the opinions expressed. For friendly counsel and other manifestations of kindly inter- est in connection with the publishing of the work I am grateful to Dean Shailer Mathews, of the University of Chicago, and Professor E. Hershey Sneath, of Yale University. Grateful acknowledgment is also made of assistance received from my friend and former pupil, Mr, E. W. K. Mould, in the preparation of the manuscript for the publishers ; from my friend Vlil ACKNOWLEDGMENT and pupil, Mr. F. W. Shorter, in the reading of the proof ; and from my sister, Miss Anna B. Macintosh, in the preparation of the index of authors. D. C. MACmTOSH. Breadalbane, Ontario, August 31st, 1916. TABLE OF CHAPTEKS Chapter I. Introductory : Philosophy and Its Principal Problems Gon- PART I: THE PBOBLEM OF IMMEDIATE KNOWLEDGE A. THE PROBLEM OF ACQUAINTANCE (EPISTEMOLOGY PROPER) 1. A Critique of Dualism Chapter II. Dualism and Avowed Agnosticism Chapter III. Dualism and Attempted Metaphysics Chapter IV. Dualism and Attempted Metaphysics eluded) .... 2. A Critique of Idealism Chapter V. Mystical and Logical Idealism Chapter VI. Psychological Idealism . Chapter VII. The Older Absolute Idealism Chapter VIII. The Newer Absolute Idealism r Chapter IX. The Disintegration of Idealism A Critique of the New Bealism .... Chapter X. Antecedents of the New Realism Chapter XI. The Neo-Realistic Doctrine of Secondary Qualities ' . Chapter XII. The Neo-Realistic Doctrine of Consciousness Chapter XIII. The Neo-Realistic Doctrine of Relations, Universals, and Values 11 11 11 13 36 67 72 72 92 126 154 181 211 211 232 259 4. Constructive Statement 310 Chapter XIV. Critical Monism in Epistemology . . 310 B. PROBLEMS OF THE WAYS AND MEANS OF KNOWING (MORPHOLOGY OF KNOWLEDGE, AND GENETIC LOGIC) 336 Chapter XV. The Morphology of Knowledge . . .336 Chapter XVI. The Genesis of the Apriori . . . .351 PART II: THE PBOBLEM OF MEDIATE KNOWLEDGE . 367 A. THE PROBLEM OF TRUTH (LOGICAL THEORY) . 367 Chapter XVIL A Critique of Intellectualism . . .369 Chapter XVIII. A Critique of Anti-Intellectualism . .401 Chapter XIX. Critical Monism in Logical Theory . . 438 B. THE PROBLEM OF PROOF (METHODOLOGY) . . 459 Chapter XX. The Problem of Scientific Method . , 459 ANALYTICAL TABLE OF CONTENTS Introduction (1) : a. Philosophy in general (1). Its definition (1). Its relation to the special sciences (1). Problems of values and of reality (3). 6. The Problem of Knowledge (6). Typical attitudes toward the problem (7) : Woodbridge (7), J. Watson (8), Ladd (8). Subdivisions of the problem (10). PART I: THE PBOBLEM OF IMMEDIATE KNOWLEDGE (11). Division A. THE PROBLEM OF ACQUAINTANCE (EPISTE- MOLOGY PROPER (11). Introductory definitions (13). Subdivision I. A Critique of Absolute Epistemological Dualism (Epistemological Dualism and Critical Realism). An absolute duality of the experienced and the independently real (14). Agnosticism logically involved (14). a. The agnosticism avowed (14). Locke (a forerunner) (14), Kant (17), Schulze (24), Hamilton (25), Spencer (27), cf. Bradley and Hodg- son (30), Riehl (31), Dilthey (33). h. Attempted metaphysics ; agnosticism not fully acknowl- edged (36). (1) Jacobi (37), Reinhold (37), Fries (38), Neo- Friesianism (41), Herbart (42), Lotze (44), Ladd (50), Pringle-Pattison (51), Strong (52), Love joy (56). (2) Schelling (57), Schopenhauer (58), Hartmann (60), Volkelt (63), Cornelius {m), Kuelpe (68), B. Russell's earlier view (70) . Subdivision II. A Critique of Idealistic Absolute Epistemological Monism (Epistemological Monism and Dogmatic Idealism). The view that (physical) reality is nothing beyond idea, offered as a solution of the problem of knowledge (72) . Preliminary distinctions (73). 1. The Integration of Idealism. Development of idealistic thought from simpler to more composite forms (74) . xil ANALYTICAL TABLE OF CONTENTS A. Mystical Idealism. The interpretation of physical reality as mere idea, suggested by the mystical experience (75). The Upanishad philosophers (75), Albertus Magnus (79), Eckhart et al. (79), Mrs. Eddy (79). B. Logical Idealism. The interpretation of reality as the logical idea (the abstract absolute idea, the absolutely satisfactory predicate, or definition), suggested by the dialectic (81). Plato (81). C. Mystical-Logical Idealism. A dyadic compound of ele- mental forms of idealism, the mystical and the logical (89). Plotinus (89). D. Psychological Idealism. The interpretation of the phys- ical object as mere content of consciousness, under the influence of the psychological point of view (92). The logical fallacies involved (93). a. Undisguised psychological idealism : (1) Berkeley (97), Hume (98), J. S. Mill (100), Clifford (100), Pearson (102), Marshall (103). (2) Fichte (104), Fouill^e (105). (3) Lipps (106); Vaihinger (107), Poincar^ (108). 6. Disguised psychological idealism. (Pure empiri- cism ; immediate empiricism ; the philosophy of pure experience) (109). Mach (110), Avenarius (111), Petzoldt (112), Wundt (112), Fullerton (114), Hodgson (115), James (116), Dewey (117), Mead (118), A. W. Moore (119), Bawden (119). E. Mystical-Psychological Idealism. A dyadic compound of elemental forms of idealism, the mystical and the psychological (120). Bergson (120). F. Logical-Psychological Idealism. A dyadic compound of elemental forms of idealism, the logical and the psy- chological. Reality as the concrete absolute idea, the concrete uaiversal, the logical within the psychological (126). The monistic (singularistic) form : Absolute Idealism (128). a. The pre-Bradleian constructive movement; the thesis (130). (1) Intellectualistic absolute idealism. The ab- solute idea determined by rational criti- cism (130). ANALYTICAL TABLE OF CONTENTS xui Hegel (130), Stirling and Wallace (131), J. Caird (134), Green et al. (134), E. Caird (137), J. Watson (138). (2) Voluntaristic absolute idealism. The abso- lute idea determined by purpose (141). Royce (141). /S. The Bradleian criticism ; XhQ antithesis {\^Q). Bradley (146). 7. The post-Bradleian reconstruction ; the synthesis (154). (1) Intellectualistic absolute idealism (154). Bosanquet (154). (2) Voluntaristic absolute idealism (159). Taylor (159). (3) Mystical absolute idealism, or G. Mystical-Logical-Psychological Idealism. The triadic compound of the elemental forms of idealism (161). Hocking (161). 2. The Disintegration of Idealism. Tendencies from the more integrated forms of idealism back to the more elemental forms (181). A. Toward Personal (Pluralistic) and Psychological Ideal- ism (181). (1) Monistic Theistic (Psychological) Idealism (182). Paulsen (182). (2) Semi-Pluralistic Theistic Idealism. Lotzian ideal- ism (184). (3) Pluralistic Theistic (Psychological) Idealism (184). Leibniz (184), Ward (185), Rashdall (187), Schil- ler (188), Renouvier (188). (4) Pluralistic Semi-Theistic (Logical-Psychological) Idealism. Howison (189). (5) Pluralistic Atheistic (Logical-Psychological) Ideal- ism. McTaggart (190). B. Toward Abstract and Logical Idealism (192). (1) Psychological-positivistic abstract idealism (192). Fullerton's earlier view (192). (2) Critical-positivistic (intellectualistic) abstract idealism (193). Transition to this view. Lange (193), Liebmann (195). Typical representatives. The Marburg School : H. Cohen, Natorp, Cassirer (195). (3) Critical-transcendental (voluntaristic) abstract idealism (198). Windelband (198), Rickert (200), Munsterberg (201), Miincb (201). XIV ANALYTICAL TABLE OF CONTENTS (4) Logical-transcendental abstract idealism (201). Husserl (202), Meinong (203). A and B. Partial combination of Personal and Abstract Idealism. Bakewell (206). C. Toward Mystical Idealism (208). Spiritual or Keligious Idealism. Theoretical idealism retained for its supposed spiritual and especially reli- gious value (208) . Eucken (208), Boyce Gibson (209). Subdivision III. A Critique of Healistic Absolute Epistemological Monism (Epistemological Monism and Dogmatic Realism). The attempt to maintain the independent reality of the total object of experience or thought (211). 1. Antecedents of the New Realism (211). a. Naive realism (212). b. The Scottish realism (213). Reid (213). c. Movements leading to the new realism (217). (1) Disguised psychological idealism, as transitional to physical realism (219). Hodgson (221), James (222), Bush (224), Dewey (224), Boodin (227), Fullerton (229). (2) Disguised logical idealism, as leading to logical realism (230). 2. The New Realism (232). a. The neo-realistic doctrine of secondary qualities (232). Woodbridge (233), Alexander (234), G. E. Moore (236), McGilvary (238), Nunn (240), B. Russell (242), Hob- house (244), Wolf (246) , Fullerton (247), Boodin (247), Marvin (248), Spaulding (249), Perry (250), Pitkin (251), Holt (251), Montague (254). b. The neo-realistic doctrine of consciousness (259). (1) The English School (259). Influence of Hodgson (259). Hobhouse (260), Mc- Dougall (260), Moore (261), Russell (262), Alex- ander (262), Stout (264), Wolf (266). (2) The American School (266). Influence of James (266). Bush (267), Fullerton (268). Woodbridge's introspective view (269). Views of Woodworth (270) and Dunlap (271), criticised by Angell (272). The behaviorist view (272). Influence of Dewey (273). Thorndike (273), J. B. Watson (274), Frost (275), Singer and Woodbridge (276). Criticism by Miller (277). ANALYTICAL TABLE OF CONTENTS XV Views of McGilvary (277), Boodin (279), Marvin (280), Holt (280), Spaulding (282), Pitkin (282), Perry (283), Holt (later statement) (286), Mon- tague (287). (3) Comparison and general Criticism of views (289). C. The neo-realistic doctrine of relations (293). Nunn (293), Alexander (294), Stout (294), Russell (294), Montague (295), Holt (296), Marvin (296), Spaulding (297), Pitkin (298), Perry (300). d. The neo-realistic doctrine of universals (302). Relation to Platonism (302). Views of several neo- realists (302). e. The neo-realistic doctrine of values (306). Moore (306), Russell (306), McGilvary (307), Alexander (307), Montague (307), Perry (308). General statement and criticism of the new realism (309). Subdivision IV. Constructive Statement : Critical Realistic Episte- mological Monism (Epistemological Monism and Critical Realism : Critical Monism in Epistemology). Normally, the experienced object and the real object numerically identical, the real object existing, with some of its qualities and relations, when not experi- enced (310). Prior immediate knowledge the only explanation of mediate knowledge (311). Activistic view of sensation (312). Activistic explanation of the nature of consciousness (314). Activistic con- ception of causality (316) . Approaches to this point of view (316). Argument for activism from freedom (317). The subject-matter of psychology (318). Explanation of hallucination, etc. (320). Primary and secondary qualities and relations (322). Partial knowledge of the thing-in-itself (326). Tertiary qualities and relations (327). Values (328). Anticipation of further problems (330). Contrast between our doctrine and the Kantian (331). Solution of the problem of the externality or internality of rela- tions (332). Reflections on the status of epistemology in general (333). Divisions B and C. SPECIAL PROBLEMS OE THE WAYS AND MEANS OF KNOWING (MORPHOLOGY OF KNOWLEDGE, AND GENETIC LOGIC) (336). Division B. The Fundamental Mode of Cognition (Morphology of Knowledge) (336). Subdivision I. A Critique of Absolute Morphological Dualism. Perception and conception as two irreducibly different modes of cognition (336). Subdivision II. A Critique of Conceptualistic Absolute Morpho- logical Monism. Pure conception the only ultimate mode of cog- nition. Platonism (336). XVI ANALYTICAL TABLE OF CONTENTS Subdivision III. A Critique of Perceptualistic Absolute Morpholog- ical Monism. Pure perception the only ultimate mode of cogni- tion. Bergson (337). Subdivision IV. Constructive Statement. Critical Perceptualistic Morphological Monism {Critical Monism in the Morphology of Knowledge). Perception the one typ- ical and fundamental mode of cognition (337). Criticism of approximations to this view (338). Kant (338), Royce (339), James (340). Processes not ultimately per- ceptual, non-cognitive (341). Cognitive processes ultimately perceptual (342). Perception in a complex (343). Introspec- tion (344). The new (perceptual) intuitionism (346). Division C. The Genesis of the " Apriori " Element in Cognition (Ge- netic Logic) (351). Subdivision I. A Critique of Absolute Genetic Dualism. Absolute rationalism with reference to cognitive forms and absolute empir- icism with reference to cognitive contents. Kant (351). Subdivision II. A Critique of Rationalistic Absolute Genetic Mo- nism. Absolute rationalism in the account of the contents as well . as the forms of cognitive experience (352). Approximations to this view in Descartes and Leibniz (352). Fichte (353), Hegel (353), Green (353), H. Cohen et al. (353). Subdivision III. A Critique of Empirical Absolute Genetic Monism. Absolute passive empiricism in the account of the forms as well as the contents of cognitive experience (353) . Hume (353), Jas. and J. S. Mill (354), Spencer (354). Subdivision IV. Constructive Statement : Critical Empirical Genetic Monism (Critical Monism in Genetic Logic). An activistic em- pirical account of the genesis of the forms as well as the contents of cognitive experience (355). Approximation to this view by James (355). Three defensible theories of the apriori which would make possible a critical realistic epistemological monism : (a) Inheritance of some active psychical characters empirically acquired by ancestors (359). (6) Rapid individual acquirement from experience on the basis of a special mutation (359). (c) A combination of (a) and (6) (359). Criticism of extreme aposteriorism (363). Summary (364). PART II. THE PBOBLEM OF MEDIATE KNOWLEDGE (367). Division A. The Problem of Truth. (Logical Theory, or Philosophical Logic) (367). Introductory (369). Subdivision I. A Critique of Absolute Logical Dualism. Two irre- ducibly different criteria and definitions of truth, the one intel- lectualistic, the other anti-intellectualistic, e.g. practical (Kant) (370). ANALYTICAL TABLE OF CONTENTS XVU Subdivision II. A Critique of Intellectualistic Absolute Logical Monism. The criterion and definition of truth purely intellectu- alistic (identity with reality) (371), A. Absolute Intellectualism in combination with Epistemological Dualism (371). Locke (371), Leibniz (371), Lotze (372), Lovejoy (372), Boyce Gibson (373). B. Absolute Intellectualism in combination with Idealistic Episte- mological Monism (374), Hegel (374), J. Watson (375), Bradley (375), Bosanquet (381), Joachim (383), Koyce (384). C. Absolute Intellectualism in combination with Realistic Episte- mological Monism (391). Aristotle (391), Aquinas (391), McGilvary (392), Perry (393), Montague (393), Holt (394), Marvin (394), Boodin (395), B. Russell (396). " The Law of Significant Assertion " (398). Subdivision III. A Critique of Anti-Intellectualistic Logical Monism. The criterion of truth not discoverable and truth not definable intellectualistically, but only in an anti-intellectualistic way (401). a. Anti-conceptualism. Truth to be learned by reverting to pure intuition, without concepts (401). Bergson (401). Cf. James (406). /3. Current Pragmatism (407). Essential pragmatism. Practical value of some sort the crite- rion of truth (407) . 1. Approximations to essential pragmatism (409). A. Semi-pragmatism. Movements in the direction of pragmatism, which yet fall short of the pragmatic criterion of truth (410). Peirce (410), Baldwin (411), Boodin (412), Royce (413), Hocking (413). B. Quasi-pragmatism. Practical values the measure of what, for practical purposes, and even in science, we are justified in taking as truth ; but this not necessarily real truth (414). Mach (414), Poin- car^ (414), Vaihinger (415), Bergson (415). Appreciation, criticism and explanation (416). 2. Developments beyond essential pragmatism (417). A. Pseudo-pragmatism. All practical value of ideas or judgments an indication or proof of their truth (417) . James (418), Schiller (419), the Chicago School (420). B. Hyper-pragmatism. Practical value (usefulness) or the process of working (verification) the essential nature of the truth (422), James (422), Schiller (423), Dewey (423), A. W. Moore (423), Bawden (423), Boodin (423), The conversion of J. E. Russell (423). Criticism and explanation (426). xviii ANALYTICAL TABLE OF CONTENTS Further problems for the essential pragmatist (431) : The avoidance of ultra-utilitarianism (431), and ultra-individualism (433). The interpretation of consistency and the theoretical interest (435). Subdivision IV. Constructive statement : Critical Pragmatic Logical Monism {Critical Monism in Logical Theory). The proximate genus in the definition of truth derived from intellectualism, and the differentia of the species from pragmatism (438). The defect of presentation and need of representation (439). The criterion of truth found in practical identity ( functional equiva- lence) of predicate w^ith subject, of idea with reality, or with fur- ther experience of reality (440) . Definition of truth according to representational pragmatism (443). The ideal element in truth (446) . The essentially human character of truth as defined (446) . Further definition of the ideal (448). The permanence of truth (449). The need of transcending mere representational prag- matism (450). Definition of truth according to revised (intu- itional or scientific) representational pragmatism (452). Further tests of representational pragmatism (455). Division B. The Problem of Proof (Methodology) (459). a. The problem of certainty (459). 6. The problem of proof as the problem of the production of logical certainty, the problem of scientific method (461). Subdivision I. A Critique of Absolute Methodological Dualism. Deduction and induction as tyfo irreducibly different scientific methods of proof (461). ' Subdivision 11. A Critique of Rationalistic Absolute Methodological Monism. Deduction the only scientific method of proof (462). a. Pre-Kantian rationalism. Views of Descartes, Spinoza, and Leibniz (462). h. Rationalism of the " logisticians. " Views of Dedekind, Cantor, Frege, Russell, Couturat, Royce, et al. (462). c. The rationalistic dialectic. Views of Hegel and McTaggart (470). Subdivision III. A Critique of Empirical Absolute Methodological Monism. Induction the only scientific method of proof (472). a. Pre-Kantian empiricism. Bacon to Hume (472). h. Post-Kantian empiricism (473). Comte (473), J. S. Mill (474), Bergson (475). Subdivision IV. Constructive Statement : Critical Empirical Meth- odological Monism. {Critical Monism in Methodology.) The one scientific method of proof both inductive and deductive (476). Views of Kant (476) and Poincar^ (477). Scientific procedure (487) ; definitions (488) ; assumptions (488) ; empirical data (490) ; fundamental principle (490) ; laws and inductive methods (490) ; theory (492) . General Conclusion. Critical Monism (494). THE PROBLEM OF KNOWLEDGE THE PROBLEM OF KNOWLEDGE CHAPTER I Introductory: Philosophy and its Principal Problems Of all intellectual enterprises philosophy is perhaps the most difficult to define. A glance over the course of what is called the history of philosophy reveals not only a disappoint- ing transitoriness of solutions and lack of unanimity among philosophers as to methods and presuppositions, but what is much more disconcerting, an almost total shift from time to time in the problems themselves. It will not do, however, to conclude at once that the grouping together of the problems generally called philosophical has been purely arbitrary. The resort to etymology for purposes of definition is com- monly of doubtful wisdom ; and yet in the present instance it puts into our hands a clew which may conduct us through the maze of historical transformations to our desired definition. The philosopher has been from the first, as his name proclaims him, a lover of wisdom; and philosophy has always been, in spite of those admirably modest utterances of Pythagoras and Socrates, not the love of wisdom simply, but the best wisdom of the lover of wisdom. But one must not take too rigidly in this connection the distinction between wisdom and knowledge. In the beginning the term ''philosophy" seems to have been used to cover all such knowledge as was not either the common possession of the community or the immediate result of some special ex- perience of the individual. It was applied to whatever there existed of those organized bodies of adequately verified knowl- edge, the special sciences, including mathematics. Nor is it very long since this broader use of the term was given up. Even within the memory of persons still living the physical sciences bore the name of ''natural philosophy," and apparatus B 1 2 THE PROBLEM OF KNOWLEDGE employed for experimental purposes could be referred to as ''philosophical instruments." One might almost say that originally philosophy held all the special sciences in solution, and that of late these sciences have been crystallizing out and taking on a relatively independent existence. The analogy is somewhat misleading, however; in both philosophy and the sciences there has been from the beginning a process of growth, of creative becoming. The special sciences, which have been differentiating them- selves out from the matrix of philosophy, are commonly classified as abstract, descriptive, and normative. Of these three groups the first and last are most readily understood in relation to the second, the descriptive sciences. These are constituted of generalizations as to the relations of quaHties and processes in experienced objects or groups of objects. Astronomy and chemistry, biology and anthropology, psychology and sociology, will serve to represent the class. The abstract sciences, such as arithmetic, geometry, and even mechanics, deal with isolated aspects of reality. Their laws are accurate, but essentially hypothetical; they state what would be verified in experience if the ideal conditions which they assume were ever actualized. The abstractness, however, is only relative ; all generalization, even such as occurs in the descriptive sciences, is more or less abstract, and the propositions of even the most abstract sciences are descriptive — with certain provisos — of reality. This is true even where, as in the case of the non-Euclidean geometries, the provisos are contrary to experience, thus making difficult the empirical verification of their conclusions.^ The normative sciences, finally, are made up of generalizations, selected from the results of the descriptive sciences and organized into a system of rules for the realizing of an end. The normative sciences which figure most largely in relation to philosophy are logic, aesthetics, and ethics, the laws of which are rules for the realization of truth (or at least of consistency, which is hypo- thetical truth) , beauty, and moral goodness, respectively. Thus, while the laws of descriptive sciences are categorical, and those of abstract sciences hypothetical, the laws of normative sciences are always either categorically or hypothetically im- 1 See Ch. XX infra. INTRODUCTORY 3 perative. The ''applied sciences" may be regarded in general as complex and loosely organized normative sciences. But, as the special sciences multiply, what is to become of philosophy? As new sciences continue to detach themselves from the parent body of philosophy, the question is naturally raised as to whether it is not to be expected that as the sciences increase, philosophy must decrease, and even as to whether in fact philosophy is to be regarded as anything more than the rapidly disappearing remainder of prescientific thought? It would almost seem that, to quote the words of Windelband, ''philosophy is like King Lear, who divided all his goods among his children, and it must now befall him to be cast out as a beggar upon the street." ^ But another interpretation of the present situation is possible. May it not be that philosophy, as the characteristic wisdom of the lover of wisdom, has been finding out by a process of elimina- tion, as the sciences develop, just what are its own proper and persistent problems? Is it not discovering that its peculiar task is not to be a science — not even a "science of the sciences," which would be simply another special science, however im- portant — but that its main business is to arrive at a wise estimate of the world we live in, of ourselves and our ideas, and of the wise man's way of living? The philosopher still finds much — perhaps more than ever — to occupy his thought in questions concerning reality in its broader aspects, con- cerning life and its ideals, and concerning the relation between these two, reality and ideals. It has been recently claimed by the Danish philosopher, Harald Hoeffding, that the persistent problems of philosophy are four : the problem of consciousness, the problem of knowl- edge, the problem of being, and the problem of values. ^ But these problems are not all mutually exclusive, and the really ultimate problems of philosophy may be reduced to a simpler classification. The problem of knowledge, in so far as it tran- scends psychology and logic, belongs to the problem of values ; it is concerned with estimating intellectual value. The prob- lem of consciousness in turn, in so far as it transcends empirical ^Praeludien, 4th ed., 1911, Vol. I, p. 19. « Hoeffding, The Problems of Philosophy, Eng. Tr., New York, 1905. 4 THE PROBLEM OF KNOWLEDGE psychology, is reducible in part to the problem of being, and in part to the problem of knowledge ; it has to determine on the one hand what, in general, consciousness is, i.e. what place it has in the realm of being, and on the other hand what it is as awareness, i.e. as knowledge. Manifestly, then, Hoeffding's philosophical problems may be reduced to the problem of being and the problem of values, or, in other words, to metaphysics and criticism. That all the problems left over from the sciences for " wis- dom" or philosophy to deal with are problems of either meta- physics or criticism may be confirmed by an examination of the historic problems of philosophy. Corresponding to each of the normative sciences there is an elementary branch of critical philosophy. Thus philosophical logic discusses the nature of the ideal or value which the rules of logic as a norma- tive science subserve ; its problem is that of the nature of truth. Similarly ethics as a branch of critical philosophy is concerned with the question. What is moral goodness ? And philosoph- ical aesthetics with the problem. What is beauty? One might even go farther and speak of the problem of philosophical eco- nomics, or. What is wealth ? the problem of philosophical poli- tics, or, What is good government? and so on throughout the entire list of philosophical counterparts to the normative sci- ences. They are all problems of value. There are other problems of value, however, besides these relatively simple questions as to the nature of valid ideals. There are the problems as to the value of certain complex actualities, such as human knowledge, human religion, and human development in general. These problems give rise to those branches of critical philosophy known as epistemology, the philosophy of religion, and the philosophy of history. It is characteristic of these complex branches of critical philosophy, these philosophical critiques of phases of actuaHty, that on the one hand they make use of certain sciences, and on the other hand they are each intimately related to metaphysics. Thus epistemology, in so far as it is concerned with investigating what ''knowledge" is, must necessarily make use of psychology and logic ; and in so far as it is concerned with estimating the worth of our best ''knowledge," it must reach a positive INTRODUCTORY 5 and favorable conclusion, if our metaphysics is not to be re- duced to vain imagination. The philosophy of religion makes fundamental use of the history and psychology of religion in solving the fundamental problems of the nature of the reli- gious value or ideal and the essence of religion. It also in- cludes, besides this historico-psychological part, an ethical part, undertaking to estimate the moral value of religion, and an epistemological part, dealing with the value of religious experi- ence for knowledge, and especially for knowledge of a religious Object. But if the epistemological philosophy of religion should result in establishing the validity of religious knowledge, a further development of the philosophy of religion would be called for, viz. a metaphysical discipline, undertaking to formu- late our knowledge of the religious Object in systematic unity with the rest of our metaphysical knowledge. The philosophy of history also, while drawing upon a vast number of special sciences, both descriptive and normative, and especially upon descriptive historj^, may be either a branch of critical philos- ophy or a metaphysical discipline, or both. Not only is it concerned with a critique of progress, the norm of which is made up of the ideals established in the other branches of critical philosophy; it may include a metaphysical explana- tion of this progress as due to some theological or ontological principle, such as the will of God, or the evolution of the "Absolute Idea." But while some philosophical disciplines are thus partly critical and partly metaphysical, it would seem that there are, over and above problems of criticism, no philosophical problems which are not problems of metaphysics, the theory of being, or reality. In undertaking to classify these problems of being, there is still a good deal to be said for the classic subdivisions, viz. ontology, or the philosophy of being in general, and the three parts dealing with particular forms of being, viz. psy- chology, which, as a branch of metaphysics, deals with the nature of the self (soul, mind, spirit) ; cosmology, which deals with the fundamental nature of the universe; and theologj^ undertaking to set forth the nature of the religious Object. More commonly nowadays, however, the problems of meta- physics are not grouped together in this manner, but stated 6 THE PROBLEM OF KNOWLEDGE separately, as, for example, the problem of mind and matter, the problem of contingency and order, the problem of mechan- ism and teleology, the problem of the One and the many, the problem of good and evil. We do not at this point undertake to say whether or not these metaphysical problems can be solved without any dependence upon the philosophy of values ; but it is quite evident that, with the partial exception of the last, they are not themselves problems of value, but problems of reality, and such, moreover, as are not capable of being ade- quately dealt with in any of the special sciences, whether descriptive, normative, or abstract. In the present volume our concern is simply with the prob- lem of knowledge. Before attacking the particular problems, critical and constructive, into which this general problem naturally falls apart, a few introductory remarks may be offered. Epistemology, let it be frankly admitted at the out- set, may be construed either as a science or as a department of critical philosophy, or as an aggregate of both. As a descrip- tive science it would assume, as all such sciences do, the actual- ity of its subject-matter. Assuming, then, that there is such a thing as knowledge, and that it is sufficiently accessible and distinguishable to be recognized and described, the science of epistemology would simply undertake to state the observed nature of the various types of knowledge-process, and thus to arrive at an adequate empirical definition of knowledge. It would necessarily make use of much of the materials also employed by logic as a normative science, and might even itself be turned into the normative form, in which case its fundamental assump- tion would be the possibility of realizing knowledge as an ideal. As descriptive science its question is. What is knowledge? or, How do we know? As a normative science its question would be, What must we do (or experience) in order to know? But a further question may also be raised, viz.. Is what we have and call our knowledge really what we take it to be? Is our ''knowledge'' really knowledge? Is knowledge a human possi- bility? Now this question cannot be adequately treated by merely calling attention to the fact that he who asserts that no knowledge is possible has already tacitly assumed what he explicitly denies (viz. that some knowledge is possible, if only INTRODUCTORY 7 the knowledge that knowledge is impossible) . There are many undogmatic agnostics who desire to be reassured that our best human knowledge-values are genuine and may be taken at their face value. What is called for is a branch of philosophical criticism, the critical evaluation of knowledge-claims. This critical philosophy of knowledge is lightly esteemed by some recent writers. F. J. E. Woodbridge, for example, while making ample room for the theory of perception as an experimental science, asserts that the function of philosophical epistemology is moral and spiritual only ; it can broaden one's spiritual vision and thus modify character, but it can make no difference to our knowledge.^ But even if we grant that it is quite possible for one to know without knowing that he knows, we are not obliged by this admission to subscribe to the generali- zation that our knowledge can never be affected by either our knowing that we know or our doubting that we know. We must not anticipate here the outcome of the discussion upon which we are entering, but if there is ever such a thing as knowl- edge without the ability to prove what one knows, it would seem quite possible that one who originally did know may come to doubt his knowledge until it ceases to be knowledge. If epistemology can remove such doubts with reference to genuine knowledge, it is calculated to affect not only the degree of certainty, but ultimately even the content of our knowledge. Woodbridge's denial of the knowledge-value of this philos- ophy of knowledge-values may be viewed as a rather violent reaction against the abuse of critical epistemology which has been characteristic of much recent philosophy. For a genera- tion or two it has been the custom to saddle upon critical epis- temology the task of bearing up a whole system of metaphysics. Epistemology may rightly enough pass judgment upon the question of the possibility of metaphysics, but to prescribe to metaphysics what must be its conclusions as to the nature of reality — this is another matter, and does not so manifestly lie within the province of a theory of knowledge. As typical instances of this too common tendency to exploit epistemology in the interests of a particular metaphysical doctrine, we may 1 "Perception and Epistemology" in Essays Philosophical and Psychological in Honor of William James, 1908, especially pp. 140, 151-7, 163-6. 8 THE PROBLEM OF KNOWLEDGE cite the philosophical arguments of John Watson and G. T. Ladd. In the philosophy of the former, whose position fairly represents the absolute idealism recently dominant among Brit- ish-American philosophers, epistemology may almost be said to be reduced to the old Platonic exposure of the self -refuting character of that dogmatic absolute agnosticism which main- tains that no knowledge whatever is possible. Not only is the fact overlooked that at the root of even the most self-contra- dictory statements of dogmatic agnosticism there is real un- certainty as to the genuine validity of what we call our knowl- edge, and that the essence of agnosticism lies in this inexpugnable uncertainty rather than in the dogmatic denial ; what is more to the point is the fallacious interpretation of the denial of the universal negative as justifying the definite affirmative that the universe is intelligible. On this basis it is concluded further, by virtue of the ambiguity of one or other of the terms ''in- telligible" and ''rational," that the universe is rational and, as such, spiritual.^ In the philosophy of Ladd the dependence of the content of metaphysics upon epistemology is still more marked. His epistemology is more elaborately developed than that of Watson. Besides the psychological investigation of what it is to know, he would include in it an investigation of the guaran- ties, limits, underlying logical principles, and metaphysical presuppositions of knowledge. ^ It is in undertaking to state these "presuppositions or implicates" of knowledge that Ladd is led into that sort of dogmatism which has invited such reac- tions, almost equally extreme, as that of Woodbridge to which we have referred. Starting with the highly dubious assumption of a mutually exclusive relationship between the immediately experienced and the independently real, this writer, in repudi- ating agnosticism, foredooms his theory inevitably (as will appear more fully in the chapters immediately following) to a dogmatism as absolute but as unnecessary as that dualism whose undesirable consequences it attempts to remedy. "To know" can only be, from this point of view, "to make an onto- 1 An Outline of Philosophy, 1898, Pref., p. vi; p. 37; The Interpretation of Religious Experience, 1912, Vol. I, p. 74. * Knowledge, Life and Reality, 1909, p. 57. INTRODUCTORY 9 logical leap, a spring from the charmed circle of pure subjec- tivity into the mystery of the real." This dogmatism is veiled under such expressions as that he who claims to know exemplifies the confidence of human reason in itself, and pre- supposes that something is real and that innumerable real selves and real things are known to be existent and to be actually related in one world. What we immediately experience, it seems to be assumed, is only our conscious experience, i.e. the Hfe of the self with its conscious content, no part of which can exist independently of the self. Assuming then that we can know the independent world at all, we must conclude that it is like what we (subjectively) experience ; it must be apprehended in terms of ^'a Personal Life." Thus a metaphysical interpre- tation of the independently real world is based upon the sup- posed necessity of a particular presupposition of the theory of knowledge.^ But to put the problem of epistemology thus, How is it possible to know what is beyond myself, when any- thing, in order to be presented to me, must enter into my con- sciousness and thus become a part of my mind? is, as will be shown more fully presently, to raise an insoluble problem. To suppose that what is forever beyond my subjective experi- ence must be like what I subjectively experience, and therefore in the last analysis a Personal Life, is simply to advance a metaphysical dogma, while leaving the epistemological problem unsolved. But the insoluble epistemological problem is surely not the true one ; it must surely be due to a confusion of thought, to a faulty analysis of the nature of consciousness and the knowl- edge-relation. It is a problem not to be taken seriously, but to be gotten rid of; its solution can be accompHshed only in its dissolution. We would surmise, then, that while epistemology must not be used as a cloak for metaphysical dogmatism, there is never- theless place for a critical philosophy of knowledge which shall criticise wrong ways of stating the epistemological problem, as well as wrong solutions of the problem when rightly stated, and which shall also undertake, if such a thing should prove possible consistently with intellectual integrity, to vindicate the jus- ^lUd., pp. 119-24, 154, 159, 199-206; The Philosophy of Knowledge, 1897, pp. 22, 226-7, 366, 571 ; A Theory of Reality, 1899, Chs. XV, XX. 10 THE PROBLEM OF KNOWLEDGE tice of the natural human postulate that it is possible for us to learn to know reaHty and the truth about it. Without epistemology we may know, and not know that we know ; or we may not know^, and not know that we do not know. Epis- temology will have vindicated its right to exist if it enables us to know that we know, when we do know, and to know that we do not know, when we do not know. The problem of knowledge has two main subdivisions, the problem of immediate knowledge and the problem of mediate knowledge.^ The former is mainly concerned with the problem of acquaintance with reaUty, which is the subject-matter of epistemology proper. The latter, the problem of mediate knowledge, includes the problem of truth and the problem of its proof. All three problems involve the criticism of intellectual values, the problem of truth being the main content of logical theory, or logic as a branch of philosophical criticism. In dealing with the problem of acquaintance, however, much use must needs be made of the psychology of perception, while the psychology of judgment enters largely into the discussion of the problem of truth, and the psychology of reasoning, as well as the normative science of logic, into the problem of proof. It may be noted that corresponding to the problem of knowl- edge in general, and to each of its subordinate problems, there is a special problem of knowledge — the problem of the knowl- edge-value of religious experience and thought. It seems not unreasonable to suppose that an investigation of this special problem would prove to be the most interesting part of episte- mology ; but we shall not touch upon it in this volume. Our present concern is simply with the problem of knowledge in general. ^ Compare Wm. James's distinction between "knowledge of acquaintance" and "knowledge-about," The Principles of Psychology, 1890, Vol. I, pp. 221-2. PART I: THE PROBLEM OF IMMEDIATE KNOWLEDGE A. THE PROBLEM OF ACQUAINTANCE (EPISTEMOLOGY PROPER) 1. A CRITIQUE OF DUALISM CHAPTER II Dualism and Avowed Agnosticism In dealing with the problem of immediate acquaintance with reality, our procedure will be at j&rst critical. Before attempt- ing to set forth our own view, we shall undertake an examina- tion of current epistemological theories. A theory of knowledge may be either monistic or dualistic, and it may be either realistic or idealistic. Epistemological monism is the doctrine that the experienced object and the real object are, at the moment of perception, numerically one. Epistemological dualism is the doctrine that the experienced object and the real object are, at the moment of perception, numerically two. Epistemological realism is the doctrine that the real object can exist at other moments than the moment of perception, or of any other conscious experience, and independently of any such experience. Epistemological idealism is the doctrine that the real object can- not exist at other moments than the moment of perception, or of some other conscious experience, nor independently of such experience. The combinations of these doctrines which figure largely in contemporary philosophical discussion are episte- mological dualism and realism, epistemological monism and idealism, and epistemological monism and realism.^ Inasmuch, however, as we shall have to distinguish sharply our own point of view, which is a form of epistemological monism and realism, from the current forms of that doctrine, we shall commonly employ a slightly different terminology. In distinction from the view to be set forth constructively, which may be called critical epistemological monism, the doctrines to be criticised may be designated, respectively, absolute epistemological dual- ^ This is the terminology which was employed in the Report of the Com- mittee on Definitions at the 1911 meeting of the American Philosophical Asso- ciation. The definitions offered above also closely approximate those given in that report. See the Journal of Philosophy, Psychology and Scientific Methods, Vol. VIII, 1911, p. 703. 13 14 THE PROBLEM OF KNOWLEDGE ism, idealistic absolute epistemological monism, and realistic absolute epistemological monism. This terminology has an ad- vantage, moreover, in that it indicates more correctly than the other the parallel relations of the two types of absolute monism to the absolute dualism, and also to the critical monism. If it should seem desirable to avoid the term "absolute" in the designations employed, the distinctions between the views in question might be indicated by the terms epistemological mon- ism and critical realism, epistemological dualism and (critical) realism, epistemological monism and (dogmatic) idealism, and epistemological monism and dogmatic realism. Absolute epistemological dualism, then, is the doctrine 'Hhat perceived objects and real objects are never the same, though the former may be representative of the latter" ; or more fully, "that the perceived object and the real object are at the mo- ment of perception numerically two, and that the real object can exist at other moments independently of any perception," ^ or, we may add, of any other conscious experience. It must be quite evident that this absolute dualism cannot promise much as a theory of knowledge. If what is immedi- ately experienced is never independent reality, and independent reality is therefore never immediately experienced, how can the subject of immediate experience ever know any independent reality? Any absolute duaHsm in epistemology is foredoomed, it would seem, to agnosticism. By some epistemological dual- ists the agnosticism is openly acknowledged and stoutly de- fended; others, however, seek to evade this consequence by one device or another. In the present chapter we shall con- sider some outstanding and typical instances of an absolute epistemological dualism accompanied by an avowed agnosticism. The most illustrious exponent of absolute epistemological dualism is Immanuel Kant. But in order to understand the historic foundations of this modern point of view we must go back at least as far as John Locke. Locke's Essay concerning Human Understanding embodies the result of an inquiry into "the original, certainty, and extent of human knowledge," un- dertaken, the author informs us, on the supposition "that the 1 Report of above-mentioned Committee, Journal of Philosophy, etc., Vol. VIII, 1911, p. 703. I DUALISM AND AVOWED AGNOSTICISM 15 first step towards satisfying several inquiries the mind of man was very apt to run into, was to take a survej^ of our own under- standings, examine our own powers, and see to what things they were adapted." ^ It is assumed at the outset that "the object of the understanding when a man thinks," what the mind is employed about in thinking, is always an idea in the mind of the thinker.^ It is subsequently argued that these ideas, of which we are immediately aware, are either simple products of the action of external things upon the senses, con- veyed by those senses to the mind as a passive receptacle, or else such combinations of these simple ideas as result from the activity of the mind in reflecting upon the ideas received through the senses.^ But it must not be supposed, Locke hastens to warn us, that the simple ideas of sensation always exactly re- semble the qualities of the external bodies which act upon our organs of sense. Only the primary qualities of bodies, qualities like soHdity, extension, i&gure, motion or rest, and number, wliich are ''utterly inseparable from the body, in what estate soever it be," are resembled by our ideas of those qualities. All colors, sounds, tastes, smells, and other secondary or sensible qualities are ''nothing in the objects themselves, but powers to produce various sensations in us by their primary qualities." ^ When Locke comes to state the results of this point of view in so far as they bear upon his problem of the nature, extent, and reality of human knowledge, we find that it is only with difficulty that he wins even the appearance of an escape from agnosticism. He really has two definitions of knowledge, one being the agree- ment of our own ideas with each other, and the other, which he illogically makes a subclass of the first, being the agreement of our ideas with real existence.^ It is with knowledge in the second sense that we are here concerned. Our knowledge of real existence, he claims, is of three sorts, viz. : intuitive, of our own existence; demonstrative, of God's existence; and sensi- tive, "of the existence of particular external objects by that perception and consciousness we have of the actual entrance of ideas from them." ^ It is with this last that we have occasion 1 Essay, Bk. I, Ch. I, §§ 2, 7. 2 76., § 8. 3 Bk. II, Ch. I. 4 Bk. II, Ch. Vin, §§ 8-10, 15. e Bk. IV, Ch. I, §§ 2, 7. e Bk. IV, Ch. II, § 14 ; Ch. III. 16 THE PROBLEM OF KNOWLEDGE to deal in the present connection. That Locke felt keenly the problem as to the possibility, from his point of view, of this *' sensitive knowledge," is apparent from his own words. ''It is evident," he says, ''that the mind knows not things im- mediately, but only by the intervention of the ideas it has of them. Our knowledge therefore is real only so far as there is a conformity between our ideas and the reality of things." "But," he asks, "how shall the mind, when it perceives nothing but its own ideas, know that they agree with things them- selves?"^ Referring to "simple ideas" derived from sensa- tion, his answer is that since these are the product of things operating on the mind in a natural way, they produce therein "those perceptions which by the wisdom and will of our Maker they are ordained and adapted to," and thus necessarily "carry with them all the conformity which is intended, or which our state requires." ^ Complex ideas of substances, being our own product, can only be known to be true when they are made up of such simple ideas as are known to coexist in nature. Even here, then, he claims, we have knowledge which, while not very extensive, is nevertheless real; our ideas, though not, perhaps, very exact, are yet true copies.^ Nevertheless it is manifestly not as the outcome of his view of the nature of the mind and its ideas, but in spite of it, and by reason of his sound common sense, that he lets either simple or complex ideas of external substances "pass under the name of 'knowledge.'" While "going beyond bare probability," all assurance as to external objects, since, according to his theory, it falls short of either intuitive or demonstrative certainty, "is but faith or opinion." ^ That there can be no knowledge of the qualities, or even of the existence of independent physical bodies on the basis of the complete epistemological dualism and passive empiricism of Locke's theory was soon made evident through later develop- ments of English philosophy. Starting out from the Lockian view that all the materials of knowledge are passively received by the mind from without in sensation, Berkeley roundly denied the necessity of assuming any independent physical 1 Essay, Bk. IV, Ch. IV, § 3. ^ lb., ^ 4. 8 lb., § 12. 4 Bk. IV, Ch. II, § 14 ; Chs. XI, XV, § 3. DUALISM AND AVOWED AGNOSTICISM 17 things whatsoever. Hume in turn showed — as far as the sceptic can show anything — that upon Locke's principles no genuine knowledge of independent, external reality is possible at all. '^'Tis impossible upon any system/' he declares, ''to defend either our understanding or senses ; and we but expose them farther when we endeavor to justify them." "Careless- ness and inattention alone can afford us any remedy." ^ This agnosticism was the logical outcome of Locke's absolute episte- mological dualism. Hume's problem was inherited by Kant. ''I confess it freely," the latter writes in the oft-quoted passage in the Prolegomena to any Future Metaphysic, ''the remembrance of David Hume was the first thing which many years ago inter- rupted my dogmatic slumber, and gave to my investigations in the field of speculative philosophy a quite new direction." Thenceforward Kant's problem was how to conserve at the same time the good in rationalism without its dogmatism, and the good in empiricism without its scepticism. He never gave up the rationalistic conviction that validity in things human is always to be determined by agreement with those universal forms of rational consciousness which — as he continued to believe — do not originate from experience, but are inherent, a priori. Still, to avoid dogmatism, he recognized as just the scientific principle that nothing should be admitted as knowl- edge but what has been verified within human experience. His chosen philosophical method, by which dogmatism and scepticism were both to be avoided, was what he termed " criticism," the search for the a priori or rational elements by which validity is imparted to empirical judgments. His philosophy as a whole is best understood from this point of view. It is, in all its parts, the rationalistic criticism of ex- perienced values, intellectual, moral, aesthetic, and religious. In the Critique of Pure Reason the aim is to vindicate the validity of experienced intellectual values by showing the a priori element involved in cognitive experience ; the Critique of Practical Reason aims to show the validity of experienced moral values, by pointing out the a priori element involved in moral experience ; for a corresponding purpose the first part ^ A Treatise of Human Nature, pp. 218, 268. C 18 THE PROBLEM OF KNOWLEDGE of the Critique of Judgment is concerned with the a priori element in aesthetic experience, while the second part of the same work, together with the volume entitled Religion within the Limits of Mere Reason, attempts the same thing for the values ex- perienced in religion, both ''natural" and ''revealed." In Kant's critical philosophy, then, as Hans Vaihinger remarks, the empiricism is rationalistic, and the rationalism empirically conditioned.^ In the present discussion our concern is with this combina- tion of rationalism and empiricism, in so far as it affected the problem of know^ledge. Here it was maintained with the rationalist as against the sceptical empiricist that genuine knowledge is possible in mathematics and natural science. To the sceptic it was conceded, however, in opposition to the dogmatic rationalist, that metaphysical knowledge is im- possible. The possibility of mathematical, scientific, and metaphysical judgments is explained in rationalistic fashion as dependent upon the synthetic activity of reason with its a priori forms, principles, and fundamental ideas. The validity of mathematical and scientific judgments as knowl- edge is explained in empirical fashion as due either to the fact that they synthesize what is given in sense-experience, as in the natural sciences, or else to the circumstance that they set forth w^hat, according to the inherent constitution of the per- ceptive and thinking faculties, are the necessary forms of all possible sense-experience, as in geonaetry. The impossibility of arriving at valid metaphysical knowledge, however, is explained as due to the fact that the constructions of rational psychology, cosmology, and theology go beyond all possible human experience and are therefore, as the empirical sceptic maintains, pure dogma. Thus there could be no judgments at all without the activity of a priori factors. But, on the other hand, these judgments do not become knowledge save as they embody the materials of sense-experience, or operate with the necessary forms of all possible sense-experience. On this side of his thought it would seem that Kant not only anticipated Comte's positivistic rejection of meta- physics and identification of knowledge with science, but 1 Commentar zu Kants Kritik der reinen Vernunft, 1881, p. 55. I DUALISM AND AVOWED AGNOSTICISM 19 that he placed this positivism upon a rationahstic as well as an empirical basis. But in this fusion of continental ra- tionalism with English empiricism other far-reaching results were involved. In a sense Kant may be said to have taken over the subjectivism of the Humian empiricism. There is a partial truth in Vaihinger's contention that whereas the older rational- ism had been combined with objectivism, claiming to be able by means of reason to transcend the limits of human experience, and while the former empiricism in its final form was combined with subjectivism, the philosophy of Kant was a combination of rationalism with subjectivism.^ From this point of view the only objects known are one's own ideas, the contents of one's own consciousness. The rationalistic element gives a measure of relief, for while the physical objects known are de- pendent upon one's own sensations for their existence, they are also dependent upon an activity of mind which operates accord- ing to universal cognitive forms, so that to this extent we ex- perience our objects according to forms which are universally necessary for all possible human experience. It gives us a certain measure of consolation in our subjectivity to know that all others are in the same predicament; ''misery likes com- pany." But it is especially important to note that, on the basis of this combination of rationalism with the subjectivism of em- piricism, not simply the explicit judgment about objects is de- pendent upon an activity of mind; experience and all its objects, the world of nature itself, — all are products of the constructive activity of mind working upon the materials furnished through the senses. It becomes necessary to assume that all cognitive activity is a synthesis, a construction of its object. The understanding does not create its object out- right ; materials are furnished it to work upon. But out of these materials the human mind constructs the world of nature 1 Commentar, pp. 50, 52. Vaihinger remarks that the fourth combination, that of empiricism with objectivism, is an illogical and impossible one, as was demonstrated, he claims, by the outcome of Locke's attempt. This is un- doubtedly true when the empiricism is passive, as Locke's was essentially ; but whether there may not be developed another critical philosophy, different from the Kantian, which will combine, without violence to logic, an activistic empir- icism with an objectivistic or realistic epistemology, remains to be seen. 20 THE PROBLEM OF KNOWLEDGE and all its laws. This, then, is the Copernican revolution in philosophy, as Kant himself called it. As it is due to our posi- tion on the earth that the heavenly bodies seem to move around us, so it is because of the nature of the a priori forms of our sensibility and understanding that we have a world of objects existing in space and time, and governed in accordance with uniform laws. Or, in other words, just as, according to Coperni- cus, it is the movement of the earth in the solar system that accounts for apparent motions of the heavenly bodies, so it is the activity of our own reason in the world of the senses which accounts for the way the world of nature appears to us.^ But the question may well be asked whether the Kantian revolution in philosophy was as scientific and final as the Coper- nican revolution in astronomy. A truce was arranged between rationalism and empiricism, but the prospects for a final and satisfactory solution of the problem of knowledge were Httle brighter as a result of the Kantian criticism. The Humian scepticism reappeared in the ultimate agnosticism of the Kantian system. Even granting the truth of Kuno Fischer's contention ^ that the dependence of the world of nature upon human reason was intended to apply only to man as the subject, not to man as the object, of knowledge, this is not a real escape from subjectivism, but a mere cloak to conceal it ; man as ob- ject of knowledge includes all we can say about him as subject of knowledge. Kant's failure was also partly obscured by his introduction of the concept of "possible experience," which, instead of independent reahty, he makes the object of scientific knowledge. But '^ possible experience," so far as it goes beyond actual experience, is not reality at all ; and Kant was not able to dismiss from consideration an independent reality, a ''thing- in-itself," which could not be interpreted in terms of possible experience, but which must be assumed as the cause of the sense-impressions given to the mind from without.^ 1 See Kuno Fischer's Immanuel Kant, 4th ed., 1898, Vol. I, pp. 8, 9, and H. Hoeffding's History of Modern Philosophy, Eng. Tr., Vol. II, pp. 45-6. 2 Op. cit., Vol. II, p. 541. ' See, for example, Kritik der reinen Vernunft, 1st ed., p. 565 ; 2d ed., p. 537 ; Watson's Selections, p. 184. Also in the Metaphysic of Morality (Hartenstein, Vol. IV, 307 ; Watson's Selections, p. 258), the passage, where it is asserted that a certain contradiction "disappears if we say that behind phenomena there are DUALISM AND AVOWED AGNOSTICISM 21 According to the principles of the Kantian criticism, this reaUty, as forever beyond possible experience, can never be known; and yet, while not the object of possible knowledge, it is the object of a necessary question. We know that it is, and yet can never have valid knowledge of what it is. Thus we see that although Kant avoided Locke's view of the complete passivity of the mind in perception, he was forced into an episte- mological dualism more absolute and an agnosticism more critical but more pronounced and complete than that of his English predecessor. Even Locke's ''primary qualities" of bodies wefe relegated to the realm of mere appearance — al- though, within the limits of human experience, of appearance universally. The completeness of Kant's epistemological dualism and the reahty of his realism are perhaps best brought out by the ex- pression of his doctrine of the thing-in-itself in terms analogous to those which he himself applies to intuitions of sense and ob- jects of actual and possible experience. Intuitions and phe- nomena, according to Kant, are empirically real (real in experi- ence), but transcendentally ideal (not real beyond experience). Things-in-themselves, however, in the Kantian doctrine, are transcendentally real (real beyond experience), but empirically ideal (not real within experience ; nothing but empty ideas, so far as experience is concerned).^ Neo-Kantian idealists commonly try to eliminate Kant's reahsm as a non-essential feature, holding that the thing-in- itself is to be interpreted as having been intended as a mere ''limiting concept," marking the end of the possibility of ex- perience and knowledge. But this is true of the thing-in-itself only from the standpoint of experience; it is indeed empirically ideal. But none the less it is regarded by Kant as transcenden- tally real. If neo-Kantianism, i.e. Kantianism without the thing- things in themselves, which, though they are hidden from us, are the condition of phenomena." From all non-realistic points of view " possible experience " is not reality, but simply an abstraction representing past or future experience. From the realistic point of view the term may be taken as a not very accurate way of denoting independent reality, viz. as that which is, whether experienced or not, but which we ordinarily think of as it would be if it were experienced. 1 At this point I am indebted to C. M. Walsh's articles in Mind, N.S,, XII, 1903, pp. 454-72 ; XIII, 1904, pp. 54-71. Cf. K. Fischer, op. cit., Vol. II, p. 551. 22 THE PROBLEM OF KNOWLEDGE in-itself, were established as valid philosophy, it would be legitimate in dealing with opponents of idealism to regard this neo-Kantianism as the essence of Kantianism. But if neo- Kantianism is itself, like other forms of idealism, as we shall maintain (Chs. 'V-IX), untenable, then, in dealing with the idealist, the good essence of Kantianism is its realism, whereas, when one is defending monistic realism, the had essence of Kan- tianism is its dualism and consequent agnosticism. The application of the term '' noimienon" to the thing-in-itself is to be interpreted in agreement with what has been said. The term is not to be interpreted in the Platonic sense as signifying reality known by pure reason, although Kant borrows the term from Plato. It is rather to be taken as what Plato's noumenon becomes in the Kantian dualistic epistemology, viz. mere non-phenomenon from the standpoint of human knowledge, and yet what might be known through intellectual intuition by some superhuman mind. So interpreted, it becomes at once evident that the term ^' noumenon '^ is appropriate to designate the thing- in-itself, which, while transcendentally real, is empirically ideal.^ We conclude, then, that Kant's doctrine is an epistemological dualism so absolute as to leave the sphere of reality and the sphere of knowledge coincident at not a single point. Kant himself made a very notable attempt to overcome this agnostic dualism, at least sufficiently for the needs of the moral life, in his Critique of Practical Reason. This, however, was at best a palhative measure. It sought to relieve one dualism by introducing another, viz. the logical dualism of two fundamen- 1 In the second edition of the Critique of Pure Reason (see Miiller's transla- tion, p. 789) Kant says : "If by noumenon we mean a thing so far as it is not an object of our sensuous intuition, and make abstraction of our mode of intui- tion, it may be called a noumenon in a negative sense. If, however, we mean by it an object of a non-sensuous intuition, we admit thereby a peculiar mode of intuition, namely, the intellectual, which, however, is not our own, nor one of which we can understand even the possibility. This would be noumenon in a positive sense." See also Miiller's translation, pp. 206 ff. and 541-2 ; K. Fischer, op. cit., Vol. I, pp. 455-9 ; E. Caird, The Critical Philosophy of Kant, Vol. I, pp. 318, 649 ; Vol. II, p. 633 ; F. Paulsen, Immanuel Kant (Eng. Tr.), p. 200. Riehl seems not to appreciate sufficiently Kant's special use of the Platonic terminology. He says: "The idea of noumenon is a practical ideal concept, in using which Kant is in evident contradiction with his doctrine of the un- knowableness of the thing-in-itself." {Der philosophische Kriticismus, 1879, Vol. II, Part II, p. 29.) DUALISM AND AVOWED AGNOSTICISM 23 tally different kinds of truth, the dualism of reason and faith, or of theory and practice. For practical purposes we must act as if there were realities which, theoretically speaking, we can never know. Kant made application of this doctrine only in the realm of the moral consciousness. Still, the genuineness of his realism, and therefore of his dualism, is indicated by his giving to the practical reason the primacy over the pure or theoretical reason. A thoroughgoing application of this point of view throughout the whole domain of practical intellection was not made, prob- ably for the reason that it would have seriously discredited the fundamental assumptions of the Critique of Pure Reason, and the so-called Copernican revolution would have had to be followed by a counter-revolution, which might even have amounted to a return to epistemological monism and realism over a pathway similar to that being taken by some modern pragmatists. As it is, the final word, so far as exposition is concerned, would seem to be that Kant was a remarkably con- sistent dualist — so consistent, indeed, that he even ventures to contradict himself. What is true from one point of view (theoretical reason) is false from another (practical reason) ; theoretically we have not, practically we have, knowledge of independent reality ; not to have contradicted himself at certain points would therefore have been in Kant a mark of incon- sistency with his logical and epistemological dualism. The secret of the Kantian agnosticism lies in the will to be a realist, even before any fallacy in subjectivism has been dis- covered. It is the lingering presence of the Humian sceptical empiricism, which Kant's formal rationalism was never able fully to overcome. It is the result of a natural suggestion arising from insufficiently critical thought, whose fallaciousness may be exposed by its being expressed in syllogistic fashion. Either of the following syllogisms may be taken as fairly representative of the reasoning that commonly leads to philosophical agnosti- cism. What I suppose to be experience of independent reality is included within what I experience; but mere sense-im- pressions which I do not know to be valid of independent reality are also included in what I experience ; therefore what I sup- pose to be experience of independent reality is mere sense-im- pression, which I do not know to be vaUd of independent reality. 24 THE PROBLEM OF KNOWLEDGE Again, what I suppose I know is included in what I think; but what I merely think and do not know is also included in what I think ; therefore what I suppose I know is what I do not know, but merely think. The fallacy in both syllogisms is that of reasoning by means of an undistributed middle term, and thoroughgoing agnosticism is the inevitable result. A measure of apparent relief is obtained by the device to which we have already referred, the use of the abstract concept of ''possible experience" as a substitute for that of ''reality"; this, however, simply covers up, but does not solve, the theoreti- cal problem. It amounts to the dogged determination to put up with the lack of real objectivity in our knowledge, since every other human being is obliged to do the same. It is the regarding of subjectivity as if it were objectivity, simply on the ground that it is a necessary and universal subjectivity. There is one further criticism of the Kantian dualism and agnosticism which may well be mentioned here, and which amounts to the charge that if one will be as agnostic as Kant, he must logically be more agnostic still. This was virtually the position taken by G. E. Schulze, one of the earliest of Kant's critics, who maintained ^ not only that if the categories are not to be applied to things-in-themselves, we can have no certain knowledge of the existence of such things, but that moreover we can have, on Kantian principles, no such knowledge of the a priori conditions of human experience as the Kantian criticism assumes to be possible. If there is knowledge only when there is empirical intuition, there can be no knowledge of an absolutely a priori activity; the supposed absolutely a priori conditions of experience can never be objects of experience and so are to be regarded as unknowable things-in-themselves also. This seems to be a valid redudio ad absurdum of the Kantian combina- tion of an absolute apriorism with an absolute metaphysical agnosticism. One or the other, and in the end probably both, must go. The influence of Kant upon the development of philosophical doctrine since his day has been tremendous. From the per- spective of our own day the fundamental divisions of the history of modern philosophy are the pre-Kantian, the Kantian, and 1 ^nesidemus, 1792. DUALISM AND AVOWED AGNOSTICISM 25 the post-Kantian. The great bulk of post-Kantian philosophy, moreover, is perhaps most instructively represented as falling under one or another of the three following general characteriza- tions : first, variations within the limits of the original Kantian dualism, with the acceptance of its inevitable agnosticism; second, a series of idealistic movements, stimulated and largely guided by the Kantian analysis of the a 'priori conditions of experience, and claiming to transcend the agnostic difficulty; and third, a development of the reahstic side of the Kantian dualism, as an expression of the desire to escape from Kant's absolute agnosticism as to independent reality. Each of these movements represents a way of attacking the problem of the thing-in-itself, which Kant bequeathed to those who should come after him. The more extremely duaUstic and agnostic type of thought has retained the thing-in-itself, holding that while we cannot know what it is, we must believe that it is. The idealistic movement in its earlier post-Kantian form, typi- cally represented by Fichte and Hegel, interpreted the thing-in- itself as being just rational thought, with the result that nothing else could be so surely regarded as progressively knowable by man as could this same thing-in-itself; latterly, however, idealism has rejected the thing-in-itself, on the ground that we cannot know what it is, and therefore cannot know that it is. The less agnostic realistic movement not only retains but seeks to describe the thing-in-itself ; it holds that we both know that it is, and may gain some real knowledge — indirectly, by in- ference from our empirical knowledge — as to what it is. The more frankly agnostic tendency we shall deal with in the present chapter. The two following chapters will take up the more realistic and less agnostic movement. Thereafter we shall immediately turn to a critical examination of the various types of theoretical idealism. Among those, who, following Kant more or less closely, have brought into bold relief the agnostic implications of epistemo- logical dualism, Hamilton and Spencer, in England, and Riehl and Dilthey, in Germany, will chiefly occupy our attention. The philosophy of Sir William Hamilton shows the influence of Reid's "philosophy of common sense," and in still greater degree that of the Kantian dualism of phenomenon and thing-in- 26 THE PROBLEM OP KNOWLEDGE itself. The philosophies of Reid and Kant, however, have about as much aifinity for each other as oil and water, and in the Hamiltonian doctrine we see sometimes the one ingredient and sometimes the other, but never a compound of the two.^ De- veloping the Kantian doctrine of the constructive function of the understanding in cognition, Hamilton insists that since to think is to condition and limit the object of thought, and since, of course, thought is involved in all human cognition, the only knowable objects are such as are conditioned, limited, modified in and through the process of becoming known. Reality, as it would be apart from the spatial, temporal, qualitative, causal, and other conditions imposed by human thought, is thus for man forever unknowable. The following strongly agnostic expressions are typical: ''AH qualities, both of mind and of matter, are ... only known to us as relations ; we know nothing in itself."^ ''Of things absolutely or in themselves ... we know nothing, or know them only as incognizable. . . . All that we know is . . . phenomenal, — phenomenal of the unknown." ^ "We may suppose existence to have a thousand modes ; but these thousand modes are all to us as zero, unless we possess faculties accommodated to their apprehension. But were the number of our faculties coextensive with the modes of being, — had we, for each of these thousand modes, a separate organ competent to make it known to us, — still would our whole knowledge be, as it is at present, only relative. Of exist- ence absolutely and in itself we should then be as ignorant as we are now." ^ But while we cannot know what the Absolute or Uncondi- tioned is, that it is is a conviction from which we cannot escape. As Reid maintained, the original pronouncements of conscious- ness, underlying, as they do, all human thought, must be accepted as true; and one of these original pronouncements is the inescapable conviction that a world exists independently of consciousness. "By a wonderful revelation we are, in the very consciousness of our inability to conceive aught above 1 Cf. A. Seth (Pringle-Pattison), The Scottish Philosophy, 1885, p. 149. 2 Hamilton, Reid' s Collected Writings, 6th ed., 1863, p. 965. 8 Discussions, p. 608. * Lectures, 1st ed., I, 153; 1874 ed., I, 107. DUALISM AND AVOWED AGNOSTICISM 27 the relative and finite, inspired with a belief in the existence of something unconditioned beyond the sphere of all comprehen- sible reality." ^ It is claimed, however — strangely enough, as it may seem — that while the categories of human thought are not applicable to the Unconditioned, the principles of formal logic nevertheless may and must be applied to our thought of that unthinkable Reality. According to the Law of Excluded Middle the Un- conditioned must be either limited or unlimited, and our choice between these two'alternatives is rightly determined by practical, ethical considerations in favor of the unlimited Unconditioned, or infinite Absolute. Such an Object, while it cannot be con- ceived as it is, may be conceived in its relation to the finite and conditioned as the human soul is related to the human body. But in concluding that we may think of the unconditioned Unthinkable as related, and therefore conditioned, Hamilton surely comes perilously near to furnishing the redudio ad absurdum of his own philosophy .^ Henry Mansel, the most important of the close followers of Hamilton, is chiefly remark- able for the way in which he developed his master's doctrine on the theological side, making the agnosticism an argument for a reason-defying, traditionalistic religious dogmatism. Herbert Spencer's agnosticism, while not altogether un- original, shows the influence of the views of Hamilton and Mansel on the one hand, and of the more characteristically English thought of John Stuart Mill on the other. The philos- ophy of Mill, in turn, might almost be regarded as a synthesis of the sceptical empiricism of Hume and the positivism of Comte. This important French philosopher should not, how- ever, be considered as in any pronounced sense an agnostic or a dualist in epistemology. He rejected theology and meta- physics, not because he was interested to maintain the un- knowableness of God or of ultimate Reality, but rather because he regarded both these forms of thought as antiquated and in- adequate methods of interpreting the world of nature and of man, which is eminently accessible to experience and knowable ^Discussions, p. 15. Cf. Reid's Collected Writings, pp. 747a, 750a, 7616; Lectures, 1st ed., I, 220; IV, 62. « Cf. Hoeffding, History of Modern Philosophy, Eng. Tr., Vol. II, p. 390. 28 THE PROBLEM OF KNOWLEDGE by the methods of empirical science. In Spencer, then, we seem able to trace the sceptically inclined English empiricism, the positivistic reduction of philosophy to empirical science, and, as mediated by Hamilton and Mansel, the Kantian agnosticism with reference to absolute, independent Reality. In the light of these antecedents it is easy to appreciate Spencer's arrangement of his thoughts on the "first principles" of philosophy under the two heads of 'Hhe Unknowable" and the '^Knowable." In his philosophy of the Unknowable, Spencer maintains that the conflict between science and religion has been partly due to the dogmatizing of scientists beyond the proper sphere of science. If the scientist is sufficiently critical of his own fundamental concepts, "he, more than any other, truly knows that in its ultimate nature nothing can be known." "Ultimate scientific ideas are all representative of realities that cannot be comprehended." Space and time are wholly incomprehensible. Taken objectively, they can be conceived neither as entities nor as attributes of entities, nor yet as non-entities. Taken subjectively, they would be the mere forms of intuition, but we have the direct testimony of consciousness that they enter into the objective content of intuition. The case is similar with the concepts of force and matter. We must, and yet we cannot, think of matter as acting on matter through empty space. When we consider the concept of consciousness we are again face to face with an inscrutable enigma. Objective and sub- jective things are alike inscrutable in their substance and genesis.^ But that human intelligence is utterly incapable of knowing the reality which exists behind all appearances, may be exhibited, continues Spencer, in other ways besides this experimental testing of the ultimate ideas of science, and showing from the alternative impossibilities of thought invariably involved that all such ideas are mere symbols of the actual, not cognitions of it. The same conclusion as to the relativity of knowledge may be proved analytically. An analysis of the product of scientific thought shows that the particular is always explained by the more general, leaving the most general necessarily inexplicable. An analysis of the process of thought shows that we know by ^ First Principles, §§15-21. DUALISM AND AVOWED AGNOSTICISM 29 distinguishing relations, differences, and similarities; from which it may be inferred that the Absolute, as that of which no necessary relation can be predicated, is unknowable. Once more, the same conclusion also follows from the biological view of mind. What is true of life in general is true of intellectual life in particular. It is a continuous adjustment of internal relations to external relations ; each act of knowing is the forma- tion of a relation in consciousness answering to a relation in the environment, so that the external agency itself is never what is within consciousness. But then, all that is required for the purposes of life is that the internal actions should correspond with the external actions in their coexistences and sequences ; knowledge of what the things are in themselves is quite un- necessary.^ But what Spencer means is not simply that anything beyond the relative would be unknowable ; he is equally insistent that we must believe that something beyond the relative actually exists. ''In the very denial of our power to learn what the Absolute is, there lies hidden the assumption that it is ; and the making of this assumption proves that the Absolute has been present to the mind, not as a nothing, but as a something. '^ To say that our knowledge is limited to appearances necessarily involves the thought of a Reality of which they are the appear- ances, and the very demonstration that a definite consciousness of the Absolute is impossible to us unavoidably presupposes an indefinite consciousness of it. There is, indeed, as an indefinite thought formed by the coalescence of a series of thoughts, and forming the basis of our intelligence, an ever present sense of real existence, a nascent consciousness of space, for instance, beyond those bounds which we definitely imagine, or of a cause behind that cause which we have definitely in mind. From the impossibility of getting rid of the consciousness of an Actuality lying behind appearances, there results our indestructible belief in that Actuality .^ It was soon pointed out, in criticism of Spencer, that his doc- trine of an unknowable Reality behind Appearance was self- contradictory, in that the saying what anything is not always involves, in some measure, saying what it is. Spencer was 1/6., §§22-5. 2/6., §26. 30 THE PROBLEM OF KNOWLEDGE forced to admit the justice of this criticism; but in reply he could only reiterate his former contention, that we cannot say anything concerning the non-relative without carrying into our propositions meanings connoted by words moulded on the relative.^ He was almost within sight of the real solution of the problem, however, when he said, ''Unless a real Non- Relative or Absolute be postulated, the Relative itseK becomes absolute." ^ Why should we not regard the distinction between the Relative and the Absolute as itself relative rather than absolute? Even an all-inclusive Whole must necessarily exist in relations — to its parts. There is, by reason of the imperfection of our knowledge, appearance which is to be distinguished from reaUty, and thus a relative which is not the absolute ; but is there any Absolute which is not essentially relative ? What we mean to imply is not subjectivism. The circumstance of anything's being relative does not mean that all its being is dependent upon its being in the relation of being known by a subject. Its being absolute may be relative to some human purpose, but its being is not necessarily relative to human purpose. What it is, however, does not need to be completely independent of all of its relations. If, then, it has not been shown that an Absolute cannot be at the same time relative, the a priori arguments of Hamilton, Mansel, and Spencer for the unknowableness of the Absolute fall to the ground. Before turning to a consideration of recent agnostic reaHsm among German Kantians, brief reference may be made to two EngHsh thinkers whose philosophical views will be discussed more fully in other connections, viz. F. H. Bradley and S. H. Hodgson. Bradley is noteworthy in this connection as having driven absolute idealism, under the lash of logical criticism, almost to the verge of the Spencerian agnosticism. His Ab- solute, so strongly contrasted with all appearances, is all but identical with Spencer's ''Unknowable." Hodgson, with his conception of experience becoming what we later recognize as reality, made an earnest effort to establish his metaphysics upon the ground of an epistemological monism ; but he did not quite succeed. On the one hand he argues that in the process of consciousness the object of consciousness is formed, but on 1 First Principles, Postscript to Part I. ^ lb., § 26. DUALISM AND AVOWED AGNOSTICISM 31 the other hand he is forced to admit that matter, as we know it, has conditions beyond those of our own consciousness, and that therefore sense-data are evidence of a reahty that is non-con- sciousness.^ So long as he adheres to epistemological monism, he is committed, as we shall see (Ch. VI), to what is virtually a disguised subjective idealism; so soon as he acknowledges realism, he lapses into epistemological dualism. He is thus forced to hover perpetually in unstable equilibrium between subjectivism and agnosticism. Among contemporary exponents of Kantian doctrine there is perhaps no one who so faithfully clings to the essentials of his master's position as does Alois Riehl. He frankly assumes realism in combination with an absolute dualism in epistemology at the outset, and adheres to this point of view with remark- able consistency throughout the entire course of his thought. *'I take the realistic hypothesis," he says, ''as my point of departure; I assume that something different from and independent of consciousness exists." ^ This is assumed as founded in a feeling of real existence other than appearance, that cannot be driven from even the most elementary form of our conscious life.^ But it is involved in this realistic assump- tion that the objects of our experience, which are, as such, dependent upon our consciousness, are doubly dependent, because consciousness is itself an appearance of something beyond it. Objects, then, are functions of functions, appear- ances within an appearance.^ Or we might proceed the other way about. Finding the marks of relativity upon both consciousness and the objects appearing within it, we would be compelled to assert an existence beyond consciousness. The evidence for the existence of relative forms is necessarily at the same time the proof of an existence which is not relative, i.e. of the Absolute. The idea of a thing-in-itself is indispensable for one who does not wish to regard his sensuous presentations as groundless.^ This thing-in-itself, moreover, is quite unknowable. Through the phenomenon of presenta- ^ "Matter," Proceedings of the Aristotelian Society, Vol. II, Part I, 1891-2, pp. 20, 24. 2 Der philosophische Kriticismus, 1879, Vol. II, Part I, p. 18. 3 /6., Part II, pp. 60-1. * 76., Part I, p. 18. »/&., Part I, pp. 18, 19; Part II, pp. 28-9. 32 THE PROBLEM OF KNOWLEDGE tion we are always necessarily separated from everything as it is in itself.^ We never experience or know the physical apart from the psychical, nor the psychical apart from the physical. Every relation perceived or presupposed among things is primarily a relation among our sensations.^ Indeed the thing, as we know it, is a constant group of sensations.^ Physical laws are funda- mentally laws of our sense-experience — not mine simply, but ours ; they state the experientially permanent similar conditions under which we obtain certain sense-experiences. They give no information regarding independent reality, for our different sensations are not signs of a process taking place in the thing-in- itself ; they are signs only of each other .^ Process and place are themselves simply phenomenal and relative.^ Atoms are the products of thought abstracting from the particular con- ditions of perception ; it is only through careless thinking that they are regarded as things-in-themselves.^ Indeed, although we have absolute knowledge that the thing-in-itself is, we are left in absolute ignorance of what it is. All our knowledge of prop- erties is relative. Properties are dependent upon conscious- ness, but existence is not dependent upon it; rather is con- sciousness dependent upon existence. Of the being of the object, as of the being of the subject, we have absolute knowledge. Cogito ergo sum et est. But of the object's being object, as of the subject's being subject, our knowledge is but relative.'^ It follows, of course, that Riehl's attitude toward science on the one side and metaphysics on the other is what might be described as critical or Kantian positivism.^ Metaphysics, as knowledge of ultimate reality, cannot be obtained by the methods of induction, which apply to phenomena only. There is no place for metaphysical hypotheses, for it is only in ex- perience that hypotheses can be verified. Metaphysical knowl- edge, therefore, can come only through pure reason, if at all. But an examination of metaphysical attempts reveals the fact that it is always some prominent individual characteristic of ^ Der philosophische Kriticismtis, Part II, p. 29. ^ 75.^ p. 30. » lb., Part I, p. 202. * lb., Part II, pp. 33, 151. 8/6., p. 312. « 76., pp. 31-2. 7 lb., pp. 130, 147, 150, 153, 297. « Cf. op. cit., Part II, p. 149. DUALISM AND AVOWED AGNOSTICISM 33 thought or of experienced reahty which is raised to the status of a metaphysical idea and made all dominant in the system. Metaphysical hypotheses, therefore, producing, as they do, the illusion of an all-comprehensive knowledge, are simply opiates for the understanding.^ Metaphysical systems are philosophi- cal romances ; the heart, not the understanding, is their special creator ; they belong to faith, not to science.^ In reality, science and a valid theoretical philosophy are one and the same.^ In so far as there is any task which is peculiarly philosophical, it is the winning of a scientific knowledge of scientific knowledge itself.'^ In general, the philosophical task of our time is the elevation of science itself to philosophy, the making of science philosophical and philosophy scientific.^ Riehl is consistent, as we have admitted ; but, in assuming a position necessarily agnostic, he is fundamentally dogmatic. Obviously agnosticism is not to be accepted, if it can be legiti- mately avoided ; and RiehFs system, however interesting and instructive, is not to be chosen if any non-agnostic realism equally or more tenable can be discovered. Moreover, such assertions as that the objects of experience are, as such, de- pendent on our consciousness,^ and that things are constant groups of sensations,^ show that Riehl is driven to agnostic realism — as a substitute for subjective idealism, which is undesirable, and for a non-agnostic realism, which is unattain- able — by his having fallen a victim to the fallacious suggestion that what is experienced must be itself experience, that what is thought about can itself be nothing but thought.^ Wilhelm Dilthey, although making room for a philosophy of reality, nevertheless occupies essentially the same agnostic position as Riehl ; all theoretical supports of metaphysical con- struction are, in his opinion, worthless. In his Einleitung in die Geisteswissenschaften he indicates his negative attitude toward metaphysics, which, he claims, ''does not overcome the rela- 1 lb., pp. 85-6. 2 Ueber wissenschaftliche und nicht-wissenschaftliche Philosophie, 1883, pp. 8, 12. » Der phil. Krit, Vol. II, Part II, p. 120. * Ueber wissenschaftliche, etc., p. 36. 5 Der phil. Krit, Vol. II, Part II, p. 120. « 76., Part I, p. 18. ^ 75.^ p, 202. » Cf. supra, pp. 23-4. D 34 THE PROBLEM OF KNOWLEDGE tivity of the sphere of experience" and "the subjectivity of the psychicaHife." ^ ''Epistemology is the end of the course of metaphysics," ^ for surely "no one can even want to know how the external object appears when no one takes it up in his con- sciousness." ^ In his essay entitled, "Das Wesen der Philoso- phie" ^ he gives to philosophy a place alongside of poetry and religion, as dealing with the same riddles of the world and of life.^ Philosophy, as metaphysics, differs from religion and poetry, in that it attempts to raise some particular world-view to universal validity,^ but its task is insoluble.'^ Individuality, circumstances, nation, and period influence the philosopher as well as the poet and the religionist.^ Materialism, objective idealism, and the idealism of freedom are the chief forms of world- view,^ but none of these is demonstrable. A restless dialectic drives the thinker on from one of these views to the other.^" In the end only a personal, heart-felt conviction remains to sup- port any philosophical system,^^ and that in turn is largely de- termined by the system of culture environing the individual.^^ When fully critical, then, philosophy becomes simply Weltan- schauung slehre, a discipline which is essentially akin to the history of philosophy, and whose task it is to solve the contra- diction between the claim of philosophical systems to universal validity and the endless individualism of such systems, by bring- ing to light the relation of the human spirit and its experiences to the riddle of the world and of life.^^ In closing this discussion of the frankly agnostic epistemologi- cal dualists, then, we may state the epistemological problem as follows : Is agnosticism the necessary implicate of epistemologi- cal dualism? This question we shall have to consider further, in the light of a critique of those systems which claim to avoid the agnosticism while retaining the dualism.^'^ If it should turn 1 Pp. 513-14. 2 p. 516. 3 p. 502. 4 Die Kultur der Gegenwart, I, Part VI. b p. 35. 6 Pp. 55, 57. 7 p, 60. 8 P. 57. ' P. 59. Cf . Die Typen der Weltanschauung und ihre Ausbildung in den meta- physischen Systemen in Max Frischeisen-Kohler's (ed.) Weltanschauung. 10 Die Typen, etc., p. 50. " P. 51. 12 Das Wesen der Philosophic, p. 68. 13/6., pp. 37-8, 62. See Max Frischeisen-Kohler's "Wilhelm Dilthey als Philosoph," Logos, III, 1912, pp. 29-58. 1* Chs. Ill and IV, infra. DUALISM AND AVOWED AGNOSTICISM 35 out that agnosticism is necessarily involved in the dualism, manifestly, then, the dualism itself ought not to be accepted, provided it can be avoided with intellectual honesty and without the necessit}^ of a still more undesirable alternative. The prob- lem will then be to discover some better alternative. CHAPTER III Dualism and Attempted Metaphysics Besides the followers of Kant who, like those whose doctrines we have examined in the preceding chapter, frankly confess the agnosticism which seems to be involved in the epis- temological dualism, and those others, to be dealt with in our critique of idealism, who undertake to eliminate the agnosticism by cancelling the thing-in-itself, thus denying the dualism and at the same time the realism, there is the third class of followers, who seek to avoid the agnostic conclusion while holding on to the dualistic premises. Although assuming that we never have immediate experience of any reaUty which exists independently, they maintain that the thing-in-itself is not entirely beyond our knowledge, but that we are in a position to know not only that it is, but also to some extent what it is. This group may be sub- divided into two minor groups. In one of these would be in- cluded philosophers who, when taken either individually, or two or more together, represent a movement of thought begin- ning with a pronounced realism and seeking to overcome the agnosticism of dualism by proceeding in the direction of idealistic metaphysics. On the other hand there are those who begin by paying their respects to the idealistic side of the Kantian thought, but then proceed to develop a positive doctrine of the thing-in-itself in the direction of non-idealistic metaphysics. Our best illustration of the one movement will be found in the systems of Herbart and Lotze, taken together, and of the other either in the transition from Schopenhauer to von Hartmann or in that from Wundt to Kuelpe. In the present chapter we shall be concerned with the former movement, leaving the latter to be dealt with in the chapter following. Before entering upon our examination of the systems of Her- bart and Lotze we may refer briefly to certain thinkers who, either in their criticism or in their further development of the 36 DUALISM AND ATTEMPTED METAPHYSICS 37 Kantian realism, seem to have been feeling after a positive knowl- edge of the thing-in-itself . We shall speak of Jacobi, Reinhold, and Fries. Jacobi would substitute for the Kantian theoretical agnosticism and claim of practical knowledge or moral faith with reference to the thing-in-itself, a speculative faith with reference to ultimate reality. He was the first to attack the Kantian combination of realism with agnosticism. Without the realistic postulate of a causal nexus between the subject and a reality beyond experience, one could not, he claimed, enter into the Kantian system; but if one were to develop the implica- tions of that idea of a causal nexus, he could not remain a Kan- tian.^ He held that we could not demonstrate even the existence of the thing-in-itself, and yet he regarded it as a self -destructive course to will to believe simply what one needs to believe. He advanced the view, however, that through a faculty which at first he called faith as opposed to reason, and later reason as opposed to the understanding, we have an immediate conviction or apprehension of the suprasensible. Formally this was a repudiation of epistemological dualism, but it did not quite amount to a realistic epistemological monism. Rational faith might be assured of independent reality, but the understanding was necessarily sceptical ; and what was claimed was not an immediate experience of independent reality, but an immediate conviction of a reality forever transcending experience. It amounted to little more than a dogmatic declaration of faith. Reinhold's view of the nature of consciousness had certain realistic implications which might have led him to claim positive knowledge of the thing-in-itself, had it not been for his ac- ceptance of the Kantian account of the ''form" and "ma- terials" of consciousness. In beginning his Versuch einer neuen Theorie des menschlichen Vorstellungsvermogens,^ he claims that since before we can expect to have a universally convincing philosophy we must have one that is universally valid, there is suggested the necessity of inquiring how universally valid knowledge is possible. But prior to answering this we must ask within what limits knowledge is possible at all ; and before answering this in turn, what one is to understand by knowing and the ability to know.^ This, then, is the most fundamental 1 Werke, II, 304. 2 x789. » §§ i_5. 38 THE PROBLEM OF KNOWLEDGE philosophical problem, and Reinhold's solution of it is offered as ''elementary philosophy." This solution is to the effect that knowing is, in all its forms, an activity of a subject with reference to an object, a presentation or representation, which, as an activity, is to be distinguished from the representing sub- ject and the represented object.^ The Kantian *' sensibility, '^ ''understanding," and "reason" are to be interpreted as variant forms of this representation of an objective reality by an equally real subject. ^ But this promising beginning was hindered from becoming a positive or non-agnostic realism by Reinhold's acceptance of the Kantian view of the subjective origin of the "form," or "primary qualities" of objects, and the objective origin of their sens.e-materials, or "secondary quali- ties." ^ Since the subject with its forms cannot produce the matter of the representation of the object, there must be the so-called thing-in-itself to account for that sense-material. This thing-in-itself, however, being simply the cause of the data of sense, is no more representable than is the representing subject itself.^ But that Reinhold himself was not satisfied with this agnostic conclusion is shown by his later adherence to the sub- jective idealism of Fichte, then to Jacobi's combination of epis- temological dualism with dogmatic realism, and finally to the dogmatic, rationalistic realism of Bardili, who claimed that the laws of nature on the one hand and the laws of the association of ideas in man's logical thinking on the other, are the necessarily corresponding manifestations of the one Absolute Reason which is fundamental to both the objective world and the consciousness of man. J. F. Fries is another of the earher followers of Kant who might have developed a positive or non-agnostic critical realism, if it had not been for an inherited element in his philosophical creed which made such an issue impossible. He adopted the rationalistic Kantian and pre-Kantian doctrine of a non-intuitive and yet unmediated, and, therefore, supposedly, absolutely a 'priori element in all our knowledge. For example, one knows a priori and with absolute certainty that every change must have a cause; this is not an intellectual intuition, because it does not come explicitly to consciousness without reflection; 1 § 7. 2 §§ 9-11, 48, 67, 77. « §§ 15, 16, 18-20. < § 17. DUALISM AND ATTEMPTED METAPHYSICS 39 and yet the knowledge does not originate out of the reflective process.^ But while agreeing with Kant that such knowledge is a priori, Fries differed from his master in maintaining that these a priori forms could be discovered only a posteriori, by psychological observation and abstraction, resulting in the for- mation of concepts corresponding to the a priori forms of ab- soluteh'- certain knowledge. Here space and time are included, as well as substance, cause, and the other categories of Kant's list. Thus the true critique of reason would be one of the em- pirical sciences, a sort of inner anthropology .^ As against the more orthodox Kantians, who held that the critique of reason must not be made an empirical science, for the reason that validity can be guaranteed only to the empirical sciences by means of the critique of reason. Fries objected that the question as to the possibility of knowledge was not a proper theme for any theory or discussion whatsoever. The supreme principle of all processes of human judgment, he claimed, was that of the self-trust of human reason ; this is involved in the critique of reason as necessarily as in any other scientific investigation.^ Indeed, in opposition to Kant, Fries held that the objective validity of experience cannot be proved ; we must use certain categories, but there is no way of showing further that we are justified in doing this. He agrees with Kant, however, that human knowledge is never transcendent, but always purely immanent, empirical. We have no knowledge of anything beyond the sensible ; the a priori forms are simply imposed upon the sense-material, and if reality is known at all, it is only as it is given in sensuous intuition. Truth is simply the agreement of mediate or discursive knowledge with that of immediate per- ception; it is not agreement of our mediate knowledge with existence. Our spatial, temporal, and causal concepts give us no completed series; we never transcend the relative and limited, the world of phenomena. Indeed, even the existence of the thing-in-itself is not known. Nor are we to seek refuge in practical postulates ; the primacy of the practical reason is not ^ "Ueber das Verhaltnis der empirischen Psychologie zur Metaphysik," Psychologisches Magazin, III, 1798, p. 181 ; Neue Kritik der Vernunft, § 95. Cf, Leonard Nelson, Ueber das sogenannte Erkenntnisproblem, 1908, § 162. 2 " Ueber das Verhaltnis," etc., pp. 175-6, 181. Cf. Nelson, op. cit., §§ 154-5. 3 Neue Kritik, §§ 89, 131. Cf. Nelson, op. cit., §§ 158, 163, 165. 40 THE PROBLEM OP KNOWLEDGE given in sense-experience, and is therefore no part of our knowl- edge.^ But while Kant held that the thing-in-itself is simply the ob- ject of a necessary question, unknowable save as we are enabled to postulate certain beliefs on practical grounds, Fries main- tained that we have an assured speculative faith as to its exist- ence. What it is, however, we can only describe in negative terms. By negating the positive categories involved in our knowledge of the finite, we arrive at as many negative ideas as there are positive categories; and taking all these together, we get the purely negative idea of the unlimited, the infinite. Thus, by simply conceiving the limits of our knowledge of reality transcended, or, in other words, by thinking of reality (which we know under the necessary limitations imposed by our experience) as we do not know it, viz. as an absolute totality, we arrive at our speculative faith in the existence of the Abso- lute, or Thing-in-itself.2 But while the sensible is the object of knowledge, and the suprasensible the object of faith. Fries adds that we have a feeling, or presentiment (Ahndung), of the manifestation of the suprasensible in the sensible. This presentiment is present in both the aesthetic and the religious consciousness ; the beautiful and the adorable are taken as a revelation of the infinite in the finite. It must not be concluded, however, that the lan- guage of art or of religion can ever be literally true of the Ab- solute. It is mere symbol, figure of speech; it describes the suprasensible in terms of the sensible. To regard this as knowl- edge is to construct a mythology. And yet it is one and the same reality which is known by science as the finite world of phenomena, truly thought of by speculative faith as in itself unlimited and therefore not positively conceivable, and repre- sented symbolically by religion and art as if it were an object of sensuous experience.^ The philosophy of Fries thus turns out to be thoroughly agnostic with reference to independent reality. We cannot even know, 1 Neue Kritik, §§ 123, 129; Wissen, Glaube und Ahndung, pp. 67 £f., 72 ff., 155 ff., 164 f. Cf. Nelson, op. cit., § 164-5. ^ Neue Kritik, §§ 123, 124, 129; Wissen, Glaube und Ahndung, passim. ' Glaube, Wissen und Ahndung. DUALISM AND ATTEMPTED METAPHYSICS 41 although we do undoubtedly believe, that this Absolute exists. The remedy for this agnostic conclusion might have been found in a certain departure from the presuppositions, as well as from the method, of Kant. We have no quarrel with Fries for using the empirical method in seeking to discover the a priori element in human knowledge; our objection is to the presupposition of both Kant and Fries that it is absolutely a priori, that neither in the individual nor in the race has it come to be what it is, viz. relatively a priori, as .a result of past experience. It would not help to suppose that the a priori forms had been passively received by man from without ; but if these forms of cognition by the human subject have been actively moulded upon the in- dependent reality of the environment, their value for the knowl- edge of that reality can be maintained. The further develop- ment of this view belongs, however, to later chapters ; ^ all we are here interested to maintain is that every absolute apriorism of the categories, when combined with a critical rather than a dogmatic attitude, leads to agnosticism, just as inevitably as does the old Lockian empiricism, with its notion of the complete passivity of the mind in perception.^ Among the early disciples of Fries were Schleiden, Apelt, and the theologian, DeWette. At the present time a note- worthy attempt is being made, under the leadership of Leonard Nelson, of Gottingen, to revive the Friesian philosophy. Nel- son differs from his chosen master chiefly in his understanding of the nature of the reflective process through which the a priori elements in human knowledge are discovered. What he objects to is the description of reflection as a sort of self-observa- tion, or inner experience, by means of which intuitive knowledge is repeated, or originally obscure consciousness brought to light. He claims that Fries virtually reasons that since we first be- come aware of our knowledge through inner perception, we must therefore proceed psychologically in philosophy. Strictly 1 XIV and XVI, infra. 2 There is an absolute a priori, of which, as a factor, experience is the result, but it is not the a priori meant by Kant ; much less is it that of Fries. It is the absolutely new and creative factor in experience. The a posteriori is the old, the result of experience, and includes within itself the oldest, the relatively a priori. It is this relatively a priori alone that is in some sense innate, pre- existent. See Chs. XIV and XVI, infra. 42 THE PROBLEM OF KNOWLEDGE interpreted, this would not only make metaphysics as well as criticism purely psychological; it would even make criticism itself unnecessary. True and false judgments, critical and dogmatic assertions, are all alike psychological. According to Nelson the process of investigation whereby the general a priori element in particular acts of knowledge is discovered, while empirical and inductive, as Fries maintains, is fundamentally logical rather than purely psychological in its character. In this way he would undertake to restore something of the Kantian transcendentalism, thus to relieve the Friesian criticism of its undue psychologism. It is a modification in the realistic direction, a securing of the epistemological dualism of the sys- tem, as distinguished from all forms of idealistic epistemological monism. Thus Jacobi, Reinhold, and Fries, each by a different path, sought escape from the Kantian agnostic dualism. They all set out in the direction of a positive or non-agnostic realism, whereby it might be maintained that knowledge of independent reality is possible. They all failed to reach their goal, however ; they conceded too much to Kant at the outset, and meta- physical agnosticism clings to them still at their journey's end. We now turn to the typical representative of the older critical realism, J. F. Herbart. While retaining the Kantian episte- mological duahsm of appearance and reality, phenomenon and thing-in-itself, Herbart claimed to pass from experience to independent reality, thus relieving the agnosticism of the earlier critical philosophy. Whereas Kant had essayed to cross the gulf between appearance and reality on the postulates of the moral consciousness, and Jacobi and Fries on one form or another of speculative faith, Herbart laid claim to speculative knowledge of the thing-in-itself, on the basis of a rational criticism of our empirical knowledge of phenomena. His method, he insists, is simply to follow out more thoroughly the procedure of the natural sciences, correcting the contradictory character of what is experienced by positing, back of phenomena, a reahty which is itself free from contradiction. But whereas the sciences are still content to work with forms and categories that lead to antinomies of thought, we must eliminate space, time, change, and multiplicity of attributes in one substance, as involved in DUALISM AND ATTEMPTED METAPHYSICS 43 contradiction and thus shown to be ultimately unreal. Phenom- enal space and time, it may be argued, are both finite and infinite ; change and multiplicity of attributes involve the asser- tion that a thing can be what it is not. Reality must therefore be thought of as made up of a large number of absolutely inde- pendent and unchangeable real beings, each having but one quality and existing in changeable external relations in an "in- telligible" order, as distinguished from the spatial -temporal order of phenomena. What appears to us as a substance with many qualities is in reality a combination of many independent substances of one quality apiece. What seems to be a change of substance is in reality but a change in the external relations of substances ; there are no relations internal to, or belonging to the essence of, any substance. Continuity is but a false ap- pearance of the eternally and absolutely discrete. One of these independent "reals" is the individual human soul. Sen- sations and other " Vorstellungen " are the forms of its appear- ance as it maintains itself in its changing relations with other real beings. The soul is not to be identified with the ego with which psychology deals ; this latter is a combination, and there- fore a result, of those representations which are themselves the product of the soul.^ But apart from the assertion of plurality, simplicity, un- changing essence, and changing relations, Herbart is agnostic with reference to independent reality. We know that these reals exist, he claims, because appearances exist, and there cannot be appearances without there being something which appears. But what the peculiar quality of any one of these in- dependent reals is, we are never able to say. Even of the soul all we know is that it is one of the independent reals ; we have no knowledge of what it is in distinction from any other being. But one may go further and say that even this slight escape from the agnosticism of epistemological dualism is itself a mere appearance, and not reality. As has been often remarked, the independent reals are mere products of abstraction from all particular qualities of phenomenal objects, and as such residues of abstraction we have no sufficient reason to affirm their ^ Einleiiung in die Philosophie; Hauptpunkte der Metaphysik ; Allgemeine Metaphysik. 44 THE PROBLEM OF KNOWLEDGE reality. Herbart's metaphysics, in view of his admission that we have no direct experience of reahty, is simply a return to the rationalistic dogmatism of the eighteenth century. If we never have any direct experience of a reality which exists indepen- dently of our experience of it, we have no means of verifying our speculations concerning the thing-in-itself. To eliminate con- tradiction from our speculations is only to establish hypothetical possibility, not actuality. Any theory that enters not in through the door of a bona fide experience of reality, but climbs up some purely speculative way, is a thief and a robber when it takes to itself the name of knowledge. We said that Herbart's apparent escape from agnosticism was very limited; we may now say that, in view of his initial absolute dualism of reality and appearance, he makes no legitimate escape from agnosticism at all. R. H. Lotze will be mentioned in another connection as illus- trating the transition from monistic absolute idealism to pluralis- tic personal idealism. But his fundamental position in epis- temology is dualistic realism. His main philosophical interest, however, seems to have been metaphysical rather than episte- mological. He is not concerned to dispute the main features of the critical philosophy, but he does not think it necessary that we should undertake a critique of human reason before venturing to use our rational powers in the attempt to discover the nature of realit}^ Broadly speaking, his assumptions, his purpose, and his method are those of Herbart. Dissatisfied with Her- bart's results, especially with his valueless view of existence, Lotze would do the work over again. He endeavors to overcome, more fully than his predecessor was able to do, the agnosticism of epistemological dualism, and at the same time to provide in his world-view for the preservation of the values of the spir- itual life. His method is neither deduction from a set of first principles, nor mere empirical investigation, nor even the formal adoption of the dialectical procedure. He starts with the realism of the plain man, revised by the sciences, and under- takes by the method of first criticising fundamental concepts and eliminating contradictions, and then offering analogies drawn from personal life as the only way of escape from agnosti- cism, to ''ascertain the impalpable real basis of the possibility DUALISM AND ATTEMPTED METAPHYSICS 45 of all phenomena, and of the necessity of their concatenation." ^ Our present interest in his thought will centre in the question to what extent he has succeeded where Herbart failed, in the endeavor to avoid the besetting agnosticism of absolute epistemo- logical dualism and to attain to a genuine knowledge of the nature of reality. He begins by investigating what we mean when we say that things are. In opposition to Berkeley he takes the realistic view : to be does not mean to be perceived ; on the contrary, it means to exist independently of the knowing relation. But in opposition to what he understands to be the contention of the realist Herbart, he insists that to be does not mean to exist independently of all relations. A thing which neither exists in any place nor at any time, and which neither does nor suffers anything, is as if it were not. To be does not mean to stand in the particular relation of being perceived, but it does mean to stand in relations.^ In opposition to Herbart, again, a thing is not to be identified with a single quality, any more than with a sum of perceived qualities. Qualities are ascribed to things, and when we say that a thing some of whose perceptible qualities have changed is still the same thing, this is not, as Herbart maintains, a self-contradiction; however it may accentuate our problem, it must be acknowledged that a thing is a unit}^ in multiplicity, a permanent identity of essence in the midst of the changing qualities.^ Still, we must not, with Herbart, adopt the substantive conception of the Real pure and simple. Real is an adjectival conception, a title belonging to everything that changes in a regular order. Reality is simply a form in which content actually exists ; it can be nothing apart from content. The essence of the thing is only to be found in a law according to which its changeable states are con- nected with each other. What is meant is not a general law, nor yet a merely conceived unity ; it is a real and individual law of a series of phenomenal changes.^ Thus in avoiding the ab- stractness of the Herbartian realism Lotze seems on the verge of abandoning the realistic position altogether. But he quickly recovers himself and begins to move definitely ^Metaphysics, Eng. Tr., Vol. I, p. 12. ^Metaphysics, §§ 1-14. 3/6., §§ 15-30. *Ib., §§31-6. 46 THE PROBLEM OF KNOWLEDGE in the direction of spiritual realism. As against Herbart's explanation of identity in the midst of change by the theory of unalterable elements in fluctuating external relations, Lotze's view is that being is itself but a particular form of becoming. But if, on the one hand, to be is to stand in relations, so that a change in relations means a change in the thing itself ; and if, on the other hand, to be is to change in a definite and orderly fashion, it may be concluded that, if there are any existing things at all, the mutual relations in which they have their being are relations in orderly correspondence with changes in other things. But since for becoming the only sufficient reason is an efficient cause, it may be concluded that to be, to stand in relations, to change in orderly correspondence with changes in other things, is to exchange actions.^ But "transeunt" action, this interchange of actions between independent things, presents difficulties for critical thought, and in the face of these difficulties a further transformation of the common view is suggested. It is inconceivable that a state or event should detach itself from one thing, make its way independently to another thing and enter into it.^ "Im- manent" action, however, cannot be denied; in experience of our own development we have indisputable evidence that in one and the same being the reality of one state is the condition of the realization of another. May it not be that what appears to be transeunt action is in reality immanent, that instead of a multiplicity of independent things, all elements are parts of a single real Being? In view of the contradiction involved in holding that independent beings can be influenced by, and thus dependent upon, each other, this transition from pluralism to monism is set forth as the only rational possibility for our thought. All individual beings are included in an Absolute Being, and only thus are they able to act upon each other.^ Now these individual things, it will be remembered, were found to be in continuous becoming, and yet to preserve their unity and identity throughout the whole process. But in experience we find but one being, ''the spiritual subject, which exercises the wonderful function not merely of distinguishing sensations^ ideas, f eeHngs from itself, but at the same time of knowing them 1 Metaphysics, §§ 40-1, 44-5. 2 lb., §§ 55-6. ^ 75., §§ 68-71, 81. DUALISM AND ATTEMPTED METAPHYSICS 47 as its own, as its states, and which by means of its own unity connects the series of successive events in the compass of memory." Hence ''if there are to be things with the properties we demand of things, they must be more than things. . . . They can only be unities if they oppose themselves, as such, to the multiphcity of their states.'' Two points are essential, *' one, the existence of spiritual beings like ourselves . . . feeling their states and opposing themselves to those states as the unity that feels, . . . the other, the unity of that Being in which these subjects in turn have the ground of their existence, the source of their pecuhar nature, and which is the true activity at work in them." Any world of things over and above this it is not necessary to assume.^ Ultimately this one Being is in- terpreted, on the basis of the analogy of the human spirit, as the personal God who constantly creates the mechanical pro- cesses of Nature for the realization of his purposes. ^ Lotze's philosophy has gained many friends, but more, one suspects, for the spiritually satisfactory character of the results at which he supposed himself to have arrived than for the really conclusive character of the processes of his thought. He con- ceded too much to Kant ever to be able to make much progress in metaphysics. Space is regarded as purely subjective, a per- ceived relation of which the cause is a changeable "intelligible" — but unknowable — relation existing between realities.^ Events which occur in the non-spatial real world cause sensa- tions which we construct into the purely subjective and there- fore unreal spatial world, which is the only world directly acces- sible to us.^ The problem then is, how to learn the nature of the real world, from which we are shut off by spatial phenomena. Time, it must be admitted, is treated with more respect ; while the idea of the totality of empty time is regarded as only a subjective form of apprehension, there is a real succession in- volved in the operation which is of the very essence of reality.^ But "the completely human subjectivity of all our knowledge" is asserted as unavoidable, in view of the fact that no mind which does not include all reality within itself can ever gain " a 1/6., §§96-7. 2 /&., §§ 229, 230. Cf. Outlines of the Philosophy of Religion, Ch. 4. ^Metaphysics, §§ 114, 116. * lb., § 217. Ub., § 156. 48 THE PROBLEM OF KNOWLEDGE view of the objects of its knowledge as they would seem if it did not see them." Our only refuge is the confidence of Reason in itself, or the belief that the all-inclusive Reality '^has given our spirit only such necessities of thought as harmonize with the world." ^ Thus, like many another before him, Lotze has recourse to rationalistic dogmatism to avoid the agnosticism logically involved in epistemological dualism; only, in this case what is claimed is not that we know reality as it is, but only, in effect, that we do not know that what we have is not knowl- edge. Although we can never know that we have knowledge of an independent reality which we never directly experience, we can nevertheless trust — for this is what Lotze seems to mean — that what we have, and call our "knowledge," is either knowledge or a satisfactory substitute for it. The fundamental agnosticism of Lotze's position becomes evident when we examine the way he interprets the nature of things and of the World-Ground by falling back upon analogy. Assuming that we -have knowledge of ourselves as spirit, he claims that we cannot know what reality is, unless we interpret it as essentially spirit. What this means, evidently, is that we are offered as alternatives spiritualism and agnosticism, the choice between them being not rationally determined, but left arbitrary. But if we must choose between spiritualism and agnosticism, and we can not know which we must choose in order to have the truth, we cannot know, manifestly, what reahty is. To leave as ulti- mate alternatives agnosticism and any other position whatso- ever is to give the victory to agnosticism. But the weakness of Lotze's argument at this crucial point is but symptomatic of further disorders in his philosophical system. Indeed one finds that the transitions of thought upon which his argument chiefly depends are by no means rationally necessary. Let us begin with his criticism of Herbart's view of independent Reals. This he condemns as self -contradictory, on the ground that independence means absence of all relations, whereas Herbart, as he proceeds, has to speak of the Reals as related in some ways to each other. But what Herbart means by independence was surely not the absence of all relations, but simply the lack of dependence for existence upon anything ^ Metaphysics, § 94. DUALISM AND ATTEMPTED METAPHYSICS 49 else. Lotze's criticism is valid only on the assumption that for anything to be in any relation is for that thing to depend upon that relation for its own essential nature ; or, in other words, that for anything to be in any relation is always for that relation to be in it. This doctrine of the necessary internality of re- lations, like the closely related proposition upon which so much is made to hinge, viz. : ''To be is to stand in relations," Lotze himself assumes upon no other basis, apparently, than the mere fact that, if there are more things than one, all things must stand in some relation to each other. Where existence is plural, to be manifestly involves standing in relations; but this does not mean that ''standing in relations" is an adequate definition of being. Furthermore, while what a thing is sometimes does depend to some extent upon some particular relation in which it stands, there are cases where it does not so depend. Whether it does or not is determined by the practical purpose back of the question. If the relation makes a difference in the object for our purposes, it is, for us, a relation internal to the object; if it makes no difference, it is external. Apart from some special purpose for which it makes a difference, a thing's relations to other things, other than its relation as the effect of a cause, are incidental, not essential; external, not internal. This con- sideration undermines that particular argument for metaphysical idealism which rests upon combining with the doctrine that to he is identical in meaning with to stand in relations, the relic of Kantian subjectivism, "Relations are the work of thought." A similar criticism may be made against the view that no knowable reality except spirit exists, because we know no other reality which can remain identical in the midst of changing states. Whether or not what remains, after some quality has been changed, is to be regarded as the same thing as existed before the change, depends upon the purpose in relation to which the question is considered. We cannot conclude, there- fore, from the presupposition of identity in the midst of change, that the reality is spiritual, but onl}^ that it is being considered by some conscious being with reference to some purpose. Again, it is not invariably true that to stand in relations is to exchange actions. There are other relations between things besides the causal relation and such relations as are established 50 THE PROBLEM OP KNOWLEDGE by thinking the two things together. For example, to cite an extreme instance, it surely cannot be maintained that to stand in the relation of non-interaction is to exchange actions. Moreover, the ingenious dialectic by means of which a numer- ical ontological monism is supposed to be established through a synthesis of the empirical actuality of interaction with its theoretical inconceivability, also fails to convince. According to Lotze's own principle, we have not, in metaphysics, to ask why there should be a world at all, or how reality can be what it is ; we have to take it as it is, to find out what it is.^ There is mystery in all ultimate existence, in all real productivity, all action, as well as in interaction. We should no more argue that interaction is impossible, because mysterious, than that there is no real becoming, because real becoming is an ultimate mys- tery. But with the disappearance of any contradiction of the ultimate reality of interaction, the synthesis of the two anti- thetical propositions, viz. Lotze's numerical monism, also falls to the ground — at least so far as this argument is concerned. We can scarcely avoid the conclusion, therefore, that Lotze has not succeeded in his attempt to develop a positive meta- physic on the basis of a dualistic epistemology. At practically every crucial point his argument is fallacious, or at least incon- clusive. G. T. Ladd, to whose conception of epistemology we have already referred,^ is a disciple of Lotze whose Lotzianism is tinged with influences from the modified Scottish philosophy of Noah Porter. Consequently, while much of our criticism of Lotze would apply to the doctrines of Ladd, certain features of the latter philosopher's discussion of the problem of knowl- edge invite special attention. He is very insistent that cogni- tion always transcends experience.^ But his acceptance of the Kantian criticism leaves this assertion of ontological validity little more than a dogmatic appeal to ''consciousness," after the manner of Reid. This element in his thought appears in the following quotations: "Experience is . . . truly ontologi- cal. To tell how such experience is possible, this was the prob- lem of the Critique of Pure Reason. But because its answer 1 Cf. F. C. S. Schiller, "Lotze's Monism," in Humanism, p. 66. » Ch. I, supra. » The Philosophy of Knowledge, pp. 325, 332, 341, etc. DUALISM AND ATTEMPTED METAPHYSICS 51 laid all the emphasis on the analysis of the subject, the knower, and did not share the undying confidence of men that the object, that which is known, belongs in all its complicated structure to the world of reality, this Critique failed to satisfy the demands of consciousness. '\ ''The cognition of the world of things by the human mind actually takes place with the passionate and determined assumption of a right to know what things really are. The admission of this right extends and validates our system of concepts relating to things. It is, therefore, an assumption of the highest epistemological value. We shall return to it again." ^ But merely to assert the fact of ontological knowledge on the basis of the right to know, and in spite of a critical view which would naturally lead to ontological agnosti- cism, without showing how such knowledge is possible, is dogma- tism. A. Seth Pringle-Pattison is perhaps most widely known for his revolt from the Hegelian absolutism in the interests of moral personality in God and man. Each self, he is concerned to maintain, ''resists invasion"; it is "a unique existence, which is perfectly impervious."^ But this transition from monism to plurahsm, while not necessitating the giving up of meta- physical idealism, made it necessary to maintain, from the standpoint of the individual subject, an epistemological realism. He maintains that the metaphysical dualism of mind and matter may be avoided by developing in its stead the epistemological dualism of the world of real things and the individual's world of consciousness. The special interest attaching to his thought in the present connection, then, lies in three things : the offer of epistemological dualism as a substitute for metaphysical dualism; the absoluteness of that epistemological dualism; and finally, the dogmatic claim to know reality, notwithstand- ing the absolute dualism of his theory of knowledge.^ There are indeed two worlds, says Pringle-Pattison, but they are not mind and matter, respectively; the one is the world of consciousness; the other the world of independently 1 A Theory of Reality, p. 8 ; The Philosophy of Knowledge, p. 227. The italics, except in the word "right," are mine. ^ Hegelianism and Personality, 1887, p. 216; 2d ed., 1893, p. 227. 3 Cf. A. H. Jones, "Professor Pringle-Pattison' s Epistemological Realism," Philosophical Review, XX, 1911, pp. 405-21. 52 THE PROBLEM OF KNOWLEDGE real things, of " epistemological things-in-themselves." The two worlds are mutually exclusive. The mind is never in immediate relation to things. All objects, from those which are in immediate contact with the organism to the remotest star, are completely and inexorably outside the individual's world of consciousness.^ It is maintained, however, that the world of real things is known to thought; objects and subjects are completely sundered in experience, but they are related to each other as members of one w^orld, metaphysically speaking,^ and '' knowledge points beyond itself to a reahty whose repre- sentation or symbol it is." ^ But the same old question returns. If we never have any direct experience of the real world, how do we know what it is, or even that it is? Pringle-Pattison holds that while we can prove neither the one nor the other directly, because we can never get behind our own knowledge,^ there is nevertheless an indirect proof to be found in the instinctive belief of all mankind and the failure of non-reahstic theories to avoid practical ab- surdity.^ It may be remarked, however, that the instinctive realistic belief of mankind is not in epistemological dualism, but in a reahstic epistemological monism. Moreover, the failure of idealism does not mean the establishment of dualistic reahsm, unless this can be shown to be the only other possible theory — which, however, is not the case. Nor is dualistic reaUsm, with the agnosticism logically involved, desirable, if any essentially monistic realism can be found to admit of adequate rational defence. C. A. Strong has given us a detailed exposition of epistemo- logical dualism and critical realism, in combination with ideal- ism, or panpsychism, in metaphysics. The position as a whole is supported by some new arguments, a special feature of the discussion being the attention given to the problems set by physiological psychology. Strong acknowledges indebtedness to William James and D. S. Miller in arriving at the conviction that cognition is nothing but having a feeHng which so resem- 1 Philosophical Review, I, 1892, pp. 514-16. 2/6., I, 1892, pp. 145, 513; III, 1894, p. 61. ' 76., I, 1892, p. 504. * lb., Ill, 1894, p. 59. Cf. Hegelianism and Personality, pp. 84-5 ; 2d ed., p. 90. 6 Philosophical Review, I, pp. 507, 511-12. DUALISM AND ATTEMPTED METAPHYSICS 53 bles reality that we are enabled to operate upon it. In develop- ing further his theory of perception and of cognition in general, Strong seems to have found a clew in the nature of memory, where the represented object, like the representing image, is psychical. He adopts as his hypothesis the view that the reahty to which thought refers is not something different in nature from thought and more real than it, but simply other experience than that which constitutes the reality of thought.^ It then becomes his task to work out the details of this theory in the hght of scientific knowledge, and to defend it against objec- tions and rival interpretations. There is, he claims, a twofold existence of the object ; the object of which I am immediately conscious cannot be the object which acts on my senses and calls forth the perceptional brain-event ; it is a modification of my own consciousness, and at best a mental duplicate of the stimulus.^ The real object produces an image in the brain, an image which, abstracted from our consciousness, is projected into space as the physical object. The physical order, as made up of projected images, is symbolic of a real order, of which our sensations are effects. The real object is known, therefore, substitutionally, through the medium of its symbol, the physical object or projected image.^ This physical object, as a projected image, a modifi- cation of consciousness, while an existence distinct from the object of which we are immediately conscious, is still an object in the same world ; the real world is itself psychical. It is made up of minds and their actual and possible experiences.* Strong tries to minimize the dualistic, and, therefore, logically agnostic features of his epistemology, by insisting that it is to be distinguished from the representative theory of knowledge, which holds that the thing known primarily in sense-perception is the image, the real object being known only by inference and 1 " A Naturalistic Theory of the Reference of Thought to Reality," Journal of Philosophy, etc., Vol. I, 1904, pp. 253-4, 259-60; Why the Mind has a Body, 1903, pp. 221-2. 2 Why the Mind, etc., pp. 172, 178. 3 Why the Mind, etc., pp. 195, 251; " Substitutionalism " in Essays . . . in Honor of Wm. James, 1908, pp. 170, etc. ; Journal of Philosophy, etc., IX, 1912, pp. 598-9. Cf. D. Drake, Journal of Philosophy, etc., VIII, 1911, pp. 365 ff. ; IX, 1912, pp. 149 ff. ; Mind, N.S., XXIV, 1915, pp. 29-36. * Why the Mind, etc., pp. 228-9 ; Journal of Philosophy, IX, p. 533. 54 THE PROBLEM OF KNOWLEDGE representation. His own view is that the image is not primarily the object of knowledge at all, but its medium, or vehicle ; the object is known directly, although not immediately.^ But, when we get back of these verbal distinctions, we find that what the theory amounts to is that the real object, which is psychical, produces a cerebral image, which, however dissimilar to the real object, still in some sense represents it. This image, how- ever, is projected, so that ''the image is taken as being where it is not and what it is not." ^ That is, what we know directly and immediately is the projected (and, therefore, changed) cerebral image, which we call the physical object ; and this is called knowing — directly but mediately and symbolically — the real psychical object which produced the cerebral image. But, we would remark, to project a representative image — which is actually but to treat it as if it had been projected — does not take away its representative character, although it may make it a more useful representative. Whether improved or not from the practical point of view, theoretically — accord- ing to the logic of Strong's theory — it leaves our fancied knowl- edge doubly removed from direct cognition of the real object. What we know directl}'- is a distorted product of the object, nothing more. Substitutionalism does not offer us genuine knowledge, but a substitute for it, upon which the trade-mark of knowledge has been stamped. We would agree with Strong that what we have and use deserves to be called knowledge; but that is because what we have does not fall under his de- scriptive formula. What he describes would not be knowledge. But even if Strong were to concede that on his view what we ordinarily call knowledge is simply a practical makeshift, might he not be able at least to maintain that the genuine knowledge of reality is that contained in his metaphysical doc- trine of things-in-themselves other than human and animal minds, but themselves also psychical in nature ? ^ He offers three ''proofs" of his doctrine, the cosmological, the physio- logical, and the evolutionary. The cosmological is to the effect ^ Essays . . . in Honor of Wm. James, pp. 171-2 ; Journal of Philosophy, IX, p. 540. * Journal of Philosophy, IX, p. 599. ' Cf. D. Drake, The Problem of Things-in-Themselves, 1911. DUALISM AND ATTEMPTED METAPHYSICS 55 that things-in-themselves must be assumed in order to fill in the gaps between individual minds, and to give coherence and intelligibility to our conception of the universe.^ The '^ physio- logical proof" is to the effect that since our perceptions are phj^siologically conditioned, we are able to triangulate, as it were, to things-in-themselves as their causes.^ Now we are not concerned to attack the view that there are things-in-them- selves; but if we assume that, as Strong teaches, no human being ever has had or can have immediate experience of these realities which can exist independently of their being humanly experienced, the above arguments for such independent things are by no means conclusive. It might be that the whole con- tent of experience is produced, as Leibniz maintained, as a result of the inner constitution of the individual. The '^ evolutionary proof" is used to support the view that these things-in-themselves are psychical. The older argument for this doctrine has been that since consciousness is the only reality of which we have any immediate knowledge, and there- fore our only sample of what reality is like, we cannot have even a conception of any reality which is not psychical.^ This, of course, is at best simply an alternative to agnosticism, and so not a proof. The evolutionary argument, which is advanced as '^absolutely conclusive," is that things-in-themselves must be mental in their nature, because individual minds arise out of them by evolution.^ But, in the first place, this argument rests upon the presupposition of the existence of things-in- themselves, which presupposition we found to be inconclu- sively established, provided we assume the validity of Strong's contention that we can have no immediate knowledge of any reality not dependent upon our own consciousness. More- over, if we hold to creative evolution, the '^ evolutionary proof" loses all force from any point of view; individual minds may be thought of as having arisen, not as new variations of pre- viously existing reality, through mere rearrangement of ele- ments, but as new variations from it. We conclude, then, that Strong's presuppositions would compel him to believe that our 1 Why the Mind, etc., pp. 252, 259. « 75.^ p. 264. ' /&., p. 294. 4 /&., pp. 268, 292. Cf. W. K. Clifford, "On the Nature of Things in Them- selves," Humboldt Library of Science, No. 145, p. 35. 56 THE PROBLEM OF KNOWLEDGE ordinary knowledge is not genuine, and that his own panpsy- chism is simply an unproved speculation. A. O. Lovejoy's most important contributions to epistemology have been his criticisms of the absolute epistemological monism and realism of the neo-realists. His own position, however, is epistemological dualism and metaphysical temporalistic idealism. Finding it easy, in opposition to what he regards as the neo-realistic view, to show that knowledge is sometimes mediate, he goes on to state that since there can be mediate knowledge, there is no reason why knowledge should not al- ways be mediate. He feels free, therefore, to hold that the existence and some of the attributes of things can be known, although always only mediately, since the perceived object and the real object are always numerically different, although they may be quahtatively identical in part.^ What is over- looked here is the possibility that it might be just because there is such a thing as immediate knowledge, that mediate knowledge becomes possible at all. Moreover, he does not sufficiently canvass the possibilities in the way of a less extreme realistic epistemological monism than that of the neo-realists. His critical arguments, which are largety valid as against the abso- lute epistemological monism of the new realists, do not neces- sarily apply to that critical realistic epistemological monism which we shall defend in a later chapter,^ and which would maintain that the experienced object and the independently existing thing may be numerically identical, even if to some extent qualitatively different.^ 1 Journal of Philosophy, etc., X, 1913, pp. 568-9, etc. 2 See Ch. XIV, infra. 3 Lovejoy gives passing notice to the intermediate views of Schuppe and Wolf, and he may be right enough in hinting that these theories perhaps amount to no more than "a weak and untenable compromise." ("On the Existence of Ideas," The Johns Hopkins University Circular, 1914, No. 3, pp. 49-52.) But it does not follow that the same must be true of all possible theories of knowl- edge between absolute epistemological dualism and a realistic epistemological monism so absolute as to be debarred from making any distinction between appearance and reality. CHAPTER IV Dualism and Attempted Metaphysics {Concluded) We turn now to some epistemological dualists who, like those just considered, claim to be able to arrive at some posi- tive knowledge of reality which is not dependent upon human experience, but who, unlike them, exhibit, when taken either singly or in groups, a tendency away from idealism, metaphysi- cal as well as epistemological, and in the general direction of a non-idealistic metaphysics. We shall be interested to learn whether they are able, any more successfully than those just examined, to overcome the apparently agnostic implications of absolute epistemological dualism. In post-Kantian philosophy one of the most conspicuous examples of this development is to be found in the philosophies of Schopenhauer and von Hartmann. But, as the philosopher who, perhaps next to Kant, most influenced Schopenhauer, and who himself illustrates, in his relation to Fichte, the movement from subjectivism to objectivism, we shall briefly refer to Schelling. In reaction from his early adherence to the doctrines of Fichte, Schelling deliberately undertook to ''break through" this closed system of subjective idealism ''into the free open field of objective science," with its realistic acceptance of the independent existence of the real world. ^ The result was the working out of his philosophy of nature, in which nature is viewed as creative, and the subjective as being added to the objective. Later, indeed, he worked out a transcendental philosophy, in which spirit is the creative factor, and the objective is repre- sented as being added to the subjective ; but this again was taken up into the philosophy of identity, according to which objective and subjective, nature and spirit, the real and the ideal, are fundamentally the same. The Absolute appears as 1 See Kuno Fischer, Geschichte der neueren Philosophie, Vol. VII, 1899, pp. 311 f. 57 58 THE PROBLEM OF KNOWLEDGE nature and spirit, but in itself it is neither the one nor the other, but the higher Unity, comprehending both. Schopenhauer's philosophy may be viewed as a further separa- tion of the idealistic and realistic elements of the Kantian dual- ism, counteracted by a more or less dogmatic assertion of iden- tity, after the manner of Schelling. If we were to take his doctrine of 'Hhe world as idea" alone, Schopenhauer would have to be classed as an epistemological monist and idealist ; taking his doctrine of *Hhe world as will" alone, we should have to call him an epistemological monist and realist; but, both sides of his thought being taken together, his doctrine of 'Hhe world as will and idea" brings him fairly within epistemological dualism and realism. His emphasis upon the idealistic element, how- ever, is very pronounced. The whole spatial, temporal, caus- ally connected world, of which the individual has experience, is interpreted as nothing more than that individual's idea; it is ''conditioned through the subject and exists only for the sub- ject." ^ This being the case, theoretical egoism, or solipsism, never can be refuted. But, on the other hand, it never can be proved; and on this ground Schopenhauer decides to ignore this theoretical possibility .^ The thing-in-itself, however, from the point of view of rational knowledge, is unknowable ; hidden under the triple veil of space, time, and causality, it never could be known if the investigator were nothing more than the pure knowing subject. But the investigator is himself rooted in the world. His body is given as idea, an object among the objects of the phenomenal world ; and yet it is also given in an entirely different way, viz. by direct apprehension, as will. The act of will and the movement of the body are one and the same, given in two entirelj^ different ways — the former through, or rather in, the most immediate inner consciousness of each of us ; the latter, in perception, for the understanding. We each of us know one thing-in-itself, viz. our own self; and we know it as will.^ This, then, is taken as the key to the metaphysical problem. Since nothing is conceivable that is not will or idea, and since 1 The World as Will and Idea, Eng. Tr., Vol. I, pp. 3 f. Cf. Introd., pp. xxv-xxvi. 2 76., pp. 135-6. 2 lb., pp. 129-30, 142, 145. DUALISM AND ATTEMPTED METAPHYSICS 59 we can find nowhere any other kind of reaHty besides will, we may judge of all phenomenal objects, and of the phenomenal world as a whole, after the analogy of our own bodies, con- cluding that the inner nature of every physical thing is the same as that in ourselves which we call will. In its inmost nature the kernel of every particular thing, and also of the Whole, is will. It is not meant that in all things this will, or striving, is consciously directed, as it is in man ; Schopenhauer simply names the genus after that one species which is directly and immediately known. The world as it is in itself is will ; as it appears in perception it is idea.^ In working out the details of this identity-philosophy, Schopenhauer's thought runs into what looks like flat self- contradiction. On the one hand it is claimed that matter is simply a human idea, and on the other that thought is a mere product of matter. But criticisms more fundamental still are to be made against the system. In the first place, is Schopen- hauer, as a radical Kantian, justified in regarding even will as anything more than phenomenon? In his later thought he became conscious of this difficulty. He insists that the knowl- edge each of us has of his own willing is neither perception nor an empty concept, but he has to admit that even in- ward experience does not give us adequate knowledge of the thing-in-itself. The act of will is only the closest and most distinct manifestation of reality; in it the thing-in-itself ap- pears in the very thinnest of veils — free from space and caus- ality, but still not quite divested of time. In the end, Scho- penhauer makes the agnostic confession: "The question what that will ultimately and absolutely is in itself . . . can never be answered, because becoming known is itself the contradic- tory of being in itself, and everything that is known is, as such, only phenomenal." ^ Here at length he becomes consistent, and lapses into the Kantian agnostic dualism. But while he is at this point at length consistent — or at least as consistent as explicit agnosticism easily can be — he is not, we would maintain, correct. As we have seen, epistemo- logical dualism is founded on confusion and defended by fallacy. It has not been shown to be necessary; and, in a later con- 1 16., pp. 136. 143. 2 76., Vol. II, pp. 405-8. 60 THE PROBLEM OF KNOWLEDGE nection ^ we shall set forth an opposite hypothesis, according to which it would be incorrect to say that we can have no gen- uine knowledge of independent realities, things-in-themselves, if you please, and that on the basis of perception and reflection. From this our own point of view we would say that Schopen- hauer's assertion, that we have direct knowledge of ourselves as will, is in itself correct, although for him inconsistent. But when he goes on to assert, ex analogia hominis, and in order to escape from agnosticism, that all reality is will, he simply lapses into dogmatism. ' From his own presuppositions, strictly inter- preted, not even the human individual, much less every real thing, could be said to be will, any more than it could be said to be anything else. Not in the philosophy of Schopenhauer, at least, do we find epistemological dualism legitimately set free from agnosticism. Among those deeply influenced by Schopenhauer, some have gone in the direction of absolute idealism, as, for example, F. Paulsen and P. Deussen. These thinkers are monistic panpsy- chists; but E. von Hartmann goes in the opposite direction and develops a philosophy of the ''Unconscious." He calls his theory of knowledge ''transcendental realism." It is what we have called absolute epistemological dualism, or epistemo- logical dualism and critical realism, and is combined with a rather highly developed system of metaphysics. This theory of knowledge he regards as the only alternative to naive real- ism on the one side and subjective idealism on the other. Naive realism is to be rejected, he holds, for its failure to see that everything we can reach with our thoughts can always be only our own thoughts, never the reality lying behind them. Sub- jective idealism he rejects for the error of denying the existence of that which is beyond the limit of thinking, for no other reason than that it is inaccessible to thought. ^ The only other pos- sibilities being thus eliminated, transcendental or dualistic epistemological realism is regarded as established. Von Hartmann, however, does not think it necessary to remain agnostic, even on the basis of this dualistic epistemology. Even at the expense of contradicting the idealistic side of Kant's 1 See Ch. XIV, infra. * The Philosophy of the Unconscious, Eng. Tr., Vol. Ill, p. 198. DUALISM AND ATTEMPTED METAPHYSICS 61 doctrine, he sets himself to develop the realistic side of Kan- tianism, to gain positive knowledge of things-in-themselves.^ In general, this is accomplished by assuming, in agreement with Schelling, 'Hhe^ homogeneity of thought and its trans- cendent-objective object," as the only supposition upon which knowledge is conceivable.^ To account for this we must assume the identity of Thought and Being. The Beyond of con- scious thinldng must be unconscious thinking, for consciousness thinks its own conscious thought, and yet supposes something else ; hence, in so far as thought is true, reality can differ from what is consciously thought only in being unconscious.^ But that we must avoid dogmatism in our transition from consciousness to the Beyond, von Hartmann himself urges. We must employ, he tells us, ''the successive inductive ascent from experience." ^ The bridge whereby we may pass induc- tively from the world immanent within our own consciousness to the transcendent is found in the fact that in sense-experience we are affected by something beyond us; it is the bridge of transcendent causality. There is a transcendent cause of our sensations, and this cause is represented in our consciousness by the ''transcendental object." ^ Causality is the only rela- tion between the immanent and the transcendent.^ In itself, to be sure, it establishes only the existence of things-in-them- selves, and has nothing to say about any similarity between the thing-in-itself and object of consciousness.^ But by means of this causal bridge the whole realm of the transcendent lies open to us.^ From the diversity of objects perceived through one sense, we must conclude that there is a 'plurality of things- in-themselves. As operative, they must be changeable, and thus exist in time. Indeed Kant was wrong in forbidding the transcendental use of the categories. With space and time, they are the existential forms of what exists, as well as the thought-forms of what is thought.^ ^ Kritische Grundlegung des transcendentalen Realismus, [3d ed. (in Ausge- wdhlte Werke, I, 2d ed.), p. 54. 2 Philosophy of the Unconscious, Vol. Ill, pp. 198-9. » 76., pp. 200, 203 ; cf. Kuelpe, Philosophy of the Present in Germany, Eng. Tr., p. 190. * lb., p. 203 ; cf. Vol. I, pp. 9-13. ^ Kritische Grundlegung, etc., I, pp. 55-7. « lb., p. 94. 7 75.^ p. 66. s Jb., p. 94. «• lb., p. 106. 62 THE PROBLEM OF KNOWLEDGE Upon this basis von Hartmann proceeds inductively to de- velop further his metaphysical theory of the Unconscious. He states his doctrine in the following terms : "Being is a product of the non-logical and the logical, of Will and Representation. Its Hhat' is posited by volition, its 'what' is the ideational content of that volition. It is thus not merely homogeneous with the Idea, but because it is itself Idea, identical in the strictest sense of the term. But the Real is distinguished from the Ideal by that which lends reality to the Ideal, by the Will. . . . The Unconscious is not the Absolute Subject, but is what alone can become Subject, just as it is what alone can become Object, simply because there is nothing beside the Unconscious." ^ That the world per se is the Unconscious, identical with the conscious, as far as the latter goes, but going far beyond it — this doctrine von Hartmann finds reenforced by such facts as those of instinct, the unconscious union of sensations in perception, and the unconscious association of ideas and pro- duction of feelings and motives. In his ''speculative results,^' however, he has gone far beyond the sober method of induc- tion which he professed to follow. We see the influence of Schelling's identity philosophy, under the guidance of which Hegel's Absolute Idea and Schopenhauer's Absolute Will are brought together in the Absolute whose two attributes are infinite Will and finite Idea. The metaphysics of the Uncon- scious is poetical, mythological, and dogmatic, rather than a simple unification of results of scientific induction. With reference to the use made of the idea of transcendent causality, it may be said that this attempt to attain to knowledge of the thing-in-itself is itself laudable, but not when associated with a thoroughgoing epistemological dualism. If we have some direct experience of things, we may indeed use the causal category as a bridge from directly known realities to causes operating beyond our immediate experience ; but if we never have direct experi- ence of independent reality, how can we know that it con- tains any causes of our sensations? Von Hartmann's funda- mental error would seem to lie in his supposition that a dualistic epistemological realism is the only alternative to naive » Philosophy of the Unconscious, Vol. Ill, pp. 200-1. DUALISM AND ATTEMPTED METAPHYSICS 63 realism and subjective idealism. It will be our task, as we have already intimated, in a later connection ^ to point out another alternative. Johannes Volkelt acknowledges a debt to Schopenhauer,^ and he has evidently been not only interested in the philosophy of von Hartmann,^ but also considerably influenced by his thought, especially by his problem. His fundamental position in epistemology — which branch of philosophy he suggestively describes as ''science without presuppositions"^ — is, in the sense in which we have used the word, dualistic.^ He virtually assumes that our experience, in so far as it is not thought, is simply experience of our own conscious states. Confronted, then, with the fundamental difficulty of establishing the objec- tivity of knowledge in spite of the subjectivity of experience, the only scientific method in epistemology, he claims, is simply to show up the conscious processes involved in what we call our knowledge.^ Within the limits indicated Volkelt proceeds with admirably critical care, and succeeds not only in avoiding inconsistency to a remarkable degree, but also in covering up to a large extent the underlying dualism of his point of view. This is accom- plished by setting over against the dualism of the immanent subjective and the transcendent objective the duality of expe- rience and thought. Indeed it would seem that the only dual- ism which the philosopher acknowledges, even to himself, at first, is this ''dualism" of experience and thought. As against the epistemological monists of thought, like Plato, Spinoza, and Hegel, and the epistemological monists of pure experience, such as Hume, Mill, and Avenarius, Volkelt confesses adherence to the innocent enough looking "dualistic" doctrine that all true knowledge is the elaboration of pure experience by thought.^ But where the duahsm really lies, appears later, when, after 1 See Ch. XIV, infra. « Arthur Schopenhauer, 1900, Preface, etc. » See article in Nord und Sud, July, 1881 ; also Das Unbewusste und der Pessimismus. * Erfahrung und Denken, 1886, Pt. I, Ch. I ; Die Quellen der menschlichen Gewissheit, 1906, p. 4. 6 Die Quellen, etc., § 2. « Erfahrung und Denken, Pt. I, Ch. II. ' Die Quellen, etc., pp. 2, 3 ; cf. Erfahrung und Denken, passim. This prob- lem we shall have to deal with in our discussion of the morphology of knowledge, Ch. XV, infra. 64 THE PROBLEM OF KNOWLEDGE the self -certainty of consciousness has been dealt with/ the ''dramatic crisis of epistemology " occurs with the raising of the question ''whether I can exhibit in my consciousness a source of certainty which allows me to transcend my conscious- ness, not of course actually and truly, but in the way of cer- tainty." 2 Experience is subjective, but thought, with its logical necessit}^, presumes to deal with the transsubjective. Having, then, as our undoubtedly certain knowledge our immediate awareness of the subjective, the question has come to be whether we have, in the necessity of thought, valid medi- ate awareness of the transsubjective.^ A critical examination of the content of our necessary thought regarding the subjec- tive reveals, as involved in our simplest judgments of fact, four minimum propositions, viz. the existence of other conscious- nesses,^ the continuous existence of transsubjective entities, the rational correlation of transsubjective entities, and the numerical oneness of the world of the senses.^ In the case of each of these propositions the transsubjective validity of the necessity of thought must be recognized, or else we are led into affirming what is manifest nonsense. And yet the necessity of thought is itself only a subjectively experienced necessity. As a form of immediate certainty which reaches out to the transsubjective, it may be called faith, and intuition ; but the faith is not irrational, and the intuition is not in opposition to the logical. Rather is it the way of being immediately conscious which characterizes the logical itself. So, then, the necessity of thought cannot itself be proved, but must be be- lieved. Thought has the character of a demand, a postulate ; being itself only an individual conscious event, it comes into contact with the transsubjective only by way of demanding transsubjective validity for its content. Of the fulfilment of 1 Die Quellen, etc., §§ 4-6. ^ lb., pp. 19, 23. 3 Erfahrung und Denken, Pt. II, Ch. I ; Pt. Ill, Ch. II ; Die Quellen, etc., §§ 8-10. * Cf. Julius Baumann, Anti-Kant, 1905, p. 2, "The inner life of others like our own ... is a sure case of the thing-in-itself." 5 Die Quellen, etc., §§ 12-16 ; cf. Erfahrung und Denken, Pt. Ill, Ch. II, A. Cf. also Ueber die Moglichkeit der Metaphysik, 1884, pp. 16, 17, where it is in- timated that while the " absolute and dogmatic " type of metaphysics may well be regarded as impossible, the same should not be concluded of metaphysics Uberhaupt. DUALISM AND ATTEMPTED METAPHYSICS 65 this demand it can never be certain. Thus we have at length Volkelt's confession that thought is ''duahstically broken."^ His synthesis of subjectivism with transsubjectivism, his union of epistemological ideahsm and epistemological reahsm, may be regarded as in the end itself only a demand. Still, the necessity of thought, with its demand of transsub- jective vahdity, exists; and, grounded in mere subjectivity as it is, its guidance may be followed, not only in the specula- tive sciences, but in metaphysics as well. In the sciences it leads to what may be called a transsubjective of the first order, in which is included as much of the inexperienceable as is neces- sary for making the given intelligible by means of a thorough- going causal connection. In metaphysics, however, search- ing into the essence of things, even under the guidance of the necessity of thought, we are not led to conclusive results ; it can ever be science only in the sense of a scientific discussion of logical possibilities. At best the hypotheses of metaphysics are perhaps well-grounded postulates, carrying us some dis- tance on the way to truth, but never, as we know from result- ing inconceivabilities, bringing us to the goal.^ Before reach- ing the end of metaphysical inquiry w^e are confronted with the completely unknown, the superrational and the irrational.^ But besides the knowledge of the transsubjective demanded in the necessity of thought, there are various forms of non- rational or non-logical intuition. Indeed, defining intuition as essentially the union of immediacy with transsubjectivity, Volkelt here applies the term exclusively to these non-logical forms. From this point of view, then, there are three kinds of certainty : the self-certainty of consciousness or pure experi- ence, which is immediate, but not transsubjective; logical certainty, which is transsubjective, but not immediate; and intuitive certainty, which is both immediate and transsubjective. Five varieties of this strictly intuitive certainty are discussed, viz. moral, religious, aesthetic, vitalistic, and naive-realistic, claiming immediate transsubjective knowledge of the moral law, of union with God, of the harmony of the world, of one's ^ Die Quellen, etc., pp. 73-6 ; cf. Erfahrung und Denken, Ft. Ill, Ch. Ill ; Ft. IV, Ch. III. 2 Die Quellen, etc., § 22. ^ lb., § 23. 66 THE PROBLEM OF KNOWLEDGE own life, and of the independent, external world, respectively.^ These intuitive certainties are not to be taken uncritically, however. If naive realism were fully right, for example, the contents of sense-experience as such would have to he the ex- ternal world. 2 Still, human needs call for a broad philosophy, which shall draw, not only upon the sciences, but also upon these intuitive sources of certainty. The more logical such a philosophy of life is, the nearer it is to science ; the more the certainty of feeling retires the scientific way of knowing, the closer it stands to pure faith. But in any case it is only in its formal and negative aspect that philosophy can be regarded as scientific.^ It would be difficult to imagine a more satisfactory treat- ment of the problem of knowledge under the self-imposed limits of epistemological dualism than this which it receives from Vol- kelt. He admits, in his doctrine of intuitive certainty, the epistemological monism and realism of our ordinary conscious- ness; and yet, as a critical realist, he recognizes the essential dogmatism of this naive point of view. He claims, as a criti- cal realist, an irreducible minimum of valid representation of the transsubjective, reached through following out the neces- sities of thought ; and yet, as an epistemological dualist, he is consistent enough to admit that, strictly speaking, we do not, even in the necessity of thought, possess transsubjective knowl- edge, but simply demand it. Thus it is the very satisfactori- ness of Volkelt's discussion that reveals the unsatisfactoriness of the dualism of his epistemology — which is thus shown to be not ''without presuppositions." N] In the course of his discussion Volkelt refers to Hans Corne- lius as one who refuses to recognize the necessity of assuming transsubjective entities, and who simply refers instead to the experientially known law-abiding character of our perceptions, with its included meaning that the contents of this law ought to be ordered by the concept of this law-abiding character."* In a later article, however, Cornelius seems to be fairly upon the ground of a dualistic epistemological realism, which claims, in spite of its dualism, to be able to overcome the 1 Die Qudlen, etc., § 24. 2 lb., p. 122. ^ lb., § 25. * lb., p. 68, referring to Cornelius : Einleitung in die Philosophie, pp. 257 £f. DUALISM AND ATTEMPTED METAPHYSICS 67 Kantian agnosticism with reference to the thing-in-itself . In this article he sets himself the task of solving the prob- lem how through the impression, which is apparently in us, we can ever know the thing, which is something outside of us.^ It is generally supposed by philosophers, he remarks, that things-in-themselves are unknowable; but, if they are unknowable, why, he asks, do we continue to speak of them ? 2 After defining the thing-in-itself, in distinction from the appearance, as the thing which continues to exist while it is not perceived,^ he goes on to say : ''The law holds good of the thing, even when I do not see it, that if I will consider it under these conditions, it will have a certain appearance, and if under those conditions, a certain different appearance. This whole law is independent of momentary perception, and so dogmatic idealism is not correct, for we know that this law is true of the object even when I do not perceive it.'' * What the natural sciences, physics and chemistry, for example, teach about things, is a network of such laws for our perceptions. Every physical and chemical property of a thing therefore denotes a law for phenomena accessible under definite conditions. Since, then, every such law gives us knowledge of the thing- in-itself, the assertion that the thing-in-itself is unknowable is a mere prejudice.^ Now, apart from the dualistic presuppositions, this claim to know at least the laws of the appearances of things which exist independently of their appearances, may be accepted as valid. From our own point of view, even more than this can be known of the thing-in-itself. But the question is, whether, if all we ever know directly is our own subjective im- pressions and constructs, we can ever know that independent things exist. Would it not be sufficient to say that the law is what is true of the phenomenal object when it comes into existence? In spite of our practical conviction that such things do exist, and that we can and do know them, the theoretical doubt would remain. This is not, of course, a criticism of the * "Die Erkenntnis der Dinge an sich," Logos, I, 1910, p. 362. » 76., p. 364. 3 lb., p. 366. i lb., p. 369 (condensed translation). » 75.^ pp, 369-70. 68 THE PROBLEM OF KNOWLEDGE belief of Cornelius, that we have knowledge of the thing-in- itself ; it is simply a criticism of his fundamental theory, which would make it forever impossible fully to justify that belief. Another contemporary German thinker remains to be con- sidered here, viz. Oswald Kuelpe, who, in his relation to his old master, Wundt, may be regarded as representing the move- ment away from idealism in the direction of an essentially non- idealistic metaphysical construction. Wundt calls his own system ideal-realism, ^ but what he has accomplished does not amount to a completely harmonious synthesis of idealistic and realistic points of view in epistemology. Rather is it, as even his disciple recognizes, a fluctuation between a disguised psychological idealism (included by Kuelpe under positivism) and the metaphysics of a reality beyond the reach of human experience (though not beyond the reach of the rational thought- processes of the special sciences) .^ Wundt, however, rejects the supposed thing-in-itself, inaccessible to experience and thought, as a mere fiction.^ Metaphysics is not onl}^ possible but necessary, and must be a synthesis of the special sciences, physical and psychological.^ Kuelpe is quite ready to follow Wundt in the attempt to make philosophy a synthesis of the sciences,^ but he claims that one cannot consistently hold to the reality of both immediate experience and the transcendent objects of thought. One must choose either a positivistic immediate empiricism, in which case all metaphysical creation and aspiration are to be con- demned as futile, or else a '^neo-rationalism," which would regard immediate experience as being itself nothing real, but a stepping-stone to reality, as in the empirical sciences.® Kuelpe himself chooses the latter alternative, the transcendental method,^ which takes non-dependence upon the experiencing subject as the mark of objective reality.^ In other words, in 1 System der Philosophie, 3d ed., 1907, Vol. I, pp. 196-7. Further attention will be given to Wundt's system of thought in Ch. VI, infra. 2 Kuelpe, Philosophy of the Present in Germany, Eng. Tr., pp. 217-19. 3 System., etc., I, p. 84. * Die Kultur der Gegenwart, I, vi, p. 132 ; System, etc., I, p. 9. ' Philosophy of the Present in Germany, p. 236. «/6., pp. 218-19, 235, 248-9. ' Erkenntnistheorie und Naturwissenschaft, 1910, p. 40. ^ lb., p. 13. DUALISM AND ATTEMPTED METAPHYSICS 69 this ''critical realism of natural science," ^ which takes the results of the natural sciences as the pattern of genuine knowl- edge,2 the criterion of reality is neither purely rational nor purely empirical, but reahty is that which is found by abstract- ing from all the subjectivities of pure experience.^ It should be noted, however, that the presuppositions of Kuelpe's epistemology are more frankly dualistic than Wundt's. Experiences as such do not show anything of an external world beyond themselves, he says; they are completely shut in to themselves.^ But thought can transcend experience and reach metaphysical reality, not only by positing the external world as the cause of our perception,^ but further by taking the elements of sense-experience as representing external corporeal elements. This view, that thought has the power of transcending ex- perience, he defends, in addition to his reference to the success of the natural sciences, negatively, in his claim that Kant did not prove the forms of thought to be of purely subjective valid- ity.^ But it is not to the physical sciences alone that Kuelpe appeals in support of his theory that rational thought, tran- scending the immediately given, can discover the independent reality of which it is at once the effect and the representation. Even psychic reality, he contends, is not immediately given in experience, but has to be sought by rational thought behind the phenomena of consciousness.^ Finally, the non-idealistic character of his metaphysics appears in his refusal to commit himself to the doctrine that all reality is to be determined after the analog}^ of the mental life ; the psychical, he insists, is no better known than the physical.^ When we raise the question whether Kuelpe has been suc- cessful in passing, without dogmatism, from epistemological dualism to knowledge of the thing-in-itself, the answer must be negative. We would agree that the procedure of the sciences ought to be taken as our best guide into the field of metaphysics, and therefore as our best guide to the solution of the problems of epistemology ; but those sciences, when left to themselves, do not assume that reality is not presented at all in immediate 1 lb., p. 22. 2 Ih., p. 34. 3 76., p. 20. * lb., p. 21. 5 76., p. 22. 6 Immanuel Kant, 3d ed., 1912, pp. 77-80. 7 Philosophy of the Present, etc., pp. 222-35. » lb., p. 235. 70 THE PROBLEM OF KNOWLEDGE experience, although they do teach that it is not completely presented there. How Kuelpe comes to know that an objec- tive reality is represented in consciousness, without having ever been presented there, is not made apparent. The real problem as to how, assuming epistemological dualism, knowledge is possible, is allowed to remain unanswered. Bertrand Russell, in his Problems of Philosophy, ^ concedes that solipsism cannot be strictly proved to be false, but claims that there is not the slightest reason to suppose that it is true. Moreover, it is a simpler hypothesis to suppose that there are external physical objects to which our perceptions and ideas correspond ; and, besides, this is the instinctive belief of all mankind. 2 The epistemology suggested here is decidedly dualistic. This dualism seems to centre in Russell's doctrine of space, influenced as that has been by the non-Euclidean geometries. While it is maintained that the time-order which events seem to have is the same as the time-order which they really have,^ it is contended that we can know nothing of what physical space in itself is like, but only that the arrangement of objects in perceptual space results from and corresponds — in its logical relations — to their relations in extra-experiential or physical space.* He is unable to reach the physical object and the space of physics except by an inference which leaves their nature unknown, and only certain of their logical relations discoverable. He acknowledges that through rational thought we know only what may be, not what is. Consequently, in his philosophy, which is abstractly developed according to the methods of the mathematical sciences, while our knowledge of what may be, i.e. of what is logically possible, is found capable of indefinite extension, our knowledge of what is is reduced to an almost insignificant minimum, and even that is still dependent upon the appeal to ''instinctive belief." ^ And 1 Since the publication of this little volume Russell's views have undergone an important modification, so that now he is able to work out, with a high degree of consistency, an extreme form of realistic epistemological monism, which does not lie open to the criticisms mentioned in the present discussion. To this later form of Russell's doctrine we shall have to refer in our examina- tion of the new realism, Chs. X to XIII, infra. » The Problems of Philosophy, pp. 27, 34-5, 37-8. 3 lb., p. 52. * lb., pp. 49, 50. » 76., pp. 37-8, 230. DUALISM AND ATTEMPTED METAPHYSICS 71 SO we see that Russell, as a result of his approximation to epistemological dualism, has a narrow escape from agnosticism, if, indeed, he may be said really to have escaped it at all. As a result, then, of our investigation of recent attempts to construct a positive metaphysical system upon the basis of an absolutely dualistic epistemology, we must conclude that no such metaphysical system can logically be regarded as knowledge, for the reason that its verification by reference to immediate experience is impossible. The only metaphysics possible for the epistemological dualist is dogmatics. 2. A CRITIQUE OF IDEALISM CHAPTER V Mystical and Logical Idealism In absolute epistemological dualism, which we have ex- amined in the three immediately preceding chapters, there is asserted an existential or numerical duality between what is perceived and what is independently real. Corresponding to the two sides in this absolute dualism, the perceptual, or con- scious, and the real, respectively, we have in recent and con- temporary philosophy two forms of absolute epistemological monism, the one idealistic and the other realistic. The realis- tic form would overcome the dualism by cancelling the percep- tual or conscious content, holding it to be nothing in addition to the independent reality. The idealistic type of epistemo- logical monism, at least in its usual forms, would avoid the dualism by eliminating the other term, the independent reality, holding the real object to be nothing in addition to a perceptual or other conscious content, as such. If either form of episte- mological monism can with reasonableness be maintained, it will prove a solid foundation for the assertion that knowledge is possible. Of the two forms of absolute epistemological monism, we shall first take up for consideration the idealistic. This ideal- istic absolute epistemological monism, or ''epistemological monism and idealism" as it was designated in the recent report of the Committee on Definitions of the American Philosophical Association, was defined by that committee as the view that ''the real object and the perceived object are, at the moment of perception, numerically one, and the real object cannot exist at other moments independently of any perception." ^ This definition needs to be supplemented, however, if it is to cover 1 Journal of Philosophy, etc., Vol. VIII, 1911, p. 703. 72 MYSTICAL AND LOGICAL IDEALISM 73 the various types of theoretical ideaUsm in their epistemological aspect. In some types of ideahsm the reahty is identified with the immediate datum of consciousness, considered as a part of consciousness ; but in other cases it is identified with a predicate, the result of an abstraction from the immediately given. We may say, then, that the idealistic form of absolute epistemological monism is the doctrine that the real object and the percept or an abstract are, at the moment of perception or of thought, numerically one, and that the real object is depend- ent for its existence in the one case upon perception, and in the other case (although the relation is partially obscured) upon thought. Before examining in detail the principal varieties of ideal- ism, it may be well to intimate that we have nothing to say against either practical idealism or a certain relative theoretical idealism. By practical idealism is meant the view that there are ideals which have valid authority over every personal life, and which one must therefore assume to be, at least ultimately and progressively, realizable. More particularly, it is the doctrine that the spiritual or ''ideal" interests are properly ends, and other interests ultimately mere means ; that the ideal, also, in so far as it gains subjective existence, is a real, efficient factor in the changes which take place in the objective world. Obviously, if our customary terminology were more accurate, this practical idealism alone would bear the name of ideal-ism. ; theoretical idealism, the doctrine that reality is essentially idea, in some sense of the word, is, strictly speaking, not ideal- ism, but idea-ism. But it should be noted, at any rate, that idea-ism (theoretical idealism) has gained much of its prestige from the ideal-ism with which it is so often confused, and whose name it bears.^ But we also mentioned a relative theoretical idealism, as not to be controverted here. By this designation we mean the view that in some cases it happens that certain qualities or relations of real objects are produced directly by thought ^ See, for example, Ladd's Knowledge, Life and Reality, p. 54, where, after emphasis upon the reality of ideals as "spiritual facts and forces," the remark is added that idealism has always been "the 'school' which has commanded the adherence of the choicest spirits, as well as the most thoughtful minds." 74 THE PROBLEM OF KNOWLEDGE and the underlying purpose. Sometimes a thing becomes what the purpose and thought of an individual or of a group take it to be, even although it does not possess the quality or relation in question apart from the thought and purpose of the said individual or group. An obvious instance is the giving of a name for the first time. The same is also true of valua- tion in some cases ; an object often comes to have a value which is fictitious without being unreal ; it depends for its existence upon the thought which thinks it, and yet, at the same time, relatively, to the individual or the group concerned, it is a real value of the object. The view which recognizes both the reality of such qualities and relations, and their status as immediately dependent upon purpose and thought, might justly enough be called a relative theoretical idealism; and whatever might prove to be the case with other varieties of theoretical idealism, this at least could be defended as undogmatic and true. It so happens, however, that the name " idealism '' is not commonly applied to this doctrine. Leaving out of consideration, then, what we have called practical idealism and relative idealism, let us turn to philo- sophical or theoretical idealism, in the common acceptation of the term, and view its varieties in relation to the problem of acquaintance with reality. In undertaking a classification of the various ideaUstic theories the most natural procedure would probably be to divide them according to their deriva- tion into (1) those based upon the subject-matter of judgments in its experienced immediacy as it enters into the conscious- ness of the individual, and (2) those based upon the predicate, the mediating element. The idealism of immediacy, however, includes two main forms, viz. mystical and psychological idealism, both of which seem to be based upon the suggestion that since we learn what objects — in particular, physical objects — are through immediate mystical or sense-experience, this their appearance in immediate experience constitutes the whole of their reality. The idealism whose appeal is to the predicate rather than to the subject-matter of the judgment, the idealism of the (logical) idea, and thus the form which has a peculiar claim to be regarded as idealism proper, may be designated logical idealism. It seems to rest upon the sugges- MYSTICAL AND LOGICAL IDEALISM 75 tion that since we learn what objects are through ideas, predi- cates, things themselves must be ideas, or combinations of ideas. We propose, then, in entering upon our critique of ideal- ism, to begin by examining these three elemental types, which, to name them in the chronological order of their becoming historically important, are mystical idealism, logical idealism, and psychological idealism, respectively. Mystical Idealism Mystical idealism is an interpretation of the physical world with its contained objects, under the influence of suggestions arising from mystical experience, as being maya, mere decep- tive appearance, mere idea in "mortal mind." Since, in the more extreme phases of the mystical state, through rapt con- centration of soul upon the religious Object, the Absolute One, distinct consciousness of the physical environment lapses, this disappearance of the material world is interpreted by the mystic as meaning the unreality of matter, especially since the mystical state is felt to have a value far transcending that of ordinary consciousness.^ This seems the most natural explanation of the religious philosophy of Yajnavalkhya and the other sages of the Hindu Upanishads. One does not forget, indeed, that Deussen seems to hold that Hindu mysticism was a practical consequence of the speculative metaphysics of the sages, rather than the source from which they received their original sugges- tions and the norm with reference to which they controlled their speculations,^ but his position is virtually assumed rather than proved. He refers to the comparative lateness both of the Yoga Upanishad, containing the practical instruc- tions for cultivating the mystical experience, and of the teach- ing concerning turiya, the fourth, or mystical state of the soul ; ' but this must not be taken as conclusive evidence. Technical instruction for reproducing mystical states would not in any case be likely to be committed to writing until after the theo- ^ Cf. Delacroix, Etudes d'histoire et de psychologie du mystidsme, p. 370 ; G. A. Coe, "The Sources of the Mystical Revelation," Hibbert Journal, VI, 1908, pp. 363-5. , 2 P. Deussen, Philosophy of the Upanishads, Eng. Tr., pp. 342, 383. » lb., p. 309. 76 . THE PROBLEM OF KNOWLEDGE logical dogmas suggested by that experience had gained con- siderable prestige, such as might have come through their being put into hterary form, as in the earlier Upanishads.^ And as for the lateness of the idea of the fourth state (turiya), the explanation may conceivably be that the mystical state was formerly meant to be included in the third state of dreamless sleep (prajna), in which, while there is no consciousness in an empirical sense, the self is not annihilated, but becomes identi- fied with the one supreme Spirit. The fourth state differs only in that the unification with the supreme Spirit is realized in a consciousness; when the false knowledge of the first two states (ordinary consciousness and dreams) and the no knowl- edge of the third state vanish, then the fourth state is reached.^ What this suggests is that prajna is, or includes, the mystic unconsciousness, while turiya is either the mystic consciousness, succeeding the condition of trance, or the trance itself, inter- preted, in the light of later reflection, as being a sort of super- consciousness. In either case the idea of turiya would be merely a later supplement to a previously existing doctrine of the mysti- cal state. But the evidence upon which we mainly rely for our con- viction that the idealistic interpretation of the physical world in the philosophy of the Upanishads was originally based, at 1 It may be, as some scholars (e.g. E. W. Hopkins, H. Oldenberg, Die Lehre der Upanishaden und die Anfdnge des Buddhismus, 1915, pp. 89, 90) assert, that the idealistic maya doctrine is not clearly discoverable in the very oldest of the Upanishads. This would only bring the date of the beginning of what we take to be the mystical idealism of these writings down closer to the time of the undisputed existence of mystical practice. But even if we should agree with these scholars in contradistinction from Deussen with reference to the explicit teaching of the oldest Upanishads, we could still point out on the one hand that these earliest philosophical writings contain a monistic (singularistic) idealism in which the maya doctrine is at least implicit, and on the other hand that there was what we may regard as a crude and primitive mystical practice and experi- ence in the shamanistic religion of the seers of the late Rig- Veda period. George F. Moore says of the methods employed by Hindu mystics for the purpose of inducing trance states, "At a later time these methods are systematised in the Yoga, hut in essentials the method is very old; it had a place in Buddhism from the beginning " (History of Religions, I, 1913, p. 278, italics mine) ; but, on the basis of what has just been said, we should judge it probable that Hindu mysticism was at least as early as Hindu idealism, and that the latter rests upon the former. 2 Deussen, op. a^, p. 311. MYSTICAL AND LOGICAL IDEALISM 77 least in large measure, upon suggestions derived from mystical experiences, we find in the remarkable correspondence between the characteristic doctrines of the Upanishads and what would be most naturally suggested by the psychological fea- tures of extreme mysticism. The doctrine of the sole reality of Brahman would naturally be suggested by the experience of rapt mystic contemplation, when all but the divine One lapses from consciousness. Moreover, it is noteworthy that it is chiefly with regard to Brahman (originally the God of mystic power and of prayer) , who during the Brahmana period gradu- ally displaced Prajapati (the Lord of creatures),^ that this doctrine was formulated. That Brahman (the Absolute) is Atman (the Absolute Self, one's own true Self) is a characteristic doctrine of the mystics ; we are reminded of Madame Guy on, whose illumination came as a consequence of the suggestion that God is to be sought within one's own heart. ^ The sole reality of the Atman and the illusory character consequently to be ascribed to the seemingly independently real world of appearance are mystical doctrines which seem to be at least incipiently present in what scholars take to be the oldest of the Upanishads.^ That knowledge of the Atman is not a means to emancipation (moksha) simply, but is emancipation, could hardly have been suggested otherwise than in a mystical experience at once of illumination and emotional uplift.* In the description of Brahman we have the usual negative theology of the mystics ; ^ any doctrine of God as being in relation to the universe is to be understood as the result of accommodation to the point of view of ignorance (avidya).^ Moreover the identification of the divine One with the syllable ''Om"^ seems almost meaningless, except in view of the fact that con- 1 lb., p. 86. 2 Of. Kathaka-upanishad, 2. 4. 1, "The wise man right within saw the Atman, Fastened his gaze on himself, seeking the Eternal." See Deussen, op. cit., p. 84, and Sacred Books of the East, Vol. XV, p. 15. ^ E.g. Brihadaranyaka-upanishad, 2.4, 3.1-4.5; v. Sacred Books of the East, Vol. XV, pp. 108-13, 121-85. * Brihadaranyaka-upanishad, 4.2-4 ; v. S.B.E., XV, pp. 158-81 ; cf. Deussen, op. cit., pp. 344-55. ^ Mundaka-upanishad, 1.1.5, 6; v. S.B.E., XV, pp. 27-8. ' Cf. Deussen, op. cit., p. 159. 7 Svetasvatara-upanishad, 4.18 ; v. S.B.E., XV, 253 ; Deussen, op. cit., p. 352. 78 THE PROBLEM OF KNOWLEDGE centration of attention upon this syllable was used as a means of inducing, by self-hypnotization, a mystical experience. Even if sacred associations may have been established in other ways, still the result of mystical practice and experience would be to bring about a more intimate relation between the syllable and the idea of the divine Being. It seems equally impossible, apart from this reference to mysticism, to appreciate the basis of Yajnavalkhya's declaration, "Brahman is bhss and knowledge."^ The depreciation of action, or works, as "that evil thing," with which those who find the Atman are "no longer stained "^ strongly suggests the influence of a quietistic tj^pe of mysticism. (It is not denied, of course, that the whole karma doctrine, of which the passage just cited seems to be an expression, has other roots besides this of mysticism.) A similar significance should probably be found in the fact that the earliest known appearance of asceticism is claimed as having been among the Indian people ; ^ fasting and other rigors, endured at first perhaps involuntarily, may have led at times to the mystic trance ; naturally, then, such practices would be adopted as a voluntary system of self-discipline, looking to a repetition of so highly valued an experience. In view, then, of the thoroughly mystical character of the doctrines associated with early Indian idealism, we feel war- ranted in taking the latter as an instance of mystical idealism, in the sense defined. Here, as elsewhere, the primitive explana- tion of experience was ontological, rather than psychological; the Hindu mysticism was a source, as well as, in its later de- velopment, a consequence of Hindu philosophy. Deussen's conclusion has not improbably been influenced by his very evi- dent interest in finding a confirmation of his own philosophy in being able to think of it as worked out, primarily by the speculative method, by the sages of ancient India, as well as by Parmenides and Plato in ancient Greece, and again inde- pendently by Kant and Schopenhauer in modern Europe.^ His bias is further shown by his statement that the thought 1 Brihadaranyaka-upanishad, 4.1 ; v. S.B.E., XV, pp. 153, 157. 2 TaiUiriya-brahmanam, 3.12.9.8; v. Deussen, op. cii., p. 373. ' Deussen, op. cit., p. 65. * lb., pp. 40-1 ; The System of the Vedanta, Eng. Tr., pp. 47-9. MYSTICAL AND LOGICAL IDEALISM 79 that the entire universe is only appearance, and not reality, is the presumption and sine qua non of all religion. ^ Among the mediaeval mystics we meet with mystical idealism again, and in some cases we may even see it in process of forma- tion. Thus Albertus Magnus writes: ''When thou pray est, shut thy door — i.e. the doors of thy senses. ... A mind free from all occupations and distractions . . . is in a manner transformed into God, for it can think of nothing, and love nothing, except God ; other creatures and itself it sees only in God. . . . Do not think about the world, nor about thy friends, nor about the past, present or future. But consider thyself to he outside of the world and alone with God, as if thy soul were al- ready separated from the body and had no longer any interest in peace or war, or the state of the world. Leave thy body and fix thy gaze on the uncreated Light. Let nothing come between thee and God. The soul in contemplation views the world from afar off.^^ ^ According to Eckhart, again, the soul can know finite and material things only by creating images of such things,^ but the mystic is one who ''has renounced all visible creatures." ^ Eckhart teaches definitely that out of God there is nothing but nonentity. The independent existence of single objects is mere appearance, having its source in human thought.^ In the Theologia Germanica, once more, we read, "The two eyes of the soul of a man cannot both perform their work at once ; but if the soul shall see with the right eye into eternity, then the left eye must close itself and refrain from working, and be as though it were dead.'' ^ In all these passages the doctrine seems to be that the would-be mystic must learn to treat the finite and material world as unreal, until it comes to seem as unreal as it really is. In modern times perhaps our best example of mystical ideal- ism, whether it be regarded as taken over from traditional 1 Philosophy of the Upanishads, p. 45. 2 De Adhcerendo Deo, 1st paragraph; italics mine. See R. M. Jones, Studies in Mystical Religion, 1909, p. 219. 3 Mystische Schriften, p. 15. * Strasbourg Sermons. 5 Pfeiffer's Deutsche Mystiker, Vol. II, pp. 207, 589 ; see Ueberweg's History o/P/it7osop%, Eng. Tr., Vol. I, pp. 475-6, ^ Ch. 7 ; italics mine. 80 THE PROBLEM OF KNOWLEDGE sources, or as originated out of mystical experiences/ is to be found in the teachings of Mrs. Mary Baker G. Eddy. She writes : ''To understand that the Ego is Mind, and that there is but one Mind . . . begins at once to destroy the errors of mortal sense." "Mortal existence is a dream without a dreamer." ''Rely not in the least on the evidences of the senses." "All is mind, there is no matter, and you are only seeing and feeling your belief." "When we say, 'I have burned my finger,' that is a correct statement, for mortal mind and not matter burns the finger." " Man is not sick ; for mind is not sick, and matter cannot be." ^ Mrs. Eddy is to be credited with having endeavored to take her mystical idealism seriously, at least in the treatment of bodily ills as non-existent. But even in this realm she had to acknowledge limitations. It is a surprising lapse in the direction of common sense, when she writes, "Until the age advancing admits the efficacy and supremacy of mind, it is better to leave the adjustment of broken bones and dislocations to a surgeon, while you are re- constructing mentally, and preventing inflammation or pro- tracted confinement. . . . Mental surgery is the highest branch of metaphysical science, and will be understood and demonstrated the last." ^ An elaborate refutation of mystical idealism is unnecessary. It rests upon no more stable foundation than the notion that what lapses from consciousness in a special state of mind is thereby shown to have been unreal, non-existent. And in practical life, as is notorious, it cannot but refute itself. 1 There are indications that Mrs. Eddy's philosophy was not something merely taken over from others or evolved speculatively, but that it was to some extent rooted in mystical or quasi-mystical experiences. When a child she often experienced auditory automatisms; she also tells of "a soft glow of in- eJBfable joy" experienced, while still a child, in response to prayer and accom- panied by physical healing. With reference to her peculiar doctrines the following quotation is significant: "When apparently near the confines of the death-valley, I learned certain truths: that all real being is the divine Mind and idea," etc. Her constant claim was that the contents of her book came to her by "revelation." See F. S. Hoffman, The Sphere of Religion, i>i). 188-210. 2 Science and Health, 1881 ed., Vol. I, pp. 68, 121, 187, 189, 226, 233. » lb., p. 220. MYSTICAL AND LOGICAL IDEALISM 81 Logical Idealism There is a second elemental type of idealism, which we may call logical idealism. In preliminary fashion it may be defined as the form of idealism suggested by reflection upon the logical or dialectical process. Its most important historic exemplifi- cation is to be found, we would say, in the system of Plato ; but inasmuch as there has come to be some divergence of opinion as to what Plato's doctrine really was, some brief indication of the interpretation we have adopted must needs be offered. The recent contention on the part of Paul Natorp and J. A. Stewart that Plato's doctrine of ideas was essentially methodological, rather than metaphysical, is worthy of seri- ous attention. The topics chosen for dialectical discussion in the dialogues indicate that Plato shared the fundamentally practical and ethical interest of his master ; but, probably to an even greater degree than for Socrates himself, the "So- cratic" method became to the pupil an independent object of interest.^ The dialogues were manifestly written not alone to set forth an ethical and political doctrine, but also very largely as illustrating the dialectical method of arriving at adequate definitions. But the problem of vindicating the possibility of knowledge had become a real one, especially in view of the sceptical no- tions propagated by the Sophists; and for Plato the answer to the epistemological question, practically at least, seems to have been virtually contained in the methodology he had learned to employ. In true judgment, and especially in the case of the adequate definition, the predicate, as we shall see,^ has value, for practical purposes, as a substitute for further immediate experience of the thing of which it is predicated. Now this functional equivalence of the predicate, or logical idea, with the reality under consideration is very far from being an ^ See, for example, Euthyphro, 5, 6, 11 ; Charmides, 159 ff. ; Laches, 191-2 ; Meno, 71-3, 97-8 ; Gorgias, 448 £f. ; Lysis, 212 ff. ; Republic, 507, 511, 533, 596, etc. ; Theaetetus, 185, 208 ; Politicus, 285 ; Laws, 965. Cf . P. Natorp, Platos Ideenlehre ; J. A, Stewart, Plato's Doctrine of Ideas, Part I, and Mind, N.S., XIX, 1910, pp. 117-21. Cf. A. E. Taylor, Mind, N.S., V, 1896, pp. 297-326, 483-507 ; VI, 1897, pp. 9-39 ; XIX, 1910, pp. 82-97 ; Plato, 1908. 2 Ch. XIX, infra. G 82 THE PROBLEM OF KNOWLEDGE absolute identity of the two ; and yet it would seem that Plato tended to confuse the one with the other. In seeking to know the reality under discussion, one was seeking its true definition with the help of which it could be more adequately known ; it was a natural, although illogical, conclusion that the reality, the real nature or essence of anything, is just its definition. The *'is" of predication was here turned into the "is" of absolute identity. Having once arrived at the adequate logi- cal idea, the absolutely satisfactory and universally valid predi- cate, it was assumed that one would have the essence, the per- manent reality, of the thing. Things could be regarded as essentially knowable, apprehensible by rational intelligence, since ideas (instead of being taken as the instruments of knowl- edge, which they primarily are) were set up as being the true and indeed the only objects of knowledge, the reality of the thing being at the same time identified with the "absolute idea^' or "universal." ^ This doctrine that the reality is the (logical) idea, making it possible, in spite of the supposed fact that only ideas are knowable, to hold that reality is knowable, may be called the Platonic, or logical, idealism. Briefly put, it is the doctrine that if things are known, they must be what they are known with, viz. ideas. Now this, to be sure, is not, as Taylor observes that it is not,^ "idealism in the modern sense of the word," if one means by modern idealism either a psychological idealism like that of Berkeley, or a combination of logical with psychological idealism, such as we have in modern absolute idealism. Greek philosophy was essentially prepsychological, and its idealism, if we may call it such, was also prepsycholog- ical. But the doctrine that reality is essentially idea, such stuff as definitions are made of, and that it is, as such, a possible object of knowledge, may rightly enough, it would seem, be called a form of idealism. If we could suppose that Plato noted and remembered that all such ideas are the result of abstraction, and that they have their true being in and for the abstracting mind, we should have no difficulty in classifying his system as a variety of epistemological idealism. If, how- ever, we are led to conclude that the ideas with which things, 1 Cratylus, 386, 439 ; Phoedo, 78 ; Philehus, 15 ; Timers, 27-8. 2 Plato, p. 43. MYSTICAL AND LOGICAL IDEALISM 83 when truly known, are to be identified are finally interpreted by Plato as realities existing independently of thinking, it will be more difficult to make the above-mentioned classification. The difficulty is due to the fact that in this logical idealism, as in all abstract idealism,^ there is the constant tendency to forget that the abstract idea is an idea. If, however, we may be allowed to correct for the "abstract idealist" this his over- sight, we can without doubt include his system under idealistic epistemological monism. But under neither of the two inter- pretations would there be any difficulty in classifying the system as an epistemological monism ; in both cases, during rational thought the real object and the object immediately present to thought are identical. But Natorp and Stewart offer a third suggestion. As is done in the former of the two interpretations suggested above, they maintain that, for Plato, reality is idea in the sense of mental construct ; but they are not willing to grant that it is an empty concept. Natorp, perhaps in order to gain further credit for his own neo-Kantian positivistic idealism, reads it back into Plato, as his master, H. Cohen, read it into Kant. Stewart, the Plato-specialist, become Plato-lover and Plato- idealizer, seeks to gain, one is tempted to guess, new apprecia- tion for the object of his veneration by showing how very credit- able, from modern points of view, is the whole philosophical system of Plato. Natorp, then, and, following him, Stewart agree that for Plato reality is a construct of the human mind, but only in the phenomenal realm, the realm of possible human experience. Thus when Natorp maintains that the Platonic "ideas" are ^^ merely the predicates of scientific judgments," ^ we must not fail to interpret this as simply one of the premises in a train of reasoning by which it is supposed that Plato taught the essentials of what the neo-Kantian beheves. The premises are the following : Things are ideas ; ideas are predicates ; predicates are thought-constructs. The conclusion is the neo- Kantian doctrine : Things are thought-constructs. But while we would reject the first of these three premises, and therewith the conclusion, Plato, as we shall maintain, would have re- jected the third; or, if not the third, the second; for, while 1 See Ch. IX, infra. 2 Qp^ ^i^^ p, 351 j italics mine. 84 THE PROBLEM OF KNOWLEDGE he held that things are ideas, he had no intention of asserting that things are mere thought-constructs. Similarly the doc- trine of the '^ participation" of the things of sense-experience in the non-sensuous, eternal ideas, which Plato confesses that he has been ''always and everywhere repeating," ^ is interpreted by Stewart as meaning that the perceived object is "constructed by the activity of mental categories."^ '''Participation' is predication," he writes in good neo-Kantian fashion,^ and Taylor applauds the assertion.^ But this attempt to make Plato a neo-Kantian will hardly do. Plato never held that the object which appears to us in sense-experience is a construct of human mental activity. What it would be true to say is that the ideas which we predi- cate of these objects of sense-experience are constructs of human thought; but even this seems to have been largely ignored by the Attic philosopher. His doctrine was not that real existence is a mental construct, nor even that the true idea is such a construct. Real existence is a discovery, something discovered, not a construct; and the true idea, the universal or absolute idea, the definition, is also a discovery, not a con- struct. And since the discovery is at once of the existence and the idea, it is assumed that there is absolutely no difference between the real existence and the true idea. The fallacious analysis here is easily exposed. As was intimated above, the assumption is that our knowledge of reality by means of an idea is simply knowledge of the idea ; whereas the idea is not as such the object but merely the instrument of knowledge. But what we are here especially interested in emphasizing is that the attempt to interpret Plato as in essential agreement with the neo-Kantian idealism is, for the reasons given, funda- mentally mistaken. Since the "universal idea," then, is, in the Platonic system, not a construct of human thought, but its discovery, it almost inevitably comes to be regarded as a permanent reality, inde- pendent of human cognition. Alongside of the Platonic or logical idealism, the doctrine that realities are ideas, there tends to develop a logical realism,, the doctrine that ideas, logi- 1 Phxdo, 100. 2 Op. cit., p. 67 ; italics mine. 3 lb., p. 77. 4 Mind, XIX, 1910, p. 82. MYSTICAL AND LOGICAL IDEALISM 85 cal entities, are independent realities. Let us see, if we can, whether the Platonic (or logical) idealism gave rise to, or even passed over into, a Platonic (or logical) realism. In affirming that reality is really a logical idea, a definition, the logical idealist is saying that reality is really something abstracted from reality. It is small wonder, then, that the position proves to be one of unstable equilibrium. Logical idealism is a form of abstract idealism which tends to pass over into a psychologi- cal idealism,^ or else into logical realism. If the abstractness of the logical idea (as related to reality) were consistently recog- nized, with the consequence of the identification of the reality with the logical idea as it is in its mental context, i.e. with the definition when and as it is thought, what we would have would be no longer the original logical idealism, but a psychological idealism, of a somewhat Fichtean or neo-Kantian type. But such was not Plato's doctrine. The Platonic logical idealism could hardly have been held if the philosopher had not ab- stracted from the fact that the logical idea is itself an abstract, actually existing only in a context of consciousness. If, on the other hand, the abstraction be taken abstractly, i.e. if the abstractness of the idea, with reference to (what we call) real objects, be abstracted from, we shall have, in the doctrine that all realities are independently real ideas, the source of the con- verse proposition which, when logically inferred, is the doctrine that some logical ideas are objective realities ; when illogically, the doctrine that all logical ideas are objectively and independ- ently real. Thus it would seem to be a plausible hypothesis that, by a process of double abstraction, Plato was led from his methodology, first to logical idealism, and then, because of his not recognizing what he had done, through this disguised logi- cal idealism to logical realism.^ But that the independently real ideas of Plato's doctrine 1 See Ch. VI, infra. 2 Inasmuch, also, as this second abstraction (to which we shall have occasion to refer again, first in our discussion of the disintegration of idealism in Ch. IX, and again in tracing the antecedents of the new realism, in Ch. X) remains unrecognized, it may be said that it, too, is taken abstractly. But since what we mean by abstraction in this connection is simply not recognizing an important actual relation, our criticism, that the fact of abstraction is itself abstracted from, does not lead us into any "indefinite regress." 86 THE PROBLEM OF KNOWLEDGE are thoughts of the divine Mind, which was the form the teach- ing took in Philo, in Plotinus, and in the earHer mediaeval philosophy, must not be read back into the thought of Plato himself. The opposite mistake, however, seems to have been made by the recent interpreters to whom we have referred, when they deny that Plato has any doctrine of the metaphysi- cally real existence of ideas. ^ It may be admitted that such interpreters as Zeller ^ and Windelband ^ may have insufficiently appreciated the methodological interest of Plato, and may have attached too much importance to the metaphysical aspects of Plato's thought. One may even suppose that Aristotle, in spite of his unique opportunity for knowing what Plato really thought, in stating as the essence of Plato's doctrine that which he felt it essential to eliminate, exaggerated the extent to which his master was concerned to insist upon the real existence of ideas beyond all possible human experience.^ His sketch is perhaps something of a caricature, as interpretations often are. But with the help of such an expositor as Gomperz,^ it is pos- sible to understand how a conclusion which seems foreign to our ways of thinking may have come to seem natural and even necessary to the mind of Plato. There are indications, however, that Plato had some mis- givings with regard to this metaphysical aspect of his doctrine. By this we do not mean that he detected any fallacy in his processes, but only that he gives evidence of dissatisfaction with the results. His logical realism was the converse of his logi- cal idealism, and when the question is raised as to whether the conversion was performed in a logical or an illogical manner, the answer should probably be that it was not one or the other simply, but both : at times the doctrine seems to be that all universal logical ideas are independently real existences; at other times it seems to be that only some of such ideas have transcendent existence. The explanation undoubtedly is that 1 Natorp, op. ciL, pp. 63-4, 70-1, 73-4, 86, 126-7, 131 ; Stewart, op. cit., pp. 37, 40, 62-5, 83. See Taylor, Mind, V, 1896, p. 505 ; Plato, p. 48 ; Mind, XIX, 1910, p. 93. 2 Plato and the Older Academy, pp. 227, 235, 247, 271-6. » History of Ancient Philosophy, pp. 193, 196; History of Philosophy, p. 118. * See Metaphysics, XII, 3. 1070a. 18 ff., 28. 6 Greek Thinkers, Eng. Tr., Vol. II, pp. 180-2 ; Vol. Ill, pp. 4-7. MYSTICAL AND LOGICAL IDEALISM 87 here Plato's thought had not reached a state of equihbrium, but continued to oscillate somewhat between the two positions. In the one instance the doctrine we have called logical idealism — itself more than doubtfully founded — was converted simply and illogically as follows : All realities are ideas ; therefore all ideas are realities. In the other instance the conversion was by limitation, and therefore logically, as follows : All reali- ties are ideas ; therefore some ideas are realities. It is remark- able that it is the formally logical converse which is made the basis of the more extreme and metaphysical doctrines of Plato, while the formally illogical converse becomes the basis of his more moderate and positivistic thought. This is doubtless because, as we shall see, the former is supported by semi-mysti- cal considerations, while the latter, the inference that all ideas are realities, is defended by confining its expHcit application to experienced objects. We shall first deal with the doctrine that some ideas are eternal and transcendent realities. Besides the oft-quoted explicit passages in the Phcedrus ^ and the Timceus,^ there is the characteristic doctrine of knowledge as '^ reminiscence," ^ with its implication that both the soul and its objects, the ideas, are eternal. Possibly, as Taylor suggests,* under the influence of the experience and thought of Socrates, Plato developed the doctrine that before birth, as well as after death, the soul con- stantly enjoys the ''beatific vision" of the eternal ideas, and during the present life only with difficulty recollects (or antici- pates) something of that experience, in a state which amounts, at its best, to a ''rapt amazement" or "sort of ecstasy." ^ We see at this point how natural was the transition from the philos- ophy of Plato to the definitely mystical idealism of Plotinus ; but it is not clear that in the case of Plato there was any dis- tinctly mystical religious experience, or anything more than the contemplation of the logical ideas until they became, through projected feeling, things of beauty, unique and unchangeable; and so, glowing with that subjective "light which never was on 1 247. 2 51-2. 3 See, for example, Phcedo, 67-8, 74-6 ; Phcedrus, 249-50. * Mind, XIX, 1910, p. 94 ; cf. Varia Socratica, 1911, pp. 16, 22-4, 30, etc. ^ Phcedo, 72 ; Phcedrus, 249-51 ; cf. Symposium, 210, 88 THE PROBLEM OF KNOWLEDGE sea or land," they were substantiated as eternally real exist- ences.^ Obviously, however, this semi-mystical contempla- tion would find place only in the case of some of the ideas, such as the idea of the Good, and its included ideas; or, in other words, the ideas of the ''eternal values." This was the element of Platonism that impressed itself upon the religious conscious- ness of later generations, and that was retained with theistic or pantheistic modifications, by Philo, Plotinus, Dionysius, and the ''Platonic realists" of the middle ages. But this semi-mystical contemplation of the Ideas would give no support to the notion of an eternal existence of the ideas of "hair, mud and dirt." ^ And yet in the Thecetetus it is taught that every object of thought must exist,^ while in the Parmenides we find any reluctance to believe that every object of sense, however mean, has its eternal idea, treated as evidence of philosophical immaturity.* Shall we conclude, then, with Stewart,^ that Plato is here simply criticising a metaphysical doctrine which he himself never held, but which was simply a common misinterpretation of his teaching on the part of his disciples, his own doctrine having been strictly and consistently methodological? Or, shall we give up the attempt to find consistency, and conclude, with Gomperz,^ that the Parmenides was written when the philosopher's mind was in a state of ferment, and that it simply considers a number of plausible objections to his own theory, without reaching a conclusive answer to them — in which case its doctrine would be compara- tively negligible? Or shall we hold, with D. G. Ritchie,' that we have here, in one of Plato's later dialogues, written perhaps under the influence of the young Aristotle himself, an approach to the Aristotelian doctrine that ideas have real existence only in minds and as the forms of the things of sense? Adopting this third interpretation, we should be able to see how, accord- ing to Plato, the ideas are to be regarded as causes, not only of 1 Cf. Stewart, op. cit, Pt. II, especially pp. 139^0, 167, 184, 186, 194, 196. 2 Parmenides, 130. 3 Thecetetus, 189; of. Republic, 476, 510; compare Hegel's "The rational is real." * Parmenides, 130 ; see also 132-5. 5 Op. cit., pp. 70-80 ; cf. Natorp, op. cit. ^ Greek Thinkers, III, pp. 150-1. "> Plato, pp. 115-19. I MYSTICAL AND LOGICAL IDEALISM 89 knowledge, but of being and essence ;'i they are the rational forms of reality, the universalia in rebus, without which the things of sense could not exist, but which are eternally real, not constructs of the activity of the human mind. In this phase of his thought Plato was the pathfinder for Aristotle.^ But for our present interest in the problem of acquaintance the essence of Plato's doctrine is the Platonic, or logical, or dialectical, idealism, the doctrine that reality is constituted of logical ideas, albeit in systematic relation to each other, and that we have thus direct acquaintance with reality in the ideas of logical thought. The fallacious reasoning upon which this variety of idealistic absolute epistemological monism is based has already been exposed. It may be well, however, to refer once more to the fact that an idea, even when it amounts to a definition, is very far from being existentially identical with that of which it is an adequate idea, or definition. The logical idea is always, as related to the reality under consider- ation, not the reality, but an abstraction from it or from others of its class. In the psychical context it is a reality, a mental product; but in the objective context it is not a reality. So then, to say that a reality is the logical idea which may be predi- cated of it, is virtually to say that the reality is not a reality, but an abstraction from reality. The inexpugnable error of logical idealism is abstractionism. Mystical-Logical Idealism The two elemental forms of idealism which we have already examined, viz. mystical and logical idealism, exist in combina- tion in what may consequently be called mystical-logical ideal- ism. Of this first composite form of idealism to demand our attention the best historical example is doubtless the philos- ophy of Plotinus. This system is built upon the Platonic dialectic ^ and mystical religion,^ an ecstatic experience which Plotinus is said to have had several times ^ and which seems to 1 Republic, 509 ; Timceus, 58. 2 It is at this point that we see how plausible is the neo-Kantian interpreta- tion of Plato's doctrine; but, as we have already insisted, that the object of sense-experience is a mental construct is wholly foreign to Plato's thought. 3 Enneades, I, iii, 3-5. " lb., VI, ix, 4, 8-11. ^ See Porphyrius, Vita Plotini, Qh. 23, 90 THE PROBLEM OF KNOWLEDGE be given credit for a certain phase of the philosophy in the prayer of Plotinus for insight into the relation of the Many to the One.^ As results of the synthesis we have, practically speaking, two Absolutes, the one mystical and the other logical, and a double basis for an idealistic interpretation of the physi- cal. The logical Absolute is Intellect (Nous), in which all things exist eternally as the world of absolute ideas, or pure essences.^ This interpretation of the logical Absolute as a metaphysical reality is a further development of the Platonic logical realism. The mystical Absolute, on the other hand, is the perfect One, the first God, contemplated by the mystic as that with which his soul seeks union. ^ Here we have mystical realism ; the Absolute of mystical experience is affirmed to be the ultimate Reality. And yet for Plotinus, fundamentally, these two Absolutes are one. The dialectician can get no nearer to ultimate Reality than as far as Intellect, the world of rational forms; the mystic penetrates further and becomes absorbed in the One. There is nothing in Intellect which is not, in some sense, in the One ; although not all of the One is in Intellect, or can be reached by intellection.^ But, more to our purpose than this synthesis of mystical and logical realism in the absolutism of Plotinus, is the way in which mystical and logical idealism are combined in his essen- tially idealistic interpretation of physical objects. As the realism of the mystics with regard to the rehgious Object has commonly been led to an idealistic interpretation of physical objects, and as the logical realism of Plato and his followers originated in what we have called a logical idealism in the inter- pretation of the things of experience, so the combined mysti- cal and logical realism of Plotinus with reference to the Abso- lute on the one hand conditioned, and on the other hand was conditioned by, an idealistic interpretation of the physical. The One, being perfect, and therefore in want of nothing, ''be- comes, as it were, overflowing, and the superplenitude of it produces something else." Its first product, or emanation, is Intellect, the Absolute of logical reahsm.^ Similarly the soul 1 Enneades, V, i, 6. ^ /&., m, ix ; V, i, 7 ; ix, 4, 8-11 ; VI, ix, 2. « lb., Ill, viii, 8, 9 ; ix, 3 ; V, i, 7 ; ii, 1 ; VI. ix, 4, 8-11. * lb., VI, ix, 4. 5/6., V,i, 6; ii, 1. MYSTICAL AND LOGICAL IDEALISM 91 is the product, by emanation, of Intellect.^ The Soul, again, produces all animals and inspires them with life.^ But the world is also an animal, comprehending within itself all ani- mals.^ Hence all things physical depend upon Soul for their existence and, inasmuch as Soul depends upon Intellect, and Intellect upon the One, all things physical depend for their existence upon Intellect (or the absolute idea), as in logical idealism ; or, to speak still more ultimately, they depend, for what being they have, upon the undifferentiated One of mystical intuition. Obviously, the criticisms which are valid against mystical and logical idealism in separation are still valid against the idealism resulting from their combination. Neither of the elemental types was incomplete merely ; each, as we have seen, was the result of positively erroneous suggestion. Hence they cannot be said each to supplement the deficiencies of the other; rather does each, by appearing to confirm the other, simply afford the mystic-philosopher a deceptive feeling of security in his twofold error. 1 76., V, i, 7. 2 75., V, i, 2. » 76., V, ix, 9. CHAPTER VI Psychological Idealism Besides mystical and logical idealism, there is a third ele- mental type of idealistic philosophy, viz. psychological idealism. This may be defined, in preliminary fashion, as the interpreta- tion of the physical object, under the influence of an erroneous suggestion arising in connection with the psychological point of view, as being essentially idea, in the psychological sense of that word, i.e. as being simply a part of consciousness, a content of conscious life which depends upon consciousness for its exist- ence. As contrasted with the other elemental forms of idealism, this psychological type is characteristically modern. This is undoubtedly connected with the fact that psychology may be said to be, almost exclusively, a modern science. Ancient and mediaeval thought were both essentially prepsychological, the former through lack of consciousness of self as soul, the latter through defect of scientific spirit. But already at the dawn of modern philosophy we find a dualism, a consciousness of prob- lem in connection with mind and matter. The new conscious- ness of self or soul, as constituting a subjective world and not as a mere element or principle of activity in the objective world, was probably due in large part to two causes. First, there was the attention given in the Christian religion and in mysticism to the soul, with the accompanying high estimate of its value and the sense of momentous importance attaching to its different states. And, secondly, there was, as seen in the Renaissance and the Reformation, as later in the Aufklarung and the Revolution, the protest of the individual against the established order. In opposing themselves to the objective social order, men became more conscious of themselves as subjects.^ 1 Cf. J. Dewey, Philosophical Review, XVIII, 1909, pp. 182-3. 92 1 PSYCHOLOGICAL IDEALISM 93 But, be the explanation what it may, the duaHsm of the psychical and the physical was especially prominent in the thought of the early modern philosophers. Its place was funda- mental in the systems of Descartes and his immediate followers among the rationalists, and in that of Locke among the empiri- cists; and the later rationalists and empiricists alike busied themselves with the problems to which it first gave prominence. These problems were chiefly two, viz. how such essentially dif- ferent substances as body and mind could interact on each other, and how mind could know extra-mental objects. The earlier solutions, apart from the Cartesian appeal to a deus ex machina, were three : absolute monism, represented by Spinoza ; monadism, represented by Leibniz ; and, finally, psychological idealism, of which Berkeley was the pioneer and a typical repre- sentative. According to absolute monism, there is no interac- tion of substances, since there is but one substance; and the test of cognitive value is something immanent, viz. rationality. According to monadism there is neither interaction nor im- mediate awareness of external reality, but only immanent action and cognition, the difficulties of the view being relieved to some extent by means of the dogma of a pre-established harmony securing the appearance of transeunt action and cogni- tion. According to psychological idealism, there is no material substance, but only minds with their ideas ; from this point of view, therefore, the problems of interaction between mind and matter and of knowledge of external reality disappear as false problems. It is with this last philosophical doctrine that we are here concerned. Now it will be seen on examination that the dialectic culmi- nating in subjective idealism should by no means be regarded as convincing. Besides the fact that other solutions offered (absolutism and monadism) are no more fantastic from the point of view of common sense, there is the question whether philosophy is justified in taking up as its task the explaining away of the appearance of interaction. May it not be that action and interaction are ultimate facts, which are tO be acknowledged rather than denied and explained away as mere appearance? Can mystery be eliminated from the fact of becoming, even without the hypothesis of interaction? And 94 THE PROBLEM OF KNOWLEDGE with reference to knowledge of external reality, may it not be that the problem here arises because of a false mode of conceiv- ing the mind, as a quasi-spatial receptacle, which can contain only mental entities, viz. ideas, in the psychological sense. Another view of mind and consciousness might cause this problem to disappear without the drastic expedient of denying the reality of material objects altogether. But it was not simply as the synthesis of apparent antinomies that psychological idealism arose. It was presented as the outcome of an analysis of experience, such as is performed in the most elementary psychological study, or, at least, as a legitimate inference from the results of this analysis. Hence, as we have intimated, the name, psychological idealism. What we mean, then, by psychological idealism is the doctrine that things are ideas in the mind, or in consciousness; that they depend for their existence upon their being in the mind, or at least in the conscious relation to some subject. That this doctrine is pure dogma will appear when it is shown that the argument in support of it, when stated as a formal inference, cannot be other than fallacious, and this because the original analysis was vitiated from the start by a natural but erroneous suggestion. Man had the problem of accounting for illusion and error, the content of which after all had some sort of reality, for it was there in experience. Since it was found not to have objective reality, itg reahty must be subjective; its existence, in so far as its illusory or erroneous features were concerned, was dependent upon its being object for some conscious subject. Thus it may be said that the consciousness of self and of the relation of objects experienced to the self naturally arise to- gether, illusory objects being subjective in a twofold sense, i.e. as dependent upon the conscious self for their existence, and as related to a self which is conscious of them. But it is noticed, at least when one begins to psychologize, that all objects of which one is conscious, whether illusory or not, are subjective, at least in the sense of being related to a self which is conscious of them ; and, further, that the psychical processes in the two cases of normal and abnormal perception are mainly the same. It is an easy step, consequently, for unclear think- ing to conclude that these objects are all subjective in the PSYCHOLOGICAL IDEALISM 95 other sense, i.e. that they are dependent upon the conscious- ness of the conscious self for their existence. The fallacy may appear as one of equivocation — the common fallacy of ''four terms" — as in the following syllogism: What is subjective (dependent on self for existence) is not externally real, but mere idea; all objects of which we are aware are subjective (related to a self which is conscious of them) ; therefore, all objects of which we are aware are not externally real, but mere ideas. Or, if the equivocation be avoided, the fallacy will re- main as that of an "undistributed middle term," as in this syllogism : The unreal objectively is subjective (related to a subject) ; similarly, all of which one is conscious is subjective (related to a subject) ; therefore, all of which one is conscious is unreal objectively (mere idea). Or, more simply, psychologi- cal idealism may be said to rest upon a fallacious conversion. From the obvious truth that all elements which depend on consciousness for their existence, such as pains, feelings, desires, etc., are in the subjective relation, i.e. are objects for a subject, it is inferred, by the fallacious process of simple conversion, that all that is in the subjective relation, all that is object for a subject, is dependent upon consciousness and this relation to consciousness for its own existence. This is the fallacy of arguing for idealism from what R. B. Perry has called "the egocentric predicament." ^ We can never be conscious of any object that is not in the relation of object of consciousness to ourselves as subject — this is the "egocentric predicament"; but, as Perry justly urges, this fact proves nothing at all as to whether there are or are not other objects not in conscious relation to ourselves, or to any other subject. D. H. MacGregor has made substantially the same point in his exposure of what he calls "the great fallacy of idealism." He points out that what idealism has proved is that "reality cannot be thought as existing, independently of thought," but that what it believes it has proved is that "reaUty cannot be thought, as existing independently of thought." ^ But the same criticism was made years before by T. H. Green, ^Journal of Philosophy, etc., VII, 1910, pp. 5-14; Present Philosophical Tendencies, pp. 129 ff. 2 Hibbert Journal, IV, 1906, p. 788. 96 THE PROBLEM OF KNOWLEDGE not against the idealistic doctrine, but against this psycho- logically ideahstic argument, as used by John Caird. The prop- osition ''that no object can be conceived as existing except in relation to a thinking subject," must not, he points out, be confused with the proposition " that it cannot exist except in that relation." ^ We are not contending that psychological idealism, can, by such logical criticism as we have urged, be proved to be false; we simply maintain that the arguments by which it was sup- posed to be proved true may be shown, by this logical criti- cism, to be worthless, so that there appears as yet no good reason why a view so artificial and so difficult of adoption in practice should be regarded as true. This applies to psycho- logical idealism, whether in its Berkeleian form, where the self is thought of as a passive recipient, and perceived objects as ''ideas," because mere sense-data; or in the form in which it was presented by Fichte, where the self is thought of as crea- tive, and perceived objects as contents in consciousness, con- structed by mental activity (Berkeley was influenced by Locke's view of mind as a tabula rasa, while Fichte followed Leibniz and Kant in emphasizing the self -activity of thought) ; or, finally, in the intermediate form of a psychological or sub- jective neo-Kantianism, in which the self is thought of as pas- sive with reference to sensations but creative with regard to relations and perceived objects, consequently as being partly datum and partly thought-construct, but in both cases mere dependent content of consciousness. Besides these Berkeleian, Fichtean, and neo-Kantian types of psychological idealism, there is another line of subdivision, which, to a certain extent, runs across the other groups, or at least across the first and the last. This is the division between what we may call the undisguised and the disguised psycho- logical idealists. The former have the courage of their convic- tions ; they acknowledge their subjectivism, emphasizing the constant subjectivity of objects. The latter, the disguised psychological idealists, seek to cover up their subjectivism, even from themselves, by means of a device which proves in the end to be merely verbal. They speak of subject and ^ Works, Vol. Ill, 1888, p. 144. Cf. p. 134 infra. PSYCHOLOGICAL IDEALISM! &7 object as being opposite poles of experience, and of the con- tent of experience as alternating between subjectivity and objectivity; normally it is objective, but under certain condi- tions it may become subjective. Or, as some prefer to put it, originally experience was neutral, neither subjective nor objec- tive, but under certain conditions subjectivity is introduced, and with it, by way of contrast, objectivity. But the experi- ence of which this is an approximate description, it should not be forgotten, is conscious experience, the experience which a self has. The objects of conscious experience are always sub- jective, in the sense of being in the conscious relation to a sub- ject; but under certain conditions we pay attention to this relation, we think of the objects as being in the conscious rela- tion ; that is, we make their subjectivity (relatedness to a sub- ject) an object of thought. But this does not make the original objects of consciousness for the first time subjective ; as objects of consciousness they were as subjective — as much related to a conscious self — when thought of simply as things, as they are now that we are thinking of them as objects-thought-about. It surely will not be maintained that the relation of being object for a subject could not exist except as that relation itself is made the object of conscious attention. But whether of the passive, the active, or the intermediate type, and whether disguised or undisguised, psychological idealism is, we shall contend, in all its forms a malady which the psychologist-philosopher needlessly inflicts upon himself; in all its forms it is an unnatural, unnecessary, and inde- fensible dogma. To further justify this statement we must set forth in some detail and in their systematic context the chief historical and contemporary varieties of this type of idealism. The earlier representatives of psychological idealism are so well known and have been so often discussed in philosophical treatises, that we may pass them with but brief mention. In Berkeley's doctrine objects are combinations of ''sensations or ideas imprinted on the sense" ; they are the things we per- ceive by sense, and as such they can be no more, it is claimed, than our own ideas or sensations, no one of which can exist unperceived; their esse is per dpi; the object and the sensa- 98 THE PROBLEM OP KNOWLEDGE tion are one and the same thing.^ Here, obviously, we have the result of the fallacious process which we have just pointed out. But in Berkeley's system the existence of 'Hhe perceiving active being" which we call mind, spirit, soul, or self, whether of God as the creator of ideas, or of man as their recipient, is assumed. Its esse is not per dpi; it is not any one of our ideas, but "si thing entirely distinct from them, wherein they exist, or, which is the same thing, whereby they are perceived." ^ But essentially the same arguments by which belief in an inde- pendent material reality was supposedly discredited would serve to discredit Berkeley's own belief in a transcendent God and substantial human souls. A more thoroughgoing psycho- logical idealism would say of God and of souls also that their esse is percipi, and Hume did not hesitate to take this further step. Hume adopted and tried to carry out to the bitter end the central thesis of psychological idealism, viz. that .''nothing is ever really present with the mind but its impressions and ideas," the latter being defined as faint images, or impressions derived from impressions; but what he means here by ''mind" is no simple and immaterial substance, but simply the successive impressions and ideas, united by certain relations, especially that of cause and effect.^ We have no idea of substance, either material or mental, he holds, except a collection of ideas united by the imagination and given a particular name.^ The idea of existence or external existence is the very same with the idea of what we conceive to be existent. On the one hand every impression or idea is conceived as existent, and on the other hand every idea of existence is some particular impression or idea.^ Here we have an originally subjective empirical idealism seeking to become self-consistent by applying its doctrine to the subject (as object) as well as to (other) objects, with the result that, verbally, it ceases to be subjective, and becomes at this point what has recently been called immediate or pure empiricism. The system thus points in the direction of a dis- ^ "Treatise on the Principles of Human Knowledge," in Fraser's Selections from Berkeley, pp. 33-6. » Ih., p. 33. 3 A Treatise of Human Nature, pp. 67, 253. 4/6., p. 16. »/6., p. 66. PSYCHOLOGICAL IDEALISM 99 guised psychological idealism; but it cannot be said really to succeed in eliminating subjectivism. It is not with mind as object, but with mind as subject, that all impressions and ideas are present; and yet it is only mind as object that can be reduced, on Hume's principles, to successive impressions and ideas ; all of the impressions and ideas, without exception, are present with the mind, which cannot, therefore, be regarded as one of those impressions. Hume himself admits that his phi- losophy encounters at this point a difficulty that seems in- superable. "All my hopes vanish," he writes, "when I come to explain the principles that unite our successive perceptions in our thought or consciousness. I cannot discover any theory which gives me satisfaction on this head. In short, there are two principles which I cannot render consistent ; nor is it in my power to renounce either of them, viz. that all our distinct per- ceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences.^' ^ In other words, as a would-be radical empiricist he does not know what to do with our evident knowledge of relations to which, apparently, no elements of sense correspond. His theory calls for a dif- ferent sense-impression for every different relation, because relations are to him simply ideas of relations, and ideas simply impressions of sense-impressions. Manifestly, then, if it can be shown that there are some relations or ideas of relations to which no impressions correspond, we have the self -refutation of passive psychological idealism as the effect of the effort to be thoroughgoing and consistent in the application of the theory. If, however, it be maintained, as by William James,^ that there are feelings of all relations of which we have any idea, the question is still to be asked whether such feelings are definite enough to account for the ideas. As a matter of fact our ideas of relations are ordinarily much more definite than our immediate feelings of those relations. Moreover, the easy identification of relations and ideas of relations in the Humian system is explicable only as it is found to rest upon the fallacious reasoning from the egocentric predicament noted above. 1 lb., pp. 635-6. 2 A Pluralistic Universe, p. 280 ; Essays in Radical Empiricism, pp. 41-3. 100 THE PROBLEM OF KNOWLEDGE According to John Stuart Mill, we know and can know no more of material objects than the senses tell us.^ What they tell us directly is simply what our own sensations are. Rela- tions between sensations may be resolved into a difference in our sensations.^ All we can know of objects, directly or in- directly, is the sensations which they actually excite, or which we imagine them exciting in ourselves.^ The conception one forms of the world as it is at any moment comprises, along with the sensations he is feeling, a countless variety of possi- bilities of sensation ; viz. the whole of those which past obser- vation tells him he could, under any supposable circumstances, experience at that moment, together with an indefinite and illimitable multitude of others which it is possible he might experience in circumstances unknown to him.^ Thus matter or external nature is nothing but the permanent possibility of sensation, which, unlike actual sensation, is common to all individuals.^ The belief in such permanent possibiHties con- tains all that is essential in the behef in substance.® Real externality to us of anything other than other minds is incapable of proof. "^ Moreover, of mind itself our knowledge is entirely relative.^ Mind, as we know it, may be regarded as nothing but a series of feelings, together with a permanent possibiHty of feeling.^ At this point, however. Mill has to admit that he experiences a final difficulty : in the case of mind the series of feelings is aware of itself as a series, extending from the past through the present into the future. ^^ This is essentially the same difficulty as Hume encountered. Even if all objects, including the subject in so far as it is object, could be reduced to feelings, there would still be the subject to which these feel- ings are present to be accounted for. Here, again, as in the case of Himae, self -refutation in the end is the penalty of fallacy in the beginning. W. K. Clifford describes the self as a stream of feehngs such that each of them is capable of a faint repetition, and that when two of them have occurred together the repetition of the one 1 An Examination of Sir William Hamilton's Philosophy, 3d ed., pp. 6, 7 ; see especially Chs. II, XI, and XII. 2 lb., p. 7. 3 75.^ p. 8. * lb., p. 222. 5 76., p. 227. 6 lb., p. 229. 7 75.^ p. 232. 8 lb., p. 235. 9 76., pp. 236, 238. "> 76., p. 242. PSYCHOLOGICAL IDEALISM 101 calls up the other, according to certain rules. ^ The object is defined as a set of changes in consciousness, and not anything out of it, whether or not there are things-in-themselves which are not objects. The physical object, whether presented or inferred, is always a part of one's own consciousness ; but the mind of another can never be an object in my consciousness. The inferred other conscious selves are ejects, things thrown out of consciousness, and recognized as not being a part of me.^ Clifford then goes on to develop his view in a way that antici- pates to some extent what we have called disguised psycho- logical idealism. A feeling is not my feeling, he maintains, until on reflection I remember it as my feeling. Thus a feeling can exist by itself, without forming part of a consciousness. Such elementary feelings, or eject-elements, might well be the true things-in-themselves. Moreover, a thoroughgoing parallel- ism of the physical and the mental is inferred from "the doc- trine of evolution," with its principle of an unbroken line of ascent, which is supposed to necessitate the conclusion that, since consciousness has been evolved, "some ejective fact or event which might be a part of consciousness" corresponds to every motion of matter.^ This panpsychism is brought out most clearly in connection with the following considerations. Let us suppose that I see a man, whom we will call A, looking at a candlestick, which I also see. The candlestick is material, but this means simply a group of my sensations, actual and possible. There is an image in A's brain, representing, i.e. corresponding, point for point, to the candlestick, which is external to him. This cerebral image, like the candlestick, is material ; but this again means simply a group of my possible sensations. But there is in A's mind an image, or perception, representing the external reality, and this mental image which A has is, of course, nothing but mind- stuff ; it is to be interpreted, not as my object, but as my eject. But if A's mental image of the candlestick is related (repre- sentatively) to the externally real candlestick which he sees, as A's (material) cerebral image is related to the material candle- 1 "Body and Mind," Humboldt Library of Science, No. 145, p. 16. 2 "On the Nature of Things-in-Themselves," Humboldt Library of Science, No. 145, pp. 28, 29, 31. s /^.^ pp. 33^ 35_6. 102 THE PROBLEM OF KNOWLEDGE stick which I see, then, since this relation in the latter case is fundamentally a relation of identity of stuff (in this case, matter), the conclusion follows that the external reality which A sees must be made up of mind-stuff, just as his mental image is. Both are my ejects. But both of my objects (A's cerebral image and the candlestick which I see), although material, are, as we have seen, nothing but (my) mind-stuff. " The universe, then, consists entirely of mind-stuff. . . . Matter is a mental picture in which mind-stuff is the thing represented." ^ In Clifford's system we have a psychological idealism resting upon the usual incorrect analysis of objects as sensations ; but his doctrine is complicated by the further application of the principle of psychological idealism to the relations of the feel- ings to the self. Consistently enough for the psychological idealist himself, but unwarrantably, since psychological ideal- ism is based upon a fallacy, it is assumed that the relation of feelings to a self which has them can exist only when there is consciousness of this relation. But, as a matter of fact, such feelings are remembered as my past feelings, although when they were actually present I was not explicitly aware of them as mine. Moreover, with reference to the evolutionary argument for panpsychism and parallelism, as was said in our critique of the similar views of C. A. Strong, whose philosophy shows the marks of Clifford's influence, if we admit the possibility of *' creative evolution," the argument loses most of its weight. Clifford's type of psychological idealism is more ingenious than most others, but it is no more demonstrative, no less fallacious and dogmatic, than those previously examined. Karl Pearson holds that ''an external object is in general a construct." He does not use this term in quite the Kantian sense, however; his affiliations are with the psychological idealism of the older English empiricism. He means by " con- struct" "si combination of immediate with past or stored sense- impressions." 2 Although he distinguishes between the ideal and the real, he does not identify the ideal with the unreal. The ideal is that which passes into reality when its perceptual 1 "On the Nature of Things-in-Themselves," Humboldt Library of Science, No. 145, pp. 36-7. 2 The Grammar of Science, 2d ed., 1900, pp. 41, 64, PSYCHOLOGICAL IDEALISM 103 equivalent is found; the unreal can never do so. Physical hypotheses as to the nature of matter are not unreal but ideal, for they do not lie absolutely outside the field of possible sense- impressions. The concepts of the metaphysicians, however, among which he includes the " thing-in-itself " of Kant and the ''mind-stuff" of Clifford, are not ideal, but unreal.^ Even physical science is a classification and analysis of the contents of the mind.^ The thinker is like the clerk in the central tele- phone exchange, who projects outside his office sounds which are really inside the office, and speaks of them as the external universe.^ According to Pearson, we must remain absolutely agnostic as to whether sense-impressions are ^'produced" by un- knowable '' things-in-themselves," or whether behind them there may not be something of their own nature.^ He thus refrains from a dogmatic denial of things-in-themselves ; but all reality of which we can ever know that it exists, he would interpret after the manner of psychological idealism. His own dogmatism lies in his assertion of a subjectivism such as leaves absolutely no possibility of knowledge of any reality which might exist in- dependently of our own subjective impressions and '' constructs." H. R. Marshall advocates the ^'thoroughgoing subjective view . . . according to which the outer world and the objects within it are complex systematized concepts which are within and part of consciousness." ^ He states his "introspective monism" in a way that makes it virtually solipsism. "The 'now' of consciousness," he says, "is all that exists, whether of me or of the universe for me." ^ Consciousness contains the self and its presentations. The presentations are consti- tuted of the ego and its objects, both objects in the outer world and activities in the nervous system.'' The objective view is convenient but inaccurate; only in the subjective view, according to which esse is percipi, have we a true philosophj^ of reality.^ Thus the natural world is simply that part of the mental order which has "out-thereness."® In Marshall's solip- sism we have the logical outcome of psychological idealism; but solipsism has been so universally taken as the redudio ad 1 76., p. 41. 2 lb., p. 52. 3 /&., pp. 61-2. ■» 76., p. 68. 6 Consciousness, 1909, p. 10. «76., p. 2. ^76., p. 6. *Ib., pp. 9-11. 9 See Journal of Philosophy, etc., IX, 1912, p. 106. 104 THE PROBLEM OF KNOWLEDGE ahsurdum of any philosophy of which it is the necessary outcome, that further comment is needless. We shall now turn to a consideration of certain representatives of a psychological idealism in which the subject of conscious- ness is not represented as the passive recipient of things as ''impressions" or ''ideas/' but the active creator of objective reality within the sphere of the conscious life. In this connec- tion we think first of Fichte. In leading up to the distinctive doctrines w^hich he was concerned finally to enunciate, Fichte commits himself to the most subjective type of psychological idealism that can well be imagined. Assuming that what we do not perceive immediately, we do not perceive at all, he goes on to assert that in all perception we perceive only our own con- dition. Strictly speaking, we do not immediately perceive external objects; we immediately perceive onty our own per- ceiving.^ Instead of saying, then, that the object is red, one should say, "I feel myself affected in the manner that I call red." 2 We extend our own sensation through space, and caU an independent reality what is a product of our own thought.^ But neither do we perceive the subject as an independent reahty ; our pure rational activity in its original and unchangeable unity is beyond possible perception, and it would even seem as though intelligence were a mere product of thought.^ But through faith that we can have such knowledge as is necessary for the fulfilling of our moral vocation,^ we may posit, as indeed we also must, both Self and Not-self as valid realities for thought.^ The object is dependent for its being upon consciousness of the object, just as consciousness in turn is dependent upon self -consciousness (consciousness of consciousness).'^ Thus, in Fichte's final philosophy, the Ego, or Intelligence, or pure rational activity, creates in consciousness the external world of experience. It is an imaginative construct of the obstacle which is posited to explain the Ego's feeling of limita- 1 "Die Bestimmung des Menschen," 1800, Fichte's Popular Works, Eng. Tr., 1889, Vol. I, pp. 357-8. 2 76., p. 360 ; cf. pp. 368, etc. 3 /^.^ pp. 368, 399. 4 76., pp. 383-4, 399. & 75.^ p. 411. ^ Various works on Wissenschaftslehre, passim. 7 " Zweite Einleitung in die Wissenschaftslehre," 1797, Werke, Vol. I, pp. 458- 63 ; cf. Kuno Fischer, Geschichte der neuern Philosophie, 3d ed., Vol. VI, p. 308. PSYCHOLOGICAL IDEALISM 105 tion in its activity/ and even to serve as a basis for the activ- ity of the ego. ''The infinite activity of the Power ... is only for the sake of evidencing, in Intuition, the Being of the Will." 2 Thus Fichte's doctrine, although activistic and ab- solutistic, remains to the end a type of subjective psychological idealism. And in essentials this subjectivism is virtually as- sumed at the outset. His analj^sis of experience, as can readily be seen, is infected with the fallacy common to the other forms of psychological idealism. The activistic interpretation does not affect this fallacious basis. Another activistic psychological idealist is Alfred Fouillee. His idealistic epistemological monism is indicated by his in- sistence that psychology has for its object realities, not mere reflections of realities.^ Internal and external phenomena are held to differ only in that the latter are, through activities of sight and touch, spatial,^ and are commonly viewed in abstrac- tion from their relation to the subject.^ The physical is an aspect of experience; it is inseparable in reality from the mental.^ Number, space, and movement are mental, phenom- ena, ideas; and psychology covers the whole field of meta- physics, in so far as it can be covered at all.^ The dualistic opposition of a world of unreal appearances and a world of reality which does not appear, is thus repudiated.^ Phenomena are simply a part of reality, which reality as a whole is a complete (psychological) experience.^ We may not be able to say that the Unknowable does not exist, it is admitted ; but, it is claimed, neither are we entitled to affirm its existence. The problem is merely one which arises when we come to the limit of the ex- perience of the subject. ^^ But, urges Fouillee, mental phenomena — and all phenomena i"Grundlage der gesammelten Wissenschaftslehre," 1794, 2d ed., 1801, Werke, Vol. I, pp. 265-70 ; cf. E. L. Schaub, Philosophical Review, XXII, 1913, pp. 18, etc. 2 "Die Wissenschaftslehre in ihrem allgemeinen Umrisse," 1810, Werke, Vol. II, pp. 706-9 ; cf. W. Wallace, Prolegomena to HegeVs Logic, p. 133. 3 La psychologic des idees-forces, 1893, p. xiii. 4 lb., p. xiv. 5 lb., pp. xv-xvi ; cf. Uavenir de la metaphysique, 1890, p. 285. ^ Uavenir, etc., p. 300. "^ La psychologic, etc., p. xi; cf. Uavenir, etc., p. 302. 8 Uavenir, etc., p. 53. » lb., pp. 53-4, 278, lo lb., pp. 281-3. 106 THE PROBLEM OF KNOWLEDGE are really mental — are not originally representations, but appetitions} Every state of consciousness is idea as including discernment, and force as including preference ; moreover, the faculty of discernment is only developed with a view to choice. ^ All psychical force is therefore ultimately volition,^ and psy- chology is essentially the study of the will. Its problem is, How does the subject act? ^ According to Fouillee, then, physi- cal objects are spatial realities immediately discerned, but dependent upon conscious will for their being what they are. His philosophy is thus, like Fichte's, an activistic psychological idealism. It lacks the Fichtean absolutism, but it retains the same fundamental fallacy and consequently the same dogmatic subjectivism. We now turn to a consideration of some psychological idealists who have been strongly influenced by Kant, and who conse- quently regard the subject as neither passive nor active, ex- clusively, in perception, but both passive and active. We shall first speak of Theodor Lipps, whose ''psychologism" is shown at once by the fact that he would make psychology the funda- mental and indeed the all-inclusive philosophical science. When psychology has fulfilled its task, it has done, he declares, the work of logic, of aesthetics, of ethics, and of the only accessible metaphysics; it deals with validity and with the real as im- mediately experienced.^ Indeed, all presentations are objects for psychological investigation.^ The non-psychological sciences simply show up the law-abiding character of the contents of experience, viewing them in abstraction from their relation to the ego, or conscious life. Objective and subjective are thus simply two aspects of the same process.^ Lipps is at con- siderable pains to interpret the apparently non-empirical as being what it is in and for the experience of the individual. There are substrates — things, self, other selves — in which the more immediate objects of sense perception, as well as those of inner and social experience, inhere, or to which they belong.^ But these are ultimately interpreted, at least in the earlier 1 La psychologie, etc., p. vii. ^ /^^ p_ x. 3 76. * lb., p. xxvi. ^ Psychologische Untersuchungen, II, 1, pp. 1-4, 22-7. ^ lb., p. 15. 7 16., p. 27 ; I, 1, p. 20 ; Leitfaden der Psychologie, 3d ed., 1909, pp. 77, 167. ^Leitfaden, etc., pp. 171-2, 222. PSYCHOLOGICAL IDEALISM 107 thought of Lipps, as mere possibiHties of conscious experience.^ What they are is what they are immediately felt to be, and in this process of immediate perception the process of Einfuhlung — the reading of one's subjective feelings into the object — plays an important part.^ Thinking is regarded as making the object — for the thinker, of course — out of the contents of immediate feeling experience.^ But this constructing activity of thought seems more arbitrary and individual in the system of Lipps than in that of Kant. According to the Kantian doctrine, thought must work, in universally necessary ways, upon a content whose temporal as well as spatial relations have already been estabhshed by ''sensibility." According to Lipps, however, thought can take a present content of a certain sort and make it past by so thinking it ; the past of which we think, it is contended, is a part of present experience.^ More- over, while for Kant the line between appearance and reality is never an arbitrary one, but always definitely fixed, whether what one is thinking of is the distinction between phenomenon and thing-in-itself, or that between what is not and what is conformable to the principles of scientific order, according to the philosophy of Lipps, when an earlier content is corrected by a later experience or Einfuhlung, it then becomes, for the first time, mere appearance ; it may persist in being, after it has been corrected, although, of course, as corrected.^ What this means is that even what on logical grounds must be rejected as unreal must be accepted, in many instances, on psychological grounds as real; and there is no way of overcoming the contradiction, because no place has been left for any metaphysics but psy- chology. This final contradiction is the penalty of the initial fallacy to which we have had occasion so frequently to refer. Hans Vaihinger has been deeply influenced by Kant, but he develops his philosophy along the lines of what we have called psychological idealism. He calls his doctrine "idealistic posi- tivism." Reality, according to this thinker, is the immediate^ given content of experience ; but over against it are to be set 1 Leitfaden, 1st ed., 1903, pp. 337-8. ^Leitfaden, 3d ed., pp. 222, 227-38; cf. Psy. Untersuchungen, II, Parts 2 and 3. ^ Psy. Untersuchungen, II, 1, pp. 13, 14; cf. Leitfaden, 3d ed., p. 225. * Psy. Untersuchungen, I, 1, pp. 43, 47. ^ Leitfaden, p. 236. 108 THE PROBLEM OF KNOWLEDGE on the one hand hypotheses, which are mental constructs repre- senting a past or a possible future content of experience, and capable, therefore, of verification and refutation; and on the other hand fictions and half-fictions, which are also mental constructs which are either highly convenient or even indis- pensable aids to thought and life, in spite of the fact that the half-fictions contradict reality (experience), while the fictions are not only contradictory of reality, but self-contradictory as well.^ Vaihinger's psychologism here makes it necessary for him also, like Lipps, to give the He to logic. The only difference is that whereas Lipps chose to maintain that what logic has excluded as not possibly real is nevertheless real, Vaihinger has chosen to defend the doctrine that even thoughts which are scientifically as well as practically indispensable, may be mere empty concepts, to which no reality corresponds. As a matter of fact both ways of defying logic are involved, logically, in the original fallacious adoption of the point of view of psychological idealism. J. H. Poincare's discussions of scientific method are worked out on the basis of a neo-Kantian psychological idealism, quite similar, fundamentally, to that of Vaihinger, or even of Lipps. Like them too he is forced in consequence to confront the prob- lem of satisfying the logical demands of scientific thought without departure from the principle that reality is to be found within the limits of psychologically describable experience. His way of dealing with this difficulty is not to discount the thought-constructs in favor of the given, as Lipps and Vaihinger both do, each in a way of his own ; rather does he discount the immediately given in favor of the constructed. External objects, he says, in Kantian fashion, are groups of sensations, cemented by a constant bond, a relation, which is the object itself. These relations, he claims, are all we know of the object ; unlike sensations they are transmissible entities, constituted by thought. ''All that is not thought is pure nothingness ; since we can think only thought . . . to say there is something other than thought is therefore an affirmation which can have no mean- ing." 2 Like Vaihinger, whose work, although written earlier, 1 Die Philosophie des Als Ob, 1911, pp. xiv-xvi, 21, 143-54, et passim. 2 The Value of Science, Eng. Tr., 1907, pp. 138, 142. PSYCHOLOGICAL IDEALISM 109 was later in making its appearance, Poincare distinguishes not only between reality, or fact, and hypothesis, but also between hypotheses and other mental constructs, which he regards, not as indispensable fictions, but, more conservatively, as symbols which are convenient, although not necessarily true. For ex- ample, he contends that the Euclidean geometry is no truer than any other ; it is only more convenient} That Poincare did not finally solve the problem is indicated in at least two ways. In the first place, this doctrine that two or more mutually contra- dictory systems can be equally true is more probably a reductio ad absurdum of something in the premises, than the paradoxical profundity its author evidently takes it to be. But in addition to this, we have to note the apparent movement, in Poincar6's later thought, in the realistic direction.^ This movement, if maintained, would eventually have undermined some of his most characteristic doctrines ; but the mere tendency is signifi- cant as marking the felt inadequacy of psychological idealism for philosophical construction, even in the hands of so ingenious a thinker as Poincar^. We shall now examine some typical instances of what we have called disguised psychological idealism. Speaking generally, it may be said that psychological idealism becomes disguised when its doctrine, that objects depend for their existence upon their being experienced as objects, is applied to the subject as one of the objects. Then, prior to self-consciousness, there is no self; experience prior to self-consciousness is ''pure" or "neutral" experience, upon their relation to which both subjects and objects, both selves and things, depend for their existence. Now this homeopathic treatment of subjective or psychological idealism leaves it the same thing in disguise. Moreover, this disguised psychologism, as we shall see, is a halfway house on the way to the new realism. It is itself a transitional form of philosophy, a position of unstable equilibrium. If contents, as it claims, are independent of any relation to a conscious subject, it seems the natural conclusion to infer that they 1 76., p. 121 ; Science and Hypothesis, Eng. Tr., 1905, pp. 38-9. 2 Le materialisme actuel, by Poincar6, Bergson, et al., 1913 ; Demises pensees, 1913, Ch. VI ; cf. Journal of Philosophy, Vol. IX, 1912, p. 308, and H. C. Brown, "The Work of Henri Poincar6," Journal of Philosophy, etc., XI, 1914, pp. 231-2. 110 THE PROBLEM OF KNOWLEDGE are real independently of their being experienced. This view, however, in which the psychological idealism is no longer ap- plied to the object, is the position of the new realism, which thus, by virtue of the thoroughgoing epistemological monism of the two forms of psychological idealism (the undisguised and the disguised) from which it has descended, begins as a thorough- going realistic epistemological monism. It starts all over again from the very beginning ; its doctrine coincides with that tacitly assumed by the most uncritical naive realist. But two wrongs do not make right. If we were justified in regarding the original, undisguised psychological idealism as founded on fallacy, then neither pure empiricism nor the new realism can be regarded as established simply because the effects of the first fallacious process have been covered up by a second similarly fallacious step. The only sure way of escaping the evils of psy- chologism is to retrace one's steps. To attempt to press on through it to some more satisfactory ground is only to render the final inevitable retreat all the more difficult. Ernst Mach's views may be regarded as transitional between an undisguised and a disguised psychologism. His works have attracted a good deal of attention as showing the results of the attempt of a physicist to express physical facts and theories in the terms of psychological idealism. Of the history of his thought he tells us that at the age of fifteen he was deeply impressed by Kant's Prolegomena, and that two or three years later the superfluous r61e of the thing-in-itself dawned upon him. Then his ego suddenly appeared to him as ''one coherent mass of sensations." He says, ''I had to struggle long and hard before I was able to retain the new concepts in my specialty (phj^sics). . . . Only by alternate studies in physics and in the physiology of the senses . . . have I attained to any con- siderable firmness in my views." ^ Thus he came to view bodies as complexes of sensations, the abiding existences which they seem to have being really nothing but thought-symbols for these complexes of sensations. ^ Molecules and atoms are regarded, not as realities behind phenomena, but as mere means for facilitating our dealing with the facts of the senses.^ So far, 1 Analysis of Sensations, Eng. Tr., p. 23. 2/6., p. 22. 3/6., pp. 154, 207. PSYCHOLOGICAL IDEALISM 111 we are on the ground of undisguised psychologism. The follow- ing doctrine, however, points to the neutral empiricism which is psychologism in disguise. There is no objective distinction, he says, between the real and the experienced. "In the sensory sphere everything is at once both physical and psychical.'' "The apparent opposition . . . lies only in the way of con- sidering." ^ What we have here is thus a system of thought founded upon the same old fallacy of reasoning from the ego- centric-predicament, but shown to be, in spite of the partial disguise of its subjective idealism, exceedingly difficult to apply in the interpretation of physical facts. In the "empiriocriticism" or " philosophy of pure experience" of Richard Avenarius^ we have one of the earliest and best illus- trations of psychological idealism in disguise. The initial assumption is that nothing exists save experience. An appear- ance of realism is given to the system by the further assumption that the fundamental characteristic of the content of ex- perience is space. But the novel result of combining these two assumptions might be more appropriately called materialis- tic idealism or idealistic materialism than realism. The main reliance for the defence of the system is placed in the exposure of the "fallacy of introjection" — a falsification of natural ex- perience, issuing in the common dualism of the physical and the psychical. The process of introjection, as Avenarius de- scribes it, is as follows : Since we see that the real objects which another observer sees — or thinks he sees — lie outside of his body, assuming that what he really sees — his perceptions — must lie within him, rather than outside of him, we conclude that he perceives, at best, the subjective counterparts of objects, not the real external objects themselves. But by analogy we must conclude the same thing about our own perceptions. Hence dualism, or, as an alternative, subjective idealism, arises. Avenarius tries to render consistent what he regards as the original natural view, by interpreting the distinction between things and thoughts as a distinction due to the one being a 1 lb., p. 195; 2d German ed. (Analyse der Empfindungen) , 1900, p. 19. 2 Kritik der reinen Erfahrung, 1888-90 ; Die menschliche Weltbegriff, 1891 ; cf. N. K. Smith, "Avenarius's Philosophy of Pure Experience," Mind^ N.S., XV, 1906, pp. 13-31, 149-60. 112 THE PROBLEM OF KNOWLEDGE first and the other a second experience ; and, further, by reducing the distinction between the relative and absolute points of view to a distinction between two kinds of the relative point of view. But it is easily seen that his whole philosophy rests upon the fallacious inference that, since it is only through experience that we can know that anything exists, therefore ''nothing exists save experience." Avenarius is undoubtedly justified in taking exception to subjectivism and dualism, and to the process of introjection, as he describes it ; but it remains to be seen whether the reality of the subject and the distinction between the psychical and the physical may not be maintained without falling into any of the errors against which he rightly enough protests. J. Petzoldt, acknowledging the influence of Mach and Avena- rius, expresses his own view as follows: '' There is no world in itself, but only a world for us. Its elements are not atoms or any other absolute existences, but ' sensations ' of color, sound, touch, space, time, etc. Still, things are not purely subjective, mere appearances in consciousness. On the contrary we must think of the constituent parts of our environment, which are made up of these elements, as continuing to exist, just as they were during perception, even when we no longer perceive tl;iem." ^ This retention of a psychological relativism in spite of the ex- plicit repudiation of psychological idealism can be understood only as an expression of the " philosophy of pure experience,'' which, as we have seen, is, notwithstanding all protests, nothing but psychological idealism disguised and masquerading in the clothes of natural realism.^ The disguised psychological idealism of Wilhelm Wundt, the intermediate position of which between idealism and reahsm is recognized in the designation ''ideal-realism," has not a Httle in common with the doctrines of Avenarius. Wundt regards the philosophj^ of Avenarius as the only consistent materialism, but he himself would avoid that conclusion by in- sisting, in Kantian fashion, upon the thought-activity of the ego. He differs conspicuously from Kant, however, in holding that all the categories have had an empirical origin. ^Das Wdtjyroblem vom Standpunkte des rdativistischen Positivismus aus, 2d ed., 1912, pp. V, etc. « See Ch. X, infra. PSYCHOLOGICAL IDEALISM 113 A first examination of Wundt's philosophical system may lead to the impression that his planned "ideal-realism" has been successfully brought to realization. It may seem that without abandoning the fundamental positions of empirical idealism he has included in his philosophy the truth of realism. He undertakes to retain the view that in experience the object is given immediately as a real thing; this reality, he claims, remains a part of knowledge, subject to no correction. For naive thought, however, according to Wundt, the given is not something which is at once subjective presentation and also object ; it is only an object with such and such characteristics. But, because of contradictions between different perceptions of the same object, one is forced to take the qualitative content of sensation back into the subject; and yet, Wundt insists, this is necessary only for the particular case in which it occurs ; it is generalized only by an arbitrary act of thought. Still, knowing is thus separated from the object, thinking is recognized as subjective activity, and every given object is seen to be given in the subject.^ Thus while he agrees with Avenarius in his view of the original natural experience, Wundt differs from him not only in his description of the process which accounts for the consciousness of self, but also in regarding that process as valid, at least to the extent of its arriving at knowledge of the ego, or cognitively active subject, really involved in all experi- ence of the world. We must maintain, then, that this apparent realism is simply an original psychological idealism, such as we might expect would appeal to a structural psychologist like Wundt, but elaborately accommodated to the point of view of the non- psychological empirical sciences — also quite as might be ex- pected of the experimental psychologist. ''The original unity of thinking and knowing" is regarded as at the same time "a unity of thinking and being." ''Our presentations," he goes on to say, "are originally the objects themselves." ^ This must not be taken as describing reality apart from its being experienced, but the content of experience (treated here as if there were no other way in which anything could exist) apart 1 System der Philosophie, 3d ed., 1907, Vol. I, pp. 78, 128-9, etc. « lb., pp. 78-9. I 114 THE PROBLEM OF KNOWLEDGE from any reflective consciousness. The fallacious dogma of psychologism is tacitly assumed, and the supposed escape from subjectivism is only formal and apparent. The view that the object depends for its existence upon reflective thought is avoided ; but no place is found for the reality of things which are not dependent for their existence upon their being given in im- mediate experience, as well as not being mere products of re- flective thought. Finally, then, while predisposed to identify the content of a psychologically describable experience with objective reahty, Wundt is compelled, nevertheless, out of deference to the physical sciences, to admit a real transcendence, resting upon the unending character of the progress of thought. Even the idea of "possible human experience" proves thus inadequate as an ultimate category of reality.^ Wundt's epistemological problem, formulated as the problem of pre- serving objective reality in spite of the subjective point of view introduced through the consciousness of illusion and of those non-objective elements of experience which have led to the consciousness of consciousness as such, and so to the develop- ment of the science of psychology, ^ must be regarded as left by him still awaiting a solution. The other more or less typical representatives of disguised psychological idealism whom we shall mention are EngHsh or American philosophers, who are also significant in other connec- tions. Of one of these, G. S. Fullerton, it need only be said at present that his System of Metaphysics, published in 1904, expresses a point of view intermediate between his original Berkeleian psychological idealism and his present realistic position. It consequently coincides at certain points with the covert or disguised psychologism with which we are here con- cerned ; but inasmuch as it falls into a certain peculiar abstrac- tionism, it will be more profitably discussed as representing a variety of abstract idealism. We shall turn, therefore, to a brief examination of the views of certain other philosophers, being concerned chiefly with S. H. Hodgson, William James, and John Dewey, all of whom, like Fullerton, are also significant 1 System der Philosophie, 3d ed., 1907, Vol. I, pp. 179, 188. 2 lb., pp. 82, 88, 91, 135 ; cf. Kuelpe, Philosophy of the Present in Germany, Eng. Tr., p. 200. PSYCHOLOGICAL IDEALISM 115 in connection with the transition from psychological idealism to the new realism. Shad worth H. Hodgson would have metaphysics based upon a subjective analysis of experience, without presuppositions.^ Experience, he insists, cannot be transcended; we cannot think of matter as a real condition without first thinking of it as a percept. 2 He rejects metaphysical idealism and all other forms of trans-empirical metaphysics,^ and even regards the idea of the thing-in-itself as not objectively valid, but simply 'Hhe name for an unrealizable attempt at thinking." ^ Even ''the bare idea of Being or Existence, as the percipi of a content of consciousness, is man's idea; that there is a universe at all is a thought of ours" ; and the perception of this truth should prevent us, he thinks, from attempting to frame a speculative theory of the universe.^ "There is no consciousness which does not reveal Being, and no Being which is not revealed in consciousness." Even ''unrevealed Being" falls under the general notion of consciousness. ^ These quotations suggest an undisguised psychological idealism; but when it is re- membered that Hodgson regards the subject as an objectifica- tion of an abstraction, viz. what is left of present experience when we abstract from all present perception of past perceptions (objects),^ subject and object are seen to be special develop- ments within, and on the basis of, pure experience. But even with the aid of this disguise, Hodgson is not, as we have else- where intimated, able to realize his ideal of an interpretation of aU reality in terms of immediate experience.^ It is con- fessed that matter has real conditions beyond all immediate human experience, so that in the end there appears the spectre of the unknowable thing-in-itself, in spite of the special pains taken to drive it away.^ 1 The Metaphysic of Experience, 1898, Vol. I, p. 18 ; Vol. IV, p. 368. 2 76., Vol. IV, pp. 263, 275. 3 75.^ pp. 37I-8I, etc. *" Method in Philosophy," Proceedings of the Aristotelian Society, 1903-04, p. 11. 5 lb., p. 11. 8 The Metaphysic of Experience, Vol. I, p. 6. ' Proceedings of the Aristotelian Society, 1903-4, p. 60 ; The Metaphysic of Experience, Vol. I, pp. 4, etc. * See Ch. II, supra. 9 Proceedings of the Aristotelian Society, Vol. II, No. 1, Part I, 1891-2, p. 24; No. 2, Part II, 1892-3, pp. 16, etc. 116 THE PROBLEM OF KNOWLEDGE William James says that his ''radical empiricism," according to which nothing is to be admitted as a fact except what can be experienced at some definite time by some experient/ has more affinities with natural realism than with the idealism of the English school (Berkeley and Mill).^ The way in which this comes about is that in his ''philosophj^ of pure experience" all relations are reduced to experienced contents ; ^ and so even the relation to the subject or consciousness is also something objective. "The peculiarity of our experiences, that they not only are, but are known, which their 'conscious' quality is invoked to explain, is better explained by their relations — these relations themselves being experiences — to one another." ^ The separation of pure experience into consciousness and con- tent is really adding to a portion of experience in one context the same portion of experience in another context.^ Thus con- sciousness cannot properly be said to exist as a different sort of stuff, or quality of being, from material reahty; it is rather to be viewed as a special (cognitive) function of certain ex- periential (real) elements, or "a series of experiences run to- gether by certain definite transitions," or "a kind of external relation" between experiential (real) terms.® But it seems difficult to harmonize the statement that all reality must be experienced by some experient, whether by one's self in the present or future, or by our neighbor, or by itself,^ with this doctrine that pure experience is prior to the distinc- tion between the mental and the phj^sical. Moreover, James says that his view does not preclude the possibility of things beyond experience,^ and he distinguishes physical things from facts of consciousness by saying that, while the latter exist only once, the former are "supposed to be permanent";^ but, on the other hand, he not only expresses the opinion that "we should be wise not to consider anything of that (extra-experien- tial) nature, and to restrict our universe of discourse to what is 1 Essays in Radical Empiricism, p. 160. 2 75.^ p, 75, 3 76., p. 185, etc. ; The Meaning of Truth, Preface, pp. xii-xiii; cf. The Will to Believe, p. 278 ; A Pluralistic Universe, p. 280. * Essays in Radical Empiricism, p. 25 ; cf. pp. 1-38. 6 lb., p. 9 ; cf. p. 75. 6 lb., pp. 3, 80, 125. 7 76., pp. 88, 160. 8 76., p. 250 ; cf. The Meaning of Truth, p. xii. ' Essays in Radical Empiricism, p. 127. PSYCHOLOGICAL IDEALISM 117 experienced, or" — note the convenient ambiguity — "at least, experienceahle," ^ but even goes so far as to repudiate the idea of "sl transphenomenal principle of energy." ^ Such evidently conflicting statements argue a wavering between differ- ent points of view, if not downright confusion ; and the explana- tion of the acrobatic movements of James's thought undoubtedly is that the disguised psychological idealism, or philosophy of pure experience, upon which he endeavors to maintain his balance, is so unstable a position that he is unable to keep from tipping now towards an undisguised psychological idealism, and again in the direction of natural, or even scientific, realism. According to Dewey's "immediate empiricism," things are what they are experienced as ; ^ but he is careful to explain that this does not mean that they are nothing but what they are known as ; in his view knowing is always mediate, i.e. it is by means of ideas, which are instruments for the reconstruction of the experienced environment. Perceptions are selected elements of experience ; perception is constituted by the func- tional transformation of the experienced environment under conditions of uncertain action (and so of subjectivity, con- sciousness) into conditions for determining an appropriate organic response (i.e. into conditions of objective experience, or reality, again) by means of the judgment, or knowing process, the reconstructive act of cognitive consciousness. Thus know- ing makes a change in things, and the changed reality is what it is experienced as, after the knowing has been accomplished.'^ In other words, the environment is pre-perceptual experience (or, what is taken to be the same thing, its contents). When the conditions for favorable organic response do not obtain, experience is thrown into subjectivity; it becomes conscious; ideas are constructed and employed in tentative judgments. When the practically satisfactory idea is found, the judgment in which it is predicated is an act of knowledge, reconstructing certain elements of experience (or the environment) into an 1 Quoted by J. Dewey, New York Times, June 9, 1912. 2 Essays in Radical Empiricism, pp. 184-5, note 2. 3 Influence of Darwin on Philosophy, 1910, pp. 226 ff, ^Journal of Philosophy, VI, 1909, p. 19; VIII, 1911, pp. 396-7; IX, 1912, p. 659; "Does Reality Possess Practical Character?" in Essays . . . in Honor of William James, 1908, pp. 51-80; Studies in Logical Theory, 1903, pp. 23-85. 118 THE PROBLEM OF KNOWLEDGE object, or a reality, a perceptual experience or perception, selected, in the manner thus described, from the formerly pre- perceptual experience, or environment. The fact that the object is said to be constituted only in small part by its being known, seems to differentiate Dewey's doctrine from idealism ; but that this is really only a disguise becomes evident when it is noticed that for the common realistic distinctions of reality, experience and judging, Dewey has substituted the idealistic terms, experience, consciousness, and knowing. In other words, no cognitive consciousness is recognized except that in which the judgment is present as an explicit act of predication; no conscious experience is recognized except the experience in which contents of experience are explicitly subjective, as my sensations, my feelings, and my ideas ; and no environment is recognized except what is immediately experienced, as if the past and the absent could have no reality but what they have as immediately experienced. Further exposition and criticism of Dewey's view may be deferred until we come to discuss the antecedents of the new realism ; ^ but from what has been said it ought to be clear that his system is properly classified as disguised psychological idealism. G. H. Mead, following Dewey, undertakes to define the psychical as a phase of experience. The objective is that content of experience with reference to which we can act ; the subjective is that with reference to which we cannot, or may not, or should not act. Moreover, it is that which is identified with the consciousness of the individual, as individual.^ In the unreflective stage the entire content of consciousness is subjective and objective at once.^ The psychical element is unessential, because purely individual.^ Not all reality is psychical, inasmuch as it would be a mistake to introject, as purely individual, a content with reference to which one was ready to act.^ Here we have a disguised psychological idealism, evidently developed in the usual way, by applying psychological idealism to the subject as object, but in the specific way ex- 1 Ch. X, infra. 2 "The Definition of the Psychical," University of Chicago Decennial Publi- cations, Vol. Ill, p. 3. ' /6., p. 20, * /6., p. 21. 6 /&., p. 28. PSYCHOLOGICAL IDEALISM 119 emplified by Dewey, viz. by taking the term "conscious" as applicable only to such experience as is explicitly self-conscious. With this is combined the tendency to interpret the objective as that which has place in social, as opposed to individual, or conscious, experience. A. W. Moore emphasizes this last point also. In repelling the charge of solipsism and subjectivism he appeals not to an independent physical world, but to a social situation, the individual consciousness being interpreted as an organic func- tion of the social world. ^ He thus gives further basis for the charge that his pragmatist doctrine of matter is in accord with a not very well-disguised psychological idealism. For individual solipsism he substitutes a social solipsism. H. H. Bawden, another disciple of Dewey, has set forth in his Principles of Pragmatism the "experience philosophy,'^ without making some of the distinctions recently emphasized by Dewey himself, and without the emphasis placed by Mead and Moore upon social tests. He uses the term " experience " as meaning the totality of things for a person's consciousness, the universe from an individual point of view.^ "There can be no sense," he declares, "in speaking of reality beyond or outside of experience, since this very judgment of transcendence or externality itself constitutes the relation which it sustains to experience." This remark, which is, in effect, that to judge a thing to be beyond experience is to bring it within experience, is a particularly fine instance of the fallacy of reasoning from the "egocentric predicament." "Reality," it is concluded, "is what is experienced — whether actually or ideally, whether as fact or as possibility."^ "To-be and to-be-experienced come to the same thing. Things are what they are experienced as.^^ 4 ''There is but one reality : the content of experience." ^ The only thing that saves this view from explicit solipsism is the interpretation of consciousness in such a way as to dis- guise — albeit but slightly — the psychological ideahsm of the fundamental position. Consciousness is described as "a cer- tain kind of adjustment which takes place between two portions 1 Pragmatism and Its Critics, 1910, pp. 220-1. 2 y/ie Principles of Pragmatism, 1910, p. 52. 3 lb., p. 53. * lb., p. 55. e 75.^ p. 55. 120 THE PROBLEM OF KNOWLEDGE of the universe^ ^ ;^ it is "the growing point of experience." ^ "Sensation, and image are merely functional phases of that intellectual reconstruction of experience which we call knowl- edge." ^ "Knowledge is not a process of representing or refer- ring to a reality beyond the act of knowledge ; it is a process going on within the object. . . . Knowledge is the totality of the object or situation undergoing reconstruction." ^ In this view, which Bawden calls pragmatic or functional ideahsm,^ there can be objectivity "only in a functional sense." ^ That part of my experience alone is objective which is brought clearly to consciousness in knowledge, and which serves as an instru- ment to control another part.^ "The doctrine of an indepen- dent and external reality must be given up." ^ It will not do for pragmatists to complain that charges of sohpsism against this type of pragmatic idealism are altogether unfair. Mystical-Psychological Idealism We shall now consider a second of the dual combinations of elemental types of ideahsm, viz. mystical-psychological ideahsm, of which the philosophy of Henri Bergson will afford us our best available illustration. We do not say that, in its final form, Bergson's doctrine is an unambiguous instance of idealism ; but what we do insist is that his final position has come to be what it is only through the use of certain ideahstic presuppositions and suggestions. His philosophical method is a psychologi- cally oriented empiricism, pushed to the mystical extreme. He would find reality in experience in its most radical immediacy. An immediate vision of reality — this, which the mystics claimed, Bergson would make the only true method for the metaphysician.^ Negatively, the method may be regarded as the resolute elimination from philosophy of all traces of logical idealism. All conceptual construction must be tran- scended if one would grasp, in immediate intuition, the ultimate nature of reality. Thought does not reveal the absolute ; but ^ The Principles of Pragmatism, 1910, p. 96 ; italics mine. 2 lb., p. 104. 3 76.^ p. 163. 4 Ih., p. 165. ^ 75.^ p. 261. « lb., p. 255. 7 7^,.^ p. 257. » 76., p. 255. 5 See Introduction A la metaphysique, in Revue de metaphysique et de morale, January, 1903, and Eng. Tr., 1912, passim. PSYCHOLOGICAL IDEALISM 121 rather falsifies and hides it. Intuition, on the contrary, is ^' that art of intellectual sympathy" which transcends concepts, and by which ''one transports oneself into the interior of an object in order to become harmonious with what is peculiar to it alone, and so, inexpressible." Indeed the intuitive penetration of the object is described as inserting one's self into the object's "states of mind" (etats dfame), the being identified (coincider) ^ for the time being, with the other.^ In so far as it succeeds it is, in a sense, as the mystic has always claimed, "superhuman." ^ Before proceeding farther it may be pointed out that Bergson already, in his explanation of the nature of his method, betrays the fact that he tacitly assumes, evidently in the usual fallacious way, the idealistic interpretation of things, which one commonly finds among the mystics, and which is the essential feature of psychologism. In immediate experience of anything one does not necessarily enter into it, so as to become part of it (as mysti- cism assumes), or so as to have it become part of one's own consciousness (as psychologism would have it) ; but one or the other, at least, is involved in Bergson' s descriptions of intuition. Just how mystical idealism and psychological idealism are separately fallacious has already been shown ; and there is no reason to suppose that two fallacious suggestions, when reen- forcing each other, are able to render each other innocuous and logically sound. The traces of this fallacious idealistic assumption are dis- coverable as a confusing factor throughout the various works of this remarkable philosopher. What is revealed most obvi- ously by intuitive apprehension is the fact of duration.^ But this is interpreted as an actual persistence of the past in the present.^ This tendency to regard the persisting memory of the past as the actual presence of the past can be understood only if one remembers that in psychological ideaUsm it is ^ lb., translation by Luce, pp. 3-6, 10; see also pp. 66, 81-2, 86-7; cf. translation by Hulme, pp. 1-3, 7 ; also pp. 55-6, 69, 74. 2/6. (Luce), p. 90; (Hulme), p. 77. ' Essai sur les donnees immediates de la conscience, 1889, Eng. Tr., Time and Free Will, 1910, Ch. II. * lb., Eng. Tr., pp. 100, 101, 107, 110, etc. ; cf. La perception du changement, 1911, p. 30; cf. W. E. Hocking, "The Significance of Bergson," Yale Review, N.S., III, 1914, p. 313. 122 THE PROBLEM OF KNOWLEDGE assumed that the "immediate data of consciousness" are at once parts of the consciousness and the only possible species of reality. The common mystical misinterpretation of the lapse, in the mystical state, of ordinary consciousness of time, as being an actual transcendence of temporal succession, may also have had some influence here. It is a particular instance of the idealistic assumption that to lapse from being object of consciousness is to lapse from being real. But it is especially with Bergson's idealistic interpretation of matter that we are here concerned. In order to learn the real nature of matter, we must eliminate entirely the apperceptive or memory element in perception, leaving only "pure perception" as an intuition of present reality, or matter.^ It is forthwith assumed that this pure perception is matter, which is simply more of the same.^ The psychological idealism in this is un- mistakable. Even the psychological term "images" is used to describe the nature of matter as thus intuitively perceived.^ But, it may be objected, Bergson himself denies the idealism, claiming that matter, or pure perception, this aggregate of images, is more than the idealist calls a representation as well as less than the reaUst calls a thing.^ But this is because Bergson's psychological ideahsm is of the "disguised" variety. It is significant that he claims to find nothing in his own works incompatible with the radical empiricism of James.^ In his view it is memory alone which lends to perception its sub- jectivity ; ^ so that when, as in pure perception, one transcends memory, the resulting consciousness would be no longer sub- jective. Matter, the content of pure perception, is not, to be sure, a construct of intellect in Bergson's system; there is nothing of logical idealism in the Bergsonian philosophy of ultimate reality. Nevertheless matter, as immediately known and identified with that content of pure experience in which the subject is apparently lost (as in mystical absorption), is necessarily held to be a "kind of consciousness."^ It is "a totality of images," ^ "an uninterrupted series of instantaneous ^ Matihre et memoire, 1896; Eng. Tr., 1911, pp. 26, 64, 68, 77, 80, 84-5. 2 76., p. 78. 3 jjj^ Preface, pp. vii, viii. *Ib.; cf. Introduction, etc. (Luce), p. 33; (Hulme), pp. 27-8. 5 Journal of Philosophy, VII, 1910, p. 388. • Matter and Memory, p. 80. ' lb., p. 313. * lb. PSYCHOLOGICAL IDEALISM 123 visions," which visions are *'a part of things rather than of ourselves." ^ Bergson has exphcitly repudiated the idea that Hfe transcends experience, or that absolute reality is beyond the most searching experience. ''Life," he says, "transcends intelligence, but not experience; and it apprehends itself ab- solutely in an intuition which, though actually incomplete, can go on completing itself indefinitely." ^ This double doctrine of consciousness, as identified first with subjective consciousness and then with the immediate data of non-subjective experience, is symptomatic of disguised psychological idealism, and accounts at the same time for the elusive character of Bergson's funda- mental metaphysical intuition. In fact, there are in Bergson's doctrine at least ^2;e clearly dis- tinguishable applications of the term '' consciousness," some of which do not seem to be altogether compatible with each other. To begin at the upper limit, there is that presumably "super- human" consciousness, fleeting glimpses of which are not altogether unattainable by man, viz. intuition.^ Then there is the characteristically human form of consciousness, intelligence. This was originally developed in connection with the process of adjusting the developing life to its material environment, and bears conspicuous marks of its early history. It found in the spatial form a convenient symbol of the material reality about it, and so constructed the world of spatial objects out of the immediate data of consciousness. In this it was well within its rights ; but it is incapable of dealing satisfactorily with life, which it inevitably tends to interpret mechanically, i.e. spa- tially.^ Here, it will be noted, we find an idealistic and approxi- mately Kantian interpretation of the physical object as it is for intelligence. But the most characteristic form of animal consciousness is instinct, which reaches its highest development in the Arthropods, and which is not to be viewed as unconscious because it is not, in the ordinary sense of the term, intelligent.^ But Bergson speaks of consciousness as coterminous with life, 1 76., p. 69. 2 Journal of Philosophy, VII, 1910, p. 388. ^Introduction, etc., passim; Uevolution creatrice, Eng. Tr., 1911, pp. 360-1, etc. * Introduction, passim; Creative Evolution, pp. 135-65, 186-90, 202, 206, 208, et passim; Hocking, Yale Review, N.S., III, p. 315. ^Creative Evolution, pp. 135-51, 165-76, etc. 124 THE PROBLEM OF KNOWLEDGE and so opposes it to inert matter. In fact, life is defined as consciousness using matter for its purposes. Consequently it is found necessary to explain that in the vegetable world life or consciousness is in a state of torpor ; it has become drowsy, as it were, having renounced movement, and devoted itself to the preparation of the explosive substances employed by animals to enable them to utilize matter in their movements.^ But even matter itself, as we have seen, is interpreted as ^'a kind of consciousness." What Bergson seems to mean here is that even inert matter is consciousness with but a minimum of duration or memory, which can only be known by a supreme effort of intuition, whereby the knower ''enters into" the object, so as to share its being or its being consciously perceived. It is not that, in all forms of consciousness, the reality of which there is awareness is dependent on that consciousness, but that, ultimately, it would seem, reality as life can only be interpreted as consciousness; it is inwardly felt duration. Inert matter, then, is the same thing as life, only its movement is in the opposite direction; it is life, or consciousness, ''unmaking itself." 2 Perhaps what Bergson means is that inert matter is life with but a minimum of Velan vital, so that it acts as a drag upon the central life and movement, and even seems to be moving in the opposite direction. In any case, in this mul- tiple signification of the term "consciousness," especially in its application to vegetable torpor and even to inert matter (not to dwell upon the more doubtful case of instinct), we have evidence of Bergson' s determination to abide by the consequences of his original tacit assumption of psychological idealism, an assumption that was none the less fallacious for its being disguised and propped up by means of certain concordant but equally fallacious mystical suggestions.^ ^Creative Evolution, pp. 109, 111, 113-14, 128-35, 181; "Life and Con- sciousness," Hibbert Journal, X, 1911-12, pp. 24-44. 2 Creative Evolution, pp. 245-51 ; Hibbert Journal, X, p. 37. 3 If this interpretation should have to be given up, the only plausible alterna- tive left would seem to be that in the interval between the publication of Matiere et memoire and the writing of devolution creatrice Bergson's thought suffered fundamental modification — and this, it is understood, he is himself unwilling to acknowledge — so that, while the view of formed nature remains quasi-idealistic, the marks of idealism, so far as concerns the doctrine of inert matter, would have to be said to have at length disappeared. PSYCHOLOGICAL IDEALISM 125 Still another indication of the underlying psychological idealism is found in the doctrine which Bergson has recently stated as follows : ''There are changes, but there are no things which change ; the change has no need of a support. There are movements, but there are not necessarily invariable objects which move ; movement does not imply a thing moving." ^ Here again we have simply a peculiarly rigorous application of the assumption of psychologism, that reality is nothing but the immediate data of consciousness.^ 1 La perception du changement, p. 24. 2 Incidentally, it may be remarked that Bergson's confessed uncertainty with reference to the religious implications of his system is probably due in no small part to his interpretation of "creative evolution" in a similar psychologistic and non-substantial sense. He asserts creation, but fails to interpret it as an activity of which there is any subject. His creative evolution is not evolution as the result of creative activity, but simply evolution as if it were the result of creative activity. In the last analysis — or the last intuition, rather — it is nothing but a real becoming among appearances than which there is nothing more real. CHAPTER VII The Older Absolute Idealism Logical-Psychological Idealism We spoke of psychological idealism as the most modern of the elemental types of idealism ; but the most typical idealism of the nineteenth century at least is not that which is developed under the immediate influence of the suggestions arising from the psychological view of experience; rather is it a device which has commended itself as affording a way of escape from the subjectivism which besets that psychological idealism. As soon as the human mind has passed from the natural realism of ordinary consciousness to the subjective idealism suggested by the psychological point of view, it is confronted — through a confusion of thought, as we have seen — with the problem as to how, where everything known is one's own idea, any knowledge of genuinely objective reality is possible. From this point of view the only possible solution of the problem seems to be found in an identity — if it can be maintained — between objective reality and the rational idea at which one arrives through the dialectical process. Reality, it is still assumed, as in psychological idealism, is constituted of contents, constituent elements, of consciousness ; but, since subjectivism is to be avoided, some way must be found of distinguishing between reality and those contents of consciousness which are mere subjective appearance. Assuming psychological idealism to be vahd as far as it goes, then it would seem to be only on condition of objective reality being regarded as constituted of the logical within the psychological, the universally acceptable within the contents of consciousness, that knowledge of objective reality can be said to be humanly possible. In this, the nineteenth century's most characteristic form of ideaUsm, whose earhest undoubted representative, as well 126 THE OLDER ABSOLUTE IDEALISM 127 as the most eminent and influential, was Hegel, we have a return to logical idealism as a way of escape from the subjec- tivism of psychological idealism, without giving up the essen- tials of the latter position. Being thus a synthesis of two of the elemental types of idealism, the logical and the psychological, it may be appropriately called logical-psychological idealism. As finding reality in what is not mere private feeling, but in that which, while made up of particular experiences, is shot through and through with universally acceptable and even necessary ideas, it claims to be objective, rather than merely subjective. In its simplest form this ''objective idealism" is the conclusion that the thesis that we know objective reality which is there for every one, and the opposing thesis that we can never know anything but ideas, contents, and parts of consciousness, cannot both be true unless reality is made up of universally acceptable ideas. It has thus grown up as a solu- tion of the problem of the possibility of knowledge of objective reality, proposed by and for those who cannot see their way clear to give up psychological idealism. It ought to be readily recognized that what we have here is mere dogma, rather than a vaHd and conclusive argument. As a synthesis of logical and psychological idealism, it is still vitiated by the already exposed fallacies underlying those two elemental forms of idealism. It is well to remember that there may be an abuse as well as a proper use of dialectic. At its best the dialectical process is a part of empirical analysis. Even when both of the antithetical judgments are inductions well supported by experience, unless it is certain that the synthesizing judgment exhausts all the possibilities in the case, it should be regarded as in some degree still hypothetical until it has been empirically verified. Any other use of the dialec- tical method is dogmatic. But when both of the propositions to be harmonized are dogmas resting upon fallacious reasoning, it seems the height of dogmatism to set forth the synthesis as necessarily true. The procedure in such a dialectical process is exactly parallel logically, if not morally, to the telling of one lie to support another. Such, we would claim, is the basis of logical-psychological idealism. It is a more than dubious solution of the entirely unnecessary problem of how to avoid 128 THE PROBLEM OF KNOWLEDGE scepticism, when one has made such a mistake in analysis as makes scepticism logically inevitable. If, however, there should be doubt as to whether modern objective idealism was really designed to extricate the modern philosopher from subjectivism, especially in view of the fact that Hegel starts directly with the concept of being, without any preliminary epistemological inquiry, it will be sufficient to recall that the German idealistic movement from Kant to Hegel is to be regarded as the working out of a way of escape from the Humian sceptical psychologism. Indeed the state- ment may be ventured that no modern idealism, however much it may wish to disown its subjectivistic ancestry, can justly deny the fact of that relationship. As a matter of fact, subjectivism constitutes no small part of the stock-in-trade of the typical modern ideahst; he would find it hard to do business without it. Before proceeding further with our criticism it will be well to indicate something of the relation of objective idealism and logical-psychological ideahsm, in their chief varieties, to each other. Most forms of objective idealism are logical-psycho- logical. Objective idealism may be concrete or abstract. By concrete idealism is meant the doctrine that reality is, in some sense of the word, idea, actually present in some experience. By abstract idealism is meant the doctrine that reality is, in some sense of the word, idea, but so stated that the reality is not, or cannot be held to be, all actually present in individual experiences. In one of its forms, as we shall see, abstract idealism ceases to be logical-psychological, and becomes simply logical ; it is objective without being subjective. Concrete idealism may be metaphysically monistic (singularistic), hold- ing that all reality is essentially idea, present in one all-inclusive experience; or pluralistic, holding that reality, as idea, is distributed among many mutually exclusive and ultimately real experiences. The monistic form of concrete objective idealism is usually called '' absolute idealism"; while the plu- ralistic form is often called '^ personal idealism." In some of its forms personal idealism ceases, as we shall see, to be logical- psychological, and becomes simply psychological ; it comes to be, from the point of view of the many selves, considered THE OLDER ABSOLUTE IDEALISM 129 together, no longer objective, but subjective only. Absolute idealism, on the other hand, hke concrete logical-psychological ideahsm everywhere, is not only objective, but, in a sense, subjective also. Its objectivity it gets from logical idealism. Its subjectivity — the doctrine that the object can exist only for a subject — holds with reference to the one Absolute Self (in singularism), or with reference to the many finite selves (in pluralism), but not with reference to the single finite self, for that would be soUpsism; and this subjectivity absolute idealism, like the others, gets from psychological idealism. Finally, then, logical-psychological idealism includes, as we shall see more fully in the sequel, all, or very nearly all, forms of monistic concrete idealism, and several varieties both of abstract idealism and of pluralistic concrete idealism. We shall first consider absolute idealism, or concrete logical- psychological idealism, in its monistic (singularistic) form. This monistic form differs genetically from the pluralistic in that the subjectivism with which logical idealism is united is of the solipsistic type. It is evolved as the final synthesis in a dialec- tical process, as follows : First thesis: I know objective reality. First antithesis: I know only my own ideas. First synthesis, becoming second thesis : Reality is constituted of my own ideas. Second antithesis : As a finite knower I do not know all reality. Second synthesis: It is only my finite self whose knowledge of reality is limited ; my true or absolute self must know all my own ideas, and so objective reality is to be thought of as the complete system of the ideas of my true or Absolute Self. The process may be continued as follows : Third thesis : the second synthesis just stated. Third antithesis: There are other finite selves, of whose ideas reality is composed. Third synthesis, or fourth thesis : It is not the finite self of these individuals of whose ideas reality is exclusively composed, but the true or Absolute Self. Fourth antithesis: There is but one objective reality of which the different finite selves have ideas. Fourth synthesis: The true self of all such individuals must be one and the same Absolute Self. In this dialectic the one initial error which vitiates each succeeding synthesis is the first antithesis, the dogma of psychological idealism, that I know only my own ideas. The logical idealism is explicitly introduced into this abso- 130 THE PROBLEM OF KNOWLEDGE lute idealism by a process of thought which may be thrown into overtly dialectical form, as follows : Thesis : I know objective reality. Antithesis: I am finite, and only the ideas of the Absolute Self are absolute ideas. Synthesis: Reality is the adequate logical idea, and adequately criticised or rational logical ideas are the ideas of the Absolute Self, so that in so far as I interpret my experience by means of such ideas, the Absolute Self interprets my experience in me. But while this is the latent dialectic underlying absolute idealism, that system of philosophy is generally based explicitly upon special arguments in which the above dialectic is either obscured or transcended. In this proposed basis there may be an emphasis in a one-sided and exclusive way upon rational thought-processes; or there may be added an emphasis upon purpose and will; or, finally, feeling, and especially religious feeling, may receive special emphasis. We would thus have three main types of absolute idealism, viz. intellectualistic absolute idealism, in which the Absolute Idea (with which it is maintained Absolute Reality is identical) is regarded as dis- coverable through critical intellectual processes; voluntaristic absolute idealism, in which the Absolute Idea is regarded as determined by purpose; and mystical absolute idealism, in which the Absolute Idea is regarded as being immediately experienced through feeling. We shall therefore consider absolute idealism in these its three principal types. But there is another triple division of absolute idealism which is also of great importance. After the original constructive movement in its intellectualistic and voluntaristic forms we must consider the destructive movement within absolute idealism, as represented by F. H. Bradley, and, finally, the attempts at reconstruction, intellectualistic, voluntaristic, and mystical. In the present chapter we shall deal with the original construction and with the destructive movement. Hegel's own philosophy, although of prime importance here, has been so often exhaustively expounded and discussed that it need not detain us long. The fundamental doctrine is that the real is the rational, not in the sense of mere logical idealism (with its identification of reality with the absolute idea as the abstract universal), but in the sense that, if being is interpreted, THE OLDER ABSOLUTE IDEALISM 131 after the manner of psychological ideahsm, in terms of con- sciousness, it is further determined by the principle that the real is the rational within the psychical, the universal {i.e. the universally accessible, or public) within the particular facts of any individual consciousness, the objective within the subjective — in other words, the concrete universal. Absolute Reality, in which all that is real must be included, is an absolutely rational system, in which all particulars of conscious experience are included. Hegel is commonly interpreted as intending to teach a monistic metaphysic.^ The criticisms we would be concerned to urge against this doctrine have already been indicated in the preceding paragraphs of the present chapter. After the philosophy of Hegel himself, the older EngUsh and American Hegelianism may be taken as affording perhaps the best available example of what we have called intellectual- istic absolute ideahsm. J. Hutchinson Stirhng and William Wallace in Britain and W. T. Harris in America confined their efforts for the most part to an exposition and defence of the Hegelian system, with little or no conscious deviation from the doctrine of the master. Wallace, speaking for Hegel and for himself, maintains that knowledge begins in the immediacy of sense-perception, which is a felt totality (or totality of feel- ings) ; its further task is to raise this to an intelligible totality (or totahty of intelligently ordered thoughts). It is here assumed that the contrast of subjective and objective is simply that between the earher stage of immediate feehng and the later one of the constructs of thought. Neither the contrast between feelings and that which is felt, nor that between thoughts and what is thought of, is treated as anything more than an essentially verbal distinction; '4deal-realism" or ''real-ideal- ism," the ''idealism of nature," and the "realism of mind" are the cardinal points of Hegelian doctrine.^ Let us see what we have here, and what it presupposes and ^See "Phanomenologie des Geistes," Werke, 1832, Vol. II, pp. 73-84, 131- 40; "Die Wissenschaft der Logik, II, Die Subjektive Logik," Werke, Vol. V, pp. 230-5 ; " Encyclopadie der philosophischen Wissenschaften, I, Die Logik," Werke, Vol. VI, pp. 320-21, 385; "Encyclopadie, III, Die Philosophic des Geistes," Werke, Vol. VII, Part II, p. 283 (cf. Royce, The Spirit of Modern Philosophy, p. 208, note 2), 307 ff. 2 W. Wallace, Prolegomena to HegeVs Logic, pp. 190-3, 303. 132 THE PROBLEM OF KNOWLEDGE implies. At first it would seem as though there had been simply a reversal of the point of view of common sense; the object immediately experienced being taken as subjective, and the ideas or thought-constructs of the subject as objective. But at any rate for the object as merely sensed there has been sub- stituted the sense-qualities, and these are interpreted, after the fallacious manner of psychological idealism, as being mere sensations, feelings, modes of the consciousness of a particular subject. On the other hand thought-constructs, when the thinking has been sufficiently critical, seem less private than sensations and feelings; they are in a sense transferable, universally usable, in that the words used in communication directly express these thought-constructs, and only more remotely the sensations and feelings. The thought-constructs are more universally accessible than the feelings. Accordingly, after the fallacious manner of logical idealism, but also in default of anything less dependent upon the particular subject, since psychological idealism is assumed, this imiversality of rational thought is interpreted as being itself the essence of objectivity. The thought-construct, or '^ universal," however, is not taken abstractly, but (theoretically) in all its relations, so as to include, especially, its relation to the particulars of sense ; it is a universal in the particular, the "concrete notion " or "concrete universal" which Stirling rightly speaks of as "the secret of Hegel." ^ It is involved in this view that to think of the objects of sense-perception as capable of existing either wholly or in part, independently of the relation to the immediate data of the consciousness of a subject, is to take an abstraction as a reality. Of course, it is true, as Stirling points out,^ that this doctrine of the ''concrete universal" may be regarded as having been ''implicit" in the Kantian view that the under- standing constructs nature out of the immediate data of sensi- tive consciousness ; but when one reflects that it is not nature, but at most our mental instruments for the perceiving and understanding of nature that the human mind constructs, it is seen that in referring back to the Kantian doctrine we have simply traced the confusion and dogmatism back to a point nearer its beginnings. Going still further back, we should find 1 The Secret of Hegel, 1865, Vol. I, pp. xi, Ixix. « lb., p. xi. THE OLDER ABSOLUTE IDEALISM 133 behind Kant on the one hand the sceptical empirical idealism of Hume, and beyond Hume the dogmatic subjective idealism of Berkeley, and on the other hand the equally — if not so obviously — fallacious logical idealism of Plato. The argument upon which the intellectualistic absolute idealists generally seem to depend most is that which, first assuming that the real is intelligible, and that the intelligible is rational, concludes first that the real is rational, and then, on the assumption that the rational is mental, spiritual, con- cludes further that the real is mental, spiritual. Edward Caird, in his exposition of Hegel's philosophy, expresses the argument in condensed form as follows : ''To express all in a word, 'the real is the rational or intelligible, ' i.e. it is that which is capable of being thoroughly understood by the intelligence, just because it has in it the essential nature of the intelligence, or self-con- sciousness." 1 This argument has the appearance of logical validity ; but when we examine the assumptions, we find them highly dogmatic. So far from being justified in concluding, since universal agnosticism is self -refuting, and since all thought practically assumes the possibility of knowledge, that therefore all reality is intelligible, we are warranted only in saying that some reality is intelligible. The disproof of a universal negative is no proof of a universal, but only of a particular affirmative. The other assumptions, that the intelligible is rational and that the rational is mental, or spiritual, are capable of being used as in the above argument only because of the ambiguity of the te^-m "rational." This term may mean "mental" in the sense of that phase of the mental which is constituted by the fixed and universal forms of thought, but it may also mean that in the objective realm which corresponds to this phase of the mental. This ambiguity it is which gives the appearance of logical validity to the latter of the two syllogisms under con- sideration. The sense in which rationality may be predicated of whatever is intelligible is not the same as that in which it may be predicated that it is in every case mental. In one way or another, then, the second syllogism is fallacious : either ^ Hegel (Blackwood's Philosophical Classics), p. 176; cf. John Watson, The Interpretation of Religious Experience, 1912, Vol. I, pp. 74-7 ; Vol. II, pp. 38, 60, 104. 134 THE PROBLEM OF KNOWLEDGE ^'rational" means the same thing in both premises, in which case at least one of them is clearly an unsupported dogma; or else the term has different meanings in the two premises, in which case the conclusion depends upon the fallacy of "four terms." All that we are really justified in concluding is that some reality is perhaps mental, which is less than we knew with- out the argument. There are some individual variations among the representa- tives of this philosophy that are of considerable interest and significance. John Caird asserts that to ''constitute the exist- ence of the outward world" we must think it ''as existing for thought"; we must needs presuppose a "consciousness for which and in which all objective existence is." He therefore claims that to attempt to conceive of "an existence which is prior to thought" is "self-contradictory, inasmuch as that very thing-in-itself is only conceivable by, exists only for, thought." "We must think it before we can ascribe to it even an existence outside of thought." ^ Here we have the psychological idealism which is to serve as a foundation for absolute idealism, supported by the fallacious argument from the "egocentric predicament." T. H. Green, who is generally regarded as also a neo-Hegelian, remarks of this argument of Caird, that the reader "will be asking, from page to page, what, after all, this thought is which seems to be and to do anything and everything. Instead of being duly directed for an answer to an investigation of the objective world, and the source of the relations which deter- mine its content, he is rather put on the track of an introspec- tive inquiry what or how he can or cannot conceive. ... He will charge the author with confusing . . . the proposition that a thing is only conceivable by thought . . . with the proposition that the thing only exists for thought; the prop- osition, again, that no object can be conceived as existing except in relation to a thinking subject, with the proposition that it cannot exist except in that relation." ^ Green thus repudiates the argument from the egocentric predicament. Green's own method was to seek to ascertain the nature of "that thought which Hegel declares to be the reality of things" ^ An Introduction to the Philosophy of Religion, 1904 ed., pp. 147-8. 2 Works, Vol. Ill, pp. 143-4. THE OLDER ABSOLUTE IDEALISM 135 "from analysis of the objective world, not from reflection on those processes of our intelligence which really presuppose the world." ^ Thus while John Caird and T. H. Green agreed in the end with Hegel that ''that only is real which is rational and that only is rational which is real," Caird's favorite approach to Hegelianism as a whole was along the line of the reality of the rational, while Green maintained that the only undogmatic path was that of the rationality of reality. In examining ''the constituents of that which we account real," he claimed to find "that they all imply some synthetic action which we only know as exercised by our own spirit." "Is it not true of all of them," he asks, "that they have their being in relations; and what other medium do we know of but a thinking consciousness in and through which the separate can be united in that way which constitutes a relation?" ^ But when we examine this positive argument for absolute idealism offered by Green, we find that it is not free from the fallacious assumption of psychological idealism. To cognize relations, it is assumed, is to construct the relations thus cog- nized — a doctrine which manifestly can be true only in so far as the experienced contents among which relations are cognized have reality only in and for consciousness, as parts of conscious- ness itself. This, of course, is psychological idealism. Green has no intention, however, of indorsing any view which would leave no room for the existence of any knowable reality beyond the consciousness of the finite subject. He therefore sets up again, in antithesis to the above thesis that relations are thought- constructs, the realistic doctrine that reality has relations which are not dependent upon the thought activity of the finite subject. The synthesis, depending, after the manner of Lotze's argument, upon the argument from analogy, is that there must be "a spiritual principle in nature" which constitutes the rela- tions existing independently of human consciousness. It is the fallaciousness of the assumed psychological idealism which renders necessary the dogmatism of this final synthesis. But Green's position has not been regarded as wholly sound by some other members of the so-called neo-Hegelian school. 1 Ih., p. 144. >/6., p. 145; cf. Prolegomena to Ethics, §§ 13, 20, 26-9, 37, 52, 62, 70. ■ 136 THE PROBLEM OF KNOWLEDGE His doctrine is better understood, perhaps, as an independent development from Kantianism than as an adoption of HegeHan- ism. He is related to Kant somewhat as Berkeley is related to Locke. Having cancelled the Kantian unknowable Ding an sich, he develops instead the doctrine of an eternal Conscious- ness, operating in human knowledge and activity. This Consciousness, however, he apparently hesitates to identify with the Hegelian Absolute Reason, or Thought, manifesting itself in nature and history.^ But it was chiefly for his doctrine that facts are relations that Green was criticised. The view that nature is a fixed and unalterable system of relations, and that these relations can only be explained as the work of mind, was attacked, not only by such non-idealists as A. J. Balfour, but by Bradley, Royce, Haldane, and others within the school.^ It was as if, after the similitude of Pharaoh's dream, the lean kine of relations had devoured the fat kine of qualities, only to remain at last as lean and ill-favored as ever. But Green is not the only member of this school who has come perilously near to allowing the objective absolute idealism to disintegrate into a subjective psychological idealism, or into a position which oscillates between such a subjectivism and a sort of abstractionism. J. H. Muirhead, for example, writes as follows: ''When I. say, 'What a lot of buttercups,' what I mean by buttercups is a system of judgments which I am ready to make in reference to a particular object, judgments which I am prepared to make because I have already made them."^ D. G. Ritchie goes quite as far when, after saying that the reality of things is "what we ought to think of them," he goes on to assert: "Facts are theories. . . . Sunrise is a theory, now discarded ; the reality is the rotation of the earth : and yet we are in the habit of speaking as if sunrise were the reality and the rotation of the earth the theory."^ Now whether it be asserted that nature is a system of mentally constituted relations, or a system of judgments, or a sum-total 1 On the difference between the philosophical views of Edward Caird and T. H. Green, see article by John Watson, Philosophical Review, XVIII, 1909, especially pp. 161-2. 2 See, e.g., R. B. Haldane, The Pathway to Reality, Bk. Ill, Ch. III. 3 Mind, N.S., Vol. V, 1896, p. 512. * Darwin and Hegel, with Other Philosophical Studies, 1893, pp. 87, 91. THE OLDER ABSOLUTE IDEALISM 137 of valid theories, the faulty analysis and dogmatism must be patent to every unbiassed mind. The fact is that the idealistic *' intuition" that 'Hhings are thoughts" has dictated the results of the analysis of experience, and then these " doctored " reports of experience are used in support of the original idealistic dogma. Edward Caird and John Watson are among the best and most typical representatives of a purely intellectualistic absolute idealism. What the latter says of the former is true of both : they never waver in the " conviction that the universe is rational and that its rationality can be proved." ^ According to Caird "the ontological argument ... is simply the expression of that highest unity of thought and being which all knowledge presupposes as its beginning and seeks as its end." ^ Thus idealism is regarded as the real meaning of the ontological argument. Jn another work Caird teaches that subject and object are necessarily related to each other, and necessarily distinguished from each other. Each presupposes the other, therefore neither can have produced the other; we cannot reduce the subject to a mere object among other objects nor the object to a mere phase in the life of the subject. We are there- fore forced to seek some all-embracing unity ; binding in one all being and all knowing. This unit}^, to the idea of which we have been thus dialectically conducted, is the absolute, all- comprehending Reason or Spirit.^ But is even this careful statement of the doctrine free from- dogmatism? Is it true that all thought presupposes the ''unity of thought and being" in the sense in which idealism interprets this phrase? Is it proved that object (interpreted as any real thing) and subject are ''necessarily related to each other"? Of course the object, as that which is presented to a subject, is necessarily related to the subject; but from this "egocentric predicament" nothing can be proved. And finally, is the Hege- lian Absolute the only concept which can conceivably synthesize the juxtaposed things and thinkers? Even in the moderate statements of Edward Caird, then, we find dogmatism and evidence of defective analysis. 1 Philosophical Review, XVIII, 1909, p. 161. 2 The Critical Philosophy of Kant, 1889, Vol. II, pp. 123, 128. 3 The Evolution of Religion, 2d ed., 1894, Vol. I, pp. 64-8. 138 THE PROBLEM OF KNOWLEDGE Much the same thing may be said of Watson. For an authori- tative and consistent exposition of the orthodox British Hegehan- ism one cannot do better than have recourse to the works of this philosopher, and the following passage is especially illuminating from this point of view. ''Nature, or the so-called 'external' world, is not external to mind, but only 'external' in the sense that it consists of objects outside of one another spatially, or of events external to one another in the sense of being discrete and 'marching single in an endless file.' We are, therefore, just as directly conscious of matter as of mind. Moreover, the external or material world is not given to us in our sensations ; for sensations in their singleness are not knowledge : only when they are ordered and combined under the forms of perception and thought have we any experience of nature. Now these forms do not, like sensation, vary with each individual and change upon us from moment to moment; they are identical in all men. Thus we all construct an external world which, vary as it may in its sensible aspects, is fundamentally the same in this sense, that it consists of objects in space and events in time, all of which are connected together by the bond of natural causation. This is the world which it is the business of the sciences to survey and reduce to specific laws." ^ Here again we see the identification of subjectivity with the sense-elements of the experience of the individual, while objectivity is regarded as the product of the union of these data of sense with the mental elements common to all minds. But is not the resulting "object" still essentially subjective? Such constructed objects may be similar enough in different individual experiences for a certain "imiversaHty" to attach to such perceptual experience; but it can never amount to objectivity in the sense of reality existing prior to and inde- pendently of the knowing relation. As Watson himself says, from his point of view "it is not true that facts are independent of the individual subject in so far as he is a rational intelligence." ^ What Watson does here is what the typical objective idealist (if not an epistemological dualist and therefore merely meta- physical ideaUst) always does; he substitutes subjective uni- 1 The Philosophical Basis of Religion, 1907, pp. 76-7. 2 The Interpretation of Religious Experience, 1912, Vol. II, p. 60. THE OLDER ABSOLUTE IDEALISM 139 versality, universality-in-subjectivity, for true objectivity. (For simplicity we view Watson's philosophy here at a certain stage in its dialectical unfolding : we abstract from the meta- physical monism. Strictly, what we ascribe to him here is what he would have to say, finall}^, if it were not for his monism.) It is an ancient observation that '^ misery likes company," but a numerical multiplication of an essentially subjective experi- ence, like simple multiplication in any other situation, can hardly be said to change the character of the unit multiplied. That our interpretation of the philosopher's thought is not unfair is indicated by the later statement, ''To say that this world acts upon our minds is the same as saying that a world which exists only by the activity of our ^ minds is the cause of that activit}^" ^ What then becomes of the objective idealist's supposed escape from subjective idealism by means of the postulate ''that we are capable of knowing Reality as it actually is"?^ One or other of two issues is possible. Either the appeal to the possibility of knowledge turns out to be a mere apology for the high-handed procedure involved in passing off for real objec- tivity certain common features in the subjectivity in which human and all conceivable experience is necessarily involved; or — and this is the horn of the dilemma which Watson chooses — starting from "the principle that there is one intelligible universe and one kind of intelligence,"^ one is forced to conclude that there is but one real mind or experience, the only objec- tivity being dependence for existence upon being known by the one and only mind. But, we would say, if the philosopher chooses to enter upon this path, let him have the courage of his convictions and follow it to the bitter end. Let him accept the solipsism of the Ab- solute, and the absolutely illusory character of his own individ- uality, as of all other plurality — a conclusion which few ab- solute idealists outside of India are consistent enough to draw, or frank enough to acknowledge. Assuredly, then, the objective ^ Italics here are mine. 2 j^fig Philosophical Basis of Religion, p. 81. ' J, Watson, An Outline of Philosophy, 1898, Preface, p. vi. * lb., p. 37 ; italics naine. Cf. The Interpretation of Religious Experience, Vol. I, p. 74. Watson is suspiciously reluctant, as shown by the context of the passage quoted, to admit that the oneness of the universe is an assumption. 140 THE PROBLEM OF KNOWLEDGE idealist may be charged with failure to keep his promises : we bargained for bread, but he ostentatiously presents us with a stone ; we desired to be assured of the possibility of knowing a reality whose existence did not depend upon our awareness of it, and he has answered us by a virtual denial that we ourselves as finite individuals have any real existence. Such seems to be the penalty awaiting those who step aside from the true highway of knowledge into the devious by-paths of subjectivism and maintain their course in stubborn unrepentance. It may be instructive to examine further the argument by means of which this representative Hegelian defends the exist- ence of the one Absolute Mind. It will be seen that what is called necessary implication is really dogmatism based upon equivocation. ''An intelligible system," it is asserted, ''neces- sarily implies an intelligence that is capable of grasping the system.' ' ^ "Intelligible" in the course of this sentence changes its meaning from possessing the objective conditions for being known, to possessing all conditio7is for being known, subjective as well as objective. Similarly ' ' intelligence ' ' changes in meaning — as one sees from the context, the Absolute being finally meant — from that which can know something of an intelligible system to that which can know the system as a whole. If "intelhgible" had originally been taken in the second of its two meanings, and "intelligence" Hmited to the first of the two meanings given to it, the argument would violate no logical principle. It would be an entirely accurate interpretation of the meaning of a proposition, however useless for the purposes of the ab- solute idealist. The appearance of demonstration of the exist- ence of the Absolute Mind depends upon the double equivocation just pointed out. The original postulate of "an intelligible system" is supported by arguing that "if there exists any intelligence whatever, the universe must be intelligible." ^ This may be allowed to stand if appeal is made to experience, with its immediate awareness of the identity of the experience in which we know something and the experience in which something is known by us, and to the reflective knowledge that the conditions of the possibility of the one are the conditions 1 The Interirretation of Religious Experience, Vol. I, p. 74 ; italics mine. a/6. THE OLDER ABSOLUTE IDEALISM 141 of the possibility of the other. By means of a similar appeal to empirical intuition, we find that we may admit the further postu- late that any intelligence which ''knows the universe to be in- telligible" ''must be capable of knowing that it knows the universe to be intelligible." And so, with the help of an appeal to intuition which ultimately rests on experience, and yet de- pending of necessity upon the equivocation noted above, it is concluded that "our self-consciousness . . . implies a self- conscious intelligence that comprehends within itself all modes of personal consciousness." ^ After the manner of mediaeval scholasticism at its worst, specious arguments are called in to give the appearance of, rational demonstration to what remains in its essential nature an experience-contradicting dogma. Another direction from that taken by the intellectualistic absolute idealists is what we may call voluntaristic absolute idealism. Whereas in the intellectualistic type the absolute idea with which reality is identified is regarded as being deter- mined simply by rational processes of intellection, in this voluntaristic type the absolute idea is regarded as being de- termined primarily by purpose rather than by critical thought alone. The most eminent exponent, at least recently, of this variant form of the older absolute idealism is Josiah Royce. At first thought it may seem entirely proper to include a con- sideration of Fichte's system of thought in this connection, in view both of the voluntarism of his idealism and of the intro- duction of the notion of the Absolute Subject ; but inasmuch as he can hardly be said to have introduced what we have called logical ideahsm into his activistic or voluntaristic psychological idealism, or really to have made good his escape from subjective to objective idealism, we choose not to include at this point any further examination of his doctrine, but to confine ourselves to a brief exposition and critique of Royce's voluntaristic ab- solute idealism. Royce's philosophy may be regarded as essentially an attempt to develop the very modest, undogmatic theoretical idealism which we have called "relative idealism" into an all-com- prehending absolute idealism, and that for the solution of the modern epistemological problem which has arisen out of psy- 1 Ih., pp. 74-6. 142 THE PROBLEM OF KNOWLEDGE chologism. It is true that an important factor in the develop- ment of Royce's philosophy has been the discovery of a capri- cious and irrational element in reality, a discovery which serves to differentiate his philosophy from intellectualistic absolute idealism, with its assumption that reality, without remainder, is intelligible, and in connection with which discovery, as in the case of his voluntarism, he acknowledges the influence of Schopenhauer.^ But the main foundation of Royce's con- structive philosophical work seems to be, after the psychological idealism everywhere presupposed, the generally admitted fact that our ideas are determined, to begin with, at least, by our purposes. But while, according to ''relative idealism," the idea determines the selection of the reality to be considered, and also, within very narrow limits, may be said to enter into and become a part of that reality for the particular purpose concerned,^ in Royce's philosophy on the other hand all reality is viewed as constituted, in the last analysis, by purpose. And yet, as we shall presently see more clearly, it is the underlying psychological or subjective idealism which makes it impossible, logically, to stop with relative idealism, and leads on to an idealism in every respect absolute. To this conclusion Royce leads up by a dialectical process starting, in the most important form of his argument, from the antinomy which seems to exist between what he calls the internal and the external meaning of ideas. By ''internal mean- ing" is meant purpose, "in so far as it gets a present conscious embodiment in the contents and in the form of the complex state called the idea." By "external meaning" is meant "reference (of ideas) beyond themselves to objects." ^ Or, in other words, the internal meaning of my idea is what I call my purpose ; it is what I mean in so far as I am aware of my purpose. The external meaning, on the other hand, is the total reality which I come to know as the realization of my purpose, the reality which I meant. The antinomy lies in the twofold and apparently self-contradictory character of the meaning of 1 The Spirit of Modern Philosophy, 1892, pp. 261-4 ; The Problem of Chris- tianity, 1913, Vol. I, p. xii. 2 See discussion of "tertiary qualities" in Ch. XIV, infra. 3 The World and the Individual, Vol. I, 1899, pp. 25-6. THE OLDER ABSOLUTE IDEALISM 143 meaning. At one time it seems to be a mere content in some one's mind ; at another time it appears as the subject-matter for an unlimited number of judgments.^ The solution of this antinomy is stated as follows: "Now the obvious way of stating the whole sense of these facts is to point out that what the idea always aims to find in its object is nothing whatever but the idea's own conscious purpose or will, embodied in some more determinate form than the idea by itself alone at this instant consciously possesses. When I have an idea of the world, my idea is a will, and the world of my idea is simply my own will itself determinately embodied." ^ That is to say, the external meaning is simply identical with the internal meaning ; the thing meant — in spite of any appearance to the contrary — is conscious purpose regarded as completely determined. ''The complete content of the idea's own purpose is the only object of which the idea can ever take note." ^ Stated more generally, the conclusion is that ''what is, or what is real, is as such the complete embodiment, in individual form and in final fulfilment, of the internal meaning of finite ideas." '* The object of any idea is "an individual life, present as a whole, totum simul. , . . This life is at once a system of facts, and the fulfilment of whatever purpose any finite idea, in so far as it is true to its own meaning, already fragmentarily embodies. This life is the completed will, as well as the completed ex- perience, corresponding to the will and experience of any one finite idea. ... To be, in the final sense, means to be just such a life, complete, present to experience, and conclusive of the search for perfection which every finite idea in its own measure undertakes whenever it seeks for any object." ^ Now the gist of this argument may be put as follows : What I mean (internal meaning) is my idea, purpose, plan of action. Reality is what I mean (external meaning). Therefore, reality is my idea, purpose, plan of action. Manifestly it is a case of the fallacy of "four terms," unless in some way internal and external meaning can be absolutely identified. But in view of the underlying psychological idealism, and by the introduc- tion of metaphysical monism, which identifies my real self with 1 lb., pp. 320-4. 2 76., p. 327. » lb., p. 329. 4/&., p. 339. » 76., pp. 341-2. 144 THE PROBLEM OF KNOWLEDGE the Absolute Self, it becomes possible to make this identification. Reality, then, is what I mean, my idea or purpose, as it is for my Absolute . Self . But this is to escape fallacy at the expense of assuming the dogmas of metaphysical monism (singularism) and psychological idealism. From Royce^s own point of view, it should be noted, the argument is not fallacious ; assuming a monistic or solipsistic psychological idealism, ever3rthing ''external" is internal, and so ''external meaning" is internal meaning. This psychological idealism, however, is itself fal- lacious, and is here in union with the (also fallacious) logical idealism involved in the proposition. Reality is my idea or purpose. This philosophical doctrine which Royce offers as a synthesis of mysticism or subjective empiricism, and dualistic realism, more conclusive than any that "critical rationalism" by itself is able to accomplish, may be similarly reached from many different starting-points, as we learn from an examination of the various works of this philosopher. Taking as his thesis the proposition that there is a whole truth — a proposition such that to deny it is to assume it ^ — he sets over against it the an- tithesis that for any reahty to be represented in judgments, as the ideal of truth demands, an infinite series of judgments would be required. But such a system could never be complete, while the reality of complete truth is a necessary presupposition of all judgment. The one synthesizing concept — since realism is rejected and psychological idealism is presupposed — is found in the idea of a total rational system, or absolute experience, in which the infinite is actual as a " self -representative system." ^ Even to assert the possibility of error, or the fact of one's ignorance, assumes the reality of truth, and therefore involves the same conclusion.^ The same final synthesis is involved in affirming the reality of the self,^ or of individuality,^ and even in willing the good.^ What the self is, or what any individual 1 The Philosophy of Loyalty, 1908, p. 345. 2 William James and Other Essays, 1911, Essay IV; The Philosophy of Loy- alty, Lect. VII ; The World and the Individual, Vol. I, Supplementary Essay. ' The Religious Aspect of Philosophy, 1895, Ch. XI ; The Conception of God, 1897, pp. 15-44; The Sources of Religious Insight, 1912, pp. 105-116. * Studies of Good and Evil, 1898, Ch. VI. ^ The Conception of Immortality, 1900, passim. « William James and Other Essays, 1912, Essay V. THE OLDER ABSOLUTE IDEALISM 145 is, or what the good which is willed really is, can be completely shown, in time, only in an unending series; as real, therefore, each of these involves a time-transcending but time-including absolute system or experience, in which what we mean is eter- nally real. Finally, in his recent lectures on Christianity, Royce presents his argument in yet another form. Two oarsmen believe themselves to be in one and the same boat, although this is not a direct perception, according to Royce, but an interpre- tation. Assuming the truth of psychological idealism we should have to admit this ; the contents of the perceptual consciousness of the two men are not fully identical qualitatively, much less numerically. The interpretation put upon their experiences by the two men, viz. that they are in the same boat, cannot be true from this (subjective-idealistic) point of view, unless there is an absolute, all-inclusive experience, in which what is per- ceived by the men only fragmentarily is experienced with all its relations, totum simul, as an infinite totality. ^ Thus for Royce all dialectical paths lead to the Absolute, a realization of all possible meaning, a unity of all that is or can be meant, in a single concrete experience. Royce's system, by whatever argument it may be defended, is vulnerable, both in its process and in its conclusion. If, instead of invoking the Absolute to save him from the infinite regress, Royce had learned from the pragmatist that the true and the good, and even the self and other individuals, can be defined sufficiently for such hwnan purposes as ought to he considered, he would have been saved from the necessity of giving his ad- herence to the self-contradictory 2 notion of an actual infinite total of definite qualities and relations. But the fons et origo mali in all of the above arguments is, let it be repeated, the fallacious and dogmatic assumption of psychological idealism. What do I really mean when I assert that there is something of which I am ignorant ? Is it, as some would say,^ either my own, or some one's future experience, or, as Royce asserts, a present, or better, super-temporal experience of the Absolute? Why future in the one case? And why experience in the other case? Is not all we can say without 1 The Problem of Christianity, Vol. II, pp. 241-3. 2 See pp. 462-70, infra. 3 E.g. Dewey, Journal of Philosophy, etc.. Vol. IV, 1907, p. 202. L 146 THE PROBLEM OF KNOWLEDGE dogmatism simply that it is a present reality of which I or some one else may perhaps have a future experience, and of which whatever ''Absolute'^ there may be ground for positing may, only possibly, have experience at present? The further asser- tions can only be made on the basis of psychological idealism. But even apart from any criticism of psychological idealism, or of the notion of an actual infinite, the charge is frequently made against the older absolute idealism, whether intellectualis- tic or voluntaristic, that it can be shown to be self-refuting, in that the elements of finite experience are what they are in some measure by reason of the finiteness of the experience, so that their inclusion, without modification, in an infinite or ab- solute experience is, in the nature of things, impossible. This criticism, as against Royce, has been well put by A. K. Rogers, who writes: ''What can the duplication of thought and ex- perience be like for an Absolute Being? I think of things only because direct experience is impossible for the time. ... How can we make our ignorance a part of an all-inclusive experience without denying its existence (or changing it)? Can I feel baffled and see the solution in the same experience? Is my feeling of ignorance identical with God's consciousness of ig- norance? If so, we must accept an Absolute that grows in knowledge. ... If not, there are two facts, only one of which is the experience of the Absolute ; for my feeling of ignorance dominates my consciousness, and cannot dominate God's." ^ Or, as Bosanquet remarks, if the later occurrences modify the earlier occurrences, the events cannot remain, in actual content, within a larger span of consciousness, what they were or could be within a shorter.^ But these last criticisms are but repetitions, essentially, of certain phases of the self-refutation of the older absolute idealism accomplished once for all by F. H. Bradley, chiefly in his Appearance and Reality. Absolute idealism, having always wielded the sword of intellectual criticism, seems to have been doomed to perish, at least in its older forms, by that self-same ^ "Professor Royce and Monism," Philosophical Review, XII, 1903, pp. 47 ff. Cf. A. Aliotta, The Idealistic Reaction against Science, Eng. Tr., 1914, pp. 259- 65; A. E. Taylor, Mind, N.S., XXI, 1912, pp. 540-1. 2 The Principle of Individuality and Value, 1912, pp. 387-8. THE OLDER ABSOLUTE IDEALISM 147 sword. It has found, in Bradley, one of its most formidable foes within its own household. In view, then, of this fact, and of the attempts which, as we shall see, have recently been made to reconstruct an absolute idealism in spite of the havoc wrought by this critic, we may perhaps most instructively classify all types of Anglo-American absolute idealism under three main heads, viz. the pre-Bradleian construction, the Bradleian de- struction, and the post-Bradleian reconstruction. We must now consider the second of these, the antithesis in the dialectic of modern absolutism. Bradley started as an adherent of the orthodox absolute idealism.^ He proposed to take seriously the conclusion, '' inherited from others," ^ that reality is a single experience, in which all realities with all their appearances are included. In trying to think this through, however, he comes to the conclusion that it is a self-contradictory notion that such an experience can be ordered according to the principles of reason. Of fundamental importance here is Bradley's judgment ( !) that all judgment is essentially fallacious, in that it ''attributes to a subject something other than itself, and which the subject is not." ^ Thought can never, however complete, be quite the same as reality.^ Being abstract, relational, discursive, it can never be the same even as the lower and less inclusive im- mediacy and all-togetherness of individual human experience. He thus reaffirms, in effect, in his latest utterances the well- known conclusion of an early work: "Unless thought stands for something that falls beyond mere intelligence, if 'thinking' is not used with some strange implication that never was part of the meaning of the word, a lingering scruple still forbids us to believe that reality can ever be purely rational. It may come from a failure in my metaphysics or from a weakness of the flesh which continues with me, but the notion that existence could be the same as the understanding strikes us as cold and ghostlike and as the dreariest materialism. That the glory of this world in the end is appearance leaves the world more glorious, if we feel it is a show of some fuller splendor; but 1 See Ethical Studies, 1876. 2 Essays on Truth and Reality, 1914, p. 246. 3 Appearance and Reality, 1st ed., 1893, 2d ed., 1897, p. 57. * lb., p. 554 ; cf. Essays on Truth and Reality, 1914, pp. 230-3. 148 THE PROBLEM OF KNOWLEDGE the sensuous curtain is a deception and a cheat, if it hides some colorless movement of atoms, some spectral woof of impal- pable abstractions, or unearthly ballet of bloodless categories. Though dragged to such conclusions, we cannot embrace them. Our principles may be true, but they are not reality. They no more make that whole which commands our attention than some shredded dissection of human tatters is that warm and breath- ing beauty of flesh which our hearts found delightful." ^ Not only, then, it is claimed, can thought not be identical with human experience ; far less can it be identified with the all- inclusive experience, for, as compared with this, even im- mediate human experience itself is infected with unreahty. For example, reality cannot be said to be made up of substances which have qualities, for we do not know what a substance is. It is not the qualities, nor is it anything, so far as we can know, behind the qualities. ^ Neither can we regard reality as made up of qualities in relations. Quahties are never found without relations, and cannot be conceived as existing without them; and yet, qualities with relations are no more intelligible. The qualities cannot be wholly resolved into relations, nor can any quality be found so simple that it is not made what it is to some extent by some of its relations. Similarly of relations : without their terms they are nothing; but even with their terms they are unintelHgible. If the relation is nothing to the qualities, they are not related ; the relation is a nonentity. If, however, the relation is something to the terms, there is a relation between the relation and the term ; and so on in unend- ing regress.^ And so of primary and secondary qualities,^ space and time,^ motion and change,^ causation,^ activity,^ and the self.^ The difficulty is especially great in connection with error. It cannot, exactly as it is experienced by the person in error, belong to the all-including unitary experience, in the light of which all error is corrected ; and yet the error is a fact and so cannot be excluded from reality.^'^ Even a suggestion which Bradley finds useful in accommodating his metaphysics to common sense, viz. that there are degrees of reality, is not ab- 1 Principles of Logic, 1883, p. 553. 2 Appearance and Reality, 2d ed., Ch. II. 3 76., Ch. III. 4 75.^ Ch. I. s 75.^ ch. IV. « lb., Ch. V. 7 76., Ch. VI. 8 76., Ch. VII. ^ 76., Chs. IX, X. ^ 76., Ch. XVI. THE OLDER ABSOLUTE IDEALISM 149 solutely true; everything either is, or is not, absolutely real.* And yet appearances exist, and whatever exists must belong to reality ; - but they cannot exist in the experience of the Absolute exactly as they exist in ours, because different experi- ences are from time to time discrepant with each other.^ In the end, therefore, while retaining the idea of Reality, or the Absolute, as a single, all-inclusive, and perfectly harmoni- ous experience, Bradley concludes that we can have only a vague idea of its nature. It is not personal, but super-personal ; not moral, but super-moral; not rational, but super-rational. Critical rationalism, the only method we can use, is futile, so far as any detailed positive knowledge of the Absolute Experi- ence is concerned. The only method that would be adequate, viz. immediate or mystical intuition, we cannot use; that is for the Absolute alone. Absolute idealism, then, in the strict sense of the term, is given up. Ideas and reason having only human and relative value, idealism becomes a misnomer.* Absolutism remains, but it is such as might be called absolute empiricism, or absolute immediatism. One might even say that Bradley's view is a negative or agnostic mystical ab- solutism; reality, knowable in any case only by mystical intuition, is held to be for man essentially unknowable. As a polemic against orthodox absolute idealism Bradley's criticism was highly successful. The main criticisms to be directed against his own position are perhaps two, viz. first, that he is over-sceptical as regards the power of the human understanding to make a true judgment — a question which will be taken up in our discussion of the problem of truth ; and, secondly, that he is dogmatic in retaining what he does of the absolute idealism which his criticism has shown to be self- refuting. In fact, Bradley seems not to have completely realized how far-reaching are the logical consequences of his argument. The absolute idealism which he criticises started with the postulate that reality is rational, and, in order to defend this, was led by a dialectical process to conclude that reality is a single all-inclusive experience. Bradley, originally accepting the current idealism, comes finally to see that if reality is a single 1 lb., Ch. XXIV. 2 76., pp. 132, 140. 3 lb., pp. 241, 511. * See Appearance and Reality, p. 547. 150 THE PROBLEM OF KNOWLEDGE all-inclusive experience, it cannot be rational, intelligible.^ F. C. S. Schiller holds that an all-inclusive conscious experience could only be regarded as ''morbidly dissociated, or even down- right mad." ^ Logically considered, then, the situation is this : either reality is a single all-inclusive experience, and therefore not intelligible ; or reality is not a single all-inclusive experience, in which case it may be conceived either as rational or as not rational. Bradley chooses the first of the major alternatives, that reality is a single super-rational experience; but unless a sufficient reason is given for this choice, it is essentially dogmatic. The only approach to a reason for rejecting the view that reality is not a single experience is that one of the two possible ways of inter- preting this view would require one to hold that there can be true judgments, a conclusion which Bradley imagines cannot be maintained, because the predicate is never absolutely identical with the subject. But if a judgment, to be true, need not have an absolute identity between the subject and the predicate, Bradley's objection falls to the ground, and the view that reafity is essentially intelligible, but not all one experience, is seen to be admissible. But even with Bradley's strange prejudice against judgments, why should he not choose the view that reality is not all one experience, instead of this doctrine which he ''inherited from others"? Indeed this would have been a more defensible course than to retain, as he did, a conclusion, the original basis of which he had just destroyed. But, of course, with his theory that the judgment is never possibly true, Bradley would not be justified in holding to his view; and for the same reason neither is he justified in judging his own theory to be true. As a critic points out, "the very fact that this conclusion is arrived at by judgments, which both by Mr. Bradley's own methods and his own acknowledgement are self-contradictory, is of itself quite sufficient both to invalidate it and to make his system self -refuting." ^ What Bradley him- self says is that "in the end no possible truth is quite true." Thus, while claiming that his view is ultimate for intellect, 1 Appearance and Reality, p. 554. ^Journal of Philosophy, etc., Vol. Ill, 1906, p. 482. 8 E. G. Spaulding, Philosophical Review, Vol. XIX, 1910, p. 631. THE OLDER ABSOLUTE IDEALISM 151 and that any alternative is more inconceivable, he has to confess that even what is for us absolute truth is necessarily erroneous.^ And so, while Bradley is to be credited with revealing the untenability of monistic or absolute idealism on rational grounds, we are entitled to condemn his own metaphysics as being not only dogmatic, but, for one with his presuppositions as to judg- ments, logically untenable. It should not be imagined, however, that either the incon- sistency of Bradley's metaphysics with his doctrine of the judg- ment, or the untenability of the latter nullifies the value of his criticism of absolute idealism. That criticism has sufficient foundation in the discrepancy shown to exist between many of the various existent appearances which reality presents in different human experiences. In showing that the older ab- solute idealism is irrational, and therefore not valid, he has per- formed his major service, and one which is not affected by what he has to say about his own view. But it is important to note that Bradley's sceptical conclusion does not necessarily hold for one whose presuppositions are different. Bradley in presupposing the doctrine of absolute ideaHsm, that all reaUty is included in one conscious experience and is identical with that experience, necessarily presupposes at the same time that doctrine's presuppositions, viz. psycho- logical idealism and logical idealism. The logical idealism is eliminated in the end ; reality, it is assumed, is not what is illogi- cal, self-contradictory; but, it is finally concluded, neither is it what is logical. It is not even known by means of, much less made up of, logical ideas. Rather is it immediate content of one all-inclusive conscious experience, nothing more. Now it is noteworthy that it is in connection with the initial combina- tion of metaphysical monism with psychological and logical idealism that the principal difficulties Bradley mentions with regard to substances, quafities, relations, and the rest, present themselves. Following the initial form of the thought, we have, in effect, the following argument. Relations are constituted by the thought which perceives or knows them. There can therefore be as many relations between relations as can be thought — in other words, an unending series. But if relations, 1 Op. ciL, pp. 542-7. 152 THE PROBLEM OF KNOWLEDGE as they seem to be, are real, independently of human thought, there must be as many in reaHty (the Absolute Experience) as could be constituted by thought in an unending series of mental acts — in other words, an actual infinite number, which is self-contradictory.^ Now the obvious thing to do here is to retrace one's steps, in order to find where one went so badly astray as finally to be led into the self-contradiction. It would be found that if the fallacious doctrines of idealism had been avoided, the self-contradictory conclusion would never have been forced upon the thinker. But Bradley does not choose to part with his idealistic presuppositions just yet, but to give up instead the common-sense doctrine that relations are real independently of finite thought. Reality, then, it is con- cluded, is non-relational, and therefore also non-rational. Thus the logical idealism drops out of Bradley's absolutism finally. But it has been assumed all along that reality is not irrational, self-contradictory. Bradley, therefore, since he refuses to retrace his steps again, goes on to overcome this opposition between the assumption that reaHty is not irrational and the conclusion that reality is not rational, by postulating the super- rationality of reality. If, however, when first forced to retrace his steps, Bradley had eliminated the fallacious idealism from his premises, and had thus been led to regard relations, not as thought-products (except in the case of tertiary relations) ,2 but as phases of reality of which there may be immediate ex- periences, or of which ideas (thought-products) may be formed, he would have found nothing contradictory in supposing them to be either presented in the experience or represented in the thought of the most comprehensive consciousness which really is. Such an experience, however, would not be identical with reality, nor would it necessarily include all possible experiences of reality. It may be thought of as including only such ap- pearances of reality as are necessary for the reaUzing of certain superhuman purposes. But whether such a superhuman experience exists or not is another question; it is not under debate in this volume. And what is true of relations is true of other elements of human experience relegated by Bradley to the realm of mere appearance, 1 See pp. 462-70, infra. 2 See Ch. XIV, infra. THE OLDER ABSOLUTE IDEALISM 153 because of the contradiction involved in the "infinite regress." This unending regress is primarily due not to the attempt to state the one Absolute Experience in terms of human thought, but to the attempt to state all reality as mere contents of one conscious experience. From an essentially realistic and moder- ately pragmatic point of view there can be not only a true and adequate representation of experience in judgment, but an ade- quate experience and representation of reality without the un- ending regress. If, then, a realistic point of view were once established, Bradley's baffling paradoxes would largely dis- appear; and with the adoption of an essentially pragmatic criterion of truth, such of them as might still threaten would be easily and happily avoided. CHAPTER VIII The Newer Absolute Idealism In the wake of Bradley's destructive criticism of the older absolute idealism there have appeared several attempted re- constructions of that philosophy, of which those of Bernard Bosanquet, A. E. Taylor, and W. E. Hocking may be taken as fairly representative. They embody the intellectualistic, the voluntaristic, and the mystical emphasis, respectively. The first two expHcitly take account of Bradley's work and give quite favorable consideration to some of his most characteristic views. They claim, however, that, beyond what is retained by Bradley, certain of the most essential elements of absolute ideaUsm proper can find place in the new construction. The third of the three philosophers mentioned has probably been influenced by Royce more than by Bradley ; and yet his phi- losophy stands in a peculiarly interesting relation to that of the English philosopher. While Bradley maintains that the only conceivable knowledge of Absolute Reality would have to be an immediate or mystical intuition, which, however, he regards as humanly inaccessible, Hocking claims that this intuition of the Whole is in principle present in all human consciousness, and especially in the religious experience of the mystic. Royce's voluntaristic philosophy, although given to the world in its more finished form later than the first pubhcation of Bradley's Appearance and Reality, and although itself a newer absolute idealism, as compared with the intellectualistic type, has nevertheless been regarded here as pre-Bradleian, inasmuch as it does not take seriously the difficulties raised by Bradley against the possibility of a conscious experience which is ra- tional and at the same time inclusive of all finite experiences without modification. Among the different attempts to rehabihtate absolute ideal- ism, the one which keeps closest to Bradley's own position is 154 THE NEWER ABSOLUTE IDEALISM 155 that of Bosanquet. Not only does he emphasize with Bradley the rational criterion of non-contradiction as a test of reality as distinguished from appearance ; he also accepts the negative results of the Bradleian criticism of the "thing," the "self," and of the moral and religious consciousness, takes over the doctrine of degrees of reality, and even expresses amazement at the unfavorable reception accorded to Bradley's doctrines by philosophers generally.^ Unlike Bradley, however, Bosan- quet emphasizes the positive residue of idealistic doctrine which seems still tenable, and undertakes to develop this residue into a sane and sufficient philosophy of reality.^ Bosanquet is more favorable than is Bradley to the Hegelian principle that the real is the rational. Unlike Green and others, who put their emphasis upon the predicate, maintaining that the real is the rational, in the sense of being what it does not seem to be, viz. constituted of thought-relations, Bosanquet places his emphasis upon the subject, insisting that it is reality that is rational. In other words, it is the actual — the absolute reality which is everything and with which we are in immediate relation in experience — which is rational, at least in the sense of being free from all self-contradiction.^ This realistic ten- dency in Bosanquet's thought finds especially congenial the Hegelian notion of the "concrete universal," which, it is in- sisted, means the self-complete and harmonious individual, discoverable through rational criticism of what is given in experience. Of course our fragmentary experience has to be supplemented by thought, which is able to trace out the reality in so far as it transcends what is actually given.^ As distinct from generality, which is sameness in spite of the other, and whose test is the number of subjects which can share a predicate, universality is sameness by means of the other, and its test is the number of predicates which can be attached to the subject. 1 The Principle of Individuality and Value, 1912, p. 57 ; cf . p. 40. 2 76., p. 30. 3 76., pp. 27, 41, 51, 378. "It is possible," the author significantly observes on p. 39, "that those philosophers may prove to hold the more suitable lan- guage who deny that thought can ever be one with the real. But at any rate, we are bound to follow thought . . . towards ... a fuller perfection in the certainty that if it is itself a vanishing form, it will point us the way to what lies beyond, and when necessary, introduce us to its nature." * lb., pp. 55, 257-8. 156 THE PROBLEM OF KNOWLEDGE The true embodiment of the logical universal is not an abstrac- tion, but an individual, a self-complete world. Ultimately, indeed, it is the Individual, a world whose members are worlds.^ But the system is not so free from the fallacious ideahstic analysis and consequent confusions as might be supposed from these leanings toward realistic forms of expression. The fundamental view is that reality is experience ; ^ truth means nothing different from reality ; ^ since the subject of all predica- tion is Reality, and since there are no ideas which do not quahfy this subject, ^'it follows that the truth of the ontological argu- ment is conceded in principle'';^ on the one hand "nature . . . exists only through the finite mind," ^ and matter, taken as independent non-psychical existence, is a substantiated abstraction,^ while on the other hand, thinking is in essence simply a change in a being or content, viz. its passing beyond itself,^ and inwardness is to be interpreted as meaning simply inseparable continuity,^ "All objects of the mind," it is roundly declared, "are psychical. But some are physical as well; that is, some enter into a determinate context of reactions, which forms a special part of the psychical world, which we call the physical world and contrast with the psychical. But this is an abstraction, for the physical world can never, in the last resort, put off its psychical character. A tree is beautiful and green and tall. All these qualities are, as presentations, necessarily psychi- cal ; but the tallness at least, as a character of a thing in space, is certainly physical. And this is probably the true line of de- marcation. They are all, as we said, psychical ah initio as presentations. But qua determined by a construction of ob- jects in space they all (including 'physical' beauty) become physical also. Then they are relatively opposable to the psychical. But not more than relatively. For, taking as the test of psychical nature the being destroyed if the percipient mind were destroyed, it is plain that in a degree, though only in a degree, presentations remain psychical not only as pure pres- entations, but even as qualities of spatial objects. The sub- jective mind, which has perceived and which conceives them, 1 The Principle of Individicality and Value, 1912, pp. 37, 68. ^ 75,^ p. 39. ' lb., p. 41. -• lb,, p. 80. 6 jb., pp. 359, 371. « lb., p. 73. 7 lb., p. 60. » lb., pp. 73-7. THE NEWER ABSOLUTE IDEALISM 157 being destroyed, their existence would certainly be pro tanto diminished, though not necessarily annihilated. A physical object must at least be capable of becoming psychical at any moment. If not, it so far has not full existence." ^ What we have here is not the mere outcome of a faulty analysis, with its ignoring of the difference between the object known through perception and thought, and the sense- and thought- elements through which it is known ; we have an apergu carefully preserved for its convenience in making the transition from the preliminary realistic interpretation of experience ^ to the view that the Individual which is the Whole is a single all-inclusive and absolutel}^ self-consistent experience.^ The concept of subject, while not ultimaiely true, is vaUd as a substitute for that of substance.'* But while holding that the Individual is mind,^ we must not fall into the snare of pluralism, a temptation to which we are especially exposed because of the ineradicable superstition that finite minds are substances.® We must remember that the true nature of mind is a world of experience.'' Things are not mind-dependent, but mind-component.^ In- stead of pluralism Bosanquet offers multiplicism, the view that there are various levels of experience, each possessing its peculiar range and area,^ the highest being the Absolute Experience which is identical with Absolute Reality.^^ This ''multiphcism," which corresponds to Bradley's doctrine of degrees of reality, is the conclusion to which Bosanquet is driven by his acceptance of the main results of the Bradleian criticism, together with his determination to cling to Hegelian- 1 76., p. 361. 2 cf. The Distinction between Mind and Its Objects, 1913. ' The Principle of Individuality and Value, pp. 56, 386. ^ 76., p. 284. 5 76., p. 286. 6 76., pp. 372-3. ^ 75.^ p. 287. 8 The Distinction between Mind and Its Objects, p. 42. 8 Principle of Individuality and Value, p. 373. 10 In The Value and Destiny of the Individual, 1913, pp. 59-60, Bosanquet offers a dialectical argument for a certain phase of this view, as follows : Thesis: "What we call individual finite beings are kept apart by differences of quality of feeling and also by the reciprocal shortcomings of the content of which they are composed. These differences of quality, and these shortcomings, are often held to be the secret of individuality, the secret by which I am myself and not another, because I have not his immediate feeling, and do not compre- hend his capacities within mine." Antithesis: "When I most fall short of others, and am most in discord of feeling quality with them, I am also least myself." Synthesis: "We do not experience ourselves as we really are." 158 THE PROBLEM OF KNOWLEDGE ism, rather than to adopt a bona fide reahsm. But that it is self-refuting, by Bosanquet's own principle of non-contradic- tion, can be readily shown. There is inconsistency, on the very face of it, with the monistic fact that all experience is ultimately one experience. The appearances which constitute the contents of the lower levels of experience are at least psychically real. According to Bosanquet they are both included in the Ab- solute Experience, because they are real, and at the same time excluded therefrom, as mutually conflicting appearances. An attempt is made to cover up this contradiction by appealing to the way in which the elements of our experience are trans- muted by every change of work and of scene ; so, it is claimed, the experiences of conscious units are transmuted, reenforced, and rearranged by entrance into the fuller and more extended experience of the Absolute.^ Hence 'Hhere is no reason for making . . . the transmutation of experience in accordance with the law of non-contradiction ... a fundamental diffi- culty when we come to deal with fundamental reality. . . . The Absolute is simply the high-water mark of fluctuations in experience, of which, in general, we are daily and normally aware." ^ But in criticism of this it may be pointed out that there is a difference between the two cases which destroys all the value of the analogy. The finite mind does not retain the past inadequate experience along with the present more adequate one ; the latter is a substitute, which cancels the former. The Absolute, however, as all-inclusive experience, must retain both human experiences, the earlier experience with the inadequate appearance which is its content, and the later experience with its more adequate appearance which cancels and banishes the former. In other words, when the appearance of an object changes, there is a substitution of one experience for another, and the former experience is gone beyond recall; even apart from the lapse of time, it is an experience which could no longer exist along with the other. But still, as an experience, it was as reall}^ existent as the later one, and so would have to be included along with the other in an all-inclusive Absolute Experience. Bosanquet is right as against Royce, when he contends that the inclusion together of the two experiences would modify the earlier 1 The Principle of Individuality and Value, pp. 372-3. ^ 75.^ pp. 377-8. THE NEWER ABSOLUTE IDEALISM 159 one ; ^ but as against Bosanquet himself it must be urged that the whole reality of an experience is in its actuality, and when this is 'transmuted" it is no longer reality, but a departure from it. On this showing the Absolute, as inclusive of all reality, could not be an experience and nothing more. Hence we feel justified in regarding Bosanquet's rehabilitation of absolute idealism as revealing only the more plainly how complete has been the wreck made of that once respectable philosophy by the Brad- leian criticism. If the Absolute, as Reality, were recognized as itself not a mere experience, however unified, but as a reality of which, while we have inadequate experiences, some Being may have an adequate experience, then might our philosopher be permitted to say, ''We experience the Absolute better than we experience anything else, because ... we experience the Absolute in everything "; ^ but so long as the fundamental dogma of idealism, that reality is idea or experience, is retained, a finally self -consistent philosophy seems unattainable. A. E. Taylor has been deeply influenced by Bradley's ra- tionalistic critique of rationalistic or intellectualistic al)solute idealism, and he retains the Bradleian emphasis upon the rational criterion, "Reality is not self-contradictory";^ but like Bosanquet he seeks to save as much as possible from the general wreck, and like Royce he has recourse to, and makes fundamental, the concept of purpose. Indeed, for our present purposes Taylor's system may be regarded as essentially a synthesis of the views of Bradley and Royce, and yet he makes his appeal to purpose with a difference. Royce always main- tains that reality, even in its most "external" aspects, is what is meant or purposed ; for Taylor, reality, at least at the outset of the investigation, is simply that with which our purposes are everywhere confronted, that "of which all purposes, each in its own way, must take account." * Thus while Royce identifies reality with the idea, the predicate of our judgment, Taylor identifies it with the subject of our judgment, as he interprets it, viz. with the bare immediacy of psychical experience. "The real is experience, and nothing but experience, and experience 1 lb., pp. 387-8. 2 76., p. 378. 3 Elements of Metaphysics, 1907, Bk. I, Ch. II, § 1. Journal of Philosophy, II, 1905, p. 315. NEO-REALISTIC DOCTRINE OF QUALITIES 257 virtually projected) object has because of the influence upon the brain of something other than the real extra-organic object. In other words, these latter qualities are to be regarded as mere appearances, which lose whatever reality they had with the vanishing of the perception. ^ The meaning here seems clearly to be that the only objects of perception are the ''simu- lacra of extra-cerebral objects" contained by the brain,^ but perceived as virtually extra- organic, or projected, i.e. perceived as if they were where they are not. This seems to be a posi- tion very closely approximating epistemological dualism, and logically involving agnosticism with reference to all beyond projected cerebral simulacra. If the simulacra were such as could be actually projected, it would seem too much to say that we have any experience of the physical, even of the cere- bral simulacra themselves; moreover, since ''we could not infer the physical unless we experienced the physical," ^ the result would be absolute dualism and complete agnosticism with reference to physical reality (and therefore, according to Montague's own view, agnosticism with reference to all reality) . But what Montague means is that the simulacra are physical, and that the act of projection is "not an actual act," but "a virtual act" ; "the world we perceive is (not indeed an actual but) a virtual or potential reprojection of the effects which the world projects upon us." ^ In this case, since the cerebral simulacrum is not actually projected, we may perhaps be allowed to say that we perceive a part of the physical, viz. a part of the cerebral ; but even so, remembering that there is no actual projection, we would be, to use Montague's own language, "reduced to the wretched status of an intra-cranial solipsist." For — let his words be repeated — "if we cannot get beyond our own brains in immediate perception, we cannot get be3^ond them at all." ^ According to his own statement, this physiological solipsism can be avoided only by accepting his special theory of consciousness,^ and to an examination of that theory we must turn in a later connection; but unless 1 lb., IV, 1907, p. 383 ; V, 1908, pp. 211-12. 2 Philosophical Review, XXIII, 1914, p. 61. 3 Journal of Philosophy, I, 1904, p. 294. * Philosophical Review, XXIII, 1914, p. 62. 6/6., p. 61. «/6. S 258 THE PROBLEM OF KNOWLEDGE he should make good his case at this second trial, we must regretfully see him remanded to his narrow ^'intra-cranial" prison, from which his only possible escape must be merely verbal, that is, to use his own term, virtual, and not actual. In such a case liis monistic or '^new'' realism, one might also say, would be merely virtual, and not actual. CHAPTER XII The Neo-Realistic Doctrine of Consciousness The intimate relation between the neo-realistic doctrine of secondary qualities and the views held by the new realists as to the nature of consciousness is well indicated in the words of Montague: ''As long as the secondary qualities are accepted as objectively [i.e. what we would call independently] real, there is no temptation to regard consciousness as anything but a relation." ^ In dealing with the theories of consciousness which are current in the English and American neo-realistic schools, it may be well to note that although the doctrine of the independent reality of secondary qualities may often be found ostensibly resting upon the view that consciousness is an "external" relation, the actual dependence is probably in the main in the opposite direction. We shall turn our attention first to the English school. Here it seems easy to detect the influence of the distinctions made by Shadworth Hodgson in his first presidential address as first president of the Aristotelian Society in 1887. In answer- ing negatively the question, ''Is mind synonymous with con- sciousness?" he indorsed (although from the standpoint of the philosophy of pure experience, i.e. disguised psychological idealism, rather than from that of natural realism) the distinc- tions involved in the common-sense assumption that conscious- ness is some one's consciousness of something. Res cogitanSj cogitatio, and cogitata must, he insisted, be carefully and con- stantly distinguished. As against the confusing idealistic identification of knowing with knower, he maintained that mind is a subject of attributes, and consciousness an attribute of that subject, a knowing and not a knower.^ In later ad- dresses he characterizes mind as that which we perceive as the 1 Journal of Philosophy, II, 1905, p. 315. 2 Proc. Aristot. Soc, 1887-8, pp. 6, 7, 23. 259 260 THE PROBLEM. OF KNOWLEDGE subject of consciousness, and matter as that which we perceive as the object of consciousness.^ This suggested total identi- fication of matter with the object of consciousness, and of the object of consciousness with matter, reveals on the one hand the lack of a bona fide reaHsm in Hodgson's own philosophy. (He indorses the view that not only does all consciousness reveal Being, but that all Being is revealed in consciousness. 2) But on the other hand the identification referred to seems to explain the direction taken by the later studies of the Enghsh new realists, in which the problem is as to what consciousness can be, if the object of consciousness is always in its entirety independently real. What seems to us the most defensible feature of Hodgson's doctrine of consciousness, viz. the distinction between con- sciousness and its subject (mind) on the one hand, and its object on the other, reappears in realistic form in the writings of L. T. Hobhouse and W. McDougall. The former speaks of perception as an act of consciousness referring to the object perceived, so that, as such, it is the mind's own creation. The perception, or assertion, as mental event, is to be distinguished, according to Hobhouse, from the content, as fact perceived or asserted.^ McDougall, in his work entitled Body and Mind: A History and Defense of Animism,'^ describes ''the soul" in a way that seems most obviously to imply an essentially identical view of consciousness. A soul, he says, is "a being that pos- sesses, or is, the sum of definite capacities for psychical activity and psycho-physical interaction, of which the most funda- mental are (1) the capacity of producing, in response to certain physical stimuli (the sensory processes of the brain) the whole range of sensation quahties in their whole range of intensities ; (2) the capacity of responding to certain sensation-complexes with the production of meanings ; . . . (3) the capacity of respond- ing to these sensations and these meanings with feeling and conation or effort, under the spur of which fxu-ther meanings may be brought to consciousness in accordance with the laws of reproduction of similars and of reasoning ; (4) the capacity of reacting upon the brain-processes to modify their course in a 1 Proc. Aristot. Soc, 1891-2, p. 4. 2 j&., p. 52. 8 The Theory of Knowledge, 1896, pp. 531-4. * 1911. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 261 way which we cannot clearly define, but which we may pro- visionally conceive as a process of guidance by which streams of nervous energy may be concentrated in a way that antago- nizes the tendency of all physical energy to dissipation and deg- radation." ^ But neither Hobhouse nor McDougall develops much further the implications of this general position ; indeed, the latter seems in a later work ^ to be tending in the direction of the more typically neo-realistic doctrine of the behaviorist psychologists. '^ Consciousness," he says, ''is an activity of some being which, in all cases of which we have positive knowl- edge, is a material organism, but to which we may conven- iently give the general name, subject." ^ To an examination of this later view we shall therefore have to return in another connection. The typical English neo-realistic view of consciousness is best found in the writings of G. E. Moore, B. Russell, S. Alexander, and T. P. Nunn. Moore's first utterances on the subject show him already beginning to develop Hodgson's doctrine in the realistic direction. Experience, as a kind of cog- nition, stands, he claims, for a double fact, viz. a mental state, and that of which this mental state is cognizant.^ He evi- dently holds, however, that it would be absurd to suppose that the mind could give properties to things.^ The problem then comes to be what consciousness is, if it does nothing to its object. Sensation, it is averred, is "a case of knowing, or being aware of, or experiencing something." ^ To be aware of the sensation is not to be aware of its content, but to be aware of the awareness of a sense-content.'^ But, it is confessed, *'when we try to introspect the sensation of blue, all we can see is the blue ; the other element is as if it were diaphanous." ^ But since it is insisted that the observation of a perception of red is altogether different from the perception of red,^ the problem becomes acute as to just what consciousness, which has been supposed to be the special "subject-matter of psychol- ogy," really is. Moore seems to remain in doubt as to whether 1 Body and Mind, p. 365. 2 Psychology : The Study of Behavior, 1912. 3 Op. cit, p. 60. 4 Proc. Aristot. Soc, 1902-3, p. 82. 8 lb., 1903^, p. 135. 6 Mind, N.S., XII, 1903, p. 449. 7 76. 8 ib.^ p. 450. 9 Proc. Aristot. Soc, 1905-6, p. 104. 262 THE PROBLEM OF KNOWLEDGE the mind itself, as the subject of mental acts, or any objective contents of mental activities, even sense-data, should be in- cluded in the subject-matter of psychology.^ He would include without hesitation, however, as undoubtedly mental or psychical entities, all mental acts, all qualities distinguishing mental acts from each other, and all collections of mental acts.^ As criteria of the mental the following are offered : it must be an act of con- sciousness ; it must belong to some person, or mind ; and, finally, it must, perhaps, be directly known to one person only.^ There is much that is suggestive in this view of Moore's, and not a little that will be retained in the view to be defended in a later chapter; but it must be evident that, although consciousness may very well be mental or psychical activity, it is difficult to conceive how any activity can either be ''diaph- anous" or exist without producing any change in any of the quahties of the object toward which it is directed. Little wonder, then, that at least a tendency should be manifested, to include within the subject-matter of psychology, although inconsistently, more than this neutral sort of entity to which consciousness has been reduced. Russell has thus far had comparatively Httle to say about consciousness, but he accepts the general realistic doctrines that knowing is a relation which is external to its object,^ and that only the mental act, and not the thing apprehended, is conscious.^ Similarly, Nunn subscribes to the view. that in perception there is no psychical intermediary "on the object side" between the subject and the independently real thing.^ More elaborate attention to the nature of consciousness is found in Alexander's articles ; but whether he has succeeded thereby in throwing further light upon the problem may well be doubted. He adopts the view that consciousness is mental activity, conation, which differs in different cases only in its direction toward different objects of perception or thought.^ 1 Proc. Aristot. Soc, 1909-10, pp. 51-7. 2 lb., pp. 36-51. 3 lb. * Journal of Philosophy, VIII, 1911, pp. 159-60. 5 Problems of Philosophy, p. 65. Acquaintance, according to Russell, is a dual relation between a subject and an object, which need not have any community of nature. The subject is " mental," while the object is not known to be mental, except in introspection (Monist, XXIV, 1914, p. 1 ; cf. pp. 4 and 435-53). ' ^Proc. Aristot. Soc, 1909-10, p. 202. ^ lb., 1907-8, pp. 216, 219-22. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 263 Later he acknowledges that this mental activity is accompanied by feeling, and this, like memory (the existence of the past),^ error and imagination, is, he confesses, an unsolved problem from his point of view.^ Still later it is admitted that con- sciousness, or knowing, not only is accompanied by feeling and varies in direction, but also that it varies in intensity and complexity.^ Still, we would remark, if this is all that con- sciousness is, it is difficult to understand how there can be in ''acts of consciousness" enough subject-matter for the science of psychology, "^ and on the other hand how the object can be ''saturated'' and even "vitiated'' by suggestions and infer- ences — "elements introduced into it by the mind." ^ But Alexander has other thoughts concerning the nature of consciousness. Mind consists of mental activity, which, he asserts, is located in the body ; it is a not purely physiological function of the body,^ a fortunate functional variation in the course of evolution, through which we are enabled to learn the characters of things.'^ But the identification of mind with body is carried still further. "We are directly aware," we read in a recent article, "that our mind and body are one thing, because we experience them in the same place." ^ Suitably to this conception of mind as a physical thing, there is the definition of consciousness, or knowing, or having experi- ence, as the mere " compresence," or togetherness, of two things, one of which is, and the other of which is not, a mind, i.e. a body with the empirical character of being conscious.^ This 1 Later this is explained (?) by the dogma that memory "renews the past condition," ih., 1910-11, p. 21. 2/6., 1907-8, pp. 251, 254; ih., 1908-9, p. 6. 3 Ih., 1910-U, p. 18. " Ih., p. 7. 5 75., 1909-10, pp. 19, 28. «76., 1907-8, pp. 223-4; ih., 1909-10, p. 6. 7 Mind, N.S., XXI, 1912, p. 17. ^ Ih., p. 8; cf. Proc. Aristot. Soc, 1910-11, p. 17, A rather important sidelight upon Alexander's position here is found in the following sentence : "In my own case mental activity, especially in thinking ... is accompanied by marked movements of the eyes, which are apt to change their position with each change of the thought, and whose movements, in fact, I use as a means of directing thought in different directions and controlling it " {Proc. Aristot. Soc, 1907-8, p. 216). He tells us that in all his mental conditions he is aware of movements in different directions (ih., p. 219) ; and finally he speaks of this direction of the eyes as "mental direction," and the only thing to distinguish one thought-process from another (ih., p. 220). 9 Mind, N.S., XXI, 1912, pp. 2, 318. 264 THE PROBLEM OF KNOWLEDGE is manifestly a circular definition ; to define consciousness as the relation characteristic of a conscious body with another object is not to define it at all. A similar fault characterizes the state- ment, "knowledge of an external thing ... is the thing itself in the various ways in which it reveals itself to the mind." ^ G. F. Stout, a recent convert to English neo-realism, has furnished, apparently without fully realizing it, what may be regarded either as a reductio ad ahsurdum of the doctrine that consciousness is a purely external relation, so far as the object is concerned, or else as the antithesis to the thesis that con- sciousness is a relation, leading of necessity to the synthetic judgment that consciousness is a productive activity. In his paper entitled '' The Object of Thought and Real Being " ^ writ- ten under the spell of Meinong's " Gegenstandstheorie," he dis- cusses the implications of the innocent-looking proposition that "whenever we think of anything we think of its having a being which does not merely consist in its being thought of." ^ "It seems to involve an absurdity," he continues, "to suppose that what I think of has no being except the being thought of. For how can the being of anything be merely constituted by its being related to something else?" Indeed, "when I believe, or disbelieve, or suppose, that a centaur actually exists, I must think of its actually existing." ^ In no case is "the possible severance of what really is and what is thought " to be admitted.^ Generalities, alternative possibilities, and even non-being (de- fined as otherness), since they are objects of thought, are real independently of thought.® Now, in order to defend himself against the charge of having reduced his own realism to absurdity, Stout would probably fall back upon Meinong's and Russell's distinction between existence and " subsistence." But in the paper itself practically no use is made of this distinction. The discussion throughout is in terms of what " really is." Here Stout is like Montague, who says that all relations, including consciousness as a relation, presuppose that their terms exist ^'"^ and like M. R. Cohen, who regards the distinction between existence and subsistence as ^Proc. Aristot. Soc, 1910-11, p. 19. ^ lb., 1910-11. 3 lb., p. 187. *Ib. 6 lb., p. 188. 6 lb., pp. 192, 198-9. "f Journal of Philosophy, II, 1905, p. 313. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 265 "merely a temporary or provisional makeshift," for which he would substitute the idea of a non-mental and non-physical ex- istence} But, from our point of view,^ it is not necessary, ulti- mately, to interpret '^ subsistence " otherwise than in terms either of existence, physical or mental (i.e. as mind or as depending on mind), on the one hand, or of non-existence on the other. Nothing " really is," or has " real being," save what exists and as it exists, whether physically, or mentally, or in some other form of existence, if there be any other, — and we have no right to say there is any other, until it has been empirically discovered.^ Similarly ''generahties " exist, but only in particular things which exist, or as abstract ideas in existent minds. ''Alterna- tive possibilities " again, except in the case of what depends upon free, i.e. not completely predetermined, activity, is to be reduced either to existence or to non-existence, present, past, or future, by overcoming our ignorance. And even in the case of what depends upon not completely predetermined action, we need no other categories, ultimately, besides time, existence and non- existence ; only, until the action has taken place, no one can tell which category to employ in certain cases, whether that of (future) existence or that of (future) non-existence. From our point of view, then, the moral of Stout's train of reasoning ought to be clear. If consciousness is a mere external relation of one existent object to another, so that the object of consciousness is necessarily existent independently of con- 1 /&., X, 1913, p. 199 ; XI, 1914, p. 626. 2 Cf. pp. 84-8, 202-6, 231, supra, and pp. 302-6, infra. 3 A. Wolf's remarks on the " explicit use of terms" in order to avoid confusion are very much to the point. " There is only one world of reality," he says, " and whatever is real is in it. What does not exist in the real world does not exist at aU. A material object cannot exist as a mental process, nor can a mental pro- cess be a material object. To say that a centaur exists in intellectu is simply to use the word centaur elliptically instead of * the idea of a centaur.' Similarly to assert the existence of a centaur ' in the world of mythology,' is to use the word centaur instead of 'an account of a centaur.' . . . Existing ideas of a thing, or existing accounts of a thing, all these are as such real enough ; but their reality is a very different thing from the reality of the thing itself. If the thing itself is not real, then no real ideas of it, no real descriptions of it can as such make it real. But in that case to speak of it as having logical existence, or empirical existence in some other than the real world, is simply a mysterious way of asserting the reality not of the thing in question, but of something quite dif- ferent "(r/ie Existential Import of Categorical Predication: Studies in Logic, 1905, p. 48; cf. p. 160). 266 THE PROBLEM OF KNOWLEDGE sciousness, then we cannot think of anything which does not exist independently of our thinking of it. But this is absurd ; wherefore consciousness cannot be a mere external relation, so far as the object is concerned ; it must be to some extent a productive activity, so that it can be a relation to that which depends upon itself (consciousness) for its existence. But, in any case, from any point of view. Stout's doctrine is involved in unavoidable final self-contradiction. Non-being, defined as other than all that is real, would have to be regarded by him as real. Here we have again the paradoxical " impossible objects," with regard to which Meinong and Russell are unable to come to agreement. The idea of a productive or creative psychical activity, to which Stout ought to have been led by the dialectic of his thought, had already been given a partial, but insufficient, applica- tion by A. Wolf. In criticism of Alexander's doctrine that the self is made up of transparent acts of consciousness, conation without qualitative differences, he urges that this view is applicable only in the case of normal perception. With regard to imagination, memory, and abnormal perception, it is suggested that the representative theory be applied ; here con- sciousness is a content-process, i.e. an activity in which both process and content are mental, an activity that has in it something of the nature of production, creation, and, at the very least, distortion.^ This is moving in the right direction, but it affords no point of stable equilibrium. The conscious processes in normal perception and in hallucination are, as processes (apart from their antecedents on the one hand and their independent objects on the other, neither of which are parts of the processes in question), essentially identical in kind. If there is creativeness in the one, there is creativeness in the other ; if there is none in the one, there is none in the other. The point of transition from the English neo-realistic doc- trine of consciousness to the more extreme view characteristic of the American School is nowhere better expressed than by WilHam James in his almost epoch-making essay, " Does Consciousness Exist ? " ''I beUeve," he says, referring to G. E. Moore's doctrine of consciousness as a diaphanous activ- 1 Proc. Aristot. Soc, 1908-9, pp. 164-5, 169-71. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 267 ity, '' that ' consciousness/ when once it has evaporated to this estate of pure diaphaneity, is on the point of disappearing altogether. It is the name of a nonentity, and has no right to a place among first principles. Those who still cling to it are clinging to a mere echo, the faint rumor left behind by the disappearing 'soul' upon the air of philosophy." ^ As a radi- cal empiricist he holds that "knowing can easily be explained as a particular sort of relation towards one another into which portions of pure experience may enter. The relation itself is a part of pure experience ; one of its ' terms ' becomes the sub- ject or bearer of the knowledge, the knower, the other becomes the object known." 2 Thus while consciousness is not an en- tity, it is a function discharged by certain elements of pure experience with reference to others,^ and is therefore to be regarded as a relation between these two elements of experi- ence, purely external so far as the objective or represented ex- perience is concerned. "^ Now James is right enough, we may concede, in maintaining that this is what consciousness would be in a world of pure experience; but it is another question whether the formula will still hold when the world of pure experience is translated into a realistic world, existing inde- pendently of its being known or experienced. This, however, is what the American neo-realists have tried to maintain. W. T. Bush, who, as we have seen, occupies a somewhat transitional point of view, indorses James's "dropping of con- sciousness as a metaphysical concept." ^ His own solution of the problem of consciousness is that consciousness is that content of pure experience which is the essentially private and unsharable experience of one person.^ But this is a plausible suggestion only until one begins to divide up the contents of pure experience into the conscious and the non-conscious. Pleasure and pain and all organic feelings, including kin- aesthetic sensations, would presumably be within conscious- ness, while colors and sounds would be in the non-conscious realm ; but what about the taste of a particular morsel on the 1 Journal of Philosophy, 1, 1904, p. 477 ; Essays in Radical Empiricism, p. 2. 2 Essays in Radical Empiricism, p. 4. 3 /5_^ p, 3^ 4 j^^ pp_ 23, 25. 5 Avenarius and the Standpoint of Pure Experience, p. 75 ; cf . Essays . . . in Honor of William James, 1908, p. 102 ; Journal of Philosophy, X, 1913, p. 534. ^ Avenarius and the Standpoint of Pure Experience, pp. 75, 77. 268 THE PROBLEM OF KNOWLEDGE tongue of some individual? Is that not unsharable? But is it permissible to draw any rigid line through sense-contents, marking off the objective from the subjective? If there were but one color-blind person, his visual experience would be un- sharable. Would it on that account be a conscious experi- ence ? And then if a second color-blind person should come into existence, would the visual experience of the first one suddenly cease to be conscious ? The only basis upon which the validity of such an accidentally shifting line of division could be justified would be the frank admission that there is no distinction of essential importance between the conscious and the unconscious. G. S. Fullerton, taking his writings as a whole, is also transi- tional between immediate empiricism and the new realism. In his System of Metaphysics he says, as the pure empiricist naturally would, that one's consciousness of the world and the world of which he is conscious, both exist, as symbol and thing symbohzed, within consciousness. The real external world, the thing symbolized, is a complex of conscious elements; and consciousness, the symbol, is a compound of sensational and imaginary elements, the latter largely predominating.^ The self is not a substantial substratum of conscious states, but a content of conscious experience.^ A few years later we find Fullerton defending the common-sense doctrine of the eject, arguing that while each of us knows directly his own thoughts and feelings, he is not conscious in the same way of the thoughts and feelings of others, and that it is by the bridge of an analogi- cal argument that he is conducted to them.^ This seems to be transitional between the disguised psychological idealism of the System of Metaphysics and the new realism of The World We Live In. Consciousness is being eliminated from the basis of aU reality, by being interpreted in terms of the eject. Sub- sequently we find that while the external world is still spoken of ambiguously as phenomenon, it is regarded as external to and independent of consciousness. Minds, it is maintained, are phenomena also,* although there seems to be no clear state- ment as to just what are the criteria of the mental status of 1 System of Metaphysics, 1904, pp. 114-15. 2 75.^ p. 280. 3 Journal of Philosophy, IV, 1907, pp. 506-7. * The World We Live In, 1912, pp. 85-6, 153, 156. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 269 any phenomenon, except that mental phenomena are *' ac- counted for by taking into consideration what happens to the body," while '4n the case of physical phenomena the relation to sense is ignored." ^ We can scarcely be said, therefore, to have a definition of consciousness from Fullerton ; but in so far as there are indications of one, it would seem to be that consciousness is either that part of the contents of the phenom- enal world which depends for its existence upon something which happens in the nervous system, or else that part whose dependence upon such events is not ignored, or else again, per impossihile, both ! Whether the line between the physical and the mental is fixed or shifting, or whether it is in some inscru- table way both, is not made to appear. And the reason prob- ably is that he who tries to be a realist without giving up im- mediate empiricism, is confronted, in the problem of the nature of consciousness, with one of several problems that admit of no solution from his point, or points, of view. Among American neo-realists probably no one has given more attention to the problem of consciousness than F. J. E. Woodbridge. What looks like a key to the history of his thinking on this subject is to be found in his essay, "Percep- tion and Epistemology." If the world is "made only of the stuff of consciousness, then," he writes, "consciousness is the kind of stuff that may be condensed into a lump of sugar with which to sweeten coffee." ^ This, of course, is expressive of Woodbridge's strong reaction against idealism; but it is a significant fact that in the end he himself defines consciousness in terms of the physical alone. He has persistently stood for the application of the method of the empirical sciences to the problems of consciousness and knowledge ;3 but his report of results is that when we introspect we never find anything but things in certain relations to each other.^ Knowing, or con- sciousness, then, since it is not discovered as a thing, although belonging with things in the physical order,^ is to be defined as a real relation between things.® It is a purely external rela- ^ 76., p. 117. 2 Essays . . . in Honor of William James, pp. 160-1. 3 76., pp. 140, 157, 166. * Journal of Philosophy, X, 1913, p. 608. 6 76., II, 1905, pp. 119-20. 6 Philosophical Review, XII, 1903, p. 374 ; Journal of Philosophy, II, 1905, p. 125. 270 THE PROBLEM OF KNOWLEDGE tion, making no difference to its object ; ^ a relation of together- ness, at least. 2 In consciousness there is representation, but it is the representation of things by each other.^ But, defining more closely, this external relation, which consciousness or awareness is, is the relation of meaning, or implication, existing intermittently between the objects of experience.^ But has not the empirical investigator here allowed the object of in- vestigation, consciousness, to slip, as it were, through his fingers? Do we not constantly make the distinction between meaning and consciousness of meaning? Woodbridge's defini- tion of consciousness does not allow for this distinction, which is a perfectly valid and necessary one, especially from the realistic point of view. Woodbridge has suggested another definition of conscious- ness, from the point of view of external observation rather than from that of introspection. Before entering upon that, however, it will be convenient to mention here, rather than later, some recent psychological theories which have seemed in close accord with such realistic doctrines as that of Wood- bridge when he says that introspection reveals nothing but things. This doctrine suggests two consequences for psy- chology. The one is that images, or mental duplicates of things, are not to be looked for as necessarily and invariably conditioning conscious processes. The other is that intro- spection has been greatly overrated as a source of psychological information. The former view is represented by R. S. Wood- worth's article, ^'Imageless Thought"; ^ the latter, by Knight Dunlap's papers, ''The Case against Introspection,"^ and "Images and Ideas." ^ Woodworth asserts that ''meaning is 1 Science, N.S., XX, 1904, p. 598 ; Journal of Philosophy, VII, 1910, p. 416 ; Philosophical Review, XXI, 1912, pp. 637, 639. 'Journal of Philosophy, II, 1905, p. 120; " The Problem of Consciousness," in Studies in Philosophy and Psychology (Garman Commemoration Volume), 1906, p. 155. 3 Journal of Philosophy, II, 1905, pp. 121-2 : note the influence of William James here, with the characteristic difference due to the transition from pure empiricism to realism. * Garman Commemoration Volume, pp. 159, 160-2, 164 ; Psychological Review, XV, 1908, pp. 397-8 ; Journal of Philosophy, VI, 1909, p. 449. <* Journal of Philosophy, III, 1906, pp. 701-8. ^ Psychological Review, XIX, 1912, pp. 4:0^-12. '' The Johns Hopkins University Circular, No. 3, March, 1914. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 271 not felt as the relation between an image and an object, but as the thought of the object. . . . The thought of the object is not the image, for the image may change while the same object is thought of." 1 He claims, moreover, that imageless thought is an apparent fact of introspection.^ E. L. Thorndike supports him in this : because ''we can will acts, images of whose resi- dent sensations are unobtainable, . . . pragmatically . . . the image is an irrelevant factor." ^ B^t while Wood worth claims that introspection shows that there can be thought without images, Dunlap declares that there is not the slightest evidence for the reality of introspection in the observation of conscious- ness. ''Knowing there certainly is; known, the knowing cer- tainly is not." What is supposed to be introspection, the observation of the process of observing, he insists is really only the observation of certain muscular, visceral, and other sensa- tions.^ In his later article he takes the ground that attention to the direct content of thought reveals not images, but only muscle-sensations.^ Understood as this consciousness of muscle- sensations, then, there is an important place in psychology for ''introspection." ® We shall have to deal in a later connection with the place of imagery in thought, and with the possibility of introspection, but it may be remarked at once that this discounting of the reality and value of images and introspection has naturally been regarded as a minimizing of the distinctly psychical, a tendency to reduce it to the merely physical. Still, with regard to Dun- lap's report on introspection, much of what he says may be accepted readily enough : we shall ourselves contend that con- sciousness is an activity which is not apprehended in any case as a psychical (or psychically produced) element revealed by introspection, but only in a complex of muscular and other "sensations," and represented, or at least representable, by some 1 Journal of Philosophy, III, 1906, p. 707. « lb., p. 702. 3 lb., IV, 1907, pp. 40-2. 4 Psychological Review, XIX, 1912, pp. 410-12. 6 Johns Hopkins University Circular, No. 3, 1914, pp. 35-6. Of. S. Alex- ander's report concerning his own sense of direction of the eyes (v. supra) and that of William James concerning breathing as the only content of thought or consciousness revealed through introspection, Journal of Philosophy, I, 1904, p. 491. ^ Johns Hopkins University Circular, No. 3, 1914, p. 41. 272 THE PROBLEM OF KNOWLEDGE idea, into which visual, auditory, verbal, or other imagery enters. And in opposition to the views of Woodworth and Thorndike on imagery and thought, J. R. Angell expresses the view that thought-processes are often carried on by verbal imagery so highly schematized, compressed, and automatized as to escape identification.^ Moreover, after examining the data upon which the existence of " imageless thought " is based, he denies that any real evidence has been produced for the initiation or con- trol of voluntary movement entirely without sensory or imaginal supervision.2 But, be that as it may, it is highly significant that Woodworth has now come out with a paper entitled, " A Revision of Imageless Thought," ^ in which, while the existence of imageless thought is reaffirmed, a different interpretation of the alleged phenomenon is offered, and one that runs counter to the prepossessions of the more typical American neo-realists and behaviorists. Woodworth calls his present view the " mental reaction theory," or " perceptual reaction theory," the basic idea of which is " that a percept is an inner reaction to sensation." Following sensation after an interval too short to be detected introspectively, there comes this mental or percep- tual reaction, adding new conscious content '^ which cannot be analyzed into elementary sensations," but which is the basis of the awareness of all that is afterwards recalled, including those " more remote relations and meanings," which, in the later ex- perience, " furnish the content of ' imageless thought.' " ^ But besides the method of introspection, psychology has long been using external observational methods, and we must notice in this connection the view of consciousness taken, and the estimate placed upon this phase of psychological investi- gation by some philosophers and psychologists who have adopted, explicitly or impUcitly, the neo-reahstic position. The reference here is to those philosophers {e.g. E. A. Singer and F. J. E. Woodbridge) and psychologists (e.g. E. L. Thorn- dike, J. B. Watson, and E. P. Frost) who identify, or tend to identify, consciousness, as subject-matter of psychology, with human and animal behavior. But other names may be men- tioned, of those who have contributed in important ways to 1 Psychological Review, XVIII, 1911, pp. 312-13. 2 /?,., p. 320. 3 Ih., XXII, 1915, pp. 1-27. ■» lb., pp. 22-27. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 273 the development of this point of view. One potent influence (especially in giving direction to the investigations and theories of J. B. Watson, one of the most extreme representatives of the behaviorist doctrines) has been the combined immediate em- piricism and instrumentalism of Dewey, together with the functional psychology represented by Angell, which also grew up under Dewey's influence. In some recent articles Dewey has expressed more explicitly than before (although the Cali- fornia address of 1899 should be remembered the views con- cerning consciousness which are implied in his logical doctrine. In his essay in the Columbia volume in honor of William James, he maintains that the action of what is called '' consciousness" consists in certain organic releases in the way of behavior.^ Later, in commenting upon this, he explains that he meant that ''consciousness" is an adjective of behavior, a quality attaching to it under certain conditions. Behavior may be instinctive, habitual, or conscious. Apart from behavior consciousness is a mere abstraction, just as redness is an ab- straction apart from some red object.^ J. R. Angell, in his paper on ''Behavior as a Category of Psychology,"'* while objecting pertinently, as we shall see, to the extreme doctrines which have recently been advanced in the name of behaviorism, acknowledges that he has been conducting his work as a psy- chologist from a point of view which would make entirely easy, and even seemingly worth while, the shift of emphasis involved in making psychology primarily the study of be- havior.^ E. L. Thorndike's views have had considerable influence in developing interest in the science of behavior. In his work on Animal Intelligence ® he advocates making the study of behavior, rather than introspection, the chief psychological method. Psychology, he urges, may be, at least in part, as independent of introspection as physics is.^ What he seems to advocate is the transforming of psychology into the study of "human and animal behavior, with or without conscious- 1 V. Influence of Darwin, etc., pp. 242 ff., 270, note. 2 Essays . . . in Honor of William James, p. 69, note. ^Journal of Philosophy, IX, 1912, pp. 20, 21, 544-8; cf. XI, 1914, p. 65. * Psychological Review, XX, 1913, pp. 255-70. ^ 75.^ p. 268. 6 New York, 1911 ; first edition, 1898. ^ Qp. cit., pp. 3, 5. T 274 THE PROBLEM OF KNOWLEDGE ness." ^ It would then be essentially a supplement to phys- iology.2 Thorndike's doctrine may seem extreme enough, but it is moderation itself as compared with the ideas advanced by J. B. Watson and E. P. Frost. Watson's investigations and theories began under the guiding direction of Dewey and Angell, but he has undoubtedly been deeply biassed, as he himself confesses, by an almost exclusive attention to animal psychology for some years.^ He regards the study of animal (including human) behavior as the only consistent functional psychology ^ and would discard from his procedure all intro- spection 5 and indeed all reference to consciousness, mental states, mind, content, imagery, and the like.® Consciousness, he claims, is no more an object of study in psychology than in physics.'^ He is optimistic enough to expect that the study of the relations of external stimulus and response will solve all the problems with which the introspective psychologist has concerned himself.^ Moreover, feeling that the admission that there are mental images weakens the claim of the be- haviorist,^ he proceeds to deny that there has been produced any objective experimental evidence of the existence of dif- ferent image-types. ^° Even in the case of ''implicit behavior," commonly called ''thought-processes," where explicit behavior is delayed, and where there is response only in the speech- mechanisms and in general bodily attitudes, the right or value of introspection is denied. Although, as he admits, no method of externally observing implicit behavior exists at present, such methods, he seems to expect, will yet be found. 1^ In protest against such extreme views Dewey enters a de- murrer. Behaviorism must take more than subcutaneous processes into account, he insists ; it must include the environ- 1 Animal Intelligence, 1911, pp. 6, 7. ^ Ih., p. 16. 3 See "Psychology as the Behaviorist Views It," Psychological Review, XX, 1913, pp. 159, 175. 4 lb., p. 166. 5 75., p. 158. ^ lb., pp. 163, 166, 175-6; cf. Behavior: An Introduction to Comparative Psychology, 1914, p. 7. 7 Psychological Review, XX, 1913, p. 176. s /&., p. 177, » Journal of Philosophy, X, 1913, p. 421. «> lb., p. 422. " 76., pp. 423-4, 428; Behavior, etc., 1914, pp. 16, 19, 21, 27. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 275 ment as well as the organism in its total object of study. ^ Angell's protest is more explicit and unambiguous. He finds it difficult to take literally the idea of the complete dismissal of the image from psychology. His own work, he contends, has shown, not, as Watson seems to think, the impossibihty of finding any definite imagery involved in the control of behavior, but the amazing versatility with which different kinds of imagery may be employed upon the same task.^ Moreover, he urges, the psychologist will never be able to dispense with introspection. The gap between a specific sensorial stimula- tion and a delayed response must be bridged with information gleaned from essentially introspective sources, or else left open.^ He advises the behaviorists to forego the excesses of youth, cautioning them against committing the '^ crowning absurdity" of seeming to deny any practical significance to that which is the chief distinction of human nature — ''the presence of something corresponding to the term mind — the one thing of which the fool may be as sure as the wise man." ^ It may be instructive, however, to refer to one more example of the extremes to which the behaviorist psychology has gone. E. P. Frost regards the idea of consciousness as simply a valu- able fiction ; ^ what we ought to mean by it is simply a pecul- iarly refined but purely physiological process.® It is a nervous path responding to the just previous and still partly persisting response of a nervous path to stimulation.'^ It differs from instinctive and habitual behavior in a way that has biological significance, for it modifies the machinery of behavior by vir- tue of energy stored up in the organism by past experience.^ But a nervous reaction can never be in response to itself as stimulus; in other words, introspection is impossible.^ There are no such things to be discovered as ''sensations," "images," or "feeUngs." ^^ The term "mind," when properly used, is simply ^ See Journal of Philosophy, XI, 1914, p. 66. 2/6., X, 1913, p. 609. 3 Psychological Review, XX, 1913, pp. 262, 266, 269. < Ih., pp. 268, 270. 5 76., XIX, 1912, p. 251. « Ih., p. 249. 7 Journal of Philosophy, X, 1913, p. 717 ; XI, 1914, p. 107. ^Psychological Review, XIX, 1912, p. 252; Journal of Philosophy, XI, 1914, p. 107. » Journal of Philosophy, X, 1913, p. 717. w lb., p. 716. 276 THE PROBLEM OF KNOWLEDGE the total of such reactions to immediately preceding reactions.^ Thus, unless we mean by consciousness an ''unconscious aware- ness" of {i.e. a physiological reaction toward) an immediately previous ''unconscious awareness" (physiological reaction), there is no consciousness at all ; and inasmuch as this is itself a wholly unconscious process, there is no consciousness. The difficult problem as to what consciousness is, is solved by denying that it is. The critic might almost be pardoned, one would think, if he were to refuse to take such views seriously. When a mis- taken idea is consistently worked out to such an extravagant issue, it tends to be not only harmless, but a highly salutary warning; so that refutation by another becomes an act of supererogation, a sheer waste of energy. It does not call for refutation — it accomplishes that for itself — but it does call for explanation. It seems probable that this remarkable doc- trine is to be accounted for by the original confusion of con- sciousness with self-consciousness, and the interpretation of the discovery (?) that introspection is impossible to mean that there is no self-consciousness, and therefore no conscious- ness. Such views become important as signs of the times when it is remembered that they have received the imprimatur, at least, of the editors of two of our leading psychological and philosophical journals. And indeed it is not easy to see just where in principle the doctrine in question differs from that of some of our well-known contemporary American philosophers, e.g. E. A. Singer and F. J. E. Woodbridge. Singer claims that the hypothesis of other minds has no pragmatic meaning.^ Behef in consciousness, he says, is nothing more than expecta- tion of probable behavior. Consciousness is not something inferred from behavior; it is behavior.^ Both my mind and my fellow's mind are behavior.^ As to just what sort of be- havior mind or consciousness is, Singer at first professes igno- rance ; ^ but in a later article he advances the view that mind is the teleological behavior of an organism (which is also all the time absolutely mechanical).^ Woodbridge has expressed a * Journal of Philosophy, X, 1913, p. 719. 2 "Mind as an Observable Object," Journal of Philosophy, VIII, 1911, p. 181. » lb., p. 183. " lb. 5 75., p. 184. « Journal of Philosophy, IX, 1912, pp. 213-14. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 277 similar view. Besides his view of consciousness as a relation of implication between objects, remembering that it is also an event in the world's history/ he has attempted to define it, as a purely natural event,^ in terms of behavior. He seems to identify consciousness with the ''adaptive and even prospec- tive adjustments" of the organism to its environment.^ The criticism directed by D. S. Miller against Singer will apply to all the extreme behaviorists. They do not recognize, or sufficiently regard, the ''unique togetherness" of things which exists in all cases of consciousness, or, in other words, the existence of separate "pools of conjoint phenomenality." ^ Moreover, they seem to have obstinately closed their eyes to the surely sufficiently obvious fact that no matter how intricate or special a physiological behavior-process may be, it is always an additional item of information about it, when one is told that it is accompanied hy consciousness. On the whole, then, Miller's rebuke of the neo-realists for dogmatism seems just; they "come to conclusions" and then brace themselves "to meet the problems whose solution alone could warrant any conclusion on the subject." ^ Before turning to the various conflicting views of conscious- ness advanced by the six "collaborating" neo-reahsts, we must notice the doctrines of McGilvary and Boodin. McGilvary begins promisingly by distinguishing between "subjective objects" of consciousness (e.g. pleasure), which exist only when there is awareness of them, and other objects of consciousness, which may be called "objective objects." ® There is a sensum and there is a sentire (awareness), he continues, and even though the sentire may be the effect of a physiological process, still the sensum may be the same as the sensihile which initiated the physiological process on which the sentire depends.'' Now it is just here, we would contend, that McGilvary fatally fails to make an absolutely essential distinction. To maintain that the sensum and the sensihile are numerically the same is doubtless essential to the vindication of a genuine acquaintance 1 lb., VII, 1910, pp. 413-14. 2 lb., p. 415. » lb., VI, 1909, p. 454. Cf. X, 1910, pp. 602, 608. This idea has also been developed at length in a lecture not yet published. * Journal of Philosophy, VIII, 1911, pp. 323-4. ' lb., p. 326. « 76., IV, 1907, p. 454. ^ /^.^ pp. 457^ 593. 278 THE PROBLEM OF KNOWLEDGE with physical reahty in perception; but to assume that sen- sum and sensihile can be numerically one, only if they are in all respects qualitatively identical, is to ''fall into temptation and a snare"; it is this dogmatic ''short cut" which is "the root of all (the neo-realistic) evil," and McGilvary, having erred at this point, in company with many others, has "pierced himself through with many (epistemological) sorrows." It becomes immediately necessary to regard consciousness as "diapha- nous" ^ and, strictly speaking, undefinable; and that without the idealistic excuse, that it is the summum genus of all reality. It is asserted, to be sure, that consciousness of a thing is a "relation between objects," "a unique togetherness of the thing with other things." ^ But, while it may be admitted — and the thing admitted is an important truth — that in the event of consciousness there is a unique togetherness of things, it is still doubtful at least whether it is that unique together- ness which is the consciousness, or whether it is not merely a necessary consequence of consciousness. And then, that blessed word "unique" is here simply a device by means of which one is enabled to give a formal definition where the pos- sibility of a real definition has been cut off. To say that con- sciousness is a unique togetherness is at best to define by means of the proximate genus, leaving the differentia of the species blank, offering as excuse at the same time the more than doubt- ful assertion that no intelligible differentia exists. McGilvary does say, it must be admitted, that the togetherness is an experiential one, a being felt together or experienced together ; ^ but this is to supply the defect in the definition by virtually introducing into the predicate of the definition the term to be defined. Our philosopher evidently notices this logical fault, for he continues to labor at the problem. He finally offers the doc- trine that consciousness is a "selective relation among things," ^ and that it is also a "centred, individualized relation." ^ It is, in short, "a relation which relates in just the specific way that 1 Journal of Philosophy, IV, 1907, p. 686. 2/6., VI, 1909, p. 227; VIII, 1911, pp. 511-12. • 3 lb., VIII, 1911, pp. 519, 524. ^ 75.^ ix, 1912, p. 349. 5 Philosophical Review, XXI, 1912, p. 165. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 279 brings about the specific things that we call our experiences." * Now, inasmuch as this last seems to mean no more than that consciousness is the exact kind of relation which it is, we may be complaisant enough to agree, provided we can first accept the statement that consciousness is a relation ; and yet we can- not admit that this carries us very far toward a definition. And even accepting the additional characterizations, "selec- tive," "centred," "individualized," as applicable to conscious- ness, that they do not give us completely the specific difference by which we may distinguish consciousness from all other togetherness, is virtually confessed by McGilvary himself, when he finds it necessary again to employ that useful word "unique." If an experience is a "uniquely integrated whole of objects"; 2 and consciousness, a "unique selective rela- tion,"-^ we are still left with the questions, How integrated? and. What selective relation? unanswered. In other words, we are left without a definition. Our reference to Boodin's discussion of consciousness may well be brief, inasmuch as here again we have the doctrine that consciousness is "diaphanous," with the frank but fatal admis- sion that this means that it "has no properties."^ In this Boodin virtually concedes that, from his point of view, con- sciousness is undefinable, if not, indeed, non-existent. But this conclusion, under the circumstances, is equivalent to an acknowledgment of defeat. And now, finally, we turn to look for and examine the doc- trine of the six collaborating neo-realists concerning conscious- ness. But here again it is disappointing, and especially so in view of the collaboration, to find that instead of a doctrine, we have doctrines. Among the articles of their common creed the six have not found it possible to include a definition of consciousness. The mutual relation of their views on the subject is interesting, however. There is a fair measure of agreement between Marvin, and Holt in his earlier writings, on the one hand, and among Spaulding, Pitkin, Perry, and Holt in a very recent pubfication, on the other; but Mon- tague sets forth in this connection, as before, a doctrine radically different from that of any of the others. 1 lb. 2 7b., p. 166. 3 75.^ p, 171. 4 Journal of Philosophy, V, 1908, p. 232. 280 THE PROBLEM OF KNOWLEDGE Marvin's present views on the nature of consciousness were anticipated in large measure in his doctor's dissertation/ in which he maintains that the distinction between consciousness and what is not consciousness is not to be found in the data of experience as such, but is a matter of interpretation.^ In his recently pubhshed First Book in Metaphysics, although he in- cludes in the data of psychology reactions as well as the objects correlated therewith,^ consciousness is not identified with the reactions so much as with ^Hhe nature, the complexity, and the structure of that which controls reactions." * ''A content becomes consciousness by becoming . . . the object to which an organism reacts." ^ Thus consciousness at any moment is apparently identified with the field of consciousness, and de- fined as a certain cross-section of, or collocation of entities belonging to the universe of subsistent entities, and definable as a group by its peculiar relation to our bodily reactions. ''My consciousness of this page," he writes, ''is Uterally the page, the page in certain relations." ® In the development of this doctrine Marvin has probably been considerably influenced by Holt, whose more fully elabo- rated and practically identical theory of consciousness is to be found in his recently published volume The Concept of Con- sciousness ^ and in his contribution to the volume entitled The New Realism. Holt defines consciousness or mind as "a cross- section of the universe selected by the nervous system," ^ the group of entities within the subsisting universe to which a nervous system responds.^ He compares consciousness to the cross-section of the environment illuminated by a search-light. The cross-section is spatial and includes color-qualities, but it is not in the search-light, nor are its contained objects dependent on the search-light for their substance or their being. ^^ Simi- larly "the phenomenon of response defines a cross-section of the environment without, which is a neutral manifold. Now this neutral cross-section outside of the nervous system . . . coincides exactly with the list of objects of which we say that 1 Die Giltigkeit unserer Erkenntnis der objektiven Welt, 1898. * Op. cit., p. 30. 3 A First Book in Metaphysics, p. 258. * lb., p. 259. 5 75.^ p. 261. « lb., p. 263. 7 Completed in 1908, published in 1914. s The New Realism, pp. 354-5. 9 lb., p. 373. 1" The Concept of Consciousness, p. 171. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 281 we are conscious. This neutral cross-section as defined by the specific reaction of reflex-arcs is the psychic realm : — it is the manifold of our sensations, perceptions and ideas : — it is consciousness." ^ Henceforth in his discussion this '^environ- mental cross-section" is referred to as "psychic cross-section," ''consciousness," "mind," and even "soul," while the indi- vidual members of the cross-section are called "sensations," "perceptions," "ideas," etc.^ This view of Holt and Marvin is the consequence of work- ing out the implications of a rather superficial interpretation of the reported experience that when we introspect we find only things in their relations.^ It is assumed that because conscious- ness is not revealed to us as another element alongside of the objects of the environment of which we are conscious, it must be either dismissed as non-existent, or else identified with the objects that are revealed, the only insistence being that it is as revealed that they are consciousness. The appearance of dogmatism is toned down by the slipping in of the ambiguous term, "psychic realm," as mediating between "objects," or "field of consciousness," on the one side, and "consciousness" on the other. At this point the new realism makes liberal use of the very convenient fallacy of equivocation. But it is probably vain to expect to produce a sense of logical guilt in the mind of one who can proclaim as an epistemological gospel the doctrine that his own consciousness (being conscious) of a group of objects is neither more nor less than that group of objects, as responded to by his physical organism. But while Holt (in his published volume and his essay in The New Realism) and Marvin build their essentially physical conception of consciousness upon a difficulty of introspection, the views of Spaulding, Pitkin, Perry, and finally those of Holt in his paper, " Response and Cognition," take account also of the physiological conception of consciousness which has been growing up under the influence of that behaviorist psychology which, in turn, is itself largely a product of neo-realistic in- 1/6., p. 182. '76., p. 183, et passim. For a discussion of Holt's combination of this view with behaviorism, see pp. 285-7, infra. 3 Of. re Woodbridge, supra; v. Journal of Philosophy ^ X, 1913, p. 608. 282 THE PROBLEM OF KNOWLEDGE fluence, and attempt to combine both conceptions, the phys- ical and the physiological, in one synthetic definition. (Pit- kin's emphasis, however, is almost entirely on the second of the two points of view combined.) Spaulding has not expressed himself very much in detail on the problem of the nature of consciousness, but he has main- tained that consciousness is the function of implying, knowing, and pointing to, but in no way modifying, an independently real object.^ The term "knowing" does not give us much in- formation in this connection, because it is not itself defined, but is the main part, if not all, of what, from the realistic point of view, has to be defined in the definition of consciousness. The term "implying" suggests the method of learning the nature of consciousness by what the neo-reaHst tends to sub- stitute for introspection, which he finds impossible, viz. an analysis of the objective "content" of consciousness. It is Woodbridge's definition over again, which we have already criticised. The term "pointing to," on the contrary, is seen from the context ^ to have a biological meaning, so that here we have represented, although the expression is a vague one, the type of view that results from regarding consciousness as an externally observable relation of the physical organism to other objects. The subject of the " implying " seems to be some object within the total field, or content, of consciousness ; the subject of the "pointing," on the contrary, seems to be the physical or- ganism. Apart therefore from the difficulty of conceiving how either one can be conscious, this manifest discrepancy between the subjects shows that a unitary definition has not been offered. Pitkin charges the English reaHsts with not having really attacked the problem of consciousness, inasmuch as they con- tinue to talk of "mental activity," and himself defines the problem, as it presents itself to the American new realist, as the finding of the differentia of the cognitive activity and that of the cognitive field. ^ After a preliminary statement, in which ^Journal of Philosophy, III, 1906, p. 317; VII, 1910, p. 399; VIII, 1911, p. 72. 2 lb.. Ill, 1906, p. 316. 3 The New Realism, pp. 439-41. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 283 the influence of Dewey is manifest/ to the effect that conscious- ness involves a specific environment, a directed activity and the operation of an organic structure,^ he ventures the definition that consciousness is the crucial advance of the organism toward adjustment to external entities.^ Now this is a unitary defini- tion, but it frankly reduces psychology to a study which would be related to physiology as ecology is related to the physiology of plant life : it would be a study of the externally observable behavior of the animal organism, human or other, in relation to its externally observable environment. Now one may understand how Dewey, with his immediate empiricism, or disguised psychological idealism, might have some excuse for calling this psychology, but the same privilege can scarcely be granted to a thoroughgoing realist hke Pitkin. What he has given us is a good definition of something else, important enough in its own way, but not a definition of consciousness. To inquire whether or not this '^ crucial advance of the organ- ism toward adjustment to external entities" is accompanied by consciousness of those entities is not to ask the meaningless question. Is consciousness of anything accompanied by con- sciousness of that thing? Perry's definition of consciousness is interesting as being the result of an explicit attempt to combine the points of view of introspection (or what the neo-realist calls introspection) and external observation of mind in nature and society. The former should give ''the mind within," and the latter, "the mind without"; and as these must be, somehow, in reality one, a combination of the findings of the two processes ought to give us our required definition.^ Now it turns out that when we try to introspect our own experience, we find only objects, "si chaotic manifold of fragments of the other-than-mind." ^ In order to find the common bond between these objects or fragments, the basis of their togetherness, we must turn from the method of introspection to that of external observation.® Thereupon we find the mind without (in nature and society) 1 Cf. op. cit., p. 437. 2 75.^ p. 442. 3 /&., p. 457. < Journal of Philosophy, VI, 1909, pp. 169-70, 172-5 ; Present Philosophical Tendencies, pp. 273-4. 5 Journal of Philosophy, VI, 1909, pp. 170-1. * Present Philosophical Tendencies, p. 279. 284 THE PROBLEM OF KNOWLEDGE to be behavior, a bodily complex moved by interests.^ But it appears that the objects which we discover when we ''intro- spect" and the elements of the environment to which the ''bodily complex" responds are the same. The reflex nervous system, responding to an entity in a specific way, makes it a content of consciousness,^ i.e. a content discoverable by in- trospection. Uniting, therefore, our findings by the two meth- ods, we can say that mind, or consciousness, is the environ- ment which an organism senses, or, better, it is behavior, to- gether with the objects it employs and isolates.^ The natural mind, then, is an organization possessing, as aspects, interest, nervous system, and contents, or, in other words, externally ob- servable action and independently existing contents.^ Now in criticism of all this it may be said that the objections previously offered to the separate elements of Perry's defini- tion apply with undiminished force in spite of their having been brought into some sort of combination. In spite of all Perry's precautions, he has not succeeded in corralling con- sciousness in his definition. Indeed he himself admits that all he can discover by what he calls introspection is a "manifold of fragments of the other-than-mind." And it is a notorious fact that external observation also reveals only movements of the bodily complex in relation to its environment, in other words, again nothing but "other-than-mind." Indeed in many cases the external observer knows not whether to inter- pret the behavior which he sees, as accompanied by conscious- ness or not. In adding together the results of the two methods, Perry has succeeded in "rounding up" all the important asso- dates of consciousness, but consciousness itself is not to be found in the aggregation; other-than-mind added to other- than-mind does not give other than other-than-mind.^ Nor 1 Journal of Philosophy, VI, 1909, pp. 172-3. 2 76., VII, 1910, p. 397. » 76., VI, 1909, pp. 174-5 ; Present Philosophical Tendencies, p. 303. * Present Philosophical Tendencies, p. 304. 6 We neither mean nor need to say here that other-than-a added to other- than-a never gives other than other-than-a. For instance, 2 (other than 5) added to 3 (other than 5) does give other than other-than-5. We can say this, how- ever, only because we know enough about the relation of 2 and 3 to each other to know that when taken together they are 5. But we do not know that organic behavior in response to the environment, and the objects which it employs and isolates, taken together, are mind. Indeed we would not know this, even if we NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 285 can it even be said that organic response to a selected portion of the environment is impossible without consciousness as an accompaniment, for we are well aware of results in our own experience which have come from unconscious organic response to our physical environment. To be sure, the terms "mind" and "consciousness" may be used with radically altered mean- ing, and arbitrarily applied to the aforesaid sum of elements; but it ought to be no less acceptable to the neo-realist, as it would be far less misleading, if he were to employ instead of these terms some neutral algebraic symbol. His definition would then be stripped of the false greatness that has been thrust upon it by calling it a definition of mind.^ Holt's recent paper on " Response and Cognition "^ will un- doubtedly be recognized as one of the important documents of the new "behaviorism." While cordially approving the meth- ods of investigation employed by the behaviorist psychologists, he regards their theories, and fundamentally their definition of behavior, as defective; and he sets himself, accordingly, to remedy this defect. Thus while the behaviorists tend to treat behavior as consisting of reflex activities simply. Holt insists that it is essential to note that these reflex activities have been so integrated, so organized, that in behavior proper the action, while a constant function of some object, process, or aspect of the objective environment, is not a function of the immediate stimulus. Defining behavior, then — or " the relation of specific response," as he suggests it may be called — as "any process of release [of stored energy] which is a function of factors external to the mechanism released," and assuming that the terms with which psychology deals can be adequately trans- lated into the terms of the science of behavior, he takes up for re-definition some of the more important concepts of ordinary knew that whenever the biological selection of objects and organic response thereto occur together, the mental relation is present. It might very well be that mind was the cause of both the selection of the object and the organic re- sponse, and not a mere efifect of their occurrence together, much less a mere name for their combination. 1 In criticising Perry's definition, as not borne out by our everyday knowledge of our own consciousness, Russell remarks, " In order to know that such and such a thing lies within my experience, it is not necessary to know anything about my nervous system " {Monist, XXIV, 1914, p. 184). ^Journal of Philosophy, XII, 1915, pp. 365-73; 393-409. 286 THE PROBLEM OF KNOWLEDGE psychology. In the first place, the object or " content of con- sciousness" is simply, from this point of view, the object of which the organism's behavior is a constant function. '' Voli- tion" is simply what the body does toward the environment, 'Hhe will" is the behavior function, and the subject of both voHtion and cognition is simply the body itself ; in behaviorism ''the physical organism will . . . supersede the metaphysical sub- ject." " The personality, or the soul, ... is the attitude and conduct, idem est, the purposes of the body;" ''behaviorism can rest unperturbed while the sad procession of spirits, Ghost- Souls, 'transcendental' Egos, and what not, passes by and vanishes in its own vapor." " Feeling " is simply " some modification of response which is determined by factors within the organism." "The long sought cognitive relation between 'subject' and 'object'" thus becomes simply the externally observable "behavior relation." When he comes to define attention and the stream of consciousness, Holt, in order to supplement the point of view of external observation, returns to the point of view of that which the neo-realist calls introspec- tion, but which is really only the observation of the objects of which one is conscious. " The attentive level of consciousness, that of which the ' self ' is aware," is then " that most compre- hensive environmental field to which the organism has so far at- tained (by integration) the capacity to respond." " The ' stream of consciousness,' " finally, " is nothing but . . . [the] selected procession of the environmental aspects to which the body's ever varying motor adjustments are directed." In the main the criticisms to be directed against Holt are so obvious that their elaborate statement seems superfluous. Most of what was said in criticism of Perry's view apphes to this doctrine of Holt also ; the chief difference is that Perry would acknowledge that not only the objects of which one is conscious, but the externally observable behavior also, is "other-than- mind." What Holt has given us is, in the main, a very valuable analysis of some of the physiological associates of con- sciousness ; but the identification of these, throughout, with the subject-matter of psychology proper, is both philosophically and psychologically inexcusable. Moreover, it may be re- marked incidentally. Holt assumes all too easily that behavior, NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 287 especially human behavior in its most highly conscious forms, is an absolutely constant function of the environment, or of any part of it. Montague rejects the behaviorist interpretation of con- sciousness, as a form of "panhylism," almost or quite as objectionable as panpsychism, inasmuch as both are self- refuting.i He seeks rather to set consciousness forth as a cer- tain sort of relation, having been from the first required to do so by his theory of the permanent objective existence of second- ary quahties.^ In the name of his special brand of monistic realism, which in this connection he calls "hylopsychism," ^ he holds that there is but one system of realities, and that exists in time and space, so that mental processes must be re- garded as occurring in space, and consciousness must be inter- preted as a relation between spatial objects.^ The question is. What sort of a relation is consciousness ? ^ Two sugges- tions seem to have been fruitful in shaping Montague's answer to this question, viz. the analogy of the search-light,^ which Holt has also employed, and the concept of potential energy.^ The resemblances between consciousness and potential energy are dwelt upon : sensation and energy are similar in being char- acterized by both intensity and polarity; and when sensation occurs it is at the same time and under the same conditions as mark the transformation of the kinetic energy of a neural cur- rent into potential energy .^ Moreover, both are essentially private and hidden ; both pervade space ; both are teleological.^ 1 The New Realism, pp. 270-80, 482. 2 Journal of Philosophy, II, 1905, pp. 314, 315. 3 The New Realism, pp. 279-81. According to panpsychism physical ob- jects are nothing but actual perceptions, or permanent possibilities of perception. According to panhylism consciousness is nothing but the possibility of objects, or nothing but an epiphenomenal correlate of the brain-process. Hylopsychism would eliminate the "nothing but" in both cases. 4 Journal of Philosophy, IV, 1907, p. 376 ; cf. Monist, XVIII, 1908, pp. 21-9. 5 Journal of Philosophy, IV, 1907, p. 377. 6 76., IV, 1907, p. 102 ; cf. IX, 1912, pp. 39-41, 46. ^ See American Journal of Psychology, XV, 1904, pp. 1-13 ; Journal of Philos- ophy, TV, 1907, pp. 378-82; "Consciousness a Form of Energy," Essays . . . in Honor of William James, pp. 103-34 ; The New Realism, pp. 281, etc. ; Philo- sophical Review, XXIII, 1914, pp. 57-9. » Journal of Philosophy, V, 1908, pp. 209-10. * Essays . . . in Honor of William James, pp. 126-8. 288 THE PROBLEM OF KNOWLEDGE Hence it is suggested that sensations are forms of potential energy, that consciousness is potential energy.^ Moreover, this would give a positive content to the idea of potential energy; the potentiality of the physical would here be the actuality of the psychical, just as in the afferent paths and centres of the nervous system the actuality of the physical is the potentiality of the psychical. ^ Or, in other words, con- sciousness objectively implies certain cortical forces, and is implied by them.^ The theory is that the external object possesses all the secondary as well as primary qualities ob- served in normal perception, is colored, for example ; its color is also objectively present in the light-waves, in the retina, the optic nerve, and the visual centre of the brain. When the energy, which throughout all this process retains, it is assumed, its specificity,^ becomes potential in the brain, it is transformed into consciousness. The brain becomes conscious, i.e. becomes, for the time being, a mind.^ This cerebral or conscious event has, like every other event, a self -transcending reference.^ It is the potential or implicative presence of a thing at a space or time at which it is not actually present.^ "The world that we perceive is (not indeed an actual but) a virtual or potential reprojection of the effects which the world projects upon ws." ^ In criticism of this view it may be said at the outset that it is dogmatically based upon what can hardly be regarded as more than a prejudice, viz. that there is only one sort of reality, i.e. a reality in time and space and measurable in terms of physical energy. An additional reason for hesitation is found in the confessed total absence of verification of the hypothesis (which is quite fundamental to the theory) that in normal perception the primary energies in the external bodies and in the cerebral tracts are specifically the same. Moreover it seems dogmatic and even fantastic to suppose — if this is what he means — that color-qualities are present throughout all the space traversed by the light-waves, and throughout all 1 Essays . . . in Honor of William James, p. 129. 2 The New Realism, p. 281 ; Philosophical Review, XXIII, 1914, p. 58. 3 The New Realism, p. 293 ; Philosophical Review, XXIII, 1914, pp. 57-8. * The New Realism, p. 299. 6 Philosophical Review, XXIII, 1914, p. 59. « 76., p. 57. 7 The New Realism, p. 281. 8 Philosophical Review, XXIII, 1914, p. 62. NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 289 the tracts of the brain traversed by the stimulation. This surely does not accord with the principle of parsimony. Or if it be explained that just as in the case of consciousness there is a virtual or potential or implicative presence in the extra- organic world of that which is actually in the brain, so there is only a virtual or potential or implicative presence of the qualities of the external object in the brain, so that there is simply the cancelling of one self-transcendence by means of another in the opposite direction, the external object being virtually put back where it actually belongs, but from which it had virtually strayed, one is still unable to see how such virtual presence of the external object where it is not (viz. in the brain) can actually he the virtual presence of the perceived object where it is not (viz. in the external world) in any such way as would allow for enough difference between the actually external object and the virtually-introjected-virtually-repro- jected object to explain the possihility of error, to explain which seems to have been the chief raison d^^tre of this elaborate theory. On the other hand, if there is enough difference for the possibility of error, there must necessarily be a total nu- merical difference between the two, in which case there is too much difference for the possihility of knowledge. In this case, as we have previously pointed out,^ Montague's realistic epis- temological monism would seem to be entirely a matter of faith, in the sense of believing what there is sufficient reason for disbelieving. These considerations, then, without further reference to his rejection of the idea of memory images, ideas or sensations as really existing,^ seem sufficient ground for de- cUning to accept Montague's ingenious and in some ways attractive speculations as to the nature of consciousness. In concluding this investigation of the neo-realistic doctrines of consciousness, it may be instructive briefly to compare and contrast the results arrived at by the English and American schools. In each of the largely separate developments of thought there is discoverable something of the nature of a dialectical process. The English new realists, reacting against the extreme idealistic philosophies which made consciousness the only and all-inclusive Being, took up the question as to 1 See Ch. XI, supra. 2 Philosophical Review, XXIII, 1914, p. 60. XJ 290 THE PROBLEM OF KNOWLEDGE just what existent consciousness is, if it is true that it is only an existent among other existent things. The answer was soon forthcoming that it is not an existent at all, unless it is a relation between existents, in particular a relation between a really existent subject and a world of really existent objects. But when the question was raised as to just what relation con- sciousness is, difficulties and diversities of opinion began to appear. Several largely distinct lines of thought may be regarded as developing the antithesis to the thesis that con- sciousness is a relation, and as marking the transition in the direction of the synthetic view that consciousness cannot be a relation unless it is also and at the same time a productive or creative psychical activity. For example, we have Stout'& contention that, applied to fictitious objects, the doctrine that consciousness is an external relation can only mean that such objects have genuinely independent reahty, since the whole reality of any object cannot be included in its relation to something else. That Stout does not seem to see the way out of his difficulty by means of the concept of productive activity, or even to see how antithetical to the merely relational view of consciousness the considerations he advances are, makes his contribution to the antithetical stage of the dialectic all the more impressive. Wolf, on the other hand, developing some of the antithetical considerations to the view of Moore and Alexander that consciousness is a purely diaphanous relation between subject and object, makes definite progress toward a higher synthesis. It is necessary to think of consciousness as a productive activity in some cases; we can only consent to regard it as a purely diaphanous relation in normal perception ; in all cases of error, perceptual or other, it is an activity pro- ductive of its object. This idea of consciousness as a produc- tive activity is carried much further by McDougall, although to what extent under the influence of neo-realistic thought it is not easy to say. At any rate he applies it so far, especially in the case of sense-qualities, that he cannot be regarded as one of the neo-realistic school. On the other hand he does not seem fully to appreciate the philosophical significance of the idea of consciousness which he introduces rather casually toward the close of his volume, Bodij and Mind. On the whole, NEO-REALISTIC DOCTRINE OF CONSCIOUSNESS 291 however, from the standpoint to be defended in our later con- structive attempt, it would seem that the dialectic of EngHsh neo-realism has been leading in the general direction of the true solution of the problem of consciousness. Turning to the American movement we find a parallel but strangely different phenomenon. Here too, in reaction against extreme idealistic views, the problem emerged as to what consciousness is, if we cannot say that it is the all-embracing reality. By such thinkers as James, Woodbridge, Holt, and others, arguments, critical and constructive, i.e. antithetical and synthetical, were presented, to show that consciousness is not an existent, but a mere relation among existents. The chief difference between the American and English schools at this point, however, is that while the English realists have contended that consciousness is a relation between a psychical or mental subject and physical objects, the Americans have generally maintained that it is a relation between or among physical objects. But here again, when the question was raised as to just what relation between objects consciousness is, considerations antithetical to the relational view were brought to light. For example, Woodbridge and Perry, although in different ways, have expressed the conviction that conscious- ness is not a relation between objects unless it is also an activity of one object (the physical organism or nervous system) upon other objects. This latter view has been developed at length by the behaviorists — to the bitter end, we take it, by some of them. Indeed, in the above exposition and discussion of the American neo-realistic doctrine of consciousness it has been shown, we think, that its dialectic has been leading it with resistless logic to a thoroughgoing self -refutation. Views such as those of Singer and Watson and Frost, to mention only the most extravagant developments, really constitute a reductio ad absurdum of some at least of their presuppositions. And if the question be raised as to how it can be maintained that this dialectical movement from existent to relation and from rela- tion to activity can lead in the one case (that of English neo- realism) toward a true position, and in the other (that of American neo-realism) to a reductio ad absurdum, our answer would be that the most plausible explanation of the difference 292 THE PROBLEM OF KNOWLEDGE seems to be that it is due to the one conspicuous difference between the presuppositions of the two schools. That is, it must be because the Enghsh new reahsts, speaking generally, have regarded the subject of consciousness as psychical, mental, spiritual, while the Americans have quite as unani- mously insisted upon viewing it as physical. But further discussion of this point we must defer until we turn from crit- icism to construction. CHAPTER XIII The Neo-Realistic Doctrine of Relations, Universals, AND Values Our critique of the new realism has thus far centred about its position with reference to secondary quahties and the nature of consciousness, but some further grounds of objection may be found, in our opinion, from an examination of its doc- trine as to relations, universals, and values. To such an ex- amination we now turn. With reference to the neo-reaUstic doctrine of the externality of relations, it should be understood that it is a further generali- zation (in the interests of system and for the sake of deductive epistemology) of the doctrine of the externality of the know- ing or conscious relation, or, in other words, of the known object's absolute independence of the circumstance of its being known. For it is evident that if it could be maintained that all relations are external to the terms related, one could deduce the externality of the knowing relation, so far as the object is concerned, or the independence of the known object from the knowing relation. On the other hand, even if it can only be shown that some other relations are external to one or both of their terms, there will be a certain added plausibility in the view that being known is an external relation, and the known object, therefore, independently real. The English neo-realists, with the exception of Bertrand Russell, do not seem to have gone into the question of the internality or externality of relations in any very thorough- going manner. T. P. Nunn admits that some relations make a difference to the object observed,^ but he gives us neither a catalogue of those which make a difference, and so are pre- sumably internal, nor a method by which we can distinguish 1 Proc. Aristot. Soc, 1909-10, p. 206. 293 294 THE PROBLEM OF KNOWLEDGE internal relations from those which are external. Alexander remarks that if we mean by the internality of a relation that it cannot exist independently of its terms, then in this sense relations are internal to their terms. ^ This is undoubtedly a highly defensible position to take, but it does not deal with the question which is of primary concern to the neo-realist. It would be coming closer to the real question to ask whether a term can exist independently of its relations ; but the exact point of dispute is whether the correct thing to say is that all relations make a difference to their terms, or that some do and some do not, or that none do. Stout, in his remark that no being can be entirely constituted by its relation to something else,2 assumes that every term must be at least partially in- dependent of any one of its relations; but, in view of the possibility of creative causality being one of the relations, does not even this comparatively modest expression seem unwar- ranted? If there is creative activity, some being or quality must depend for its existence upon something else being re- lated to it in this particular relation of creative causality. . Russell's doctrine of relations is explicit and highly pertinent to the question as to the basis of realism. In the first place he holds that relatedness does not imply any corresponding complexity in the terms related, so that the anti-realistic argu- ment on the basis of the internality of all relations is not validly founded.^ This may be regarded as safe ground to occupy, but it leaves unanswered the question as to whether there are not some relations which are internal to their terms, and es- pecially, this question being answered in the affirmative, the further question how some relations can be internal and others external to the terms which they relate. RusselFs doctrine that relations are real entities (not existences but subsistences) apart from any terms, we shall examine in connection with his theory of universals; but here it may be remarked that this doctrine seems to involve the absolute externality of all rela- tions. In this case the above moderate statements would have to be taken as representing less than the whole (im- 1 Mind, N.S., XXI, 1912, p. 310. ^ Proc. Aristot. Soc, 1910-11, p. 187. 3 Journal of Philosophy, VIII, 1911, pp. 158-9. Cf. The Principles of Math- ematics, pp. 221-6. NEO-REALISTIC DOCTRINE OF RELATIONS 295 portant) truth, from his point of view, and as having been made with a view to controversial security. Among the American neo-reahsts detailed discussion of the internality or externality of relations has almost been confined to the six ''programmists." In the introductory chapter of The New Realism, which represents the views of all six, it is stated: ''Realism rejects the premise that all relations are internal. . . . The evidence at present available indicates that while all things may perhaps be related, many of these relations are not constitutive or determinative, i.e. do not enter into the explanation of the nature or existence of their terms." ^ Sometimes it is difficult to see that complete self-consistency has been maintained in the different statements that bear upon the theory of relations. Thus Montague, in one of his early statements, seems clearly to imply the externality of all re- lations. "All relations," he says, ''presuppose the existence of the terms between which they subsist," and from this the possible independent existence of the terms is inferred, so that logical priority has evidently been fallaciously interpreted to mean chronological priority, or previous (and hence indepen- dent) existence. 2 Again, much more recently, he has main- tained that the internal view, that the nature of the parts of a complex depends upon the nature of the whole complex, is fallacious, apparently because, as he sees it, this would require one to believe that knowledge of merely a part of the truth is necessarily false. ^ It would seem, however, that the nature of the parts might depend upon the nature of the whole in some respects not requiring to be contemplated in a particular judgment, which might therefore be true notwithstanding its not being the whole truth. But what we set out to show was the evident discrepancy between Montague's statements as 1 The New Realism, p. 33 ; cf. statements in Journal of Philosophy, VII, .1910, pp. 393-401. 2 Journal of Philosophy, II, 1905, p. 313. What Montague is concerned to maintain here is the mutual implication of realism and the relational view of consciousness. " If consciousness is a relation, objects of consciousness must be real independently of their standing in that relation, while conversely, if objects are real independently of a consciousness or knowledge of them, then that con- sciousness or knowledge can not be anything other than a relation between them." « The New Realism, p. 299. 296 THE PROBLEM OF KNOWLEDGE referred to, and the implication of his scornful question in another connection : ''What kind of an object would it be, for- sooth, which remained completely unaltered by the relations in which it stood?" ^ Perhaps Montague would explain away the discrepancy by saying that only what is true as a particular, and not what is universally true, is thus dependent upon par- ticular relations ; but this is all that the opponent of the new realism ordinarily maintains. But inasmuch as Montague evidently holds ^ that the universal includes all the particulars, each in its own particular relation, he cannot consistently main- tain that he knows a thing as it really is, when part of what it really is depends upon an unknown relation. He thinks of it as if it were not what it really is. Holt and Marvin also make statements that lead them into evident self-contradiction. Holt makes the general statement that the entities of the universe are related by external re- lations,^ and yet on the same page he admits that it is seldom possible to say just where the object itself terminates and its relations to other entities commence. Now if we are going to insist that nothing must be interpreted as a relation which is not wholly external to the term, what seem to be relations and yet are internal to the term will have to be interpreted as part of the object, and not as a relation. But Holt defines mind as that cross-section of the environment, a neutral mani- fold, which is defined or selected by the response of the nervous system.^ Now this seems to mean that mind is a neutral mani- fold as selected, or in the relation of being selected, by an organ- ism, and not when not thus selected. Here then we have that which is what it is (viz. mind) solely by virtue of its relation to something else. But this is for that relation to be internal to its term (mind), and it would but thinly disguise the break- down of the theory to say that it is difficult here to say where the object terminates and the relation begins. Essentially similar is the criticism to be made against Marvin, who states the doctrine of the externality of relations in universal terms,^ and then makes the statement about consciousness 1 Journal of Philosophy, V, 1908, p. 211. 2 76. 3 The New Realism, p. 372. * The Concept of Consciousness, p. 182. 5 Journal of Philosophy, VII, 1910, p. 395. NEO-REALISTIC DOCTRINE OF RELATIONS 297 to which we have already referred : ''My consciousness of this page is hterally the page, the page in certain relations ... A field of consciousness is a certain cross-section, a certain col- lection, of entities, belonging to the universe of subsistent entities and definable as a group by its peculiar relation to our bodily reactions." ^ Here, then, is a relation, the relation which makes the subsistent, or existent, into the mental, or conscious- ness, and* which is therefore internal so far as consciousness is concerned — a manifest exception to the doctrine of the uni- versal externality of relations. The other three of the six, viz. Spaulding, Pitkin, and Perry, have argued more at length for the external view of relations. Spaulding builds his realism upon this view. The external view, he claims, is self-consistent, and is therefore established, as against all other systems which are based upon the internal view, which is self-refuting.^ He maintains that the supporter of the internal view tacitly or surreptitiously employs the ex- ternal view with reference to his own system, in the supposition that reality, apart from any relation to the knower, is what the internal view takes it to be; and that the internal view thus presupposes its own contradictory, and so refutes itself. But while the internal view, if not applied to itself, is seen to be self- refuting, when it is so applied, the result is, ''on the one hand, that by his own theory his own knowledge of his own theory is a knowledge only of that which is appearance, and yet on the other hand, that he can never know whether this is real ap- pearance or not, because the modifying effect of knowledge can never be eliminated. And again, by his own theory he cannot know that even all this is the real state of affairs, and so on in an infinite regress." ^ This is valuable criticism, but it is valid only against the view that all relations must always be taken as internal, and of course fails to show that all relations must always be taken as external. Now it may be that the opponents of realism make, as Spaulding charges, an "arbitrary use of the 'internal view' for certain purposes, and ... of the 'external view' in other 1 A First Book in Metaphysics, p. 263. 2 Journal of Philosophy, VII, 1910, p. 400. 3 Philosophical Review, XIX, 1910, pp. 276-7, 299-300, 620-1. 298 THE PROBLEM OF KNOWLEDGE connections," ^ but that does not mean that there may not be a justifiable principle according to which one may regard cer- tain relations as either internal or external. And that Spaulding himself cannot regard all relations as always external is evident from his guarded statement that "a term may stand in one or in many relations to one or many other terms," and that ''any of these terms and . . . some ^ of these relations could be ab- sent . . . without there being any resulting modification of the remaining . . . terms or relations." ^ In thus regarding some relations only as external, without showing the principle involved in this selection, is not Spaulding also ''arbitrary"? In his contribution to the volume entitled The New Realism, again, he argues successfully against those who would dispute the possibility of analysis, or the validity of its results, resting their contention on the presupposition that independence and relatedness are mutually exclusive ; but the limits of what he accomplishes are indicated in the statement: "The question in which we are chiefly interested is not whether the internal theory has no application, but simply whether this apphcation can be universal." ^ He criticises the upholder of the internal view as unable to be consistent, since he cannot make his theory universal ; ^ but he himseK neither makes the external view universal, nor gives us any adequate principle by which the internahty or externahty of relations may be determined. It is true that he undertakes an empirical investigation in order to discover just when, in cases of actual synthesis, new prop- erties appear ; and this is good as far as it goes. But besides calling attention to the fact that he feels obliged to leave it an open question whether or not the parts always remain un- changed by the synthesis,^ we would criticise his whole treat- ment of the topic as not taking into account the fact that our ordinarily practical necessities require us to treat one and the same relation sometimes as internal and sometimes as external. Pitkin, in his essay entitled "Some Reahstic Implications of Biology," admits that the realist cannot count his case won 1 Philosophical Review, XIX, 1910, p. 621. 2 Italics mine. 3 Journal of Philosophy, VII, 1910, p. 400. 4 The New Realism, pp. 165, 167. P/6., p. 168. «/6., p. 241. NEO-REALISTIC DOCTRINE OF RELATIONS 299 until he has disproved the anti-reaUstic inferences drawn from the biological sciences by Driesch, Bergson, and Dewey. The first mentioned, on the basis of a vitalistic interpretation of biological facts, concludes that the Kantian view of the con- struction of the entire '^ content" of experience by the Ego is vindicated. Bergson contends that all discreteness is pro- duced by the ''vital force" ; while Dewey regards theories and ideas as genuine constructions of the thinker. ^ Pitkin's op- position to Dewey at this point will come in for consideration in our discussion of the neo-reahstic doctrine of universals. More- over, his destructive efforts directed toward the Kantian super- structure which Driesch would erect upon his vitalistic founda- tion, we can view with a large measure of complacency and even approval. We may admit also that Bergson speculates away beyond his empirical data in his doctrine of a cosmic vital force fundamental to matter as well as idea. But when Pitkin claims to have made the ''discovery" that "organic parts do not depend upon the whole in which they naturally occur, except in an empty sense of the verb," ^ we must object that he claims too much. He cites, in support of the above proposition, the re- sults of some grafting experiments, in which he says there was no trace of mutual influence determining the development of the two parts. But, we would reply, not only has Driesch taken account of such phenomena ; but even as long ago as 1891 he yer- formed the experiment of killing one of the first two cleavage- cells of the egg of a sea-urchin, and found not that one-half of an embryo was reared out of the surviving cell, but a com- plete embryo of one-half the normal size. In this way he demon- strated that the range of possibilities for the development of a cleavage-cell was much wider than its prospective value under normal conditions of development — a fact which seems to in- dicate the inadequacy of a purely mechanistic theory of em- bryonic development. 3 It may be, as Pitkin claims, that there is some kind or measure of "organic pluralism," ^ but in view of facts of the sort just cited, he cannot convincingly argue for the universal externahty of biological relations, especially that 1 The New Realism, pp. 378-80. 2 75.^ pp. 422-3. 3 H. Driesch, The Problem of Individuality, 1914, pp. 10 ff. * Op. cit., p. 425. 300 THE PROBLEM OF KNOWLEDGE of the whole developing organism to its parts. This being the case, much less has he the right to defend the universal external- ity of relations, which seems to be the desired major premise for inferring the universal externality of the knowing relation, in the sense of "the complete independence of all things thought of," ^ or otherwise cognized. All that is really shown is that monistic reaUsm may be regarded, a priori, as conceivable, because of the discovery that some things are sometimes inde- pendent of some of the relations in which they stand, or that some relations are sometimes external. But this much, the ideaHst to the contrary notwithstanding, may be admitted as a part of common, everyday knowledge. Moreover, Pitkin incidentally makes the same damaging admission as that of Holt and Marvin referred to above. The ''indiscernibility of seeming from being," he says, is to be attributed to the relations in which they stand. In one set of relations they are ''indis- cernibles," while in other relations they are readily distinguished from each other.^ Here it would seem that the neo-realist, in his anxiety to explain the possibility of error, has allowed him- self to admit that a mere matter of relation can make so great a difference as that between seeming and being. Such a rela- tion is surely not external; the mere difference in the object between seeming and not seeming, would be enough, it might be argued, if we were very exacting, to show the awareness- relation between subject and object to be not absolutely external to the object; but the difference between mere seeming and being is a much greater difference, a difference only less, so far as these categories are concerned, than that between being and not seeming. Perry, like the others, argues for reaHsm by defending, as far as he logically can, if not farther, the doctrine of the ex- ternahty of relations.^ His earHer statement of the theory as the doctrine that "terms acquire from their new relations an added character, which does not either condition or neces- sarily alter the character which they already possess," ^ seems 1 The New ReaUsm, p. 380. 2 Ih., pp. 466-7. 3 Present Philosophical Tendencies, pp. 319 f. ; "A Realistic Theory of Inde- pendence," in The New Realism, pp. 126-51. * Present Philosophical Tendencies, p. 319. NEO-REALISTIC DOCTRINE OF RELATIONS 301 to dodge the main point of dispute. Of course a term, when related in any specific way, has the character of being thus re- lated, but the question is whether, because it is thus related, it ever comes to be different from what it was, or needs to be treated as different. Perry says that in the complex (a)R(6) the term a does not derive its content from R(6) ; ^ but before we can assent we must know what is meant. If it is meant that the content which a had before it was related to h is not derived from its relation to 6, this may be accepted as a mere truism ; but if it is meant that a in relation to h need never be treated as different from a when not in that relation, it is contradicted by our experience every day. Perry says 'Hhe content of things is in no case made up of relations beyond themselves" ; ^ but here we find the same ambiguity. If it is meant that the content of a thing is never made up, even in part, of its external relations, we can only say, Of course not ; the relations would not be ''external" if they made up any part of the content. If on the contrary what is meant is that we need never include any of the relations in which an object stands, in order to know what it is to be taken as, practically, again we must say, on the basis of everyday experience. This is by no means true. In his contribution to The New Realism, however, Perry has done much to clarify the situation with reference to relations. He sets out to give the neo-reahstic theory of independence. By independence he means not non-relation, but simply non-de- pendence.^ But while all relation is not denied, certain relations are declared to be absent when one entity is said to be inde- pendent of another.^ These relations are those of containing and being contained, causing and being exclusively caused, and implying and being exclusively implied.^ When not re- lated in any of these ways, two entities are, according to Perry, independent of each other. This would seem to be true when independent means not dependent for being existent, or true; but if independent means not dependent for its significance, it is obvious that a thing may often be regarded as independent of some of the above relations, while it is often dependent — ■ and this is the fact that is damaging to the theory in which 1 lb., p. 319. 2 75.^ p. 320. 3 The New Realism, p. 113. */6., p. 117. 6/5., pp. 113, 151. 302 THE PROBLEM OF KNOWLEDGE Perry is interested — upon relations not included in the above list. In summing up our criticism of the neo-reaUstic doctrine of relations, then, we may say that while the new reaUsts are successful in showing that the doctrine that all relations are al- ways internal is not correct, and while they are thus able to undermine one of the stock arguments for idealism (that drawn from the alleged internality, to any term, of the relation of being known), they are not able to show that all relations are always external, and so cannot by this means prove the reaHstic theory of knowledge. At most they can show that the known object may perhaps in some cases exist independently, and essentially without change through the circumstance of its being known. But even so, they leave the whole subject in a very unsatisfactory condition, for they fail to mention any adequate criterion by which it may be determined whether or not a particular relation is or is not in any case external. We now pass to an examination of the neo-realistic doctrine of universals. At this point the approximation of the new philosophy to Platonic doctrine has been remarked by many, and is acknowledged by the realists themselves. Bertrand Russell, in treating of entities which have "a being in some way different from that of physical objects, and also different from that of minds and from that of sense-data," acknowledges that his theory is "largely Plato's, with merely such modifications as time has shown to be necessary." ^ Inasmuch as close simi- larity to Plato's doctrine, as they interpret it, is also claimed by Alexander ^ and by the six authors of The New Realism, who assert that the neo-realist is also a Platonic realist,^ it may be well to refer again to what we take to be the relation between the Platonic and the neo-realistic doctrines. The original and fundamental Platonism, as we have already pointed out, was the doctrine that the true nature of reality is to be found in the universal or logical idea. But, as we have seen, by a process of conversion, fallacious or other, supported by ab- 1 Problems of Philosophy, pp. 142-3. 2 Proc. Aristot. Soc, 1909-10, p. 33. 3 The New Realism, p. 35 ; cf. also Montague, Essays . . . in Honor of Will- iam Jam.es, pp. 113-14; Perry, Journal of Philosophy, VII, 1910, p. 345; Marvin, A First Book in Metaphysics, pp. 108 fif. NEO-REALISTIC DOCTRINE OF RELATIONS 303 stracting from the abstractness of logical idealism and thus dis- guising it, various forms of logical realism were evolved. Plato himself, as we have seen, predicated some sort, or sorts, of reality of some, or even of all, universal ideas ; some were real in the eternal world, all were real either there or in the things of human experience. But that the essence of Platonism, as of the only sound philosophy, was the predicating of eternal reality, usually not distinguished from existence, of all universals, was main- tained by the mediaeval Platonic realists. The neo-realists, like these mediaeval realists, are interested in maintaining the full "ontological status," or reality, of all universals, al- though with some individual variations. The most charac- teristic doctrine is that which has been developed by Bertrand Russell, who, influenced by his mathematical studies, asserts the reality of a world which is neither mental nor physical, but made up of those entities which are the objects of a priori knowledge.^ These entities, he admits, cannot be properly said to exist; they subsist, rather, i.e. they have timeless being in the unchangeable world of universals, with which the mathe- matician and the logician deal. They are not thoughts, though when known they are the objects of thought. ^ Several of the American realists agree with Russell in thus attributing to universals only timeless '^ subsistence," whereas existence in time is reserved for particulars, whether physical, mental, or neutral.^ Alexander, however, interprets Plato as having taught that ideas are real existences, and makes bold to agree 1 The Problems of Philosophy, pp. 139-40. 2 lb., pp. 155-6. In a discussion published some years previously (Mind, N.S., XIV, 1905, V. pp. 398, 399), Russell uses the term "existence" with reference to the entities dealt with in mathematics, explaining, however, that this does not mean existence in the sense in which it is used in philosophy and in common life. Rather does it mean reality for mathematics, the being which all classes have, whether they possess any members or not. But he seems to mean something rather more positive than the mere freedom from contradic- tion which Poincare gives as the meaning of mathematical existence. Like the "Cantorians," Russell and his followers seem to be, in their doctrine of the abstract entities of mathematics, realists (v. Poincar6, Derni^res Pensees, pp. 146, 157-8). 3 E.g. Holt, Journal of Philosophy, VII, 1910, p. 394 ; The New Realism, pp. 366, 372; The Concept of Consciousness, passim; Spaulding, Journal of Philosophy, VIII, 1911, pp. 576-7; The New Realism, p. 180; Marvin, Perry, and Montague, loo. cit. S04 THE PROBLEM OF KNOWLEDGE with him when so interpreted, adding only that the particulars of sense are equally existent, equally real.^ Pitkin also seems incHned to take this apparently more radical, but perhaps more defensible, position.^ It must not be supposed, however, that the neo-reahsts have deliberately set themselves to revive Platonism. Their doctrine is to be explained rather as the product of the neo-reahstic doctrine of consciousness (itself, as we have seen, a result of the neo-realistic doctrine of secondary as well as primary quali- ties), and of certain suggestions along the line of a disguised logical idealism (which amounts almost to logical realism), supported by and even derived from the impression made by abstract mathematical studies, such as have been pursued by Russell. The latter transition we have already referred to.^ In elucidation of the former, we may say that it seems natural to suppose that if consciousness is a mere external relation in the case of physical objects, it cannot well be more in the case of the entities with which mathematics deals, and the universal vahdity of the propositions of pure mathematics is readily in- terpreted ontologically as meaning an eternal reality of the *' universals " or abstract entities with which it is concerned. So long as we are thinking about any object of thought, even the unreal, we must treat it, to some extent and momentarily, as if it were real; and the fallacy of substantiating an abstraction 1 Proc. Aristot. Soc, 1909-10, p. 33. One is tempted to ask why Alexander, as a good member of an Aristotelian Society, did not take his departure from Aristotle rather than from Plato. He would then have confined himself to asserting the existence of universals in the particulars. But, in the light of his representation of the categories as fundamental characters of things {Mind, N.S., XXI, 1912, p. 11), one may conjecture that it has already dawned upon him that this is what he means. McGilvary's procedure and doctrine seem to be more Aristotelian. See Philosophical Review, XXI, 1912, pp. 153 ff. 2 "The Empirical Status of Geometrical Entities," Journal of Philosophy, X, 1913, pp. 393-403. At this point Morris R. Cohen claims to find himself in agreement with the neo-realists. He indorses "a realism of relations or uni- versals like Plato's," which he not very accurately takes to be "the essence of the historic form of idealism." He objects to the distinction here between existence and subsistence, and would accord to mathematical entities full reality, including causal efficiency. He desiderates "a complete theory of categories, or types of existence, to take the place of the rather inadequate distinction between existence and subsistence." Journal of Philosophy, X, 1913, pp. 198-200, 209 ; XI, 1914, pp. 626-7. 3 ChSf y ^nd X, sujyra. NEO-REALISTIC DOCTRINE OF RELATIONS 305 is simply a special case of doing this, and forgetting or per- manently ignoring the nature of what we are doing; it is, as we have said before, abstracting from the abstractness of the universal. The results are in some instances remarkable enough. Among the many creations of thought, or abstractions, which are taken as independently real are Stout's generalities, alternative possi- bilities, non-being, centaurs and other fictions,^ Holt's contra- dictions,2 and Russell's abstract relations, or universals named by verbs and prepositions. For example, ''north of," though it does not exist apart from its terms in space or time, is regarded as eternally subsisting; it ''belongs to the independent world which thought apprehends but does not create."^ Similarly, according to Russell, an infinite aggregate, in spite of the fact that it contradicts the principles of "mathematical induction," on which all our arithmetical operations are based, must be ac- cepted as real.* These entities are thought of, and therefore, it is claimed, they are not thoughts, but objects of thought, real independently of and prior to thinking.^ It ought not to be surprising, then, to find Russell reviving the old doctrine of Reid as to the needlessness of ideas. We do not need ideas, he claims, in order to know, even otherwise than perceptually.^ And yet Russell has felt obliged to make the important ad- mission that "what idealists have said about the creative ac- ^ Proc. Aristot. Soc, 1910-11, p. 187 £f. " Whatever is thought, in so far as it is thought, is therefore real," p. 199. Cf . Montague : " If consciousness is a re- lation, objects of consciousness must be real independently of their standing in that relation." Journal of Philosophy, II, 1905, p. 313. 2 The New Realism, pp. 482-3. 3 Problems of Philosophy, pp. 147, 152-6. 4 Principles of Mathematics, pp. 142-3, 260, 357, 368, etc. Russell, by his rejection of the idea of possibility as an ultimate metaphysical category (Monist, XXIV, 1914, p. 179), makes it necessary to hold to the actuality of the infinite. We would maintain that the only infinite is unending possibility, given unending time, Russell seems to have ruled this out unnecessarily, inasmuch as he holds to the reality of time. 5 Problems of Philosophy, pp. 155-6 ; Principles of Mathematics, p. 46. Russell does not mean that these objects of thought necessarily exist in time, but only that they have timeless being. The new realism here goes to the opposite extreme from the idealistic argument from the egocentric predicament, in which it is assumed, roughly speaking, that what is thought of depends upon thought for its existence. 8 Monist, XXIV, 1914, p. 171. X 306 THE PROBLEM OP KNOWLEDGE tivity of mind, about relations being due to our relating synthe- sis, and so on, seems to be true in the case of error." ^ But, apart from the fact that it seems rather paradoxical to regard all correct thought as absolutely non-productive, and erroneous thought alone as productive, it seems quite arbitrary to inter- pret so differently the nature of mental activities that differ only in what is beyond them altogether, viz. in the reaUty existing prior to and independently of them. If there is mental pro- ductivity in the case of error, there is probably mental pro- ductivity in other cases also. And if so, non-fictitious universals can be adequately interpreted as existing independently, in space and time, in ^particulars, and as represented by ideas, which are products of mental activity and exist only "in" and for consciousness, while fictitious objects (including, we would claim, irrational quantities and the "infinite aggregate") are sufficiently explained as existing only as products of thought. The realm of subsistence is not required, save as itself a con- venient fiction, the product, fundamentally, of abstracting thought.2 Before summing up our criticism of the new reaHsm we must briefly refer to its treatment of the problem of values. Here the question of chief interest will be whether, in accordance with the view that consciousness is an external relation, it will be maintained that value is independent of consciousness of value. That it is thus independent is stated by Moore and Russell. Moore argues that goodness is a quality attaching to things independently of consciousness,^ and Russell makes the general statement that values are independent of conscious- ness.^ This view, however, while it is the logical one for the neo-realist, is not easily tenable, in view of the many values that seem to arise and fluctuate and disappear according to the con- scious attitude of the individual or of society toward the objects concerned. Four of the neo-realists, viz. McGilvary, Alexander, Montague, and Perry, have more or less definitely addressed themselves to the difficult task of constructing a theory of values to harmonize with the facts in the case, and also with the doc- 1 Monist, XXIV, 1914, p. 174. 2 Cf. pp. 84-8, 201-6, 231, 264-5, supra. 3 Principia Ethica, pp. 6, 137. ^ Philosophical Essays, pp. 4—15. NEO-REALISTIC DOCTRINE OF RELATIONS 307 trines of neo-realism. Let us see whether or not they have succeeded in their undertaking. McGilvary's treatment of the subject is rather incidental, but his expedient is to define value as a relation — "sl certain specific relation between the valuable thing and our desires and interests." ^ But this does not seem to be true to the facts. The value of an object may depend upon its relation to some other thing or process, but it is, on the face of it, a quality of that object. Alexander, it would almost seem, has recognized this; at any rate he makes room for values when he speaks of the ap- pearance of the object containing elements introduced into it by the mind. These elements, it would appear, need not vitiate the appearance, provided they are not unduly personal ; but, having been made qualities of the object in its appearance, they are, as such, non-mental.^ In this Alexander is, in our opinion, essentially correct, as far as he has gone ; but it con- stitutes as much a departure from the essential principles of neo-reahsm as do Wolf's theory of hallucination and Russell's theory of error. To the American neo-realists, moreover, with their rejection of the idea of ''mental activity," Alexander's solution of the problem of values would be unthinkable. Montague has chosen, as lying between the definite concepts of quality and relation, the more ambiguous concept of status: value, he says, is the status acquired by any object, existent or non-existent, in virtue of its capacity to satisfy an interest. An object that has the value-status he calls a value. ^ Now if we use the term ''value," as Montague seems to do here, to mean simply an independently existing object, viewed as capable of satisfying an interest, it becomes possible to hold that this "value" exists independently of consciousness, but the triumph is merely verbal. No provision is made for answering the question as to whether the presence of the valuing consciousness, or the existence of the interest, is essential to the object having this value-status. If so, a quality (for that is what "status" really means) of the object depends upon its relation to some- ^ Philosophical Review, XX, 1911, p. 162. 2 Proc. Aristot. Soc, 1909-10, p. 28. 3 Philosophical Review, XXIII, 1914, p. 185. 308 THE PROBLEM OF KNOWLEDGE thing else — another case of internality of relations for the neo- reahst to reckon with. If not, however, one would have to conclude that all objects have, actually and permanently, all the values that they could ever be experienced to have for all possible interests and consciousnesses, including many which logically contradict and cancel each other. The impossibility of escaping this dilemma seems to indicate that Montague has not succeeded in solving, for neo-realists, the problem of values. Perry has discussed the problem very elaborately. His main propositions are that value is the fulfilment of interest ^ and that values are neither dependent upon judgments of value not independent of desire. ^ Or, drawing this last distinction still more finely, value may exist without being known or discovered, if, as seems to be the case, there can be a desire without its being known to be a desire ; ^ and yet, apart from consciousness (as desire, or interest) there can be no value.'* These distinctions we would accept as largely valid, although it seems necessary, further, to make some distinctions, which Perry does not make, in order to gain for the distinctions he does make the measure of acceptance they deserve. For instance, are there not some values, however insignificant and arbitrary they may be, that depend upon explicit awareness, or cognition, for their existence? Again, may there not be unconscious teleological processes, such as the vital processes, by virtue of which certain objects have values which they would not otherwise possess? But in any case — and this is the point of special interest here — the value appears as a quality produced in an object, known or unknown, real or unreal, by a teleological process, whether of mere thought, or of mere desire, or of both, or conceivably of neither. But this is virtually to agree with Alexander, at least to the extent of admitting that some values are the products of consciousness, and this seems essentially the same thing as to say that they are, to some extent, the products of mental activity. Manifestly the ?;aZi^e-producing consciousness is no purely external relation. 1 Journal of Philosophy, XI, 1914, pp. 156-8. 2 The New Realism, pp. 148-9 ; Present Philosophical Tendencies, p. 332. 3 The New Realism, p. 141. 4 Present Philosophical Tendencies, p. 332 ; The New Realism, p. 140. NEO-REALISTIC DOCTRINE OF RELATIONS 309 We are now at the end of our detailed examination of the new reahsm, viewed as a theory of knowledge. The result of our critique seems to be that, in spite of valuable elements in the doctrine, which will be incorporated in our constructive view, there is an undue dogmatism with reference to the extent to which that which is presented to knowledge is real independently of consciousness. The characteristic of dogmatism is frankly admitted by Marvin ; ^ but his plea that science is necessarily dogmatic does not excuse the extent to which the new philos- ophy carries this dogmatism. The physical sciences need to assume dogmatically what, as we would undertake to show, can be philosophically vindicated ; the new realism asserts a larger independent content (at least normal secondary qualities, and in some cases all errors and contradictions and all values) than is scientifically necessary, or than can be philosophically vindicated. Ideally, it is absolute epistemological monism, denying any difference between the object as presented and as independently real (except that the independent reality includes more than is actually presented) , asserting that there is no idea but the independent thing itself, that consciousness, as a rela- tion, is absolutely external, or that, if it is viewed as a mental activity, it produces nothing. Just because it is not absolutely dogmatic, but has undertaken to be critical, it has had to con- tent itself in every case (even in that of Russell; note his admission with reference to error) with affirming something less than this ideal ; and reasons have been given in the discus- sion for believing that, in spite of this departure from their original ideal, what is still affirmed by the neo-realist is con- siderably more than is warranted on critical grounds. It may very well turn out, however, that what the neo-realists have been fundamentally interested in maintaining, viz. the fact of immediate awareness of independent reality in normal human experience, can be vindicated on adequatelj^ critical grounds. It may be that we shall discover that for the experienced object and the independently real object to be numerically the same, it is not necessary that they be qualitatively, even in normal perception, absolutely identical. 1 "Dogmatism versus Criticism," Journal of Philosophy, IX, 1912, pp. 309- 17. 4. CONSTRUCTIVE STATEMENT CHAPTER XIV Critical Monism in Epistemology We have seen reason to reject absolute dualism and an idealistic absolute monism in epistemology, as resting upon in- correct analyses and fallacious processes of reasoning, with their unsatisfactory consequences, against which the former theory struggles in vain, while the latter accepts them and tries to make the best of a bad situation. On the other hand we have not found ourselves able to go all the way with the realistic absolute monists, because of their dogmatizing beyond what is critically justified or necessary, and also because of the many insoluble difficulties into which their doctrine leads them. We seem driven therefore to seek another point of view, from which we shall be able to avoid the fallacies, the subjectivisms, and abstractionisms of idealism in its various forms and the fallacies and final agnosticism of dualism, without falling into the un- warranted dogmatism and insoluble puzzles of neo-realism. The critics of the new realism have scored several points in their attack upon the neo-realistic doctrines of illusion, hallu- cination, and error, and in their criticisms of the view that consciousness is an absolutely external relation. And yet it is not so clear, by any means, that any possible theory within the limits of the accepted definition of realistic epistemological monism has thereby been shown, as one of the most successful of these critics has claimed, to be ''inadmissible." ^ What we are to defend here might perhaps be called epistemological monism and critical realism (critical reahstic epistemological monism), as opposed to the epistemological monism and dog- matic realism (realistic absolute epistemological monism) of the typical neo-realist. By this is meant the doctrine that the object perceived is existentially, or numerically, identical 1 A. O. Lovejoy, Journal of Philosophy, X, 1913, p. 43. 310 CRITICAL MONISM IN EPISTEMOLOGY 311 with the real object at the moment of perception, although the real object may have qualities that are not perceived at that moment ; and also that this same object may exist when unperceived, although not necessarily with all the qualities which it possesses when perceived. Other appropriate but simpler designations for this position are critical realistic monism, critical epistemological monism, and critical monism in episte- mology. It is important to note at the outset that there is no necessary contradiction between Lovejoy's statement ^ that there is mediate and yet valid knowledge and Perry's contention ^ that there cannot be knowledge at all unless there is immediate knowledge of reality. May it not be that there is mediate knowledge, because, and only because, there is first immedi- ate knowledge? If, as we shall maintain, this much of the neo- realist's thesis is defensible, that in ordinary perception there is immediate knowledge of reality which is not dependent for its existence upon its being perceived, it may also be said, in the light of experience, that we often have repeated immediate knowledge of repeated, essentially identical, independent pro- cesses. Indeed, in countless instances we come to be able to predict the later stages of a process of which we have, in this particular instance of its occurrence, immediately experienced only the beginning. Again, we are often practically certain that a process of which we have immediately experienced only the beginning and the end has been essentially identical with what at other times we have had under our immediate observa- tion throughout its entire course. If, then, we define knowledge as certainty of the nature of reality, either in its immediate givenness or in true judgments,^ sufficient for all proper practical purposes, it will be readily apparent that if there is immediate knowledge of independent reality in normal perception, there may also be mediate knowledge of independent reality through the processes of thought, and that the immediate knowledge has made the mediate knowledge possible. 1 Journal of Philosophy, IX, 1912, pp. 681-4 ; X, 1913, pp. 561-72. 2 lb., VI, 1909, pp. 29 fif., 169 fif. ; VII, 1910, pp. 342-3 ; Present Philosophical Tendencies, 1912, pp. 311-13. 3 For definition of truth, see Ch. XIX, infra. 312 THE PROBLEM OF KNOWLEDGE Immediate knowledge of independent reality, then, would make mediate knowledge of the same also possible; and it does not appear what else could do it. Hence it would seem as though, unless realistic epistemological monism can be established as a tenable theory, we should have to face the dilemma of absolute solipsism or absolute agnosticism. Ideahs- tic epistemological monism, at least in any form that avoids abstractionism, cannot logically escape solipsism. Realistic epistemological dualism cannot logically escape agnosticism. Realistic epistemological monism would logically escape both. We must therefore raise the question, Is immediate knowledge of independent reaHty in perception possible ? This question, in the hght of the hypothesis which it shall be our present task to develop, we would answer in the affirmative. The epistemological dualist maintains that what we perceive is existentially and in part qualitatively distinct from the in- dependently existing subject ; it is a second object, at best only somewhat similar to the first. The typical neo-reahst tries to hold that what we perceive is existentially identical with the independent reality, and also qualitatively identical, to the full extent of the perceptual content; it is not a second and perhaps somewhat similar object, but the very same object, with no additional qualities due to its being perceived. A critical reahstic monism would combine the partial truths of both antithetical positions. Bearing in mind that in the self- identity, for us, of physical objects at different times and in spite of certain changes, there is a subjective factor (our pur- pose) and an objective factor (e.g. continuity of physical energy and of certain teleological functions other than our own), we would maintain with the neo-realist that what we perceive is existentially identical with the independent reality, and with the epistemological dualist we would say that it has, when being per- ceived, certain qualities — notably the sense-qualities — which it does not possess when not perceived. In order to be able to maintain this position it is simply neces- sary to apply to sensation the view of consciousness which Bergson applies in his doctrine of memory. In passing from perception to memory, according to Bergson, we definitely abandon matter for spirit ; memory, importing, as it were, the CRITICAL MONISM IN EPISTEMOLOGY 313 past into the present, bringing into the present experience what would not otherwise be there, is a creative activity of spirit.^ Bergson's description is too much in terms of psychological idealism; memory does not really import the past into the present, but creates representational elements in the content of the present experience which stand for past sense-elements ; but even when thus translated into realistic terms, the con- cept of creative activity still remains valid. But what we are concerned to maintain here is that in sensation, as truly as in memory, there is a creative activity of spirit — or of whatever we may choose to call the psychical subject. Upon occasion of certain stimulations, sense-qualities — particular colors, sounds, odors, tastes, and the like — are creatively produced by each psychical subject for itself, and in many cases located with more or less accuracy in or upon the very object in the environ- ment from which the stimulation proceeded. It is a case of coordination of activities, in the first instance those of objects of the environment with those of nervous centres involved in the not purely passive process of being stimulated, and ultimately of such environmental processes as radiation with such psychical activities as are involved in the production of the various color-qualities of objects. The theory is not identical with projectionism and Lotze's "local sign" theory; the sense- qualities are not first ''in the mind," or intraorganic, and then ''projected"; they are created, in each case of sensing, in the particular location in which they are found. Sometimes the qualities produced are not placed accurately upon the object from which the stimulation first proceeded. This is especially the case with heavenly bodies, whose visual qualities are placed not only in the line of the direction of the rays as they enter the eye, but at no very great distance from the observer, just a little beyond human reach from the highest trees or buildings or mountains. This is doubtless because, in the history of the race and the individual, it has worked just as well to have visual qualities thus placed, as it would have, had they been more accurately located. This extension of the activistic interpre- tation of consciousness to sensation as well as to memory and the higher thought-processes would at least have the merit of 1 Matter and Memory, Eng. Tr., pp. 80, 313. 314 THE PROBLEM OF KNOWLEDGE getting rid of Bergson's paradoxical identification of pure per- ception with matter ; the colors and other components of what Bergson calls "images" are not ultimate data, but products of subjective activity. It may be objected that such creative sense-activity is mysterious, and so it is ; but no philosopher will ever succeed in driving mystery out of the processes of life and consciousness, or from any other phase of real existence; the best we can hope to do is to get the mystery properly cornered, correctly located. This mystery of creative psychical activity is simply a special instance of the universal mystery of being, and especially of becoming. With Walter Pater ^ we may say that ''color is a spirit upon things, by which they become expressive to the spirit"; it is at any rate the created product of spirit, if the sensing subject is spirit. But it may also be said that we have here what looks Hke a solution of the problem of the nature of consciousness. For some time, as we have seen, this has been one of the most troublesome of our philosophical problems — especially to the neo-realists. It had long been a commonplace among philos- ophers that the one impregnable foundation for philosophical construction was the proposition, '' Consciousness is." But when the question was raised. If it is so certain that consciousness is, just ivhat is it ? the answer was not readily forthcoming. As has been indicated in our study of the new realism, many of the recent replies to the question may be viewed as constituting a dialectical progress of thought from the concept of consciousness as an existent entity, or quality, to the concept of consciousness as behavior, or activity. The movement of thought in the American neo-realism, however, can scarcely be called a typical synthetic dialectic; the earHer thesis is not included in the later, but excluded from it. Consciousness cannot be a quality of things, it is claimed, because it is not empirically discoverable as such ; therefore it must be a relation between objects. But since it is so difficult to determine just what relation between objects consciousness can be, the suggestion is put forward by some that it is a special kind of activity of the body or nervous system as subject upon the environment as object. Our reasons for rejecting these successive "solutions" of the prob- 1 Quoted by B. Bosanquet, The Principle of Individimlity and Value, p. 63. CRITICAL MONISM IN EPISTEMOLOGY 315 lem have been set forth above; but in connection with our present constructive attempt the general movement toward an interpretation of consciousness in terms of activity is significant. The movement of thought among the EngUsh neo-reaHsts, although it has concerned itself less, perhaps, with the question as to the nature of consciousness, than has that of the Ameri- can school, has come nearer to a satisfactory solution of the problem. They have avoided the handicap of virtually assuming that there can be none but physical existences, and the dialecti- cal movement discoverable in their thinking has consequently been more genuinely synthetic. Consciousness, it is from the first maintained, is a quality of the psychical subject, rather than of the physical object. But it is soon discovered that consciousness cannot be a quality of the subject, unless it is also a relation between the subject and the object ; and further, that it cannot be a relation between subject and object, unless it is at the same time an activity of the subject upon the object. Thus far we can agree. But the English realists seem to be at a loss when they attempt to state the nature of this activity. Moore and Alexander, as we have seen, cover their failure with the seemingly unintelligible, because self-contradictory, notion of a diaphanous activity, an activity in which, apparently, nothing is produced. Wolf is to be credited with having had the courage to depart far enough from the beaten track of the neo-realists to maintain that, in the case of hallucination and illusion, consciousness is a productive activity. But since, as has been shown above, in making the sensing process radi- cally different, psychologically, in normal and abnormal per- ception, in order to explain its different logical value in the two cases. Wolf's doctrine runs counter to well-known psychological facts, his position is one of unstable equilibrium, and as such, untenable. What wx here suggest is that it is possible to inter- pret consciousness, in sensation everywhere and always, as well as in its other forms, as being a productive activity, and that of a unique — but not indefinable — sort. The psychical subject, which we may consent with W. McDougall ^ to call once more the soul, creatively produces — each individual for itself alone, and on condition of certain stimulations — all the various 1 Body and Mind, 1911, passim. 316 THE PROBLEM OF KNOWLEDGE sense-elements which it is able to discover in the surrounding world of physical objects. Something like this seems to be McDougall's view, although he has not developed it in his published works to any great extent. Moreover, he does not seem to have discovered a way of combining this activistic in- terpretation of consciousness with an epistemologically monistic realism. Let it not be objected that in making use of the idea of creation we are reverting to a discredited concept. It is this idea of real productivity which is the original meaning of causahty. The real cause is not a mere ''unconditional, invariable ante- cedent," which does nothing to anything, but which is mys- teriously followed by a mere ''consequent," similarly inert. That, as was noted above, is simply what causality would he, if a psychological idealism or phenomenalism, of the type held by Mill and others, were true. Causation, on the contrary, is productive activity.^ The regularly antecedent event merely gives a clew to the real cause, although for some practical purposes it may be treated as if it were itself the cause. As Reid long ago pointed out, it is not the cause, but a "sign" of the cause.2 The cause is something which does something to something else, and what it does, "the difference it makes,'^ what it creatively produces, is the effect. This whole point of view, applied, as has here been done, to the psychical causes, may be called activistic realism. This conception of consciousness as a unique productive or creative activity of a non-physical subject (an activity further definable in terms of its products) is by no means so strange to philosophical ways of thinking as some might be led by recent ^ A correspondent, referring to the view presented in this chapter, writes : "Your theory. . . meets the facts, solves the puzzles — those of an empirical order. Your hypothesis would do everything — so it strikes me on a first reading. It is as to the admissibility of the hypothesis that my difficulties arise." He then goes on to suggest that the hypothesis be criticised from the Humian point of view. Now I would readily admit that, from the point of view of the Humian or any other thoroughgoing phenomenahsm, the hypothesis of creative causality is inadmissible. But what if all such pure phenomenalism is itself unnecessary, essentially fallacious, and therefore inadmissible ? Indeed, if it were true that only from another point of view could an hypothesis be framed which would "meet the facts, solve the puzzles," would not this circumstance in itself be very good evidence of the essential correctness of that other point of view ? 2 Collected Writings, p. 122a. CRITICAL MONISM IN EPISTEMOLOGY 317 discussions to imagine. The idea of teleological and quasi- teleological united with efficient causahty has been famiUar ever since Aristotle promulgated his doctrine of "entelechy," and recently it has been impressively set forth by Driesch and other vitalists, as well as by Bergson. The earlier attempts of such philosophers as Schelling and Fechner to construe the universe ultimately in terms of organism rather than mechanism point in the same general direction. Indeed, it is worth noting that opposite aspects of activistic realism one-sidedly developed, are even to be found on the one hand in Locke's doctrine of the activity in sensation of external things only, and on the other hand in Leibniz's doctrine of force acting only immanently. The synthetic activity of thought was not only sufficiently emphasized, but given a somewhat mistaken and exaggerated application by Kant, and still more by his neo-Kantian succes- sors, T. H. Green, H. Cohen, Howison, and others. In the systems of Fichte, Herbart, Schopenhauer, and Lotze, not to mention Hegel, the concept of psychical activity figures variously, but in all cases largely. The reconstructive function exercised by means of ideas in judgment has been rather more than ade- quately emphasized by Dewey and his followers. In psychology vindications of the reality of interaction (Ladd, McDougall) and of mental activity (Wundt, Paulhan,^ Angell) have a large and respectable place. Woodworth's new theory of " perceptual reaction " also has some very important points of contact with the view advocated here, and it would almost seem as if his next step might well be the adoption of an activistic interpre- tation of sensation. 2 Moreover, the concept of creation has been reintroduced into philosophy by Renouvier, while es- sentially activistic interpretations of ''free will" have been ably defended in recent years by William James, Eucken, Boyce Gibson, F. C. S. Schiller, Boutroux, and Bergson. Indeed even Ostwald's ''energetics," while not in itself an expression of ac- tivistic philosophy, in its interpretation of the ultimate nature of matter brings important grist to the activistic mill. The argument for an activistic view derivable from human freedom is worth elaborating. It is a very real motive for ^ L'adivite mentale et les elements de V esprit, 2d ed., 1913. * Psychological Review, XXII, 1915, pp. 1-27. See p. 272, supra. 318 THE PROBLEM OF KNOWLEDGE adopting the category of creative causality, and one, we would contend, which is not without logical value in a final synthesis, that without the employment of this category every act of man's life would have to be traced back indefinitely beyond the beginning of that hfe for every factor in its causal explana- tion. The man himself could not be regarded as the ultimate cause of anything; and so he would be logically justified in repudiating all responsibility for his acts. But in view of our intuitive and practical certainty that we are not morally justi- fied in repudiating all moral responsibility, we must adhere to its logical implicate, viz. some measure of ultimate origination on man's part. But if it has been rendered practically certain that there is such a process as creative activity, it does not necessarily involve a violation of the principle of parsimony to suppose that it is present in certain other processes also. In- deed, in varying degrees and forms may it not be present in every instance of becoming? The activistic view of consciousness has the further merit of furnishing the solution of several old puzzles. For instance, it enables us to define psychology, giving to it a subject-matter distinct from that of any other science. Psychology is the science which undertakes to study the psychical subject (soul, or mind) in the light of what it does. It is descriptive of psychical activities. It is not concerned with the sensible qualities of objects, as such, but with sense-quaHties simply as products of psychical activity. But, besides sensing, it under- takes to describe apperceiving, remembering, imagining, con- ceiving, judging, reasoning, feeling, wilhng — in short, all the activities of the psychical subject. Even the troublesome prob- lem as to the "subconscious" becomes, from this point of view, a httle less troublesome. It is at least conceivable that there should be genuinely psychical activities, of which the products are at first dissociated so completely from the prod- ucts of other and possibly simultaneous psychical activities, that the subject may not be, in these latter activities, aware of the former, or of their products ; and also that when some of the later after-effects of these "subconscious" activities should come to be associated with the contents of the ordinary "stream of consciousness," it may be without any memory of the CRITICAL MONISM IN EPISTEMOLOGY 319 earlier activities of which they are the consequence. Psy- chology thus becomes — "pace William James and the now old ''new psychology" — the science of the soul. It studies the psychical subject not apart from, but in, its activities, and these in the complex of their products (sense-qualities, ideas, bodily movements, etc.). The reason why sensing has been so uniformly omitted from the recognized list of psychical activities is probably that it is rela- tively static, as compared with the various "thinking" activities. Sensing is to other psychical activities as the motion of the earth is to the motions of objects on or near the surface of the earth. For ordinary purposes it is not necessary to attend to the mo- tion of the earth, or to the psychical activity of sensing ; with the result that we ordinarily take them as static. In the light of this one easily understands how it is that G. E. Moore, although in his paper on "The Subject-Matter of Psychology" ^ he moved a certain distance in this direction, was prevented from reaching a unitary point of view, according to which "acts of conscious- ness" may be viewed as constituting the entire subject-matter of psychology. He was prevented from attaining to this result by his rejection of the idea that the psychical subject can give properties to things.^ It is very much the point of view advocated here, however, that we find expressed toward the end of McDougall's Body and Mind.^ It is true that in his later work. Psychology, the Study of Behavior, he advocates an extension of the signification of psychology, such as would make it include not only the study of these psychical activities, but also the correlated processes of physiological "behavior." Angell, among other reasonably conservative psychologists, inclines to a similar view. But this is no violation of what we have set forth as the nature of psy- chology in the strict sense of the term. For various good and sufficient practical reasons it may have become expedient to use the term " psychology " to include, besides psychology in the proper and narrower sense of the term, the scientific study of the "behavior" of organisms. 1 Proc. Aristot. Soc, 1909-10, pp. 36-62. 2 76., 1903-4, p. 135. » Pp. 364-5, quoted in Ch. XII, supra. 320 THE PROBLEM OF KNOWLEDGE But the theory of consciousness which we have advanced must be judged very largely by its serviceabiUty in clearing up the philosophical puzzles that have been associated with such phenomena as after-images, hallucination, illusion, color- blindness, etc. Positive after-images are creatively produced on occasion of the continuation, for a brief period after the extra-organic stimulation has ceased, of the same sort of intra- organic stimulation as was dominated by the extra-organic stimulus. Negative after-images are creatively produced on occasion of the stimulation from certain areas in the sense-organ (which are coordinated with those stimulated extra-organically) finally becoming stronger than that continuing to come from the relatively exhausted areas originally stimulated. The effects of color-contrast are quahties of the object creatively produced by the psychical subject, on occasion of the spreading of stimu- lation from the physiological units originally active to others, presumably either in their proximity, or with which these partic- ular psychical activities are coordinated, or both. Dewey's reconstruction of the ''reflex arc" concept^ is carried still further. There is a coordination, successively, of a series of pairs of coordinated physiological and psychical activities. In some of these pairs the physiological is more prominent; in others, the psychical. The simultaneous coordination vindi- cates the partial truth of paralleHsm; but the coordination, both simultaneous and successive, is explained only by the hypothesis of interaction, at least originally, not between the coordinated events, but in all cases between relatively inde- pendent beings, physical and psychical, some of which are beings within other beings (e.g. organisms). The phenomenon of color-blindness is due to a lack of inheritance of the capacity for certain psychical activities. This incapacity, of course, is physiologically conditioned. In short, the whole process of sensing, i.e. of creatively producing certain sense-qualities in objects of the environment on occasion of certain kinds of stimu- lation, is to be viewed as the inherited result of what was first achieved in the lower animals from which the human race has ascended. Moreover, this consideration throws some Ught 1 "The Reflex Arc Concept in Psychology," Psychological Review, Vol. Ill, 1896, pp. 357 ff. CRITICAL MONISM IN EPISTEMOLOGY 321 upon the otherwise puzzHng question as to whether the colors seen by two apparently normal individuals are qualitatively identical, or whether they are qualitatively different, with corresponding differences, simply. Since both individuals inherit their sensing capacity from a common ancestry, it is entirely probable that its products are qualitatively the same, except where there has been a failure to inherit, as with the color-blind. A further sidelight upon our theory, and support of it, may be derived from paleontology. It is a well-known fact that the brilliantly colored — or shall we say colorable? — • flowering plants did not appear — and many extant species would not have survived if they had appeared — before there were animal forms, such as insects, to be stimulated by their selective reflection of light, and so to clothe them — according to our theory — with gay colors, by means of which they might be guided to them, sustain themselves with their honey, and incidentally assist in their pollination. Color in the flowers without the presence of any honey-seeking animal form would have had no biological function that we can discover, and there seems no scientific ground to suppose it existed. Assuming a color-producing capacity on the part of the honey-eating insect, however, we can account for the survival of both the animal and the plant. Hence the principle of parsimony would seem to favor, however slightly, the latter hypothesis. Hallucination is readily explained as the creation of certain sense-quahties, and ordinarily their being placed in real space, on occasion of a stimulation, similar or practically identical, so far as the last stage of its m^ra-organic history is concerned, with the ordinary stimulation, but not proceeding from the usual 6a;^ra-organic cause. The color, or other sense-quality, is, in such cases, put upon the usual cause, as might be expected, for it is with the action of this cause that that particular sensing activity is habitually coordinated. Illusion, being partial hallucination, is similarly explicable. The illusory elements and haUucinatory objects, although really existent and, if spatial at all, located in real space, are existences created by the iadividual sensing subject for itself alone, as is the case with sense-qualities in general, and so are not independently real. Usually hallucination and illusion are not regarded as ''error," Y 322 THE PROBLEM OF KNOWLEDGE unless they are later discovered to have interfered with some of the purposes of the subject. According to our theory there is also an explanation of the fact that in dreams, as compared with normal perception, the sense-qualities are less prominent than the size and shape and other '' primary qualities." A correspondent testifies that in his experience color is never given in dreams, although his sight is exceptionally sensitive to color. In this case it is simply necessary to remember what has been pointed out so clearly by Bergson in his recent discussion of dreams,^ viz. that dreaming is the result of the union of memory products with those sub- dued "sensations'^ which persist during sleep, such as the idio- retinal light ; the relative absence of color in dreams being then explained by the absence of very distinct and permanent colors in the visual field when the eyes are closed, as compared with the colors produced under the stimulus of the rays of light which enter the wide-open eyes of the subject when awake. The further exposition of epistemological monism and critical reahsm leads us to speak more definitely of the distinction be- tween 'primary and secondary qualities. In the light of scientific progress Locke's fist of the primary quahties of physical objects (the qualities which they must be thought of as possessing in- dependently of the incident of their entering into the relation of being sensed and perceived) needs revision ; but it is in con- nection with the production of secondary or sense-qualities that the sharpest deviation from the Lockian philosophy is necessary. Sense-qualities are not produced by external things and lodged in an essentially passive mental receptacle, as Locke thought; neither are they, as Lotze maintained, first produced within an inner field of consciousness and then ''projected" into outer space, with the aid of "local signs"; they are creatively pro- duced, by the activity of the subject, in things or in the individ- ual's own body, just where they are experienced as being. The mind does not passively receive impressions, but, as we shall see, it actively takes impressions of surrounding objects by means of sensation and thought. Sense-qualities are private marks, the production of which was learned by the animal race, as we shall see more clearly later, in a sort of involuntary trial-and- 1 The Independent, New York, Oct. 23 and 30, 1913. CRITICAL MONISM IN EPISTEMOLOGY 323 error process; this capacity has been transmitted to the indi- vidual, so that by a series of inherited and involuntary, but creative, psychical acts, he is able to clothe environing objects with their various sense-qualities. The result is that a more favorable adjustment to the situation than could have existed without it is made possible, and so the sensing capacity proves to have a very decisive survival-value in the struggle for existence. A special class of sense-qualities is made up of the feeling and emotion qualities which are creatively produced, and more or less vaguely located throughout the body, on occasion, as Dewey has pointed out,^ of the return stimulation due to partial inhibition of motor impulses. In the beginning it is most imperative, biologically, that aversion-producing feelings, such as various "sensations" of pain and the general feeling of discomfort, should be the ones produced ; and it is quite evident that such must have been the case with animals that have survived for any considerable length of time. But this does not necessarily lead to the pessimistic inference drawn by Schopen- hauer. If the natural history of feeling were written, it would of course appear that with successful adjustment to the en- vironment stimulus to painful feelings would cease to be active. With the cessation of the pain, accordingly, some sort of "sense of rehef " would naturally and even necessarily be produced to register for the organism the changed situation. But it is not necessary to suppose, with Schopenhauer, that this "sense of relief," creatively produced, as we would say, by the psychical subject, must always be purely negative. The victory over dis- turbing conditions may be signalized by the creation of sense- or feeling-qualities so decidedly pleasant as to more than counterbalance the pain necessarily produced in the previous situation, so that it is by no means necessary to suppose the sum of pains to be greater than the sum of pleasures. The secondary qualities are created, then, and thereby the primary qualities are revealed. Through being clothed with the secondary qualities of sense, material things with their primary quahties, their spatial and temporal location, their comparative extension in space and duration in time, and the quantity, dis- ^ " The Significance of Emotions," Psychological Review, Vol. II, 1896, pp. 13 ff. 324 THE PROBLEM OF KNOWLEDGE tribution, and transformation of their energy, are made available for human knowledge. If it be objected that there is always more or less of temporal and spatial dislocation between the independently real object of the environment and what is im- mediately sensed, that at best what is revealed is the indepen- dent object where it was and as it was when the process which has acted as a stimulus started from it, we must admit that this is true. But it remains to be seen whether this destroys the knowl- edge-value of sense-experience with reference to independent reality. There are two different sorts of cases, viz. those in which the object is at a short distance only from the observer, and those in which it is at a very great distance. In the former case, where an object has been observed at rest in any particular location and in any particular condition for any considerable length of time, we can be certain that with the exception of a small fraction of the last second, it has really been where it still seems to be; and where objects are moving or changing, the slight degree of illusion ordinarily present can generally be allowed for and practically counteracted by thought — which assumes, however, that it is very commonly possible to perceive things where and as they are. It is a further consideration in favor of this assumption in the case of objects in close proximity to the observer that what we mean by the present in practical life has a time span, so that in observing what and where a not very distant and not very rapidly changing object was a very small fraction of a second ago, we are observing what and where it is now, — in the ordinary sense of that word. The case of very distant objects is somewhat different. Here, as has been intimated before, the sense-qualities are placed upon sub- stitute objects, or in substitute locations, which, by reason of their practically equal inaccessibility to touch, for example, may represent the more distant real objects well enough for ordinary practical purposes. Here too, then, practically speak- ing, by means of secondary qualities created, primary qualities are revealed. The distinction between primary and secondary qualities is attacked, of course, from both sides. Extreme realists ask why psychical creativity should be extended so far as to include sense-qualities, while extreme idealists ask why it should not CRITICAL MONISM IN EPISTEMOLOGY 325 be extended further, so as to include primary qualities as well. To the former question it is to be rephed in the first place that while there must be something objective prior to perception to start the stimulation which is the precondition of perception, this does not necessarily mean the prior existence of all or any of the secondary qualities. Furthermore, in science — which is, in principle, simply common sense become sufficiently critical for the more specialized purposes which man has recently de- veloped — there is no need whatever to assume the independent existence of the sense-quaHties, while the independent existence of the primary qualities is necessary for a rational explanation of the causation of the contents of sense-experience. On the other hand, again, the permanent, independent reality of all sense-qualities permanently perceivable by different persons in the same object seems inconceivable, because in many cases these are mutually contradictory and exclusive. An unbearable burden of proof evidently rests upon him who would affirm that what science does not need to assert, and what leads one into contradictory statements about external reality, is nevertheless true. With reference to the idealistic question as to why human psj^chical creativity should not be viewed as furnishing the explanation of primary qualities, as well as secondary, it may be remarked, to begin with, that this conclusion is to be avoided, if logically possible, if we have any interest at all in objectivity of knowledge, as opposed to agnosticism with reference to the reality which stimulates our sensing activities. What it is of special importance for epistemological theory to be able to maintain is that sense-qualities are located not only in the body of the subject, but also often in external objects, so that some of the primary qualities, such as shape, relative size, location, are as directly present to the subject as are the secondary qualities themselves. In opposition to this view it has been common, ever since the time of Berkeley, to object that such supposedly primary qualities as shape are dependent upon the kind of sense-organs we happen to possess ; that if the lenses in our eyes were cylindrical, for instance, we should see objects as very different in shape from what they are in our present experience. To this the answer is that if our eyes were provided 326 THE PROBLEM OF KNOWLEDGE with non-symmetrical lenses we should be able to detect the illusion, and might perhaps even learn to ignore it, as we do the double imagery in all parts except the centre of the field of vision. In the perception, under the other conditions supposed, of what we now see as a square object, for instance, we could easily correct the illusion, in any one of several ways. We could do it either by using the hand to measure the length of its sides, and seeing that the hand, which would be felt to remain the same, varied in visual appearance as did the square; or by turning the object through ninety degrees and back again, and reflecting, from our experience with the parts of our own body as sensed in touch, that the mere turning of a soHd object does not alter its real shape ; or by dropping one's head to one side through an angle of ninety degrees, and finding the appear- ance of the object to change when nothing has been done to it, but only something to the body of the subject. And when it is inferred from the changing aspects presented by the primary qualities of objects in various perspectives, that there is no one shape or size or location that is more real than any other, that all are alike subjective appearance only, we would still contend that it is easily possible to vindicate the truth of the achieve- ment of common-sense knowledge, that no object can have at any instant more than one real shape or size or location; or during any period of time more than one real series of motions or changes ; or any change whatever, save as it is produced by energetic causes. The real object does not change its shape when we change our perspective. All the objective change resulting is in the shape of the projection of the object on a plane perpendicular to the line of our vision ; and we soon learn to perceive an object as square, for instance, even when this projection may not be absolutely square. We would still maintain, therefore, that through the creation of secondary qualities and their location in the body or on other independently real objects of the physical world, its environ- ment, certain primary qualities of these objects are immediately revealed, thus making it possible to hold to realistic epistemo- logical monism and to avoid absolute agnosticism. Primary qualities are transcendently real ; but some of them are some- times empirically real, and this circumstance makes all the dif- CRITICAL MONISM IN EPISTEMOLOGY 327 ference between helpless total ignorance of reality and knowledge capable of almost unlimited progress. The thin g-in-it self is knowable in part : we are practically certain that things exist with their primary qualities, even when they are not known by any human subject. The question of the possibility of knowl- edge of the thing-in-itself is the question of finding in the thing qualities with reference to which the relation of being either perceived or thought of is external. By thing-in-itself is meant here not the thing as it is when not in any relations whatsoever ; that, of course, is an Unding. By thing-in-itself we simply mean the thing as it is when neither perceived nor thought of by any human being, or even, as we may surmise, the thing as it is, essentially unaffected by any mere perception or mere thought, whether human or infrahuman or superhuman. Existence outside of all relations and existence without dependence upon being the object of perception or thought can be identified only on the assumption that all relations are the work of thought. What we maintain is that it is not necessary to assume this. If, finally, any one should be inclined to quibble over the question as to whether, even on our theory as thus presented, any primary qualities are immediately known, since it would always be by means of secondary qualities, the reply is that the perception of these primary qualities is practically, i.e. for all proper practical purposes — and therefore, as we shall see, truly — immediate : it is clearly distinguishable from knowledge of qualities not thus present, such as may be gained through memory or inference. It would be equally possible, if one were inclined to quibble, to maintain that secondary or sense- quahties are known only by means of primary qualities, since their existence is made possible only through the presence of primary qualities of something; or, again, that neither primary nor secondary qualities are immediately perceived, since they are perceived by means of psychical activity. For the sake of completeness at this point it may be said further that the qualities of physical objects are not exhaustively classified as primary and secondary; there are what may be called tertiary qualities also. Primary qualities are those qualities of physical objects which are discovered through sense- activity, but not produced by it. Secondary qualities are dis- 328 THE PROBLEM OF KNOWLEDGE covered in the object only because produced and put there by the subject of sense-activity. By tertiary quahties we mean such quahties as neither exist in the thing prior to the psychical activity of the subject nor are the inimediate product of sense- activity; they are placed in the object, not by sense, but by purposive, though purely psychical, activity of the subject. Or, more briefly, primary qualities are found by sense and thought ; secondary qualities are made by sense and found by thought ; tertiary qualities are made by thought. It must not be sup- posed, however, when thought influences sense-qualities through a series of physiological changes, as when it increases or di- minishes pain, that the resulting sense-quality is a tertiary quality. Corresponding to these primary, secondary, and tertiary qualities, there are primary, secondary, and tertiary relations. Primary relations are such as are independent of their being sensed ; secondary relations would be such as exist only in and through their being sensed, or felt; and tertiary relations would be such as are first established by the thought that thinks them, and for the purpose which that thought serves. It may be well to refer here to values also, for while most if not all tertiary qualities may be regarded as values, it by no means follows that all values are tertiary qualities. A value is a quality which any object has by virtue of its relation to a teleological or quasi-teleological process. Negative values are qualities possessed by objects by virtue of their being obstacles to the processes in question ; positive values attach to objects by virtue of their being either ends or means. Primary values are such as obtain independently of consciousness; secondary values are dependent upon feeling consciousness, but independent of mere thought ; tertiary values are dependent upon thought alone. But there is an ambiguity here which may be misleading, for what is simply a secondary or even a tertiary value so far as a community is concerned becomes a primary value to the smaller included community or to the included individual. Moreover, in the case of individual values, it is sometimes difficult to distinguish secondary from tertiary values. When through contemplation an object is idealized,^ it 1 Cf . J. A. Stewart, Plato's Doctrine of Ideas, Pt. II, pp. 139-40, et passim. CRITICAL MONISM IN EPISTEMOLOGY 329 is not always easy to say how much of its value for the individual is felt, and how much is merely posited by thought ; and indeed certain values may pass back and forth from the one to the other. It may be felt by some that in reverting to the distinction between primary and secondary qualities we are adopting a view so commonplace as to be thereby discredited. But sense- perception is itself very commonplace, and it need not be very surprising if the solution of some of its problems should turn out to be somewhat commonplace too. Indeed it would be rather disheartening if much of the truth about the common things with which philosophy deals should not be found to wear the garb of common life. It is to be questioned whether there is not something not quite wholesome in the tendency to put a premium upon novelty in philosophy. May it not possibly be to the credit of the view presented, rather than the reverse, that it is heretical from the point of view of the philosophies of the day, in that it keeps closer than most of them do to the conservative, critical revision of common sense which is charac- teristic of scientific ways of thinking. But, to return to our immediate topic, it is to be noted that the distinction between tertiary qualities and ideas of primary qualities is especially important. Objects are complexes of primary, secondary, and tertiary qualities (including values) and relations, and not of ideas of these. There are not different "degrees of reality'' ; whatever is real at all is as real as anything can be, although there are many kinds of reality, and reality in and dependent upon many different relations; and although, also, what is real in one relation, e.g. what one dreams, is unreal in another. But while an idea in its psychical relations is as real as anything else, a logical idea, as such, is not a reality at all, but an abstraction from reality. It is not an object but a representation, a proxy re-presentation of an object, or of some quality of an object, or of some relation between objects or relations, functioning vicariously for the presence, the actual presentation, of the object or quality or relation in question. Ideation, the production of these ideas, is a creative psychical activity. In attentive analysis of the presented object, thought- elements are brought into association with the qualities of the object, and thus the way is prepared for the production, when 330 THE PROBLEM OF KNOWLEDGE the object is no longer immediately present, of ideas or repre- sentations of the object, or of some of its qualities, either primary or secondary or tertiary. Now it may happen on a later occa- sion, when the same object is again sensed and thus presented, that some of these ideas of qualities may function instead of the actual presentation of those qualities. In so far as this is the case, the psychical activity is apperception. Now this apperceptive activity may, in familiar situations, very largely anticipate attentive analysis, thus rendering it unnecessary; it makes possible an economy of sensing or of analytic attention. At this point there begin to emerge problems the adequate consideration of which would carry us far beyond the limits of a merely constructive statement, and yet a solution of which is essential to an adequate treatment of the problem of knowledge. In the first place, perception is the only cognitive mode with which we have been particularly concerned ; but when we begin to consider the function of ideas and the possibilities of their manipulation, the question arises as to whether, even granted that there is genuine cognition in perception, all modes of cogni- tion are to be thought of as essentially or fundamentally per- ceptual, or whether there may not also be some altogether different way of knowing reality. This topic must be dealt with in a separate chapter.^ But besides the question as to whether there may not be some way of knowing reality independently of perception, there is a far-reaching consideration which would raise a serious problem as to whether '' perception" itself can be genuinely cognitive after all. In view of the doctrine that certain absolutely a priori forms are necessarily involved in perception, and that these a priori forms are what determine the form, i.e. the ''primary qualities" of objects perceived (or of "phenomena"), rather than the qualities of any independently existing object, it becomes necessary, in order to defend the validity of perceptual knowl- edge, to raise definitely the question of the genesis of these funda- mental forms of what we have supposed to be real cognition, and, in particular, the question of their genetic relation to what- ever independent reality may be supposed to exist. This in- vestigation also will require a separate chapter.^ i Ch. XV. 2 Ch. XVI. CRITICAL MONISM IN EPISTEMOLOGY 331 But even supposing these questions satisfactorily answered — supposing it shown that all cognition is ultimately essentially perceptual, and that perception is genuinely cognitive — it would remain a fact that there is the difference to whi^h we have found it necessary to refer, between what might be regarded as presentation and what would have to be viewed as representa- tion; explicit ideas are indispensable, and the judgments in which they are employed as predicates claim to embody true knowledge. We shall therefore have to investigate the problem of truth, ^ and, finally, the problem as to how one must pro- ceed in order to produce not only j udgments that shall be true, but also, in a way that shall be universally valid, an adequate certainty of this truth, — in other words, the problem of proof .^ In the meantime, however, assuming that critical realistic epistemological monism, which has thus far succeeded where all other epistemological theories have failed, will be shown able to endure all these further tests, we may proceed to make explicit some of the further implications of this theory which we have been endeavoring to expound and defend. One of the things most characteristic of it — and this will become increas- ingly manifest as we proceed to the later investigations to which we have alluded — is its consistent opposition to the long reg- nant Kantian doctrine. At the present stage of our discussion, in addition to the way in which in general our theory avoids and would expose, as unnecessary, the extreme dualism and con- sequent agnosticism of the Kantian doctrine (and that without falling into the extreme one-sidedness of either the traditional idealisms or the new realism), some minor contrasts may be pointed out. Our theory is the opposite of Kant's in that it regards the primary qualities and relations of the object not as the contribution of the subject, the product of its relating activ- ity, but as furnished from the objective side; while the sec- ondary quaHties are regarded, not, with Kant, as the contribu- tion of the object, or, better, of the supposedly unknowable thing-in -itself, but as the contribution of the sense-activity of the psychical subject. Because he made the properly primary qualities of his merely phenomenal and not independently real object subjective, and the secondary qualities relatively 1 Chs. XVII to XIX, infra. 2 Ch. XX, infra. 332 THE PROBLEM OF KNOWLEDGE objective, Kant foredoomed himself to permanent imprison- ment within the walls of agnostic dualism. In view of what we have said and have yet to say in exposition and defence of the view that the primary qualities of objects perceived have also independent objective existence, and that their secondary or sense-qualities are relatively subjective, we are able to maintain that a genuine cognitive acquaintance with independent reality is not only possible, but actual, in normal perceptual experience. Moreover, from our point of view, analytic judgment becomes relatively more important than in the Kantian system. Analy- sis is not of ideas, simply, or of previous mental constructs, but of preexistent and independently existent realities. And analytic judgments, just because they are thus directed toward things and not toward mere ideas, are productive of new informa- tion. Synthetic judgments, on the other hand, are constructive of ideas primarily, not of the objects of perception. The only judgments which are constructive of things — except as prod- ucts of thought are, as such, regarded as (psychical) things — are those comparatively unimportant judgments through which there are added to objects their comparatively unimportant tertiary qualities. Finally, it may be noted that with the solution, here given, of the fundamental problem of epistemology (as well as of metaphysical psychology) the way is opened up for the solu- tion of what we have seen to be the same problem in its most generalized form, viz. the problem as to the internality or externality of relations. We have found that what the object is depends largely upon whether it is sensed or not ; many of its qualities thus depend upon its relation to the conscious subject. But these qualities may, for some particular purpose, be of no importance whatever, and in such a case the knowledge- relation is external to the object. Generally speaking, the knowledge-relation, when a relation of present consciousness, is internal so far as the subject is concerned, and external so far as the object is concerned. That is, for most purposes one may ignore the difference made in the object by its being known and thought of by one's self or others, whereas knowledge is not likely to be sought, or even recalled to mind, unless there is felt to be some practical difference between the subject with and CRITICAL MONISM IN EPISTEMOLOGY 333 the subject without the knowledge in question. When the knowledge-relation is not, however, at the same time a relation of present consciousness, for most purposes it makes no differ- ence to either subject or object ; it is an external relation. But, on the other hand, there is probably no actual relation which might not become important for some conceivable purpose, in which case it would become internal to one or more of the terms related. Whether relations are to be regarded as internal or external to the terms related thus depends upon the purpose with reference to which the question is raised. Theoretically there is no relation which may not be either internal or external. The existence of relations does not commonly depend upon purpose — it does so, directly, only in the case of tertiary rela- tions — but the internahty or externality of those relations does depend upon purpose. In any particular situation, for the pres- ent explicit purpose or purposes of the subject most of the ob- ject's actual relations are external. The doctrine that all re- lations are always internal to all the terms related could be maintained only by establishing the existence of a knowing-will- ing subject for which all conceivable purposes — even the most trivial and the most mutually contradictory — were always being purposed and never reaching fulfilment. But no such ''mad Absolute'' can be rationally supposed to exist; and so there must, from any actual point of view, be some external relations. Second in importance, therefore, to our extension of the conception of creative psychical activity to sensation, as a device for showing the rational possibility of a sufficiently critical epistemology which shall combine realism with epistemological monism, we would place a more than ordinary dependence upon considerations of purpose in the attempted solution of philo- sophical problems in general, and of the problems of epistemology in particular. In bringing to a close this division of our subject, may we be permitted to indulge in some reflections on the status of epis- temology in general ? For more than a hundred years now the probl/gm of knowledge has been the uppermost problem of philos- ophy ; and for those who incline to idealistic ways of thinking, it has come to be regarded as the necessary preliminary, if not to all the sciences, at least to metaphysics. The neo-realists, 334 THE PROBLEM OF KNOWLEDGE on the other hand, tend to discount the importance of episte- mology and even the reahty of the problem ; although all the while their own philosophical discussions are mainly episte- mological. May it not be that the truth they have perhaps but dimly apprehended is that the solution of a problem ought ordinarily to mean the disappearance of that problem, and that the ideaUsts, on the other hand, have made the mistake of supposing that the problem as to the possibility of knowledge (which, in that form of it with which we have been made most familiar, arose incidentally out of fallacious reasoning and the resulting unnecessary confusion of thought) must be per- manently made the propaedeutic to all other philosophical problems — a mistake which is principally responsible for the fact that for some generations epistemology has been made the cockpit of philosophers? The method of idealistic episte- mology is like that of the quack physician ; it first administers a drug which makes the patient's ailment chronic, thus making its own further services seem permanently indispensable. The scientific epistemology which we would recommend pre- scribes a natural regimen for the sceptic, including exercise; it would help the philosophical novice through a crisis incidental to the development of his system of thought, and thus soon makes its further services unnecessary. We would make the statement, then, even if somewhat ten- tatively in view of the further problems to be considered, still with considerable confidence in view of the fatal objections that we have found ourselves compelled to urge against absolute epistemological duaHsm and against absolute epistemological monism, whether idealistic or realistic, that a tenable and the only tenable position with reference to the epistemological problem is that which we have designated a critical epistemologi- cal monism, or, more explicitly, critical reahstic epistemological monism. It regards the achievements of practical knowledge as foundations for further advances. It defines knowledge so as to make it include something which we already had before we began to philosophize. Its results are therefore not offered as the first knowledge, but as a vindication of previous knowl- edge. To reject it is to choose fallacy, or agnosticism; to go beyond it is to dogmatize overmuch. It is not offered as a CRITICAL MONISM IN EPISTEMOLOGY 335 finished demonstration, but as the most reasonable hypothesis in view of all the facts, and as continuing the practical certainty characteristic of the point of view of common sense and common science. If, then, critical monism is indeed the solution of the philosophical problem of knowledge, the thinker ought to find himself able to proceed with his metaphysical tasks very much as if this particular question had never been raised at all ; un- less, indeed, the solution of the problem should incidentally reveal the fact that either more or less than he had previously supposed is entitled to come under the designation of knowledge.^ 1 In this chapter, as also in Ch. XVI, infra, I have included, without the use of quotation marks, some excerpts from my article, "Is Realistic Epi- stemological Monism Inadmissible?" in the Journal of Philosophy, X, 1913, pp. 701-10. B. PROBLEMS OF THE WAYS AND MEANS OF KNOWING (MORPHOLOGY OF KNOWL- EDGE, AND GENETIC LOGIC) CHAPTER XV The Morphology of Knowledge The problem of acquaintance, or epistemology proper, leads naturally over into the problem of the way, or ways, of knowing, or into what may be called the morphology of knowledge. Here the particular problem is whether the different ways of knowing are, in principle and fundamentally, one ; or whether there are modes of cognition which are radically distinct, and between which no real continuity can be traced. This problem has been set for thinkers by the popular prevalence of what may per- haps be termed an absolute morphological duahsm, according to which there are two radically different ways of knowing, viz. experience and reasoning, or pure thought ; or, more exphcitly, perception and conception. The question naturally arises as to why there should be two fundamentally different ways of accompUshing the same end, and the search for a unitary view of the cognitive process begins. As might have been antici- pated, out of this more primitive duahsm there developed an antithesis between two extreme or absolute morphological monisms, the one conceptuahstic and the other perceptuaHstic. For our best illustration of absolute conceptuahstic monism we have to turn to a certain phase of Platonism. Plato, who derived his conception of science from mathematics, as is indicated in the well-known passage, "By the power of the dialectic, reason, using hypotheses ... as steps and points of departure, . . . may soar to the first principle of the whole, and ... by successive steps she descends again, with- out the help of any sensible object, from ideas, through 336 THE MORPHOLOGY OF KNOWLEDGE 337 ideas and in ideas she ends," ^ declares that perception has no part in science or knowledge or the attainment of truth. Per- haps what he means to reject is mere perception ; but he says, ^'We no longer seek knowledge in perception at all, but in that other process, however called, in which the mind is alone and engaged with being," a process which he variously calls thinking, reasoning, or opining.^ Our most instructive example of absolute perceptualistic monism will be found in the philosophy of Bergson. He ob- jects to the platonizing attempt to gain knowledge of reality by means of an examination of human concepts, as taking an artificial and inadequate imitation for the reality,^ which is adequately knowable only in a purely perceptual process, a sensuous and supra-intellectual intuition.^ He uses the term ''knowledge" in speaking of ''analysis" or the conceptual mode ; but this analysis is "knowing" the thing as it is not, but as, for practical purposes, it is convenient to take it. Only in- tuition is knowing the thing as it really is.^ The inadequacy of Bergson' s one-sided perceptualism will be pointed out at length in our critique of anti-conceptualism ; ® so that for the present it will be sufficient to point out, first of all, that, as Bergson himself acknowledges,^ absolutely "pure perception" is psychologically impossible (except, perhaps, in first conscious- ness, or, more doubtfully, in certain rather abnormal states, such as those of extreme mysticism) ; and, in the second place, that much of what Bergson calls intuition in connection with scien- tific discovery is simply hypothesis, born so rich in verifying material, previously accumulated, that it does not need to "work." In the positions just described we find illustrated again in connection with the morphology of knowledge that truth of which we have had such abundant evidence in our investi- » Republic, 511 ; cf. 507 ; see A. E. Taylor, Plato, p. 49. 2 Thecetetus, 185-7. 3 Introduction to Metaphysics, translation by Hulme, p. 75 ; translation by Luce, p. 88. * Matter and Memory, pp. 84-5, et passim; Introduction, passim; Creative Evolution, p. 360. 5 Introduction, passim. « See Ch. XVIII, infra. '' Matter and Memory, p. 26. 338 THE PROBLEM OF KNOWLEDGE gation of epistemology proper, viz. that there has been alto- gether too much ahsolutism in philosophy. Absolute morpho- logical duaHsm, and absolute morphological monism, whether of the conceptual or the perceptual sort, must alike give place to a view which will be critical enough to make room for the meas- ure of truth included in each of these one-sided views, and ex- cluded from the others. And it must be acknowledged that here we receive much help from Kant. Indeed, when our interest is in the morphological problem, the essence of Kantianism is to be found in just that beginning of a critical morphological monism which is perhaps his greatest contribution to philosophy. "Concepts without percepts (intuition) are empty; percepts (intuition) without concepts are bhnd." In the position ex- pressed in this dictum the Konigsberg philosopher, without reverting to absolute morphological dualism, avoided the one- sidedness of both absolute conceptualism and absolute per- ceptuahsm. He showed the necessity of mental activity for all developed perceptual knowledge, and yet insisted upon the necessity of the immediacy of experience, inner or outer, as the touchstone of all that claims to be knowledge. But Kant's critical morphological monism was not fully satisfactory. On the one hand, while perception without defi- nite conception is comparatively blind, if the most original and primitive perception had absolutely no cognitive value, it would seem difficult to account for the fact of such value in later experiences. If what is retained from the first and brought into the second experience is cognitive, it seems rather dog- matic to deny that there was anything cognitive in that first experience. But objection to the other side of Kant's doctrine has been much more frequent, and is more readily supported. From the days of Fichte, Schelhng, and Hegel, on to the present, there have always been some to insist that Kant's phenomenal- ism and metaphysical agnosticism show that he went too far in his injunction against the appHcation of the categories of thought beyond the limits of human experience. Hegel especially emphasized the capacity of thought, beginning indeed in sense- perception, but proceeding, according to its own inner move- ment, adequately to know the nature of ultimate reality. In this he was followed by the various branches of the Hegelian THE MORPHOLOGY OF KNOWLEDGE 339 school, and conspicuously by McTaggart, who claims that a complete metaphysical system of knowledge can be evolved by a purely conceptual dialectic, with no other dependence upon the data of experience than such as is just sufficient to estab- lish content for the most primitive of the categories, that of being. ^ This must be regarded as retrogression, rather than progress from Kant's critical monism; and much the same thing must be said of the neo-Kantian movement, as repre- sented by Hermann Cohen and his school. It does not lapse into a Platonic absolute anti-perceptualism ; but, in its inter- pretation of all perceptual elements as the products of thought- activity, it fails to do justice to the non-conceptual element involved in the foundations of knowledge. Much more valuable, as leading toward the much needed supplementation of the Kantian morphology of knowledge, are Herbart's well-known doctrine of apperception and Royce's recent philosophical discussion of ''interpretation." Royce, objecting both to what he takes to be the Platonic theory of cognition by pure conception and to the Bergsonian theory of cognition by pure perception,^ claims to be able to show defi- nitely how these one-sided views may be synthesized. In spite of his insistence that we human beings are never possessed of either pure perception or pure conception,^ he offers a triadic classification of the types of knowing process, which apparently leaves perception and conception standing as genuinely cog- nitive processes, in distinction from the process in which knowl- edge has its culmination, viz. interpretation ^ In view of this doctrine of three different processes of cogni- tion, the question might weU be raised as to whether we have here anything that deserves to be called monistic in the mor- phology of knowledge, whether it is not to be regarded as still more objectionable than dualism. But the answer to this latter query must be negative, for, since the work of Fichte and Hegel, we can never forget that the triadic may be far more monistic than the dyadic. Where there are but two, there is often hopeless conflict ; but where there is a 1 Studies in Hegelian Dialectic, 1896, p. 46. 2 The Problem of Christianity, 1913, Vol. II, pp. 117-23. 3 lb., p. 121. * lb., pp. 124, 149-52. 340 THE PROBLEM OF KNOWLEDGE third, there is hope of mediation and final unity. And so we would see in Royce's concept of ''interpretation" the prom- ise of further progress beyond both absolute duahsm and the two absolute monisms in the direction of a satisfactory critical monism. The trouble is that Royce seems not to have effected a complete synthesis of perception and con- ception; ''interpretation" seems to be a third process added to the other two, rather than the one all-inclusive mode of cog- nition. The reason for this failure is doubtless to be found, in part at least, in the pecuHar way in which Royce — obviously for the sake of leading up to certain conclusions in the phi- losophy of the history of religion and in metaphysics in which he is interested — defines interpretation. He maintains that it is always a triadic relation, involving an original expression of meaning (a sign), an interpreter, and one to whom it is interpreted.^ Thus it is not only an essentially social process, but also "calls, in ideal, for an infinite sequence of interpreta- tions. For every interpretation, being addressed to somebody, demands interpretation from the one to whom it is addressed." ^ Manifestly, Royce is here defining his term with a view to the metaphysical structure he intends to erect upon it, rather than with a view to the facts to be represented. Interpreta- tion is not necessarily, in the exact sense of the term, a social process ; we often make things to be signs for ourselves. But, in any case, interpretation, itself interpreted as Royce inter- prets it, cannot be made the one typical mode of cognition. And so, while Royce leads us to where we can gain a glimpse of a satisfactory critical methodological monism, he does not lead us into that promised land. WilHam James might perhaps be called a critical percep- tuahst, although his enthusiasm for Bergson ^ has carried him far in the direction of an absolute perceptualism. In Some Problems of Philosophy, however, while holding to the "in- superability of sensation," he admits that concepts give real knowledge, however inadequate to the fulness of reality they may be,* and even insists that the "eternal" truths contained 1 The Problem of Christianity, 1913, Vol. II, pp. 140 ff. 2 Ih., p. 150. 3 ^ Pluralistic Universe, 1909, Lecture VI. 4 Some Problems of Philosophy, 1911, pp. 78-9, 100. THE MORPHOLOGY OF KNOWLEDGE 341 in the map framed by the mind out of concepts would have to be acknowledged, were the world of sense annihilated.^ When we undertake to see whether it is not possible to realize the ideal of a critical monism in the morphology of knowledge, the question occurs whether Royce's three-fold classification (perception, conception, and interpretation) is not capable of being further simplified. In the light of what has been said in our constructive discussion of the problem of acquaintance, undoubtedly the claims of perception to be regarded as genuinely cognitive cannot be gainsaid. Through the activity of sense and whatever mental activity may further be necessary, there is an awareness of the existence and to some extent of the nature of some reality or realities. The question which must be considered, if a position as monistic as is compatible with a thoroughly critical attitude is to be established, is the question to what extent conception and interpretation are either not cognitive at all, or else reducible to practical identity with perception. Taking up conception first, our contention would be that this form of mental activity by itself never amounts to cogni- tion. Conception without perception has no connection with independent reality; its products are but empty forms, ab- stract, cut off from being. So, too, mere imagination, as that form of conception, broadly speaking — or better, of ideation, or thinking — which is least abstract, so far as qualitative detail is concerned, is non-cognitive. Judgment also, in the form of the mere supposition, assumption, hypothesis, tentative generalization, is manifestly not cognition; nor can ratiocina- tion on the basis of such assumption of itself give us knowledge. Its final conclusion is as tentative as its first assumption, until verified. Neither can even the possession of traditional teach- ing with reference to any fact, or as to the truth of any opinion, be regarded as amounting for us to knowledge of that fact or that truth. Merely to think, to have an opinion, or even to have true opinion, as Plato himself insisted,^ is not to know. All of these forms of mental activity, taken by themselves, can give us no more than preparation for cognition; they develop and manipulate the instruments of knowledge, but 1 lb., pp. 73-4. 2 Meno, 98. 342 THE PROBLEM OF KNOWLEDGE they all lack that immediate sensing or awareness of the pres- ence of reality, which constitutes the cognitive core of all perceptual experience. So far as their present cognitive status is concerned, they are related to indisputably cognitive pro- cesses in general as the having of after-images and other forms of hallucinatory sense-experience are related to normal imme- diate perception. They are detached from reality, and so, non-cognitive. In fact, they belong with dreams rather than with cognitions, save that they are more purposively governed and, as a result, more useful. ''Day-dreams," however, are intermediate between useful non-cognitive thinking and the uncontrolled dreams of sleep. Interpretation, on the other hand, we may regard as a form of cognition which is fundamentally identical with perception. The most primitive cognition may perhaps have been, strictly speaking, pure perception ; but it is generally agreed, and that on very good grounds, that perception without apperception is, or would be, comparatively — indeed, almost totally — blind. And what we are here concerned to suggest is that interpretation is simply apperception long drawn out, that apperception is nothing hut an extremely facile interpretation. We have sug- gested here, then, a critical perceptualistic morphological mon- ism; it will be our remaining task in this chapter, therefore, to investigate how far all genuinely cognitive forms of conscious- ness may be viewed as essentially identical with perception. Memory, for example, is generally recognized as being, when normal, genuinely cognitive. But that which distinguishes it from mere imagination is that the representations involved in memory always form part of the " apperceptive mass " in a more or less marginal awareness of the present conscious self. It is a representation of a part of the past life of the present perceived self, or, as James puts it, " the knowledge of an event, or fact, of which meantime we have not been thinking, with the additional consciousness that we have thought or experienced it before."^ And so it conforms essen- tially to the perceptual type of consciousness, while mere imag- ination does not. Historical information, again, is knowledge, while mere tradition is not; it has been brought sufficiently 1 The Principles of Psychology, Vol. I, p. 648. THE MORPHOLOGY OF KNOWLEDGE 343 into relation with our strictly perceptual knowledge to become a part of what is, broadly speaking, our perception of the real world in which we stand. History is community- or race- memory. Verified judgment is, of course, cognitive ; and this, too, viewed as verified, i.e. in association with the sensed or felt reality of which it is the interpretation, is essentially per- ceptual ; it is in direct experience that its verification takes place, the verification-process afterwards taking its place among the facts of memory. Generalization by itself is, as we have seen, mere hypothesis, and as such it is essentially con- ceptual and non-cognitive; but when inductive and viewed in conjunction with the verifying facts as experienced, it is cog- nitive, an interpretation of what is sensed or felt, and so, essentially perceptual. Ratiocination also, on the basis of verified judgments, is simply a drawing out further of the inter- pretative or essentially apperceptive process, and so includes the perceptual feature necessary to entitle it to be regarded as leading to genuine knowledge. It is important to note that in many cases our knowledge of the presence of a certain reality can only be what may be called perception in a complex. We are unable to clothe the reality in question directly and immediately with any one sense-quality ; but by the creation of various sense- and feeling- qualities and by their appropriate location, the presence of that reality may be readily detected, perceived. This is ob- viously the case with the perception of the fact of motion ; we detect it only in connection with our perception of a complex of other realities in successively different spatial relations. And so it is with change in general,^ and with such special changing realities as activity, life, and consciousness. When we perceive the body in certain changing relations with its environment, we perceive a living organism ; we perceive — not as a separately sensed object, but as an object sensed in this complex — the life and activity, and even the conscious- ness, of the individual soul which has the body. Indeed it is not too much to say that, in a somewhat broader sense of the term perception, we perceive the subject of this activity, the "entelechy" (the vitalistic principle in morphogenesis, as ^ Cf. Bergson, La perception du changement, 1911, passim. 344 THE PROBLEM OF KNOWLEDGE Driesch designates it), or "psychoid" (the vitahstic principle in the discharge of function i), or soul, or mind, or self, as the case may be. This perception of change in a complex is simply a special case of the perception of relations. Wilham James has labored to show that, if we are to be conscious of any relation, we must have an elementary feeling of that relation. ^ Now undoubt- edly we do have feelings of certain relations, and possibly of all; but it seems altogether too much to say that we know relations primarily by means of these feelings of relations. Rather do we know relations as included in a complex, which complex we know always ultimately by what is, in a broader or a narrower sense of the word, perception. Of special importance is the fact of the perception of conscious- ness as a unique creative activity, in the complex of an organ- ism of a certain type and its environment, the products of which activity are sense-qualities, memory and other images, ideas, feelings, volitions, etc. In the first place, it should be said that while mere imagination, conception, assumption, inference from mere assumption, the possession of traditional teaching, like illusion, hallucination, dreaming, and erroneous processes of thought generally, are not really cognitive, the perception of any of these processes of imagination, conception, and the rest, is a genuine process of cognition. But it is of still greater importance to note that we seem to have here the means of solving the old puzzles as to the nature and possibility of introspection. By ordinary definition, introspection is consciousness of one's own consciousness, or more exactly, if such a thing can be said to be possible at all, consciousness, preferably immediate, of one's own present consciousness. Now by our definition of consciousness as a unique creative activity, the products of which are sense- qualities, ideas, feelings, and the like, it might seem quite clear that when psychical products are created for the sake of cog- nizing our own psychical products, those which we seek to perceive are always necessarily different from and prior to those by means of which we would perceive them ; so that all * H. Driesch, The Science and Philosophy of the Organism, 1908, passim. 2 Essays in Radical Empiricism, passim. THE MORPHOLOGY OF KNOWLEDGE 345 introspection would seem to be, of necessity, retrospection. This, if true at all, would be most obviously true of psychical ''elements," the products of that creative psychical activity which is the real nature of consciousness. The question of introspection, as the question of an immediate perception of consciousness, must ask whether we can perceive the activity itself. To this the answer would seem to be that while there is no special psychical product, or element, which reveals the presence of consciousness, except a vague feeling of activity, which may be at least plausibly regarded as a feeling of bodily attitude and condition — a circumstance which has led to the notion that we have no right to say there is any consciousness, because we cannot discover it by introspection ^ — it is never- theless true that we do perceive our own consciousness as an activity amid the complex of our bodily life and our physical and social environment. Moreover, according to the view we have set forth, we may be said in a broad sense to perceive our past, in so far as we really remember it, and in a narrower sense to perceive the ''specious present," i.e. the present moment in its relation to a going past and a coming future, by means of perceptual or apperceptual elements which are psychical products which themselves endure with but partial and gradual change for an appreciable time, thus bridging over the temporal transition; and because of this the rigid contrast between introspection and retrospection disappears. Broadly speaking, we perceive our own conscious life and activity as having its place in the past and present and up to the very border of the still uncreated future ; and even in a narrower sense, we perceive our own consciousness in the complex of independent realities and prod- ucts of consciousness which fall within the "specious present." And this can be maintained even if, by the chronometer, the psychical activity directed toward perceiving the present con- sciousness comes after other elements comprised within the unity of this specious present. But if we adopt this solution of the problem, another ques- tion immediately presents itself. It is simply a special in- 1 Cf. Wm. James, "Does Consciousness Exist?" Essays in Radical Empiri- cism, Essay I. 346 THE PROBLEM OF KNOWLEDGE stance of the "egocentric predicament," that we can never introspect an experience which, as a content of the specious present, does not therewith come to be an introspected experi- ence ; but does this not mean that introspection changes essen- tially the character of what it seeks to investigate? To this question, however, our theory of the internality and externality of relations enables us to answer that it is quite conceivable that the fact of introspection should, in many cases, be "ex- ternal" to the remaining content, so far as the purposes which need to be taken into account are concerned. Reduced to a minimum, to introspect is to think of my present experience as my experience or my activity. It is true that so thinking may be the occasion of further psychical changes, but whether or not these changes are sufficient to thwart the purpose to introspect can only be determined by the consequences in each particular instance. But possibly this somewhat elaborate treatment of the problem of introspection is unnecessary. Have we not an intuitive awareness of our own conscious activity? May it not be plausibly contended that every act of consciousness is invariably self-presenting, and that the common confusion on this point is due to the fact that it is never self -representing ? This compels us to raise definitely the question as to the nature of intuition, including its relation to perception, a question which would in any case demand our attention in connection with the attempt to establish, in the morphology of knowledge, a critical monism, and especially so when that monism is a critical perceptualistic monism. What we are concerned with here is the new, or perceptual, intuitionism, rather than the old, or conceptual, variety of intuitional philosophy. This perceptual intuitionism, the doctrine that in sense-experience, or feeling, or both, there is a direct awareness of independent reality of some sort, is most compatible with a realistic monism in epistemology ; but there are certain approaches to it among some of the dualistic and idealistic philosophers. Kant, for instance, has his doctrine of intuition, by which he seems to mean the content of experi- ence at the extreme limit of pure, or non-conceptual, percep- tion. ThiS; however, is regarded as a product of independent THE MORPHOLOGY OF KNOWLEDGE 347 reality, rather than its presentation. Fries and his followers have recognized the psychical fact of an ostensibly immediate awareness, through feeling, of the nature of an independent reality which is never directly presented ; but they virtually deny it any genuinely cognitive character. Volkelt's doctrine of intuition is somewhat similar to that of the Friesians, in that it makes feeling the channel of such intuitive awareness as there is; but he differs from them in apparently attaching, although not without doubt and hesitation in some instances, some cognitive value to such intuition. His list of intuitive certainties, however, is not extensive, and his general attitude is conservative. Bergson's philosophy, as a sort of veiled psychological idealism, is at the same time incipiently realistic ; and reference has already been made to his methodological emphasis upon intuition. His doctrine is very fruitful in con- nection with a perceptualistic monism, but certain limitations are to be noted. In the first place, there is the troublesome question as to just how far he would have us regard as inde- pendently real the object of pure perception. Then there is the obvious difficulty involved in having the absolutely pure per- ception, in which alone, according to Bergson's extreme anti- concept ualism, true knowledge is to be found. Moreover, as has been said above, what is set forth as marking the place of intuition in scientific investigation seems really to be nothing more — at least ordinarily — than the produc- tion of the unifying hypothesis after facts sufficient to verify it have been accumulated. But among those not avowedly realistic in their general epistemology, perhaps no one comes nearer to a perceptual intuitionism than W. E. Hocking, with his insistence upon the cognitive function of feeling. While finding much in this that is suggestive and that seems tenable, we should have to dissent, nevertheless, from his use of this line of thought as an argument for theoretical idealism. More- over, while agreeing with him in rejecting the Bergsonian anti- conceptuaUsm, we should have to raise the query whether, in his synthesis of Hegelian idealism with Bergsonian intuition- ism,i he has not gone even more than dangerously far in the direction of the dogmatic rationalism of the older Hegelians. ^ "The Significance of Bergson," Yale Review, III, 1914, pp. 325-6. 348 THE PROBLEM OF KNOWLEDGE A similar dogmatism, as we have seen, marks the intuition- ism of the neo-reaHsts. But, within the hmits of the (as we hope) less dogmatic '' critical realistic epistemological monism" to which we have indicated our adherence, what becomes of intuition as a mode of cognition, and what bearing has our answer to this question upon our search for a critical perceptual- istic methodological monism? In the first place, it is funda- mental to our epistemological view that we have an immediate, or, if one chooses so to use the term, an intuitive awareness of the sense-qualities, feelings and other psychical products for which our own psychical nature is responsible. This in recognition of the Kantian doctrine of the '^intuition" of the manifold of sense, and of the conscious relation which is the nearest we ever come, after early infancy, to *'pure percep- tion." But ''pure perception" is not the only cognition. In perceiving sense-qualities we also have a direct or intuitive awareness of certain primary qualities of the independent realities of our environment. Moreover, in and through our feeling-consciousness we have a practically intuitive awareness of various values. Indeed it would seem that, whereas rela- tions are most commonly cognized by being analyzed out of an essentially perceptual complex, but may also be more or less definitely felt, values on the contrary are probably most com- monly cognized by being felt, i.e. in a more distinctly intuitive way, although they may also be found by analysis of a given complex. But the interest in maintaining a positive empirical intuition- ism usually centres in the doctrine that through our feelings, as distinguished from the sensations of the special senses, we can perceive not only certain existences, but to some extent the nature of those existences. What is claimed is a sort of direct or, in the narrower sense, perceptual awareness of what is ordinarily regarded as knowable only by inference, or in some other mediate way, even if this mediate cognition may also, as we have here claimed, be interpreted as itself ultimately and fundamentally perceptual. We may concede at once that hypotheses are often suggested in such a way as to be accom- panied by the feeling that they are true; but we must not allow to pass unchallenged the assumption that such feeling is THE MORPHOLOGY OF KNOWLEDGE 349 always valid ground for confidence. And yet, on the other hand, we must insist that this anticipatory feeling of the truth of an hypothesis is not in itself always ground for suspecting its falsehood. Obviously, in the light of experience, there is no absolutely uniform relation between the feeling that an hy- pothesis is true and its being actually true ; and yet, on the other hand, this feeling cannot be dismissed as having no significance. We would claim that it sometimes gives ground for confidence, and sometimes ground for suspicion. We shall not go into this topic very fully in this connection, as it belongs also to the problem of mediate knowledge, a consideration of which is to follow. And yet it may be said here that what is ''intuitively" felt to be true is generally something which it is pleasant to beHeve ; and certain highly emotional and wilful, and perhaps somewhat uncritical and unanalytical, natures tend to affirm it as true on these psychological rather than logical grounds. But if we would be adequately critical we must recognize that even if it may sometimes be that the pleasant hypothesis is pleasant because it is useful, it may also sometimes be pleas- ant in spite of the fact that it is not useful, but quite injurious ; and further, that even if it may sometimes be that the useful hypothesis is useful because it is true, it may sometimes be useful (relatively to some proximate end) in spite of the fact that it is not true. It is only the pleasantness which is due to the usefulness which is due to truth which can be taken as an indication of truth ; and the task of distinguishing such pleas- antness from all other varieties of agreeable emotion which may be associated with the occurrence of hypotheses is by no means easy. It is important, finally, to note that the appreciation of values, which is commonly ''intuitive" and always fundamen- tally perceptual, may function in the recognition of certain realities. The work of a certain artist, for example, may be perceived as being such by the sort and degree of value which it possesses. This obvious truth may prove capable of important applications. Our conclusion, then, is that while the absolute perceptual- istic monist is over-dogmatic in affirming the purely and narrowly perceptual character of all cognition, it is nevertheless true 350 THE PROBLEM OF KNOWLEDGE that, broadly speaking, all cognition is, ultimately and funda- mentally and indeed in its innermost essence, always percep- tual. Mediate knowledge is knowledge only by virtue of the support of immediate knowledge ; and so, its relation to imme- diate knowledge being rightly regarded as an internal relation, it enters into its knowledge-status only as a part of the ma- chinery of apperception, or, to use Royce's term, of interpre- tation. Inasmuch, however, as all ordinary perception involves apperception, interpretation, it may also be said to be, in some broad sense of the term, conceptual. But knowing is never merely conceptual. Conception, we repeat, is cognitive only in interpretation, i.e. in combination with perception. CHAPTER XVI The Genesis of the A Priori Before we proceed further, our critical realistic epistemo- logical monism, if we are to be able to regard it as adequately established, must be fortified against possible attacks on the ground that it is incompatible with any current interpretation of that element in human cognitive activity which, in its causa- tion, is prior to the experience of the individual. The classical interpretations of this a priori element may be grouped under the following heads : absolute genetic dualism, rationalistic absolute genetic monism, and empirical absolute genetic mon- ism. These must be examined with reference to their com- patibiHty with our doctrine of acquaintance; and if it should appear on the one hand that none of them can be held consist- ently with our critical epistemological monism, and if, on the other hand, it should not appear that any one of the three is demonstrably valid, it will then be incumbent upon us to inquire whether or not a fourth theory, in itself tenable, and agreeing with our general epistemological theory, may not be estabHshed. Such investigations would take us into the field of psychogenesis, and at least into the borders of what has been called Genetic Logic, ''the genetic science of logical process." ^ Of absolute genetic dualism with reference to these cognitive factors which become explicit as the fundamental forms of thought, the best illustration is to be found in the philosophy of Kant, whose doctrine on this point has been immensely influential. His teaching on this subject is an absolute dual- ism, made up of an absolutely rationahstic or non-empirical doctrine of the origin of those fundamental forms of cognition which seem to be already involved in the first intelligible expe- 1 J. M. Baldwin, Thought and Things, or Genetic Logic, Vol. I, 1906, Intro- duction. 351 352 THE PROBLEM OF KNOWLEDGE rience of the individual, and an absolutely empiricist doctrine of the origin of the sense-material, or contents, of cognition. Mind, on this view, is absolutely active {i.e. creative) with refer- ence to the forms of objects, and absolutely passive with reference to their sense-quahties. An unknowable independent reahty presumably produces these sense-data within our experience, and these are worked up into objects according to necessary and universal ways of apperceiving which are not further expH- cable. The agnosticism is logically inevitable. Obviously it would be pure dogmatism to assume that the independent things-in-themselves are even Hke the objects of sense; for this would mean nothing less than the affirmation that, after the sense-materials, as products of an unexperienced cause, have been built up in a certain way into objects, i.e. radically modified, by another cause (the human mind), these constructs miraculously happen to copy one of their unknown causes, although no one can show that they do so, nor any reason why they should. That is, even representative ''knowledge" of independent reality is precluded; it could never be certain, or even probable, and so could not be knowledge, even if it did happen to be true representation. Much less, then, could there be, on the basis of the Kantian absolute genetic dualism, presentative knowledge, such as is required by our reaHstic epistemological monism. Most modern philosophical thought, both pre- and post- Kantian, has tended, with reference to this genetic problem, to an absolutely monistic position, either rationahstic or em- pirical. Of pre-Kantian absolute genetic monism in its ration- alistic form, the doctrines of Descartes and Leibniz furnish us with good examples. Descartes's pure rationaHsm led him to raise the question why we should take the necessities of thought as giving us knowledge of the nature of reality. The only solution of this problem for Descartes was to be found in the postulate of a holy and perfect God who would not deceive us ; whereas the existence of such a God, he had to admit, could be estabHshed only by means of an argument which seemed to him (if not to many others) to be rationally necessary and thus one of those very processes the vaHdity of which must remain problematic until the existence of this Perfect Being THE GENESIS OF THE A PRIORI 353 is assumed. It is a clear case of a logical circle. In Leibniz, however, we see the thoroughgoing rationalist; according to him the object of immediate apprehension is in toto the product of the individual monad which experiences and knows it. After Kant there soon occurred, notably in Fichte and Hegel, a recrudescence of this doctrine that the creative activity of the Ego, or of Thought, is sufficient to account for all the con- tents of experience, including even those sense-elements which seem most unmistakably to be "given." More recently still, many neo-Kantian thinkers, such as T. H. Green, and espe- cially H. Cohen and his school, have taught that the object is the exclusive product of the a priori or rational activities of the thinking subject. According to Green the object is made what it is by its relations, which are, in all cases, the work of thought. Cognition is always construction, according to Cohen; the sense-qualities which we know are known only as the object of the thought which constructs them, as truly as thought constructs any other object. Manifestly such a doctrine could never accommodate itself to a critical realistic monism in epistemology, but only to an idealistic interpreta- tion of the object of experience. If there is no reality which exists, or can exist, independently of thought, all question as to the possibility of an immediate experience of such an inde- pendently existing reality becomes nonsensical. As in the case of its rationalistic form, so also in its empiri- cal form absolute genetic monism has its pre-Kantian as well as its post-Kantian representatives. Of the former, at least within the modern period, Hume must be regarded as the most important. All our ideas, even those of pure mathematics, he claims, are copied from our impressions, and these sense- impressions are simply "data" passively received, the ultimate of knowledge as of experience. More absolutely empiricist than Locke, who recognized a certain activity and initiative of mind in reflection upon the simple ideas of sense, Hume made even reflection a purely passive process, the ideas, or faint im- pressions, being simply the consequents found, as a matter of fact, habitually and inexplicably to follow their inert antece- dents, the more vivid impressions of the senses.^ 1 A Treatise of Human Nature, Bk. I, Pt. I, §§ I, II; Pt. Ill, § L 2a 354 THE PROBLEM OP KNOWLEDGE Hume was followed, in this empirical form of absolute genetic monism, by the associationists, notably James Mill and his son, John Stuart Mill, according to whom all the contents of human experience and thought are simply series of passive psychological antecedents and consequents, none of which represent anything more objective or knowable than a "per- manent possibility of sensations." On this view, what is immediately experienced is never anything but what depends upon its being experienced as an essential condition of its exist- ence, whatever invariable antecedents there may be besides. An important modification of the empirical form of absolute genetic monism was developed by Herbert Spencer. The seeming insufficiency of the experience-hypothesis to explain reflex actions and instincts he explains as due to the fact that these automatic psychical connections have resulted from the registration of ''experiences continued for numberless genera- tions." He assumes that the various strengths of different psychical relations are proportionate, other things being equal, to the multiphcation of experiences. An infinity of experiences would produce an indissoluble psychical relation; and though such infinity of experiences cannot be received by a single individual, yet it may be received, it is claimed, by the succes- sion of individuals forming a race. Thus the genesis of all instinctive elements of consciousness, including the forms of intuition and of thought, is explained on the single principle of frequency of repetition of experience in the history of the race, "supplemented by the law that habitual psychical suc- cessions entail some hereditary tendency to such successions, which, under persistent conditions, will become cumulative in generation after generation."^ But even on this view, it must be admitted that our critical realistic epistemological monism would be untenable. If the data of experience are received in a purely passive manner, there is no way of knowing that we experience independent reality, or even a copy of it, whether it is the experience of the individual or of the race that is concerned. Now from the scientific point of view there is a strong pre- sumption in favor of some such natural explanation of the 1 H. Spencer, Principles of Psychology, § 207. THE GENESIS OF THE A PRIORI 355 empirical (or at least natural) genesis of what is, relatively to the individual, a priori. Human consciousness has a genesis, and it is only sensible and scientific to seek a unitary, natural explanation of all its elements. On this account empirical monism, especially in the modified and less radical form in which we shall present it, is scientifically preferable as a genetic theory, other things being equal, to an absolute genetic dualism. It is also preferable to the rationalistic form of absolute genetic monism, we would contend, as being a more obvious and less strained interpretation of the facts. An absolute empirical monism is not wholly satisfactory, however. Apart altogether from the controversial doctrine of the inheritance of acquired characters assumed in Spencer's evolutionism, there is the difficulty emphasized by William James, that "the manner in which we now become acquainted with complex objects need not in the least resemble the manner in which the original elements of our consciousness grew up." ^ We now ordinarily perceive quite readily the nature of the present object, just because we have preformed categories for all possible objects; and we have no right to assume that the mere existence of things to be known was originally sufficient to bring about a knowledge of them, because even now it is not always suffi- cient.2 James accordingly propounds his own theory, that the original elements of consciousness came into being as ''spon- taneous variations, fitted by good luck (those of them which have survived) to take cognizance of objects (that is, to steer us in our active dealings with them), without being in any intelligible sense immediate derivations from them." ^ Time and space-relations, he still holds, are impressed from without ; the same is true, he claims, of ''an immense number of our mental habitudes, many of our abstract beliefs, and all our ideas of concrete things, and of their ways of behavior." "Here the mind is passive and tributary, a servile copy, fatally and unresistingly fashioned from without." ^ But there are certain combinations, such as the forms of judgment, "which, taken per se, are not congruent either with the forms in which reaHty exists or in those in which experiences befall us," and ^ W. James, The Principles of Psychology, Vol. II, p. 630. »/6. 3/6., p. 631. *Ib.,p.Q32. 356 THE PROBLEM OF KNOWLEDGE which thus give evidence of selection, emphasis, and, it may be, other forces, unknown to us.^ James concludes therefore that the "ideal and inward relations amongst the objects of our thought which can in no intelligible sense whatever be inter- preted as reproductions of the order of outer experience,'^ and which are often ''far more interesting to us and more charming than the mere rates of frequency of their time and space-conjunctions," are all ''secondary and brain-born, 'spontaneous variations,' most of them, of our sensibiUty, whereby certain elements of experience, and certain arrange- ments in time and space, have acquired an agreeableness which otherwise would not have been felt." ^ "The theoretic part of our organic mental structure . . . can be due neither to our own nor to our ancestors' experience." ^ With reference to this theory of James, it is to be noted in the first place that, while still strongly empirical, it is not an absolute empirical genetic monism. It offers a purely empirical explanation of time- and space-relations; but of all proposi- tions which express the results of a comparison it gives a sort of subordinately rationalistic interpretation, although always within the limits of the natural or " naturaHstic " explanation in terms of spontaneous variation and the survival of the fittest,^ his doctrine being in general consonance with that of the neo-Darwinians and opposed to that of the neo-Lamarck- ians.^ The view as a whole may be regarded as representing a critical empirical genetic monism, although not necessarily the only, or even the most satisfactory, form of such doctrine. It is fundamentally empirical and seeks to adhere as closely as possible to the empiricist doctrine, but, within the limits of a unitary theory, it makes great concessions to the rationaUstic view of the a priori. From the point of view of our epistemological interest, how- ever, the view of James, even if it should be felt to be in itself highly defensible, is not to be left free from attack. Scarcely more than the theories previously examined is it compatible with the critical reaKstic epistemological monism which we 1 W. James, The Principles of Psychology, Vol. II, pp. 633-4. 2 lb., p. 639. 3 76., pp. 677-8. * lb., pp. 644, 676-8. « 76., pp. 678-88. THE GENESIS OF THE A PRIORI 357 seem to have found to be the only tenable positive solution of the problem of acquaintance. On the one hand, in so far as a passive empiricism is retained with reference to some elements, the difficulties we have already urged against absolute empirical genetic monism remain. On the other hand, if any consider- able number of the necessary forms of thought or, more particu- larly, if all the more fundamental ones, except, perhaps, those of space and time, are merely fortunate spontaneous variations which enable us to adjust ourselves satisfactorily to our envi- ronment, then, even if a realistic epistemological monism should happen to be true, we should never be able to know it. We could never know that an independently real environment had come to be within immediate experience ; we could never know that the product of the combination of the immediate data of sense, dependent for their existence upon consciousness, with the products of the activity of thought, or at least of apperceptive consciousness, was not all the reality to be either experienced or believed in. Consequently, if we chose to be guided by the principle of parsimony, we should have to reject the hypothesis of a realistic epistemological monism, even if we could not refute the suggestion of its truth. But, disre- garding for the moment this "law" of parsimony, and suppos- ing the unprovable doctrine of realistic epistemological mon- ism true, we should have to try to explain in some way the marvellous continuous coincidence of the construct with the independent reality. The hypothesis of "accidental" varia- tion and natural selection would then seem, if we calculated the chances of such an "accident," according to the "law of proba- bilities," probably less plausible than that of either old-fashioned teleology or Bergson's "creative evolution." Some non- mechanical factor would seem necessary adequately to account for the appearance of the required forms of mental activity. But suppose we test James's hypothesis of the a priori (as made up of spontaneous psychical variations selected by the environment) on the assumed ground of absolute epistemological dualism. There could, of course, be no immediate knowledge ; could there be any knowledge at all? At this point we shall have to anticipate to some extent the results of our discussion of mediate knowledge. If we define truth as representation of 358 THE PROBLEM OF KNOWLEDGE reality sufficient for our practical purposes in the situation in which we predicate an idea of reality, may it not be maintained that, if we feel no insufficiency of the judgment for our prac- tical purposes, we have, in spite of our never having any imme- diate knowledge of independent reality, what may be called mediate knowledge of what may be called truth ? The answer is that it would be possible to define mediate knowledge in such a way as would allow an affirmative reply. We could call knowledge readiness to act on ideas that work satisfactorily, but calling it so would not make it so. Might it not also be maintained that what we would have would be simply a prac- tical substitute for knowledge where real knowledge is impos- sible, and that nothing but an extreme pragmatism would identify the one with the other? We have known all along that we have either knowledge, or some practical makeshift for it ; but to recognize this is not to solve the problem as to whether we have knowledge or not. The trouble with an abso- lute epistemological dualism is that we do not know that the independent reality exists ; we have assumed it, to be sure, but on reflection we find that an idealistic epistemological monism seems equally defensible. If we never experience immediately a (physical) reality which has existed independently of our experiencing it, how do we know that any such reality exists? But if we do not know enough about the independent reality to know that it exists, we do not know anything about it at all. The proposed definition of mediate knowledge, then, according to which the epistemological dualist could have knowledge of independent reality, even though no such reality had ever been experienced by him, must be pronounced inadequate. Let us then pass over to the suggested idealistic point of view, and ask whether one holding James's theory of the a priori could consistently hold to the possibility of knowledge. We see at once that he could have not only immediate knowl- edge of reality, but, also, if the pragmatic view of truth and knowledge be justified, mediate knowledge as well. Shall we not take this then as an indication, or even as a proof, that idealism (even if it should have to become a disguised idealism, eventually) and, incidentally, pragmatism, are true, even if attempts to prove idealism directly can seem to succeed always THE GENESIS OF THE A PRIORI 359 only at the expense of fallacy? Can we solve our problem of the possibility of knowledge on James's theory of the a priori, and at the same time find for the first time a good argument for idealism, not to speak of pragmatism? We can answer in the affirmative, if at all, only if no other theory of the a priori that is in itself tenable can be found compatible with the possi- bility of knowledge. Is any such alternative theory tenable? Our reply is that in the present state of our knowledge there are three possible theories, no one of which can as yet be de- clared untenable, and any one of which would serve as a basis for asserting the possibility of knowledge, both immediate and mediate, on the basis of a realistic epistemological monism. These theories are as follows : first, that under certain condi- tions, at least in the psychical realm, the transmission of an acquired character to later generations may take place, and that certain of the most fundamental of our mental ^'forms'' of thought are, as related to the experience of the individual, a priori, but, as related to the experience of the race, the result of impressions of the general nature of reality which have been taken by mind in its exploring activities (sensing and other creative psychical activities) ; second, that one of the spon- taneous variations, or mutations, which has occurred and be- come hereditary in the course of evolution is such a high degree of mental alertness and impressionableness as would make possible the very rapid learning, on the part of each individual, of the most fundamental ''forms" and relations of reality, and so of what ought to be, or must be, the fundamental forms of thought — so rapid, indeed, that the process may seem to be either one of inheritance of an ''acquired" character or one of simple participation in a character universally native to mind, and not one of learning, or "trial and error," at all ; and third, a combination of the mutually compatible elements of the two theories just stated. Among these three theories there can be found, we would claim, a theory which is at least as defensible in itself as James's theory of the origination of the *'a priori'* as a mere "spontaneous variation" to be preserved by natural selection ; a theory, moreover, which would be entirely compatible with our doctrine of immediate knowledge on a reaHstic basis. 360 THE PROBLEM OF KNOWLEDGE It will thus be seen that, while our epistemological doctrine would be defensible on the theory of the transmissibility of the effects of use (which theory, it is admitted, even by those who criticise it on methodological grounds as being ''not a legiti- mate hypothesis" in biolog^^, nevertheless "may be true"),^ it does not necessarily stand or fall with that theory. Even if the use-inheritance theory should become utterly discredited, the second of the three theories just mentioned would still be unrefuted and highly defensible. But the last word has not yet been said on the inheritance of acquired characters. The question has been exhaustively discussed, without definite result, with reference to gross struc- tural characters ; but, although, as the psychologist McDougall has remarked, it is in the study of behavior that our best hope lies of answering the question of the transmission of acquired characters,^ almost nothing has been done as yet systematically to investigate the problem in this field. And yet some im- pressive apparent instances of the inheritance of acquired function have been observed by chance and recorded.^ Now in connection with this idea of the inheritance of func- tional characters acquired through use, the field of investiga- tion which is most germane to our present interest is that of the instinctive elements in intellectual consciousness. And it is important to note at the outset that an increasing place is being given, by students of the subject, to consciousness, and even to cognitive consciousness, in instinctive behavior. Thus while in Hobhouse ^ we find instinct described as an adaptive but not intelligent combination of reflexes,^ and even in Berg- son the doctrine that intelligence in connection with conscious- ness is accidental and the sign of a deficit of instinct,® we find McDougall maintaining that every instinct involves knowing, as well as feeling and conation, with reference to its object.^ 1 E.g. Hugh Elliot, in Introduction to J. B. Lamarck's Zoological Philosophy, Eng. Tr., 1914, pp. xxxviii-liii. 2 Psychology : the Study of Behavior, pp. 177-80. 3 G. J. Romanes, Mental Evolution in Animals, 1884, pp. 195, 196-7 ; E. Rignano, Upon the Inheritance of Acquired Characters, 1905, Eng. Tr., 1911, pp. 162, 171 ; Jordan and Kellogg, Evolution and Animal Life, 1908, p. 202. * Mind in Evolution, p. 67. ' Cf. M. Parmelee, The Science of Human Behavior, 1913, p. 226. * Creative Evolution, p. 145. "^ Social Psychology, pp. 26-7. THE GENESIS OF THE A PRIORI 361 Again, whereas James took the once radical ground that, after its first performance by an animal with memory, an instinc- tive action ceases to be purely blind and unintelligent,^ it is now maintained by Stout that since learning by experience is itself an intelligent process, the intelligence involved in instinct cannot be purely an after-effect of learning by experience, but that it must have been present to. some extent in the first per- formance of the instinctive act.^ But it is to be noted, as Lloyd Morgan has pointed out,^ that this doctrine of the essential place of intelHgence in instinct involves, conversely, the fundamental place of instinct in intel- ligence, or, in other words, the doctrine of an inheritance in some instances, of meaning. And even Lloyd Morgan him- self, whose presuppositions, as a consistent parallelist, have always been in favor of the mechanistic, or, as he calls it, the '' physiological" interpretation of instinct,^ confesses that he is not prepared to deny the presence of inherited meaning in some cases at least. ^ But in view of the fact that meaning is normally acquired through experience,® we seem almost forced, finally, to infer the inheritance of meaning originally acquired in and through ancestral experience. It may be felt at first that H. R. Marshall goes too far when he asserts that reason is a special development of instinct ; ^ but arguments and specu- lations have recently appeared, notably in the writings of Berg- son, which make some such conclusion seem not unreasonable. In instinct, according to Bergson, there is an innate knowledge of definite objects ; but intelligence also, he claims, has knowl- edge which cannot be adequately explained by pointing to what the individual has learned as a result of his own experience, simply. Intelligence possesses innate or instinctive knowledge, not of definite things, but of relations, such as those of like to like, content to container, and cause to effect. This doctrine of innate intelligence, which Bergson himself is careful to dis- tinguish from the long-since discredited scholastic theory of ''innate ideas," is to be interpreted as meaning essentially 1 Principles of Psychology, II, p. 390. ^Manual of Psychology, 3d ed., 1913, pp. 349-54. 3 Mind, N.S., Vol. XXIII, 1914, pp. 169 S. * Instinct and Experience, p. 110. * Mind, loc. cit., p. 179. « See Stout, op. cit., pp. 169, 183-4, 385. ^ Instinct and Reason, p. 462, 362 THE PROBLEM OF KNOWLEDGE that the mind possesses innate knowledge of the most funda- mental categories required for the interpretation of nature, because and in the sense that it makes an instinctive use of them.^ But Bergson did not reap the full reward of his theory, because of his vacillation on the issue of realism and idealism. He may be largely right in discounting the categories of mechan- istic science as means of interpreting life, on the ground that our intelligence, as it leaves the hands of nature, has for its chief object not life, but the unorganized solid, so that intellect was fashioned to the form of inert matter, and as a result mis- takenly tends to impose the categories or forms of thought derived from matter in its interpretation of all objects, even the process of life itself ; ^ although even here the fact seems to be overlooked that our most fundamental concept of causality seems to owe its form to experience of a life-process, rather than to experience of inert matter. But what we would criticise especially in Bergson in this connection is the extent to which he seems willing to concede to the idealist the mental origin of the forms or relations exhibited by the material world of our experience. Because, on the one hand, as we have pointed out,^ he failed to note the psychical activity involved even in ''pure perception," in the production, namely, of the sense- qualities of objects, and being im willing, on the other hand, to accept the Kantian doctrine that all the qualities of objects save these sense-qualities are the product of mind, he was naturally led to the view that the mental forms or categories applicable to things are the result of a compromise between matter and mind; for, as he says, even assuming that the forms into which we fit matter come entirely from the mind, they can scarcely be applied constantly to objects without the latter soon leaving a mark on them, so that, if we give to matter, we probably also receive something from it.^ This would leave us in uncertainty as to just what were the qualities of objects, even of matter, independently of the products of con- scious activities. If, however, with Bergson's doctrine of direct perception of 1 Creative Evolution, pp. 147-51. * lb., pp. 153, 160, 161. 3 Ch. XV, supra. * Time and Free Will, p. 223. THE GENESIS OF THE A PRIORI 363 reality we combine our theory of the presence of creative sense- activity even in pure perception, the way is open to affirm the revelation in perception, as ''pure" as we ever have it, of the universal, preexistent forms of matter. The categories would then appear to be, as S. Alexander contends that they are, characters in the world, possessed by things as well as by mind, and first carried up from material existence into mental exist- ence ; so that once consciousness is given — as a fortunate variation, if no more — we need no further successful varia- tion in order to secure the categories, or any other a priori parts of knowledge.^ Without needing to follow Alexander further, we would be able to account for the orthogenesis of mental evolution, which is by no means adequately explained by James's theory of mere accidental variation and natural selection. An activistic theory of all conscious processes, together with the doctrine of an instinctive knowledge of cer- tain fundamental relations, accounted for by the inheritance of acquired meaning, or, to say what amounts to the same thing, of intellectual habit — or else, as an alternative view, the appearance, as a mutation, of a new form of life with a very high degree of psychical alertness and impressionableness, with the consequent yery rapid learning, by the individual, of the most fundamental relations of things — would leave to natural selection no greater task than it may very well have been able to accomplish. But we must guard against exaggerating the a posteriori character of the ''a priori.'' W. K. Wright has contended that all our fundamental categories have had a social origin, and this view he thinks necessary if we are to account for the fact that those categories, as actually employed, vary greatly in different societies and stages of culture.^ We would admit that the specific form, in which certain categories are employed by particular groups and in particular stages of cultural de- velopment, is capable, at least partly, of a social explanation. This is manifestly true, for example, of the phenomenalistic notion of causality at present dominant in the natural sciences, and through them to some extent in popular thought. But ^Mind, N.S., Vol. XXI, 1912, pp. 16, 17. * "The Genesis of the Categories," Journal of Philosophy, X, 1913, 645-57. 364 THE PROBLEM OF KNOWLEDGE with reference to the genesis of our fundamental categories in their most generic form, this appeal to social psychology is in- adequate; they are, broadly speaking, not products of social tradition, but instruments of knowledge employed by the individual to-day, either habitually, because very rapidly learned from reality by the developing individual, or else in- stinctively, because inherited from ancestors by whom they were moulded on the reality revealed through the creation of sense-qualities which were located on the objects of the environ- ment, necessarily in the form and order in which those objects existed. We return, then, to our statement that what the Kantian regards as the activity of the absolutely a priori categories of thought, is quite possibly instinctive apperceptive activity. Kant's doctrine of a 'transcendental unity of apperception," imposing upon the object a unity which it would not otherwise possess, is, from this point of view, largely false and much too simple. There is a discoverable unity in all active things, and this unity already existent may be represented in an antici- patory way by that early learned or even instinctive appercep- tion which would be, in either case, explained as having been originally, whether in the hfe of the individual or in that of his ancestors, the result of sense-activitj^ and attentive analysis directed toward other unitary objects. The only unity im- posed upon the object by the psychical subject is a tertiary quality, the unitj^ which a more or less complex content acquires by virtue of its being related to some interest, or purpose, as end {terminus ad quern or terminus a quo), as obstacle or as means. There are thus, we would maintain, various degrees of apri- ority in the apperceptive activity involved in the perception of objects, from what is most universally inherited to what has been most recently acquired by the individual ; but in no case is this an absolute apriority. It is always, in the last analysis, the result of experience. While doing justice to the elements of truth in nativistic theories, it is frankly, although critically, on the genetic side. And yet, on the other hand, since we do not interpret sense-experience as passively received, but as actively produced by the psychical subject, our doctrine of THE GENESIS OF THE A PRIORI 365 the only relatively a priori character of the ordinary formal "relatmg" activity involved in ordinary perception is very far from lapsing into the old empiricism. Ours is an activistic empiricism; all sense-activity is creative activity with refer- ence to the secondary qualities of the object, and all thought- activity is creative of the ideas and their associations, while some thought-activity is creative of tertiary qualities and rela- tions of objects as well. But more than all this, whenever a new kind of psychical activity has appeared in the history of the race, whether it be the production of a new sense-quality or the formation of a new idea, we have psychical activity which is absolutely a priori. It is not, as Kant seems to have thought, the old and universal in the way of mental activity that is absolutely a priori; rather is it the new, the original and unique. It will thus be seen that the view which we have advocated in this chapter, and which has been shown, we think, to be not only defensible, but also, in the light of data alreadj^ avail- able, well-nigh demonstrable, avoids the absolute genetic dualism of the Kantian doctrine on the one side and the two corresponding forms of absolute monism, the rationalistic and the empirical (or the nativistic and empiriogenetic), on the other. It retains a fundamentally empirical and scientific point of view, but does justice to the activistic view of mind, emphasis upon which was the great merit of rationalism. The position, which may thus be very appropriately styled critical empirical genetic monism, or critical monism in genetic logic, once successfully defended, finally secures the position taken in our discussion of the problem of acquaintance, viz. critical realistic epistemological monism, or critical monism in episte- mology proper. With this accomplished we may turn our attention from the problems of immediate knowledge to such problems as may arise in connection with the subject of mediate knowledge. PART II: THE PROBLEM OF MEDIATE KNOWLEDGE A. THE PROBLEM OF TRUTH (LOGICAL THEORY) CHAPTER XVII A Critique op Intellectualism The more formidable part of our undertaking may now, perhaps, be regarded as accomplished. But to have vindicated the fact of acquaintance with reality is not to have treated adequately the problem of knowledge. Besides the problem of immediate knowledge, there is the problem of mediate knowl- edge. If knowledge is to be communicated and to become a social possession, or even if it is to be stored up in the most effective manner for one's own future use, it must come to exist in the form of judgments. But the claim to have knowl- edge in the form of judgments involves the twofold claim that the judgments are true and that this truth is certain, and justifiably so, to those persons whose judgments they are. It will be necessary for us, therefore, to discuss both truth and what is called proof, or the production of a sufficiently critical certainty of the truth. Taking these problems in their logical order, we shall turn first to a consideration of the problem of truth. Our discussion will be in the realm of logic as a branch of critical philosophy, or of logical theory, narrowly defined. The most elemental branches of philosophical criticism have to do with those ideals, or ends, to guide to the realization of which the various normative sciences have been developed. In the normative science of logic the ideal is truth. It would be claimed by some that the logical end is mere consistency, not truth. In the practical concerns of actual life, however, to make the end of our thinking mere consistency instead of truth is regarded as indicating a lack of earnestness or else pure stubbornness and such a selfish concern for applause and the appearance of victory rather than for victory itself as can only be set down as due to pronounced selfishness and as tending to intellectual dishonesty and hypocrisy. Taking logic, then, as 2 b 369 370 THE PROBLEM OF KNOWLEDGE the normative discipline concerned with the actual thinking of practical Hfe, we would regard the logic of consistency as only a branch of the logic of truth. It is the logic of hj^po- thetical truth and simply instrumental to the logic of actual or categorical truth. Scientific logic must undertake to show how to reach truth — hypothetical truth at first, it may be, but always ultimately truth, actual categorical truth. In critical or philosophical logic, therefore, we must undertake to solve the problem of the meaning of truth, the age-long problem, What is truth? Here, again, as in the case of epistemology proper, the prob- lem for later discussions has been set by the Kantian dualism. Kant, as we saw, set up an absolute epistemological duahsm between knowable phenomena and the unknowable independ- ent reality. Then, in order to bridge this chasm as far as might be, he had to introduce another dualism — an absolute logical dualism, according to which theoretical reason was to confine itself to phenomena, while practical reason might postu- late certain practically necessary behefs concerning ultimate reality. Thus, it was claimed, a part at least of what is intel- lectually unknowable is not only practically true, but prac- tically certain ; and yet, however true and certain, practically, it can never be other than in the highest degree doubtful from the standpoint of the theoretical understanding and pure reason. It is an absolute dualism of intellectuahsm and prag- matism.^ But it seems unlikely that it should have to be admitted ultimately that there are two radically and irreducibly different criteria of truth. And so two opposite ways of overcoming the 1 In an interesting article entitled " Practical Success as the Criterion of Truth" (Philos. Rev., XXII, 1913, pp. 606-22), H. W. Wright goes beyond Kant and advocates three rather than two distinct criteria of truth, viz. intellectual consistency, technical efficiency, and emotional harmony. In some cases, he says, one criterion, and that one alone, is applicable as a test of truth ; in other cases any one of the three may be applied at will, while in still other cases it is advisable to use all three criteria together. As in the case of the morphology of knowledge (Ch. XV, supra), so here, to affirm a threefold distinction is prob- ably closer to a critically monistic position than is a mere dualism ; and each of the three criteria mentioned by Wright will be found to be recognized in our constructive statement which is to follow. But for the present let it be said that there is a strong presumption against there being three radically different criteria of what is in its meaning always one and the same. A CRITIQUE OF INTELLECTUALISM 371 dualism have been suggested. One of these would reduce the practical to the intellectual ; the other would reduce the intel- lectual to the practical. The former we may call intellectualistic absolute logical monism, and the latter, anti-intellectualistic absolute logical monism. The term ''intellectualism" has been applied to the view that neither feeling nor practical needs have anything to say, properly, in determining the truth or falsity of judgments; that the criteria of truth are purely intellectual. In the present chapter we shall deal with the absolute intellectualism, considering it first in connection with absolute epistemological dualism, and then with the idealistic and realistic forms of absolute epistemological monism in turn. First, then, let us take up the case of the intellectualistic type of absolute monism in logical theory, as it appears when conjoined with a dualistic epistemology. Here we may begin with Locke, whose epistemology, as we have seen, was, at least covertly, dualistic. While still agreeing with the scholas- tics in their intellectualistic definition of truth — ''real truth," Locke calls it — as being the agreement of ideas with things,^ he found it necessary to introduce for the phenomenalistic or subjective point of view, another definition of truth as "a right joining or separating of signs; i.e. ideas [by which he means subjective contents of consciousness] or words." ^ Here, then, we see two mutually conflicting definitions of truth (although both are intellectualistic) ; and the significance of this failure to solve the truth-problem seems to be that either the epistemological dualism or the intellectualism is at fault, or else both are, for certainly if truth is either one of the two things Locke says it is, it cannot, on his presuppositions, be the other. £ln the system of Leibniz, which, from the standpoint of the individual consciousness, is an absolute epistemological dualism, truths of reason are defined in terms of the identity of subject and predicate (both considered as ideas, or both considered as things, or states of things), whereas truths of fact are held to consist in a correspondence between the succession of the phe- nomena, or the connection of the ideas or propositions in the mind, with the succession and connection of the things in * Essay Concerning Human Understanding, Bk. IV, Ch. V, § 8. * 75^^ | 2. 372 THE PROBLEM OF KNOWLEDGE question.! Strictly speaking, however, there was no way of verifying the asserted identity of subject and predicate, con- sidered as things, nor of the asserted correspondence between phenomena and things, since all any individual ever experi- enced was supposed to be his own ''ideas." Consequently, for the maintaining of the above definitions of truths of reason and truths of fact, as well as for the defence of the assumption that metaphysical knowledge is possible at all, it became necessary to promulgate as a dogma, itself not only metaphysical, but essentially self -refuting , the notion of a preestahlished har- mony between all absolutely independent (non-interacting) indi- viduals, by one of them. And after Kant had developed the duality of appearance and reahty into an absolute duahsm more expHcit than that of either Locke or Leibniz, the difficulty of defining truth in any unitary fashion in terms of correspond- ence was still more keenly felt. The most telHng ammunition which Lotze had to use against the "copy-theory" of truth, he found in what he still retained of the Kantian duahstic episte- mology. We cannot copy, as it seemed to him, an external reality which, it is assumed, we can never immediately perceiveTj Among recent writers we may take as typical in this con- nection two who, as personal idealists, are shut off from any unequivocal epistemological monism, and so would naturally define truth in terms of correspondence, if they could, but who are compelled, by the absoluteness of their epistemological duahsm, to adopt some other expedient. A. 0. Love joy uses the term ''truth" as meaning agreement with objective reahty; but this objective reahty, qua physical, is defined in terms which would limit it, so far as we can ever know anything about it, to such contents of present and future experiences of think- ing beings as they themselves construct in common.^ In other words, truth is the correspondence of an idea in the narrow (ordinary) sense with an idea in the broader (ideahstic) sense. The idea of truth as correspondence of idea with inde- pendent reality is given up, on the ground that no independent reahty, at least of anything physical, is knowable. 1 New Essays Concerning Human Understanding, Eng. Tr., pp. 404, 422, 445, 452. '"On the Existence of Ideas," Johns Hopkins University Circular, March, 1914, p. 66. A CRITIQUE OF INTELLECTUALISM 373 Boyce Gibson's definition of truth has special interest as expressing an ingenious attempt to produce a unitary definition within the hmits of an epistemological duahsm, and yet with- out an entire abandonment of the idea of correspondence of idea with independent reahty. ''Truth," he says, ''is the unity of ideas as systematically organized through the control exer- cised by relevant fact," or, again, "the unity of thought as systematically organized through the control exercised by that aspect of Reality which is relevant to the purpose of the thinker." ^ The terms "relevant" and "purpose" are intro- duced out of respect for pragmatic considerations ; but, as will become more evident in the light of later discussions, the intro- duction of these terms does not keep the definition from fall- ing short of the essential thing in pragmatism; it remains essentially intellectualistic. It is interesting, however, as combining the realistic theory of truth, as a correspondence between idea and reahty, with the idealistic theory of truth, as the coherence of ideas among themselves. Thus, while Locke left the two incompatible definitions of truth apart, Boyce Gibson unites them in one statement — without really harmonizing them, however. The coherence theory we shall have to deal with presently, but it may be remarked here that while there undoubtedly are many unities of ideas, and while these may be formed under the controlling influence of fact, and even of "relevant fact," and while such unities, moreover, are hkely to be very useful instruments of judgment, and con- ducive to the learning of the truth, it nevertheless does not necessarily follow that this unity of ideas is itself the truth of any judgment in which it may happen to be employed. Of course, if one were to go over completely to epistemological idealism, he would perhaps find himself reduced to the necessity of accepting this coherence of ideas as the only available sub- stitute for an inaccessible truth; but unless he means to do so, Boyce Gibson is inconsistent in defining truth as both some sort of unity of ideas and fidelity to relevant fact.^ But, it must be acknowledged, the inconsistency is in his case obscured by the fact that, like all pluraUstic idealists, he is an episte- mological dualist only from the standpoint of the individual ; ^ The PTohlem of Logic, p. 1. * lb., Pref., p. ix. 374 THE PROBLEM OF KNOWLEDGE from the standpoint of the community he is an epistemological monist. But he has no logical right to shift the basis of his argument from one of these points of view to the other without explicit acknowledgment of the change; and this acknowl- edgment could not be made without the argument's lack of cogency and the untenability of the position being exposed. We shall now turn to an examination of the combination of an absolute intellectuahsm in logical theory with the ideahstic form of absolute monism in epistemology. We shall begin with Hegel. We find him making a distinction between mere cor- rectness, or truth as it is found in comimon Hfe, viz. the agree- ment, in the sense of mere formal coincidence, of an object with our conception of it,i and truth in the deeper or philo- sophical sense, which is said to be the absolute correspondence or identity of objectivity with the notion.^ In an ideahstic absolute epistemological monism this can be maintained, it would seem, because the object is interpreted as nothing but idea. Since in true judgment subject and predicate ''stand to each other in the relation of reality and notion," ^ the thorough- going ideahstic intellectuahst cannot even say that subject and predicate differ in that one is reahty and the other idea; the object which is the subject of the judgment being itself idea, truth, the coincidence of the object with its notion, re- duces to ''the coincidence of the object with itself,"^ or the "agreement of a thought-content with itself." ^ That is, the whole alone is what is true ; ^ only God or the Absolute is the Truth.'' But this means that in genuine bona fide truth the judgment disappears, and that by the cancellation of its sub- ject, i.e. reality as distinct from idea. But the truth we sought to define is a supposed quality of judgments. According to Hegel, however, no judgment can be really true. This is the logical result of taking mere identity of idea with reahty as the sole criterion and definition of truth, and persisting in this with the help of the ideahstic interpretation of reahty. One of the most thoroughly Hegehan of contemporary 1 Logic, WaUace's Tr., pp. 51, 305. 2 lb., pp. 352, 354. s lb., p. 305. * lb. 6 75., p. 52. « " Phanomenologie des Geistes," Werke, II, p. 16; Eng. Tr., p. 17. ' Logic, Eng. Tr., p. 3. A CRITIQUE OF INTELLECTUALISM 375 thinkers is John Watson, and we find in his treatment of the truth-problem what is essentially the same doctrine with all its difficulties. In spite of the admission that truth exists only in judgments/ the account of the true idea as a ''copy" of an independent real object is rejected as untenable, however plausible it may seem on the ground that the so-called ''real object" exists only in the "true idea," so that the developed idea is not different from the developed object. The object is the idea, and an idea or object cannot be a copy of itself. ^ Thus, while we may say that our ideas have some low degree of truth, in so far as they correspond to or copy the "ideal object," the object as it is for "a mind that has grasped reality as it actually is," ^ we must admit, according to Watson, that "no single judgment is absolutely true." ^ Thus the position is seen to be self -refuting : it is surely not an absolutely true judgment that is asserted when one judges that no judgment is absolutely true. And here, again, we have the setting up of truth as an essentially unrealizable and self-contradictory ideal; the ideal judgment would be no judgment at all, for the subject would have disappeared in the predicate, or idea. It may be replied that in this Hegelian position, in so far as the idea of truth is retained at all, it is transformed from some sort of identity between the reality and idea into a relation of coherence between ideas as elements of reality. This coher- ence-theory of truth, resting as it does upon the untenable dogma of idealistic epistemological monism, is itself untenable ; and the signs of its untenableness will appear when we come to examine it in what is perhaps its most highly developed form. In the meantime, however, we must call attention to the self- criticism, or self-refutation even, of idealistic intellectualism, as accomplished by F. H. Bradley. Bradley, as has been noted above,^ was at first an apparently orthodox representative of Anglo-Hegelianism, but as he him- self says, if this view ever did satisfy him entirely, there came a time when it ceased to satisfy. However immanent in each 1 Philosophical Basis of Religion, p. 159. ^ lb., p. 160; The Interpretation of Religious Experience, Vol. II, pp. 67-8; cf. T. H. Green, Works, Vol. II, p. 258. ' Interpretation, etc., II, pp. 69, 70. * Philosophical Basis, etc., p. 161. ^ Ch. VII, supra. 376 THE PROBLEM OF KNOWLEDGE part, so far as he knew, the Whole might be really, he could not persuade himself that it was everywhere inmianent recog- nizably. And especially, on the principles of the ideahstic metaphysics itself, since thought is constructive of its object, the idea of any object supplements that object; one's idea of the Whole is an addition to the Whole. ^ Now it would seem that in view of these considerations Bradley ought not simply to have rejected the ideahstic doctrine of the philosophically demonstrable immanence of the Whole in each and every part, but to have at least suspected the fundamental principles of absolute idealism itself. He seems inclined, however, to throw much of the responsibihty for his idealistic presuppositions upon his German and English predecessors from whom he inherited them.2 But while Bradley agreed with the ideahsts that all think- ing is reconstruction of its subject-matter, he refused to follow them in making that subject-matter in any case a mere product of thought. On the- contrary, he held that all judgment, in- stead of being the joining of idea to idea, is an act which refers an ideal content, i.e. a logical idea, a product of abstraction, a ^'wandering adjective," the meaning of a symbol, to a reahty or existence which is beyond the act of judging, and not itself idea.^ In all this, we would hold, he was moving in the right direction, but he failed to reap anything Hke the full reward, in constructive results, of his break with Hegehanism, because he retained the idealistic doctrine of the necessary internality of the thought relation, and indeed of all relations, interpreted in the end as estabhshed by and dependent upon the process of human thought. According to this internahst doctrine, the ideas used in judgment quahfy the reality judged about, so that it becomes different in and through the very process by which the attempt is made to know it ; apperception modifies the facts which one sets out to perceive.^ Consequently it becomes necessary for Bradley to contradict what he regards as the really fundamental axiom of the judgment, to the effect that what is true in one context is true in another.^ This 1 Essays on Truth and Reality, pp. 223-5. 2 lb., pp. 124, 246, 275. 3 Principles of Logic, pp. 10-14 ; Appearance and Reality, pp. 163-5, 168. * Essays, etc., pp. 108, 227-30, 242. 6 Principles of Logic, pp. 133, 135. A CRITIQUE OF INTELLECTUALISM 377 means that according to his presuppositions no judgment can possibly be true; the ideal of truth is self-contradictory. In the background of Bradley's thought there lurks the assump- tion, suggested perhaps by Jevons' doctrine of the proposition as the affirmation or negation of an identity, simple, partial, or limited,^ that the ideal judgment really would express an abso- lute identity between subject and predicate.^ But this ideal is, in view of Bradley's internalistic assumptions, logically unrealizable. Although absolute identity of the predi- cate with the subject would be necessary to the absolute truth of any particular judgment, this condition can never be realized, because some difference between subject and predicate is neces- sarily involved in all judgment whatsoever.^ ''There is still a difference unremoved between the subject and the predicate, a difference which, if removed, would wholly destroy the spe- cial essence of thinking." ^ No " truth '^ can be entirely true; every categorical judgment is necessarily false, and thus, theo- retically considered, a failure.^ Not only are all necessary and universal judgments regarded by Bradley — rightly, we would admit, or even contend — as essentially hypothetical ; ^ he claims that the same is true of particular judgments, and so of all judgments.^ If one's judgment is to be true as well as categorical, one must get the conditions entirely within it ; ^ and so we are driven to fill in conditions indefinitely, with the 1 W. S. Jevons, Principles of Science, ed. of 1892, pp. 37-43. The view of Jevons, however, is fundamentally realistic, epistemologically speaking. • Brad- ley radically transforms the significance of the doctrine by assuming in broadly idealistic fashion, that, though the subject is reality, rather than logical idea, it is of the nature of experience, rather than independent reality, so that it is transformable by ideas, as independent reality would not be. 2 See Principles of Logic, pp. 132-5, 344-8, and Appearance and Reality, pp. 167-70, 361-2. ^Principles, etc., pp. 23-4, 131, 346-8; Appearance, etc., pp. 167-70. * Appearance, etc., p. 361. 5 lb., pp. 361-2 ; cf. p. 396 ; Essays, etc., pp. 231-3, 251, 253, 257, 276. ^Principles, etc., pp. 47, 49. ' 76., pp. 45, etc. In his Appearance and Reality (p. 361), Bradley says he is now persuaded that it is better not to say that every judgment is hypothetical ; but this does not indicate any essential change of view. He still maintains that since what any judgment affirms is incomplete, it cannot be correctly attributed to Reality, except with a complement, and indeed one which in the end remains unknown, so that we cannot tell how, if present, it would act upon and alter the predicate. » Principles, etc., p. 99. 378 THE PROBLEM OF KNOWLEDGE result that the categorical nature of the judgment is destroyed ;^ so long as anything remains outside, assuming the absolute internality of all relations, the judgment is imperfect and its opposite is not without truth.^ And so it is not permissible to appeal to designation, i.e. the use of such indications as ^'here," "now," ^'this," ''my," in order to include conditions sufficiently for the making of an absolutely true categorical judg- ment; for the attempt to define these terms again drives one into an indefinite regress.^ There cannot even be a true cate- gorical statement of possibility, it is claimed; all possibility is merely such only because of our ignorance of existing con- ditions.^ This necessary failure of the judgment, when truth is con- ceived as absolute identity of subject and predicate, leads naturally to the formulation of the notion that truth is essen- tially coherence, rather than correspondence or identity. The realization of the ideal of truth could, on Bradley's presup- positions, mean nothing short of the disappearance of all judgment; the perfection of truth and the perfection of the reahty would be the same; what absolute truth would be, if there could be such a thing, is the coherence of all elements (which may indeed be thought of as if they could exist separately) in an all-comprehensive system, or, ultimately, in one super- relational, immediate experience.^ Ultimately non-contradic- tion, the criterion of system in the realm of judgments, is found to be realizable only in the absolute or all-comprehending im- mediate experience ; an object short of the whole tends naturally to suggest its complement, and since that suggested comple- ment is absent in fact, reality thus contradicts itself.* Bradley's position here is obviously self-refuting. He is not entitled to judge it absolutely true that no judgment can be absolutely true. In his latest work he says, "all ideas in the end, if we except those of metaphysics, lack ultimate truth." ^ But he has no logical right to exclude metaphysical judgments from his strictures on judgment in general, and in Appearance 1 Essays, etc., p. 229. 2 76., p. 233. ' lb., p. 235. * Principles, etc., pp. 186-7, 191, 194; Essays, etc., p. 233. ^Appearance, etc., p. 363; Essays, etc., pp. 113, 116, 210-11, 239. « Essays, etc., p. 241. 7 /^.^ p. 267. A CRITIQUE OF INTELLECTUALISM 379 and Reality he more consistently ( ?) admits his inconsistency in the acknowledgment that even categorical metaphysical judg- ments are logically impossible/ and that in the end even "ab- solute truth" is not absolutely true.^ Thus in spite of his having taken non-contradiction as his criterion of truth and reality, he is forced into the most glaring self-contradiction. It would be difficult to find a more appropriate object than himself against which to direct his own remark in criticism of some of his opponents, that if one is willing to be inconsistent, he can never be refuted.^ It would be a more than dubious doctrine which allowed one to atone for the sin of his incon- sistency by a mere confession of it. (Moreover, Bradley seems hardly consistent with his doctrine that the idea is always, as idea, not existent, when he speaks of the idea as approaching Reality. How can what is essentially non-existent become more and more nearly the reality?) The strength of Bradley's position, such as it is, is found only in the dialectical skill and thoroughness with which he carries out the implications of that internalistic residue of idealism which he either had not the courage to throw over- board, or was brave enough to retain. He may well challenge his critics to do any better than he has done — with the same pre- suppositions as materials. All protests and ''refutations," he says, count for nothing with him, unless they can show that on the principle adopted the conclusion drawn is wrong.^ This empty dialectical triumph Bradley's critic may be quite ready to grant him ; his conclusions are not so much to be refuted from his premises, as to be avoided, if legitimately possible. We are not concerned to show that on his own principle his conclusion is wrong ; it is easier to show that his principle is itself wrong. Indeed this has been partially accomplished already. We have shown that the arguments for idealism are not only not demon- strative, but that they are essentially fallacious ; and we have outlined as an alternative a philosophj^ which is free from fal- lacy and adequate to the facts. According to this view apper- ception is not essentially a modification of fact, but a revelation of fact. The predicate is adjectival indeed, but adjectives 1 Appearance, etc., p. 361. * lb., pp. 544-5. ^ Essays, etc., p. 235. * lb., p. 234. 380 THE PROBLEM OF KNOWLEDGE are merely representational in relation to reality ; they are not constitutive of reality in general, but only of knowledge and of what we have called the tertiary qualities of reality. Moreover, it was pointed out that the internality or externality of relations is not absolute, but relative ; it depends upon the purpose which is, or ought to be, entertained. Bradley admits indeed the rela- tive externality of relations, i.e. their externality or indifference for certain practical purposes, while he denies their absolute externality.^ In this he is right, but when he assumes that all relations are absolutely internal, he is not right ; for, as we have shown, the whole question of the internahty or externality of relations is essentially relative to purpose. Relations are neither all absolutely internal and relatively external, nor all absolutely external, nor some of them absolutely internal and others absolutely external ; they are all relatively internal and relative^ external. All existing relations are — in their own relations — absolute, but their internality or externality to their terms is always relative, not to 'Hhis or that mode of union," ^ but to this or that purpose. But, further, as we shall see more fully in the sequel, similar considerations of purpose open up a way whereby particular judgments, if not universal judgments also, may escape from a merely hypothetical to a categorical status. There seems no sense in denying the validity or truth of a particular judgment which takes account of all the conditions that need to be con- sidered for the purposes concerned, if these purposes are what they ought to be. May we not he able to get absolute categorical truth into our judgments, if first we get the categorical imperative of morality into our ^practical purposes f It may then very well be possible to ''get the conditions of the predicate into the sub- ject," sufficiently for all the purposes which ought to he considered. Moreover, the injunction against ''designation" can be defended only on the basis of the absolute internality of all relations; if relations are not absolutely internal, non-contradiction does not necessarily involve all-comprehensiveness, the subject of any one judgment need not be "Reality" as a whole, but only some reality, and this may be adequately indicated by designa- tion. Once more, it is not necessary to reduce all assertions of 1 Essays, etc., pp. 237-8. 2 Bradley, op. cit., p. 237. A CRITIQUE OF INTELLECTUALISM 381 possibility to ''suppositions founded on our real or hypothetical ignorance." ^ It is not an untenable position, at least so far as Bradley has shown, to maintain that it is here and now possible for me — all conditions, whether known or unknown, being just what they are — to act in the immediate future somewhat differently from the way in which, as a matter of fact, I shall act. Bradley has not worked without some glimpse into this pragmatic way of escape from his theoretical difficulties. He admits that primitive thought was, and apparently also that ideal thought would be, absolutely practical,^ and he even goes so far as to surmise that what works must be at least partially right ; ^ but not only will he not admit the more extreme doctrine that truth is definable simply in terms of practical effects; he refuses even to concede that intellect is so essentially related to practice that its findings can always be properly subjected to practical tests. ^ His criticisms against current pragmatism may be largely sound, and yet the failure of his splendid system to solve the problems of truth and even of reality may very easily be due to his failure to appreciate the theoretical value of practi- cal considerations, as well as the practical value of theory. And it seems not unreasonable to suppose that Bradley, who at one time seemed so close to the pragmatic path, was repelled by the crude and uncritical way in which some of the features of pragmatism were anticipated in the writings of Alexander Bain. ^ The coherence theory of truth is championed by Bosanquet, who here as elsewhere tries to retain, in synthetic unity, the essentials of both the Hegelian thesis and the Bradleian antith- esis. He holds that the elements of the judgment are a subject in Reality, the meaning of an idea, i.e. its identical reference throughout all its psychical presentations, and an identity of content between this subject and predicate. The claim to be true consists in the affirmation of the meaning as belonging to the tissue of reality at the point indicated by the subject.^ 1 Principles of Logic, p. 191 ; cf. Essays, etc., p. 233. 2 Principles, etc., p. 32 ; Essays, etc., pp. 75, 91, 141. 'Principles, etc., p. 343. ^ Essays, etc., pp. 79-89. ^ See Bradley's Principles of Logic, pp. 18 ff. ; Essays on Truth and Beauty^ p. 70. « The Essentials of Logic, pp. 69-79. 382 THE PROBLEM OF KNOWLEDGE This seems to mean that when the subject is the very same thing which the predicate means, i.e. persistently refers to, the judgment is true. But this definition surely takes insuffi- cient account of that necessary element in the judgment which makes so much trouble for Bradley, the element of difference. The result is what might have been expected. As with Bradley, the actual human judgment ^'has been gutted and finally vanishes"; ^ but it slips away more surreptitiously from Bosanquet's logical theory than from that of Bradley. All that is left is a single term, viz. ''truth" as "fact," or ''ReaUty," or "the Whole," viewed as constituted by knowledge, while the trueness of the forms of thought is their power to constitute a totality. But it will not do, Bosanquet apparently feels, to leave this truth existing or subsisting without any supporting judgment ; and so he invents "a, single, persistent and all-embracing judgment," whose content and product the "Truth" or "Real- ity" may be supposed to be.^ But no such judgment is, for man, either known or conceivably possible. It is purely imaginary, and we have no right to suppose that there is any such judgment. In any case, truth as an attribute of actual hiunan judgments has disappeared ; and, in order to distract our attention from our loss, we are exhorted to fix our attention upon the coherence of that one ultimate ReaHty which is constituted by knowledge. What Bosanquet would really be entitled to say, from his general philosophical point of view, is not that the truth of our judg- ments is the coherence of Reality, but that it is the correspon- dence of certain ideas of ours with the content of an Absolute Experience, which can never be the experience of us finite individuals, so that the test of truth can never be appHed by us. The fact is, with reference to this coherence theory, that it has taken one of the later tests of truth (coherence of judg- ments in a consistent system — a test which is vaHd enough within limits, but which can never guarantee more than hypo- thetical truth) as being itself the nature of truth — an error quite parallel, as we shall see, with the characteristic error of current pragmatism. 1 Bradley, Principles, etc., p. 27. 2 Logic, or the Morphology of Knowledge, Vol. I, p. 3. A CRITIQUE OF INTELLECTUALISM 383 H. H. Joachim acknowledges great indebtedness to Bradley and Bosanquet/ and his point of view is not essentially different from theirs, except that he seems to appreciate, at least more fully than Bosanquet, the inadequacy of the coherence-notion of truth. He assumes that all relations are essentially internal,^ and so, partly on this account, gets into trouble when he tries to think through the idea of truth as correspondence, or identity of structure. There cannot be an identity of structure between the mental and the real, because if there is no difference between the two factors, there is no correspondence, but simple identity ; whereas if there is a difference, there cannot be identity of structure.^ He is willing to regard correspondence as a symp- tom of truth, but claims that it is upon something other than the correspondence that truth depends.^ What this something else is, he seems at a loss to say ; but, as we have hinted in our criticism of Bradley, and as will appear more fully in our con- structive statement, he would have been able to find it, if, in addition to discarding his theory of the necessary internality of relations, he had given some attention to that pragmatic theory, with reference to which he almost boastingly remarks that the reader will find no mention of it in his book.^ After paying his respects to the realistic theory that truth is a quahty of independent entities, Joachim proceeds to exam- ine the coherence-theory, according to which, as he says, truth is ''that systematic coherence which is the character of a sig- nificant whole," i.e. of "an organized individual experience, seK-fulfiUing and self -fulfilled." ^ But it is confessed that there can be only one such experience, and that is not the human, so that "the truth is — from the point of view of the human in- teUigence — an Ideal, and an Ideal which can never as such, or in its completeness, be actual as human experience." ^ No single human judgment, therefore, can ever be absolutely true.^ 1 The Nature of Truth, p. 4. 2 75.^ p. 26. 3 j^,.^ pp. 24-5, 29. * lb., p. 17. 5 75., p. 3. 6 lb., p. 76. ' 76., pp. 78-9. As against such absolutists as Bradley and Joachim, as well as against absolutely dualistic epistemologists, Schiller's remark is scarcely too satirical, when he says, " Unverifiability is the distinctive mark of a consist- ently intellectualist view of truth." (Journal of Philosophy, Vol. IV, 1907, p. 493, note.) • lb., pp. 104, 113. Joachim confesses, in effect, that he has "never doubted " that there is no truth but the whole truth ("that the truth itself is one and whole 384 THE PROBLEM OF KNOWLEDGE But in that case even the judgment which embodies the cohe- rence-theory of truth cannot itself be true, and this self-refuta- tion of what he still regards as the truest possible theory of truth Joachim is himself forced to accept.^ We shall not be satisfied to share his estimate, however, until we shall have in- vestigated the possibilities of framing a non-self -refuting theory in still another way than any of those examined by Joachim. We shall have to examine the views of one more representative of idealistic absolute intellectualism, viz. Josiah Royce; for, in spite of his "absolute pragmatism," ^ this philosopher's theory of truth remains, as we will endeavor to show, essentially intellectualistic rather than pragmatic. Royce makes many concessions to the pragmatist, although, as we shall see, they all fall short of essential pragmatism. He admits that every judgment is a reaction at a particular time to an empirically given situation, a reaction expressing, and determined by, the consciousness of a need to get control over the situation. He also admits that even the most remote speculations are, for the man who engages in them, modes of conduct, and that the thinker's ideas are his own deeds, or at least his plans of action.^ But while maintaining that every opinion is a deed, intended to guide other deeds, and consequently that all truth is practical,^ he does not propose to take this practical characteristic of truth as furnishing, under any conditions, a criterion of truth; he adopts a criterion and definition of truth which are, as will ap- pear, quite non-pragmatic. But even more important for the understanding of Royce's theory of truth than his relations to pragmatism is his rejection of realism. His fallacious arguments for a non-realistic system we have already examined, but his criticism of realism is almost equally open to objection. Even in his first published volume, and complete") {The Nature of Truth, p. 178) ; but, as Schiller pertinently remarks, "Perhaps if he had been more willing, not necessarily to doubt, but, let us say, to examine, this assumption, he would not have been forced to doubt so much in the end." (Studies in Humanism, p. 167.) 1 The Nature of Truth, p. 178. 2 "The Problem of Truth in the Light of Recent Discussion," William James and Other Essays, p. 254. 3 "The Eternal and the Practical," Philosophical Review, XIII, 1904, pp. 117- 19, 141 ; cf. William James and Other Essays, pp. 223, 233. 4 Sources of Religious Insight, pp. 145-6. A CRITIQUE OF INTELLECTUALISM 385 on the ground that, as he alleges, common sense does not know what error is, he says, **Let common sense not disturb us, then, in our further search." ^ Against this summary procedure it is sufficient to remark that if common sense, without the pragmatic criterion, is unable to say what error is, it does not follow that common sense, with the pragmatic criterion, might not be able to furnish the desired solution. Moreover, in his treatment of realism, Royce commits the fallacy of assuming that the refuta- tion of a particular type of realism — and that a very extreme and indefensible kind — is a sufficient refutation of all realism. He assumes that realism must be absolutely dualistic in episte- mology, completely sundering the what from the that,^ and even supposing that what one mentally constructs and discovers as thus constructed, existed prior to that construction.^ He also assumes that 'independently real" must mean not only real independently of the knowledge relation, but real independently of all relations, a pluralism so absolute as to deny that there obtain among real things any relations except such as are so absolutely external that they are not relations at all.^ It is small wonder, then, that a seeming triumph for idealism is easily obtained. Having thus, as he supposes, shown the impossibility of a rational defence of realism, the need of finding some way of reaching objectivity becomes imperative. We need to hold that the subject of our judgments is Reality, not our idea,^ and we need that the judgments which we need for practical purposes be also true of this reality which is not our idea.® What Royce proposes in this situation, as a substitute for real- ism, is to fall back upon what is given subjectively, viz. our needs ; ^ and it is this private need of his own, as a non-realistic philosopher, this need of constructing objectivity out of our sub- jective needs, or of finding it in them, that explains his seemingly rather patronizing attitude toward contemporary pragmatism, with its emphasis upon the theoretical value of practical con- ^ The Religious Aspect of Philosophy, p. 392. 2 The World and the Individual, Vol. I, p. 107. 3 Philosophical Review, XIII, 1904, p, 125. 4 The World and the Individual, Vol. I, pp. 112, 127-136. 6 lb., pp. 95, 271-2. ^Philosophical Review, XIII, 1904, pp. 126, 141. ? lb., p. 124. 2c 386 THE PROBLEM OF KNOWLEDGE siderations. What we need in needing truth is, at the very least, according to Royce, companionship in our thinking,^ but it is more than merely this,^ The need of the moment needs to be controlled, not, as the realist imagines, by an independently existing object, but by ''some universal expression of need — an expression that simply makes conscious what the need of the moment is trying, after all, to be." ^ Our need of truth is a need of an insight such as would remain invariant for every additional point of view ; ^ and this, it is assumed, can only mean one^s own true self, including within its experience all possible points of view, and knowing that it includes them.^ Thus we need, it is claimed, the Absolute Self, other than our present finite self, as an actually and eternally existent Judge, if we are to have truth. ^ Indeed, one's true Self, as such an Absolute Judge, must exist, if there is to he truth, whether known by any finite self or not ; and this being so, there cannot be error, ^ or even ignorance,^ without the true, all-inclusive Self as Absolute Judge. Royce confesses that it is his voluntarism that is the secret of his absolutism ; ^ but, in the light of what has been said, it would perhaps be truer to say that it is his provisional subjec- tive idealism that drives him to take up voluntarism and to develop it in the direction of absolutism, a solipsism of the Absolute Self, as the only logical escape from a solipsism of the finite self. We ''acknowledge" a transcendent, truth-know- ing Absolute; we "define" the Eternal; ^^ we "appeal to" an all-comprehending insight ; ^^ and the meaning of all this is that we need the belief in such an Absolute, and so we deliber- ately take it, not tentatively as a working hypothesis, as the pragmatist might do, but absolutely and outright, as if it were our indispensable possession. 1 Philosophical Review, XIII, 1904, pp. 126, 135. 2 lb., p. 138. 3 lb., p. 131. 4 ijj^^ p. 140. 5 lb., pp. 140-2 ; William James and Other Essays, p. 236. 6 Philosophical Review, XIII, 1904, pp. 135-6, 138. 7 Religious Aspect of Philosophy, pp. 393, 424-7 ; Studies of Good and Evil, p. 165. 8 The Conception of God, pp. 28-9 ; William James and Other Essays, p. 237. ® William James and Other Essays, p. 235. 10 76., p. 236. " Sources of Religious Insight, p. 137. A CRITIQUE OF INTELLECTUALISM 387 Now it is true enough that, because we come to know things through experience, we naturally remember them as they were experienced, and imagine what we have not experienced as it would be, if experienced.^ In this sense we do indeed appeal to experience, even when we refer beyond our own present experience. But what we must not forget is that what we intend to appeal to is not what the thing is experienced as, but what it would he experienced as, if it were experienced. The superhuman experience is, so far as these necessities of thought are concerned, simply a '^regulative," not a ''constitutive" concept. What the dogmatizing rationalist does, however, is to substitute for the "would be . . . if," a simple "is," or an authoritative "must be" ; and so deft is he in his logical leger- demain that his trick imposes even upon himself, and he con- cludes that the all-inclusive experience of the Absolute Self actually is and must he. It is true enough that something besides our subjective practical needs underlying our judg- ments is needed as the foundation and measure of their claim to truth; but what is needed is not necessarily superhuman truth — although we would not argue that there is none — but independently existing reality, accessible, under whatever conditions, to human experience and knowledge. Royce, however, having rejected all such realism, and having consequently, after the fashion of logical idealists, confused the concept of truth with that of reality, is compelled to pur- sue the course we have just outlined. And in so doing, we must insist, while he still refuses the most essential and valu- able element in current pragmatism, he acts upon, and really adopts in principle, the most logically vicious element in what we shall call pseudo-pragmatism.^ Having first gotten him- self, by a philosophical mistake, into unnecessary difficulties, he finds that the only thing that will save him from the neces- sity of retracing his steps and acknowledging the error of his ways is to assume the truth of an unverifiable proposition 1 What do I mean by the object of which I know that I am ignorant ? Dewey would say,, presumably, a future content of my own experience ; and Royce, a present (or super-temporal) experience of the Absolute. But why "future"? Why "super-temporal"? Why "of the Absolute"? And why "experience"? What I mean is a present reality, of which, in some cases, I may have future experience. 2 gee Ch. XVIII, infra. 388 THE PROBLEM OF KNOWLEDGE (that an Absolute, such as he depicts, exists) ; he therefore naturally ''wills to believe" it, and immediately does so, argu- ing its certainty from its necessity, forgetting that the neces- sity is itself quite artificial and unnecessary. Royce criticises current pragmatism for its tendency to lapse into this pseudo- pragmatism as a doctrine, "identifying the truth of an asser- tion with one's own individual interest in making the asser- tion";^ but he himself seems to have done the same thing, not in theory, but in practice. His "absolute pragmatism" is a necessary theoretical veil to hide the absolute pseudo-prag- matism of his actual procedure in this particular instance. And what are the consequences of this making of agreement with the judgment of the absolute, all-knowing Judge the criterion of truth ? Evidently, that even if we avoid Bradley's conclusion that we can never possess the truth, we are not able, logically, to say that we know that we have it, because the standard of measurement is inaccessible. The completely integrated experience of the Absolute the individual man never gets before him ; ^ and since the only workings by which our assertions can be adequately judged are "their workings as experienced and estimated from the point of view of such a larger life," ^ the agnostic conclusion is logically inevitable. In spite of the assertion that the Absolute Experience is the experience of my true self, it remains a fact that I, the finite self, do not experience it. And indeed Royce seems to admit that the knowledge we get, such as it is, by adopting his abso- lutistic criterion is a knowledge of our own ignorance. ^ And if attention be called to the fact that this completely agnostic conclusion is untenable and must be given up, it ought to be sufficiently obvious that this can only mean that the absolutistic doctrine of the criterion of human truth has been shown to be self-refuting. But, strange as it may seem, Royce has no intention of mak- ing any such admission. Unlike Bradley, who consistently acknowledges the necessary inconsistency of his absolutistic position, Royce argues, against the logic of his position, that 1 William James and Other Essays, p. 232. 2 Sources of Religious Insight, p. 148. ^ lb., p. 149. 4 The Conception of God, pp. 28-9 ; William James and Other Essays, p. 237. A CRITIQUE OF INTELLECTUALISM 389 even he who has adopted the absolutist criterion can have absolute knowledge of absolute truth, because, as a matter of fact, he has such knowledge in all propositions which are such that to deny them is to assert them under a new form.^ What he has overlooked here is the fact (as we would contend it is) that even these '^ absolute truths" are reached, not by the absolutistic, but by a humanistic criterion. That in his own thinking he is really guided by some more workable principle than is to be found in his own theory is increasingly evident as we look further into his discussion of absolute truths. While it is claimed that in the realm of pure logic and pure mathematics absolute truths are accessible,^ it is admitted that all such propositions are essentially hypo- thetical;^ ^'absolute truth is not accessible to us in the em- pirical world." ^ And yet he seems to teach also that some absolute categorical truth concerning reality is derivable from these universal hypothetical truths. Hypothetical judgments, it is claimed, give us negative information about the real world,^ and in the end they tell us indirectly what is, by telling us what is not.® Moreover, they give us positive and categorical abso- lute truth about the nature of the creative will that thinks the truth.^ Indeed Royce seems to have intimations of a wider field of accessible absolute truths, when he says that, in view of the irrevocableness of every past deed, every act of judgment that calls for a deed is irrevocably (and so, absolutely) true or false. ^ This last seems to be true only on the essentially pragmatic principle that judgments which satisfy every rele- vant practical purpose that ought to be considered (the pur- pose of the scientist, which is ultimately practical, being in- cluded) are really, and therefore absolutely, true; but this is very far from being consistent with Royce's absolutistic theory of the criterion of truth. Furthermore, much of the appear- ance of contradiction in the above-cited doctrines, first of the purely hypothetical character, and then of the essentially 1 William James and Other Essays, pp. 239, 244. 2 75.^ pp. 212, 251. 3 76., p. 239 ; cf. The World and the Individual, Vol. I, p. 276. * William James, etc., p. 249. ' The World and the Individual, Vol. I, p. 274. 6 76., p. 277. 7 William James, etc., pp. 247-8. ' Sources of Religious Insight, pp. 154-7. 390 THE PROBLEM OF KNOWLEDGE categorical character, of absolute truth, could be removed if only it were explicitly understood that this essentially prag- matic criterion might be employed. Universal truths are essen- tially hypothetical, until the conditions of the subject are included in the predicate, sufficiently for all practical purposes which ought to be considered; and then they become categorical. This pragmatic absoluteness of ordinary, humanly accessible, empirical truths, is the one thing needful to make Royce's ''pragmatic absolutism" absolutely practical, and therefore presumably true. As it is, however, so far from his theory of truth being an absolute pragmatism, it remains at heart an absolute intellectualism. In the first few pages of this chapter we saw that when the intellectuaHst is an epistemological duahst, he tends to hold an idea of truth which would make it consist in a copying or correspondence or identity between idea and reality, such that ideally the difference would be simply numerical, apart from the fact of the one term being mental and the other (at least relatively to the individual thinker) non-mental. Such a relationship, however, is neither realized nor realizable in the great majority of the judgments which we are obliged to make ; the predicate regularly falls far short of the subject-matter; the idea, far short of the thing. But where the subject-matter is an independent reality, it does not appear that the predi- cate or idea could ever be known to coincide with the thing of which it is asserted, wherever the epistemological dualism is ab- solute. Evidently, then, the epistemological dualist is unable to solve the problem of truth by the way of pure intellectuahsm. We next turned our attention to the possibilities of a purely intellectualistic solution of the problem of truth, where the fundamental philosophy was an idealistic absolute epistemo- logical monism. What we found was that "truth," defined as absolute identity, numerical and qualitative, of subject and predicate, of reality and idea, was supposed to be discoverable in the subject itself, interpreted as a unitary system of all possibly coherent ideas, no one explicit judgment being left to be true. This, of course, fails to solve the problem of the truth of human judgments, for it puts real truth beyond the reach of man altogether. A CRITIQUE OF INTELLECTUALISM 391 We shall now turn to a consideration of the possibilities of a purely intellectual solution of the truth-problem for the realistic epistemological monist. And here we shall see a course pursued which in the end comes to a conclusion which is the exact opposite, in some respects, of that of the idealists. Instead of finding the absolute identity, numerical and quali- tative, of subject and predicate, of reality and idea, by inter- preting the subject as a system of ideas, they would find it by eliminating the predicate altogether; they will have no true idea as numerically distinct from the thing ; only independent reality is supposed to be left to be true, or the truth. Whether or not this is the true solution of the problem we must now in- quire, examining certain typical expressions of the neo-realistic treatment of the problem of truth. A less extreme result of the union of realism and intellectualism than that with which we shall be chiefly occupied throughout the remainder of this chapter is seen in the doctrine of truth of the Aristotelian scholastics. In mediaeval times, where the influences of mysticism and of the philosophy of Plato were strong, idea and reality tended to be identified and the contrast between appearance and reality strongly accentuated; but where common sense and the influence of Aristotle prevailed, just the opposite was true : reality and ordinary appearance tended to be identified, and the contrast between reality and idea taken as an unquestioned commonplace. It was only natural, then, that the Aristotelian definition of truth, as the quality of a judgment which exactly represents the way in which real things are conjoined or divided,^ should also have been accepted, quite as a matter of course. The statement of Thomas Aquinas that "to know Truth is to know the agree- ment of knowledge with the thing known," ^ presented no diffi- culties, for while the realism was not carried to any one-sided extreme (as is the case in the new realism), on the other hand there was no absolute epistemological dualism to be tran- scended. ^Metaphysics, Bk. IX, Ch. X, "He who thinks the separated to be sepa- rated, and the combined to be combined, has the truth." 2 Compendium of the Summa Theologica, Pars prima, Ch. XVI. For the scholastics truth was adequatio intellectus et rei. 392 THE PROBLEM OF KNOWLEDGE The representation of truth as independent reahty has been either approximated or actually performed by several of the American new realists, notably Perry, Montague, Holt, and Marvin. McGilvary's position is somewhat transitional. He has always avoided an absolute realistic epistemological mon- ism, but he has incorporated some of the dogmatic features of that view in his own system, so that we classified him with the others; he has failed to reach the most essential insights of what we have called critical realistic epistemological monism. His definition of truth consequently reflects the essentially transitional, and even unstable, position in which he has for some time been attempting to maintain his equilibrium. An idea is true, he holds, if its object is in the real universe ; false, if its object is unreal.^ At first this seems to amount to little, if any, more than the old scholastic adequatio iiitellectus et rei; it asserts that an idea (presumably an image with its mean- ing, since it is asserted that every idea corresponds to its object) ^ which represents a reality is true. Now this would be a toler- able definition of truth if in judging we began with the idea (predicate) and then looked around for a subject for it. But as a matter of fact our procedure is different; we begin with a given situation, some element of which is selected by a practi- cal interest, as the subject-matter of a possible judgment, and then we look for the proper idea, or predicate. There may be objects in the real universe such as the idea would truly repre- sent; but the question is whether this object, experienced or assumed as subject-matter of judgment, is truly represented by this idea (predicate). The logical idea, or predicate, is a ''uni- versal" ; that it is not the true predicate in this given situation (or, more accurately, the predicate in a true judgment concern- ing this assumed reality in this situation) does not mean that it may not, under some circumstances, have its place in a true judgment. McGilvary virtually treats the logical idea as a particular datum, and the reality almost as a universal. Trans- lated into terminology appropriate to the judgment as actually employed in practical life, his definition of truth amounts to no more than to say that, when there is truth, some reality is represented by an idea. But this, which may perhaps be 1 Philosophical Review, XX, 1911, p. 156. 2 75, A CRITIQUE OF INTELLECTUALISM 393 regarded as giving the proximate genus of truth, fails to state the differentia of the species ; an idea, or predicate, may some- times represent its subject, and yet not truly. Thus McGil- vary's definition is really more open to objection than is that of the scholastics; it does not provide for the adequatio as be- tween idea and the reality. In its original form it is signifi- cant, however, as expressing the neo-realistic tendency to find truth in reality, rather than in the judgment. Perry rejects the absolutistic account of truth ^ and also the current pragmatist identification of truth with the satisfying character of the practical transition from cognitive expectation to fulfilment,^ and signifies his adherence to the realistic form of intellectualism when he says that knowledge as true belongs to the context of reality, or, more conservatively, that it is verified by being found consistent with reality.^ It is proved, he claims, directly, in the elements and systematic relations of real experience, i.e. of independent reality immediately experienced.^ Truth is to be found in the thing known; it must envisage reality ; it not merely corresponds to its object ; the object plays the determining part in constituting the truth.^ In this way he gradually leads up to his rather radical definition of truth as essentially identity, or consistency, with reality.® Later, he says there is truth when a subjective manifold har- monizes with a manifold of some independent order, or, dif- ferently expressed, when a content of mind is rightly taken to be fact.^ If we gather these suggestions together, it seems evi- dent that Perry's thought has been moving decidedly in the direction of the view that truth is an identity of content in two different contexts, the one the subjective and the other the objective. Montague's definition amounts to practically the same thing. "When one content is the object of a belief, and is also a thing that exists, there subsists," he says, ''between the content as believed and the content as existing that particular form of the relation of identity which is called truth. To say that a belief when true corresponds to a reality means that the thing be- 1 Journal of Philosophy, IV, 1907, p. 370. 2 j^,, p. 371. 3 lb., p. 372. 4 lb., p. 373. 6 76.^ p. 374. 6 75., p. 422. ^ Present Philosophical Tendencies^ p. 325, 394 THE PROBLEM OF KNOWLEDGE lieved is identical with a thing that exists. . . . When we beUeve a thing that is a fact, our beHef is true/* ^ — and, we may add, this last statement is a mere truism ! Later Mon- tague says that the truth is the real, taken in a certain rela- tion, viz. as object of a possible belief or judgment. He claims that there is no more difference between what is real and what is true than between George W^ashington and President George Washington.2 Holt speaks of two senses in which the term truth may be used, viz. first, truth of correspondence, or of identity of struc- ture between an abstract system and some more concrete system ; ^ and, second, and more important, truth as the mutual consistency of propositions.^ Here we have the realis- tic counterpart of the idealistic doctrine of truth as coherence. The judgment has disappeared, so far as the question of truth is concerned ; and, the difference between concepts and reahty being interpreted as only a difference of relation, the conclusion follows that the truth is the largest system of consistent proposi- tions.^ Marvin prefers to apply the term '' correctness,'^ rather than ^* truth," to judgments, and this correctness he would define as the quaHty belonging to judgments which assert true proposi- tions. The term '' truth," which he thus appHes to propositions, he claims cannot be correctly defined in any way that does not involve a circle.^ This admission by Marvin is important for the criticism of other neo-reaHstic definitions of truth. Looking back over the previous four definitions we find that the attempt to define truth intellectuahstically, from the point of view of reaHstic absolute epistemological monism, leads to the identification, either explicitly or implicitly, of truth with reality; truth is reality, it is asserted, self-identical in different external rela- tions; and since any definition of truth must take account 1 Journal of Philosophy, VI, 1909, p. 546. ' The New Realism, p. 252. Montague holds that truth is of two kinds, immediate and mediated. The former obtains when in perception there is no distortion, either peripheral or cerebral ; the latter, when peripheral distortion is corrected by a cerebral reaction (ib., p. 292). 3 The Concept of Consciousness, p. 37. * lb., pp. 279-80. 6 76., p. 339. « A First Book in Metaphysics, p. 28. A CRITIQUE OF INTELLECTUALISM 395 of reality, for the neo-realist this is equivalent to the definition of truth in terms of truth, of reality in terms of reality. Of course, if there is no truth but reality, truth is not definable; as anything different from the facts and relations of objective reaUty, it has disappeared from the universe of discourse. The problem of truth is indeed a hard nut for the neo-realist to crack, because when properly conceived it implies a duality of idea (as mental product, an image with its function), and inde- pendent reality, a duality which is logically incompatible with an absolute epistemological monism. Neo-realism is conse- quently forced to treat ideas and truth in very cavalier fashion, with the resulting insoluble problem and the convenient ambi- guities that we have seen. J. E. Boodin's doctrine of truth is more pragmatic than that of other American realists. Not only does he contend that truth is sought from practical motives ^ and is instrumental ; ^ even though he insists that the nature and test of truth ^ are not to be confused with the practical motive which leads to the seeking of it,^ he offers, with reference to real objects, definitions of truth that are more pragmatic than intellectual- istic.^ His ideal for truth, however, seems to be intellectual- istic. What thought really means, he says, is identification.® Truth, or the validity of an idea or belief, he defines as the agreement or tallying of that idea with its reality.^ It is only because our description can never give the complete equivalent of real objects, and because so much of our thought is merely symbolical, that he is forced, in certain cases, to go beyond this simple intellectualistic conception of truth.^ But when we consider that in typical judgment the subject is some real object, Boodin's would-be simple intellectualism and enforced pragmatism really signify the breakdown of the former theory. The attempt is made, however, to save a shred at least of intellectualism, when it is claimed that whenever our knowledge is concerned with social and ideal structures, it comes to share in the ideas it would represent, and so it is no longer of reality, but is reality. Its copying of the object is 1 Truth and Reality, p. 216. 2 76., pp. 123-4, 184, 217-18. 3 lb., pp. 195-7. « lb., pp. 103, 210-11, 224. ^ 75., pp. 133, 219, 236. « lb., pp. 98-9. 7 75.^ pp. 210-11, 214, 234. ^ lb., pp. 214, 219. 396 THE PROBLEM OF KNOWLEDGE a reproducing of it; the knowing process, when it deals with psychological unities, is the nature of the object.^ But at this point Boodin's thought does not quite accurately represent the facts. When the object of a knowing process is itself a knowing process, it is ordinarily not the same knowing pro- cess; and, as we have seen,^ it is a much debated question whether there ever can be a knowing process which knows itself. Or, to speak logically rather than psychologically, in terms of ideas rather than in terms of knowing processes, even when B's judgment has for its subject A's whole judgment, then B's judgment, hke all other judgments, must be to the intellectual- ist either tautologous or not quite true; for judgments, and even general facts of impHcation, are judged about by means of predicates which are themselves, as predicates, not facts but logical ideas, so that the relation of predicate to subject cannot be one of exact identity. But if what is meant is that B's judgment has for its subject the same reahty as was taken as subject of A's judgment, and if B predicates of this same subject exactly the same idea (in the sense in which ideas of different persons can be the same) as A predicated, then B's judgment may be said to be identical with A's (except that it is B's rather than A's) ; but the identity is not the truth, for the supposed identity is not, like the relation we call truth, between the subject and the predicate of the judgment; it is between two judgments. It is thus found impracticable to deal in purely intellectual istic fashion with even the small corner of truth which Boodin has sought to reserve for such treatment. Among the English new reaHsts Bertrand Russell is the only one whose theory of truth should concern us. S. Alexander's statement that there is truth whenever the mind works so as to be in the presence of objects in the order and arrangement in which they exist,^ amounts to httle more than that there is truth whenever the mind works so as to give it. Russell's theory is elaborated at length, but in the end it looks like the last stand of a retreating and practically defeated intellectual- ism. He speaks of coherence as being often an important test of truth, but he cannot regard it as affording any infallible 1 Truth and Reality, pp. 219-21, 223. 2 ch. XIV, supra. ' Proceedings of the Aristotelian Society, 1909-10, p. 27. A CRITIQUE OF INTELLECTUALISM • 397 criterion. 1 Indeed he disclaims all ability to find any universal criterion of truth ; and yet, strangely enough, in spite of this he still ventures to define its nature.- But it would seem vain to attempt a definition of that for which there is no criterion ; the proximate genus might indeed be given, but not the differ- entia of the species. Something true about truth might be stated, but not that which distinguishes it from falsity. As a foundation for his definition of truth, however, Russell describes the judgment, or belief, to which truth or its opposite applies, as being not a dual but a multiple relation of mind to its various terms or objects.^ He uses as his illustration Othello's belief that Desdemona loves Cassio. Here believing is a relation which unites the conscious subject (Othello) as one term to the other three terms, Desdemona, loving, and Cassio. Thus the constituents of a judgment are the subject, or mind, and several objects ; and so judging is quite like every other relation in that it unites a number of terms into a com- plex whole.^ These considerations are evidently intended to support a realistic absolute monism by showing that truth can be defined, if not as identical with reality, at least as identical with a part of reality ; it is a complex of terms related in cer- tain special ways to each other. But, when we come to examine RusselFs definition of truth, we find that it does offer (as we might have anticipated, in spite of his disclaimer), although in a somewhat covert way, and however inadequately, a truth-criterion; and we also find that the author fails quite to extricate himself from the time-honored view of truth as a dual relation. ''The judgment is true," he says, ''when the relation which is one of the objects relates the other objects." ^ This somewhat cryptic expres- sion, when translated into the concrete terms of the above illustration, means that Othello's judgment is true if the "lov- ing," which is one of the objects before his mind, really does relate the other objects, "Desdemona" and "Cassio." But this is as much as to say that truth is a relation of mind (with its ideas, or of an idea or complex of ideas in or before a mind) 1 Problems of Philosophy, p. 193. 2 Philosophical Essays, p. 173, ^ lb., pp. 117 ff. ; Problems of Philosophy, pp. 193-5. * Problems of Philosophy, pp. 197-9. ' Philosophical Essays, p. 181. 398 THE PROBLEM OF KNOWLEDGE to reality, such that the relation which obtains for mind, or in idea, as uniting the terms, is the relation which unites the objects in reahty. But this is a return to that view of truth and the judgment which makes them consist in some sort of a dual relation between idea and reality, even if both the idea and the reahty are somewhat complex entities. Indeed in his later work Russell virtually confesses as much, when he says that correspondence with fact constitutes the nature of truth,^ and that a belief is true ''when it corresponds to a certain associated complex," ^ or, more simply, ''when there is a corre- sponding fact." ^ And so all the difficulties of the correspondence theory recur, and that because, as Russell confessed at the outset, there does not seem to be available, from the purely intellectuahstic point of view, any adequate criterion, which shall state the kind and especially the degree of correspondence necessary and sufficient to differentiate truth universally from its opposite. It will not do to measure the truth by the identity of the rela- tion between the mental terms and the relation between the real objects, even if there were no difficulty in conceiving that identity ; if the judgment is true, there must also be an identity or correspondence, the exact nature of which the pure intellec- tualist cannot tell, between the terms mutually related in the idea and the objects existing in real relations. For the con- sistent pure intellectualist, no true judgment can have any meaning, and no judgment which has meaning can be true. And in the case of terms and relations both, since there cannot be absolute identity between idea and reality, between predi- cate and subject, if there is to be any judgment at all, just how much identity is necessary for truth? To this question the intellectualist has no answer; he has no adequate criterion of truth. In closing this discussion of intellectualism we would sug- gest, as a counter-weight against the one-sided emphasis upon identity, the "new law of thought" formulated by E. E. Con- stance Jones. This law, which Miss Jones calls the Law of Significant Assertion, is to the effect that " any subject of predi- cation is an identity of denotation in a diversity of intension." ^Problems of Philosophy, p. 193. '^ lb., p. 201. ^ lb., p. 202. A CRITIQUE OF INTELLECTUALISM 399 That is, every significant proposition expresses a difference as well as an identity; if there is no difference the proposition is meaningless.^ But, while suggestive as emphasizing a relationship too Httle considered by most intellectualists in their attempt to define truth, this formulation does not make possible an ade- quate purely intellectualistic definition. In the first place the formula deals with the denotation and intension of the terms of a 'proposition, so that before its doctrine can be applied to the question of the truth of judgments, a certain translation is necessary. In the proposition, viewed as a dual complex of terms, the subject-term represents a reality existing independ- ently not only of this particular judgment, and of the prop- osition in which it is expressed, but independently also of its representation by the subject-^erm as well. Attempting to express this distinction in general form, we would say that the assertion that the subject is, or is represented by, the predicate, really means that that reality which is, or is repre- sented by, the subject-term (subject-idea) is, or is repre- sented by, the predicate (predicate-idea). What Miss Jones has shown amounts to no more, for our present purposes, than that the proposition which formal logic examines expresses the assertion that that which the subject-term denotes '4s'' the quality which the predicate connotes — the '4s" being taken, of course, not as expressing absolute identity, but in the sense which only the desired adequate definition of truth can state. The ''new law" may thus be regarded as virtually involving a process of deductive inference, representable by the following syllogism : That which the term (of the proposi- tion) A represents "is" (in the sense in which predication is valid) the term A ; the term A "is" the term B ; therefore that which the term A represents "is" the term B. It must be evident, then, that the change to the formal proposition from the act of judgment in which a "floating adjective" is afl&rmed of some reality does not remove for intellectualism, even with the aid of the "law of significant assertion," the puzzle as to the criterion of truth. The reahty denoted in the judgment ^ A New Law of Thought and Its Logical Bearings, 1911 ; Proc. Aristot. Soc, 1910-11, pp. 166-86. See especially pp. 166-9. Cf. ib., 1906-7, pp. 81-92. 400 THE PROBLEM OF KNOWLEDGE which the major premise above expresses is numerically identi- cal with the reahty denoted in the judgment expressed in the conclusion, and the quality connoted by the term B is different from the quality connoted by the term A; but the question still remains as to what is the exact nature of that relation in true judgments which is expressed in propositions by the copula. How can the subject-matter, an independent reality, be the predicate, a logical idea? This is the question which intellectualism is forced to face, and which intellectualism by itself is unable to answer. CHAPTER XVIII A Critique of Anti-Intellectualism Some philosophers; in order to escape the difficulties of the intellectualist, abandon the idea that truth is attainable by means of ideas, and avoid scepticism only by falling back upon immediate feeling or intuition, while others, though they regard ideas as valuable for the attainment of truth, would not find this truth in any sort of identity between subject and predi- cate, but in the purely practical value of the ideas. The former view may be called anti-conceptualism ; the latter is that of current pragmatism. Both are forms of anti-intellec- iualism, so extreme as to be properly characterized as absolute logical monism; they recognize but one criterion of genuine truth, and that not the intellectuahstic. They may therefore be designated anti-conceptuahstic absolute logical monism and pragmatic absolute logical monism, respectively. Anti-Conceptualism The one great contemporary exponent of anti-conceptualism is Bergson. He points out that the conceptual mechanism of our ordinary knowledge, and especially of our ''exact" sciences, is of a cinematographical kind. Both our images and our concepts, the latter being the lighter, more diaphanous and easily dealt with, he likens to snapshots of the passing reality, which, on appropriate occasions, we are accustomed to bring before ourselves by means of the internal movement of our processes of thought. But the movement of our thought is another movement than that of the passing reality, and just as there is no movement in the snapshots of a moving object, so there is not in our concepts of the duration and life and movement and individuality that belong to the content of 2d 401 402 THE PROBLEM OF KNOWLEDGE immediate experience.^ Furthermore, according to Bergson, the categories we habitually use in thought are, relatively to the particular phase of realitj^ we may be seeking to know, pre- existing frames, into which the moving reality is forced, so that although we use them for purposes of knowledge, we are never able bj^ means of them to discover the real nature of that pre-existing moving reality .^ The platonizing attempt to gain knowledge of the real by means of an examination of human concepts is, therefore, to take an artificial and inadequate imitation for the original, a static substitute for the living and moving reality.^ The Kantian, too, although he takes the ideas as mere relations, is in much the same position as the Platonist, who takes them to be independent things.^ True metaphysics, it is claimed, is the science of reality which would dispense with symbols; it will soar above all concepts and all relations established by thought.^ Thus Bergson not only reacts from the intellectualistic attempt of logical idealism to find knowledge in mere ideas, conceptual predicates apart from any immediately given sub- ject ; he also rejects as an undue intellectualism the idea that the forms of intellectual apprehension, even when applied in conjunction with the immediate data of consciousness, can give us the truth about realitj^ He goes to the anti-intellectualistic extreme of looking for knowledge in the bare immediacy of the subject, apart from all conceptual predicates and apart from everything which may be supposed to have been revealed by such predicates. Bergson's course here is excusable, if at all, only from the point of view of the psychological idealism which, as we have already seen,® is the underground foundation upon which his system is actually based. If the reality existing in- dependently of explicit thought is itself essentially dependent upon consciousness, if there is no essential difference between "matter" and "images," ^ then the ideas brought to the sub- 1 H. Bergson, Time and Free Will, pp. 115 ff., 228 ; Introduction to Metaphysics, tr. by Hulme, p. 67 (by Luce, p. 79) ; Creative Evolution, pp. 160, 305-6, 318, 321, 329, etc. 2 Creative Evolution, pp. x, xiv, 48-9, 197. 3 Introduction, Hulme, p. 75 ; Luce, p. 88 ; Creative Evolution, pp. 4-5, 48-9, etc. ^ Introduction, Hulme, 83-5 ; Luce, 98-100. B 76., Hulme, 9, 21 ; Luce, 12, 26. « Ch. VI, supra- 7 Matter and Memory, passim. A CRITIQUE OF ANTI-INTELLECTUALISM 403 ject-matter in judgment are, to the interest in ultimately valid knowledge, a corrupting factor; the psychical cannot receive a psychical addition without undergoing modification. Not only is it true, as Bertrand Russell contends, in his critique of Bergson's philosophy/ that it is the idealistic confusion of sub- ject and object that has led this interesting philosopher to such paradoxical doctrines as that the brain is an unimaged image, that matter is the perception of matterj and that unperceived matter is an unperceived image, i.e. an unconscious mental state ; we may add that it is because of this same underground idealism, this idealism in disguise, that Bergson is forced to conclude that the only way, if there is any way at all, of reaching absolute or independent reality is to dispense with all products of thought. But this disguised psychological idealism is simply a disguised form of a philosophy based, as we have seen, upon fallacy. In place of seeking true knowledge by means of intellection, then, Bergson would have recourse, in metaphysics at least, solely to intuition. He distinguishes between sensuous in- tuition and a supra-intellectual intuition, and it is the latter with which he is here especially concerned. ^ For example, the true nature of the self, as of duration and change, is given immediately in our own direct self-experience, whereby, in- stead of merely circling about the object in conceptual flights, we penetrate into the very heart of it and view it from within.^ Supra-intellectual intuition is a sort of artistic sympathy, by means of which one seeks to share the inner life of the object he would know, and it is able to '^suggest to us the vague feel- ing, if nothing more, of what must take the place of intellectual molds." ^ From this point of view philosophy comes to be fundamentally '^an effort to dissolve again into the Whole." ^ This is not the place where an adequate estimate of what we may perhaps call the new intuitionism in philosophy should be attempted, and we would be far from maintaining that Bergson's doctrine at this point has no value ; ^ what we are 1 Monist, Vol. XXII, 1912, pp. 343-6. 2 Creative Evolution, p. 360. 3 Introduction, Hulme, pp. 1, 9, 22, 43 ; Luce, pp. 3, 12, 27, 51 ; Creative Evo- lution, p. 176. * Creative Evolution, pp. 177, 192-3 ; cf. La perception du changement, passim. 5 Creative Evolution, p. 191. 8 An excellent appreciation of Bergson's intuitionism is to be found in W. E. 404 THE PROBLEM OF KNOWLEDGE here concerned with, primarily, is his anti-conceptuaHsm. We would maintain that the real value of his reaction against in- tellectualism is to be chiefly found, not in his negative, but in his positive, doctrine ; in his insistence upon the necessity of immediacy (and immediacy not simply of sense, but of higher types as well) for perfect knowledge, not in his assertion of the futility of mediation. In taking this position we are doing little more than to call for a repetition in our own day of what was done in principle when it was insisted that concepts without intuition are empty. But what is needed further is the com- plementary insistence that intuition without concepts is, at least comparatively, blind. Of course, as Bergson sees, to take the concept as a substitute for immediacy is hkely to mean an impoverishment of knowledge ; but properly selected concepts, properly used, in addition to immediacy, mean an enrichment of knowledge, and, it may be, as in the case of tertiary qualities and relations, an enrichment of reality itself ; just as the perception of physical reality is its consummation, not its mutilation. Moreover, in further criticism of the doc- trine of intuition in the system before us, we would ask, espe- cially in view of the too idealistic doctrine that mere thought not only sometimes can, but always does, corrupt the purity of ultimate being, does Bergson make sufficient provision against the doubt whether, even in what he cites as instances of intui- tive awareness, we really do attain to absolute reality, as he contends that we do ? ^ May there not be, even here, some residue of undetected intellection ? Bergson recognizes, of course, the practical function of intellection. He recognizes that we attach to objects various concepts, which prescribe the kind of action or attitude the object ought to suggest to us under different circumstances,^ and indeed, that for practical purposes abstract ideas are not only convenient, but indispensable, as substitutes for intui- tion ; ^ but he warns against mistaking familiarity with a concept, through habitual use, for clarity of insight,^ and Hocking's article, "The Significance of Bergson," Yale Review, N.S., Vol. Ill, 1914, pp. 303-26. 1 Introduction, Luce, pp. 4, 10, 12. * 76., p. 49 ; Creative Evolution, pp. 12, 44, etc. » Introduction, Luce, pp. 23, 59, 64. *Ih., p. 101. A CRITIQUE OF ANTI-INTELLECTUALISM 405 stigmatizes the supposed truth of our practically justified judg- ments as merely relative, and ''no more than a symbolic verity." ^ Concepts cannot give us true knowledge, but only a practical substitute for it ; even of science the function is not to show us the essence of things, but to furnish us with the best means of acting on them.^ Inasmuch, then, as Bergson sets up so radi- cal an antithesis between genuine truth and practical value, he is to be regarded as an anti-pragmatist, as well as an anti- intellectualist. Perhaps the most serious difficulty encountered by the Berg- sonian philosophy is met when attention is called to the fact that it attempts to exercise the elementary right of all philos- ophy to take shape as an explicit and coherent doctrine. Our philosopher claims that any true metaphysic must get beyond and dispense with concepts ; and yet, in so far as he states his own metaphysical position, he is forced to make use of concepts. He is himself well aware of this, of course; and at this point he advances a compromise doctrine. Metaphj^sics is wholly itself, he claims, only if it frees itself from the inflexible, ever- ready concepts, and constructs concepts entirely different from thesg — pliant, mobile, almost fluid representations, capable of following reality in all its twists and turns, ever ready to adapt themselves to, and to pictorially suggest, the fleeting forms of intuition.^ In illustration of this distinction two prop- ositions are cited : ''The child becomes the man," and "There is becoming from the child to the man." Here "becomes" is represented as masking the movement of the reality, while in the second proposition, "becoming," being the subject, comes to the front as the reality itself, so that "we now have to do with the objective movement itself, and no longer with its cinematographical imitation." ^ But this compromise in order to avoid self-refutation can hardly be considered successful. Bergson does valuable work, indeed, in criticism of some of our metaphysical concepts, but in doing so he perforce substitutes for them others which either represent the subject-matter more accurately, or else are still 1 Creative Evolution, p. 196. « 75^^ p^ 93^ , 3 Introduction, Hulme, pp. 21-2, 69-70 ; Luce, pp. 26-7, 82. * Creative Evolution, p. 313. 406 THE PROBLEM OF KNOWLEDGE more highly metaphorical, more sketchy and symbolic. It is not that we object particularly to this procedure ; some of our most valuable knowledge is contained in metaphorical judg- ments. What we know with is always necessarily somewhat different from what we know, as well as in some sense identical with it: and there may very well be an advantage, theoreti- cally as well as practically, if time, which we immediate^ ex- perience as ''duration" — to cite Bergson's own favorite example — is sometimes thought of in ''spatial" concepts. It remains, then, that anti-conceptuaHsm violates its own prin- ciple in becoming a doctrine; obviouslj^ the only consistent course for its advocate — much as we may rejoice that, like the mystics, he has refused to be consistent — would be to cease to speak, or even to think, in which case he would be- come philosophically negligible. The fact is, however, that if you scratch an anti-con ceptualist you find an intellectualist who has become so thoroughly sceptical that he has begun to advocate the giving up of the effort to make a judgment at all. His former intellectual interest persists, however, even outside the limits of the narrowly practical; and so he goes on as before, multiplying concepts and judgments, in order that he may discover and communicate the truth. William James, in his volume, A Pluralistic Universe, gives the anti-conceptualism of Bergson an anti-logical turn. He claims that in view of the impotence of the intellectualist logic of identity, we must regard human experience as funda- mentally irrational.^ He tells us that, at a certain point in his thinking, he finally found himself compelled to give up logic, fairly, squarely, and irrevocably, so far as our becoming theo- reticallj^ acquainted wdth the essential nature of reality is con- cerned. Reality, life, experience, concreteness, immediacy exceeds our logic, overflows and surrounds it.^ And the credit for this notable discovery, as it seems to him, he gives to Berg- son, without whose influence, he confesses, he would not have been so bold.^ 1 A Pluralistic Universe, p.211. ^ lb., p. 212. ^ lb., pp. 214r-15, and Lecture VI, passim; cf. article entitled, "Bradley or Bergson?" Journal of Philosophy, Vol. VII, 1910, pp. 29-33. A CRITIQUE OF ANTI-INTELLECTUALISM 407 Current Pragmatism We have thus been led to reject as unsatisfactory, on the one hand, absohite ideahstic intellectuahsm, at least in its more characteristic (epistemologically monistic) forms, because, in its account of truth, it virtually eliminates the distinctive sub- ject-matter of the judgment ; and, on the other hand, to reject both absolute realistic intellectuahsm and anti-conceptualism, because, in their account of truth, they virtually eliminate, each in its own wa}'^, the predicate. But if we assume that truth is a quality of judgments, so that both subject and predicate are required, what possible logical theorj^ is there, which may be expected to deal more fairly with both these essential ele- ments of the judgment? Now it is important at this point to note that the anti- conceptualist and all intellectualists who do not virtually deny what are ordinarily called ideas alike recognize the practical value and even necessity of the ideas which we predicate, although they do not deign, of course, to make use of this con- sideration of their practical value in discussing the problem of their truth. It may well be, however, that the stone which has been rejected by these would-be builders is to become the headstone of the corner in the temple of truth. At any rate this is the opinion of current pragmatism, which, without losing faith in the intellect, would restrain the extravagances of in- tellectuahsm,^ and which seizes upon the practicality of ideas, claiming to find in the function of truth the key to its criterion, and consequently to be able to define its essential nature. But the term '' pragmatism '' has come to stand in contempo- rary discussion for so many more or less widely different points of view, actual or imagined, that it seems highly desirable to raise the question as to just what is the essential element, or what the essential elements, in current pragmatism. On the one hand we have philosophical critics, such as Bradley, com- plaining of the ''ambiguity of pragmatism," ^ and popular writers expressing such criticisms as that ''if it is new, it is nonsense; if it is old, it is obvious" ; ^ and on the other hand 1 Cf. Schiller, Humanism, p. 6. 2 Mind, April, 1908, and Essays on Truth and Reality, pp. 127-42. «E. E. Slosson, The Independent, Feb. 21, 1907. 408 THE PROBLEM OF KNOWLEDGE we have William James himself complaining that the pragmatic movement is seldom spoken of with clear understanding. ^ In- deed, some years ago A. O. Lovejoy undertook a classification of "the thirteen pragmatisms" ; ^ and yet, in the words of A. W. Moore, '^as some pragmatists deny belonging to any of these, it seems certain that there are more." ^ Perhaps the fairest way of at least beginning the answer to this question as to what pragmatism is, would be to try to settle it pragmatically. This will involve a certain measure of anticipation of results; but then, on the other hand, it is a method to which the pragmatist himself ought to be the last person to object. In pragmatism, then, what is the practical attitude? What does pragmatism really propose to do? It surely includes more in its program than the invention of ''a new name for some old ways of thinking." One of the younger members of the school has recently said, "The mission of prag- matism is to bring philosophy into relation to real life and action" ; ^ and probably all leading pragmatists would indorse such a statement. But just what such a "mission" means to the pragmatist must be inquired more particularly. And at the outset it ought to be conceded that pragmatism, at least as represented by the leaders, has not intended to make for greater laxity of thought, but rather to introduce a more scien- tific method into philosophy, and to arrive at a more scien- tifically accurate notion of the meaning of truth. ^ And since in all scientific judgment the predicate is regarded as a mere trial-predicate, and the judgment is made purely hypotheti- cally at first, in order that by acting as if it were true it may be shown by the manner of its working whether or not the best hypothesis was chosen, the pragmatist concludes that the true way of deciding the truth or falsity of rival philosophical theo- 1 Pragmatism, p. 47. 2 Journal of Philosophy, V, 1908, pp. 5-12, 29-39. 3 Pragmatism and Its Critics, p. 1. Another writer (Max Meyer, Journal of Philosophy, V, pp. 321-6) claims that while Lovejoy's "thirteen pragmatisms" are but different aspects of the same doctrine, we may well expect to find as many pragmatisms as there are pragmatists. * D. L. Murray, Pragmatism, p. 70. 5 James, Pragmatism, pp. 51, 55, 65-6, 216-17; Schiller, Humanism, p. 105; Dewey, Influence of Darwin, etc., p. 269 ; cf. H. H. Bawden, Journal of Phi- losophy, Vol. I, 1904, pp. 62 ff. A CRITIQUE OF ANTI-INTELLECTUALISM 409 ries must be to treat them as working hypotheses, and to judge them by the way they work. If the hypothesis has been thor- oughly tested, and has worked satisfactorily, it is not only use- ful, he claims, but true.^ According to Schiller ''pragmatism as a logical method is merely the conscious application of a natural procedure of our minds in actual knowing." ^ Of essential pragmatism this may be true, but whether it has been generally true of current pragmatism is another matter. Still, the pragmatist does not necessarily claim that all judgments that are in any particular sense subjectively or tem- porarily useful are true. It is the fault of the typical absolutist critic of pragmatism that he has a passion for expressing every movement and tendency in the form of some universal prin- ciple from which it might have been deduced, and it is his mis- take that he supposes, when he has refuted this principle, that he has virtually annihilated the movement. But mere essen- tial pragmatism does not assert universally that all that is useful, or that works, is true; it merely takes as its working hypothesis in logical theory the suggestion that the true test of truth is ultimately practical, a test of working ; and it sur- mises that there is no adequate and valid test of truth that is not ultimately a test of working, the results of mere specula- tion being problematic until verified in the experiences of life.^ As Schiller has expressed it, for pragmatism the truth or falsity of an assertion is decided "by its consequences, by its bearing on the interest which prompted to the assertion, by its rela- tion to the purpose which put the question." ^ The criterion of truth, according to Moore, is always ''the fulfilment of a specific finite purpose." ^ Before undertaking to elaborate further this essential prag- matism, or to examine further into its merits and deficiencies as a logical theory, it may be well to note just what more or less closely affiliated doctrines are distinguishable from this essence, either as falling short of it, or as going beyond it. These ^ See Schiller, Humanism, pp. 91-2 ; Studies in Humanism, p. 154 ; Moore, Pragmatism, and Its Critics, p. 87. 2 Studies, etc., p. 186. 3 Cf. Schiller, Studies, pp. 7-12. * lb., p. 154 ; cf. p. 192, and Dewey, Studies in Logical Theory, p. 85. 5 In Dewey's Studies in Logical Theory, p. 372 ; cf. Pragmatism and Its Critics, pp. 14, 15. 410 THE PROBLEM OF KNOWLEDGE other theories may be grouped together, we would suggest, into four main classes, which for convenience may be labeled semi- pragmatism, quasi-pragmatism, pseudo-pragmatism, and hyper- pragmatism, respectively. By semi-pragmatism is here meant any doctrine which undertakes to supplement acknowledged deficiencies of pure intellectualism by moving in the direction of essential pragmatism, but which fails to indorse the prag- matic criterion of truth. The term quasi-pragmatism we would use to designate the view that practical value is the measure of what, for practical purposes, we take, rightly enough, as truth, but that real truth is accessible only in some other way. Pseudo-pragmatism we would define as the doctrine that all practical value of ideas or judgments is an indication or proof of their truth.^ Hyper-pragmatism we would use as the name of the doctrine that in addition to the criterion of truth being always ultimately practical, the essential nature of truth, or trueness, is just practical value, usefulness, or the process of its working, its verification.^ Thus while semi-pragmatism and quasi-pragmatism assert less, pseudo-pragmatism and hyper-pragmatism assert more than the bare content of essen- tial pragmatism. Of the representatives of semi-pragmatism, i.e. of those who stop halfway on the road to the essentially pragmatic theory of truth, the most important in connection with the history of pragmatism is Charles Sanders Peirce. He is sometimes spoken of as the founder of pragmatism, but he would be more properly regarded as its forerunner. As early as 1878, in his now celebrated paper ''How to Make Our Ideas Clear," ^ he used the term pragmatism, but it was as the name of a doctrine not of truth, but of meaning. Claiming that the whole func- tion of thought is to produce habits of action, and that what- ever there is connected with a thought but irrelevant to this purpose is an accretion to it but no part of it,* he goes on to say that if we consider what effects, which might conceivably have practical bearings, we conceive the object of our conception 1 F. C. S. Schiller has used the term "pseudo-pragmatism" in another sense, which has not gained currency. 2 Paulhan uses the term "hyper-pragmatism"' (Revue philosophigue. Vol. 67, pp. 614 ff.), but in a different sense from that in which it is used here. 3 Popular Science Monthly, Vol. XII, pp. 286-302. " lb., p. 292. A CRITIQUE OF ANTI-INTELLECTUALISM 411 to have, our conception of these effects is the whole of our concep- tion of the object.^ Years afterwards, in Baldwin's Dictionary of Philosophy and Psychology, he defines pragmatism as 'Hhe doctrine that the whole 'meaning' of a conception expresses itself in practical consequences, consequences either in the shape of conduct to be recommended or in that of experience to be expected, if the conception be true." ^ The name pragmatism was chosen for this doctrine in view of its recognition of the inseparable connection between cognition and purpose.^ But in view of the ''extremes" to which, in his opinion, James later pushed the pragmatic doctrine,* Peirce, in order to register the more emphatically his dissent, proposed the name prag- maticism for his own more conservative doctrine.^ James's doctrines of the mutability of truth and of the will to believe seem to have been what repelled him most ; ^ but in drawing back in order to avoid these features of the later development, he was kept from accepting, as the logic of his own position might otherwise have led him to accept, the essential doctrine of pragmatism, viz. that of the necessarily practical character, ultimately, of the criterion of truth (about reality). The meaning of a concept is ultimately its mean-ing, its function of being a means to certain consequences ; but it may also be said that it is those consequences toward which the concept is a means. Peirce stressed the second of these definitions of meaning, although he recognized the other. James, as we shall see, went too far, going from meaning as consequences to truth as consequences, or the process of reaching intended consequences ; but Peirce was at fault in not recognizing that since meaning is to be told by the consequences to which that which has the meaning leads, and since truth is a judgmental expression of meaning, or a quality of that expression, truth also is to be told by its consequences. The great majority of semi-pragmatists are those who, like J. M. Baldwin, stress the practical function of truth, as ex- plaining its genesis and survival, but who define truth in purely 1 lb., p. 293. 2 Cf. Monist, 1905, pp. 162, 481. 3 lb., p. 163. * See Baldwin's Dictionary, article "Pragmatism." 5 Monist, 1905, pp. 166, 481 ff. • ^Hibbert Journal, Vol. VII, Oct., 1908, p. 112. 412 THE PROBLEM OF KNOWLEDGE intellectualistic fashion as mere agreement or correspondence with reahty. In his address on "Selective Thinking," ^ Bald- win seems on the verge of passing from semi-pragmatism to essential pragmatism; he says that correspondence between the idea and the fact constitutes truth, and yet he insists that a truth is not selected because it is true, but is true because it has been selected. ^ " But the mode of expression here was rather clumsy and inaccurate ; a truth is true, not because it is selected, but because it is fit to be selected. Consequently Baldwin was compelled to retreat from the pragmatic border- territory. In the Psychological Review for July, 1903, although he claims, in pragmatic fashion, that genetic theory explains ''by what character judgments are true," he exphcitly dis- avows pragmatism.^ In his paper on ''The Limits of Pragma- tism," ^ "without prejudice to a thoroughgoing pragmatic account of the origin of the function of thinking," ^ he never- theless objects to the view that the environment is a mode of pragmatically determined reality, because it assumes the reality of mental function and development, and this in turn requires us to assume a preexisting environment.^ Since, then, we cannot have a purely active determination of reality,^ he con- cludes that the same thing must be said of truth. "The true cannot be interpreted entirely in terms of the requirements of conduct," ^ but is only definable intellectualistically as "the body of knowledge acknowledged as belonging where it does in a consistently controlled context." ^ But this conclusion is quite dogmatic, depending as it does upon a confusion of truth with fact. One might agree that the current pragmatist interpretation of reality is untenable, and yet without incon- sistency indorse the pragmatic criterion of truth. Moreover, Baldwin's definition, amounting to no more than that truth is acknowledging that something is as it really is, evidently labors under the difficulties which beset all pure inteUectuahsm. J. E. Boodin may be mentioned again in this connection, as being, although in a different way, a half-pragmatist in his ^ Psychological Review, V, 1898, pp. 1-24 ; Development and Evolution, Ch. XVII. 2 75.^ p. 251. 3 See Moore, Pragmatism and Its Critics, pp. 193-4. * Psychological Review, XI, 1904, pp. 30 S. » lb., p. 60. « lb., p. 40. 7 Thought and Things, Vol. II, p. 350. « lb., p. 357. » lb., p. 361. A CRITIQUE OF ANTI-INTELLECTUALISM 413 doctrine of truth. He tends toward even the extreme pragma- tist doctrine with reference to the nature of truth, when the subject-matter is some reahty other than a social intellectual product — although his reaUsm keeps him from going quite so far as some have done — but his doctrine of truth with reference to ideal structures is, as we have seen, quite intellec- tualistic and non-pragmatic. Truth in the former case is said to consist in ''the differences which objects make to the reflec- tive conduct of human nature, as in its evolutionary process it attempts to control and understand its world." ^ It cannot be regarded as satisfactory, however, to have two different definitions of truth, neither of which applies to all cases of true judgments. Royce's "absolute pragmatism" also falls short of essential pragmatism. His voluntaristic insistence that the idea is a plan of action, that the judgment is a precept, ^ and that any definite opinion may be compared to the counsel given by the coach to a player,^ do not go beyond an emphasis upon the practical function of truth. He nowhere definitely proposes to measure trueness in any sense by the demands of practice, and yet, to revert to the simile of the coach and the player, just as in the game the coach himself is on trial, and his advice is ultimately to be judged in the light of its consequences, so must it be with judgments generally, if, as Royce contends, they are all precepts for the guidance of action. Royce, impressed simply with ''the practical value of theory," remains on the ground of semi-pragmatism; he ignores "the theoretic value of practice," ^ and so stops short of essential pragmatism. One more example of semi-pragmatism — this again of an- other sort — is to be found in the "negative pragmatism" of W. E. Hocking. Rejecting the positive principle, "What- ever works is true," as being neither vaHd nor useful, he adopts the principle, "That which does not work is not true." ^ But 1 Truth and Reality, p. 183 ; cf. p. 219. But see p. 236, where it is said that it is not truth, but its evidence, which consists in consequences. 2 "The Eternal and the Practical," Philosophical Review, XIII, 1904, pp. 119, 131. 3 Sources of Religious Insight, p. 152. * For these concise antithetical expressions I am indebted to H. V. Knox (The Philosophy of William James, 1914, p. 94). 5 The Meaning of God in Human Experience, p. xiii. 414 THE PROBLEM OF KNOWLEDGE we would maintain that even this negative pragmatism is unwarranted, unless some sort of positive pragmatism is also true. Of course, as we shall ourselves contend, we cannot be critical and say, "All that works is true" ; but it seems very improbable that we should be correct in saying, ''Nothing that does not work is true," unless it were also true that some posi- tive relation of importance existed between working and truth, that some kind of working might rightly be regarded as a criterion of truth. The negative pragmatist must go on to find an essential pragmatism of a positive sort, or else return to nonpragmatism, the doctrine that there is no dependence whatever of the truth of a judgment upon its practical function. We shall now examine briefly some representative statements of what we have called quasi-pragmatism, the doctrine that practical value determines the proper use of concepts and judg- ments as practical substitutes for truth. For example, we have Ernst Mach's statement that even in science our theoreti- cal conceptions, such as (those of) electricity, light-waves, molecules, atoms, and energy, are mere auxiliary instruments, created to facilitate some definite purpose, and that they possess permanent value only with respect to that purpose.^ Only experience is fact ; atoms, like all substances, are things of thought ; they are mere mental expedients, designed to fill out the gaps in our experience, which comes to us as if, but only as if, these things of thought were actual facts.^ Henri Poincare develops the same doctrine further, maintain- ing that the first principles of geometry and of mechanics are mere conventions, made to enable man the more conveniently to adjust himself to the changing facts of his immediate experi- ence.^ The Euclidean geometrj^ is not truer than non-Euclid- ean systems, nor is the Copernican theory truer than the Ptolemaic; the prevailing system is simply the more con- venient.^ By natural selection our mind has adapted itself to the conditions of the external world, and in doing so it has adopted the geometrical and scientific principles most advan- 1 Analysis of Sensations, Eng. Tr., pp. 186-7. 2 Science of Mechanics, pp. 490-4. 3 Science and Hypothesis, Eng. Tr., pp. 3, 98. ■* 7&.,-pp. 53, 85. A CRITIQUE OF ANTI-INTELLECTUALISM 415 tageous to the species, because the most convenient. Our sciences are not true ; they are convenient, advantageous.^ To Le Roy's doctrine that the scientist creates the facts of his science, Poincare objects ; but this is because, unHke Le Roy, he refuses to apply to atoms and similar scientific constructs, as he regards them, the name "fact." ^ In his opinion, all the scientist creates in a fact (a content of immediate experi- ence, a phenomenon) is the language in which he enunciates it; but this 'language" includes all the conventions of scien- tific thought; the scientific fact is only the crude fact trans- lated into a convenient language.^ In his Dernier es Pensees ^ Poincare adopts for his point of view the term pragmatism, which he defines as the function which an hypothesis has, of leading to consequences which are verifiable in the facts of experience. Hans Vaihinger makes a distinction between hypotheses and fictions; the former anticipate possible experience; the latter represent what can never be experienced, but what it is convenient or even indispensable for us, for practical purposes, to think of as if they were elements in possible experience. Thus the freedom of the will, atoms, independent reality, etc., are "indispensable fictions'' — pragmatically useful and even necessary, but not true.^ Closely similar to these views is Bergson's doctrine, in so far as it relates to the judgments of science and of common life. Intelligence, he says, is the faculty of manufacturing and using artificial objects, i.e. ideas, tools which may be employed to make tools.^ Especially with reference to life and action, our customary and scientific concepts can never be more than practically useful; they never amount to more than a con- venient substitute for true knowledge, which is accessible to immediate intuition alone.^ 1 lb., p. 65. 2 The Value of Science, Eng. Tr., pp. 114-16. 3 lb., pp. 120-1. * Pp. 146 ff. 5 j){g Philosophie des Als Ob, passim. 8 Creative Evolution, pp. 1.39-40. ' lb., passim. E. Le Roy's position is practically the same as that of his master, Bergson. Scientific laws he speaks of as "practical receipts," " not true but efficacious," which " concern less our knowledge than our action " and " enable us to control the order of nature rather than to discover it." Bulletin de la societe frangaise de philosophie, 1901, p. 5 ; cf. Revue de metaphysique et de morale, 1901, pp. 141, 560. 416 THE PROBLEM OF KNOWLEDGE With reference to this quasi-pragmatism' three things need to be said; it is to be appreciated, to be adversely criticised, and to be explained. In appreciation we would say that this '^ scientific pragmatism," as some have called it, has the merit of suggesting a way of introducing the pragmatic criterion into the shaping of our judgments in a way that is strictly methodical and intellectually justifiable. It seems to give promise of a synthesis of the essentials of intellectualism and pragmatism in an intelligible and serviceable definition of truth. On the other hand, the criticism is that the judgments so constructed have, according to these philosophers, to be rejected as not really true, but only convenient, or practically necessary. At this point it is interesting to note how closely this quasi-pragmatism approximates the doctrine of Albert Schinz, which he calls '^ anti-pragmatism.'' Pragmatism, he says, will carry the day, not because it is true — for it cer- tainly is false — but because it is desirable.^ The truth is sad and dangerous, he thinks; from the social point of view, the false is preferable to the true. For practical reasons, therefore, Schinz proposes a philosophically indefensible dualism of a philosophic truth, independent of consequences, on the one hand, and a pragmatic ''truth" on the other, not really true, but the social philosophy of the people, and conducive to the well-being of society.^ But this dualism which Schinz boldly acknowledges, this opposition of the necessary and the true, is in principle implicit as a disintegrating element in the doctrines of Mach and Poincare, of Vaihinger and Bergson. And finally, the explanation of this theory of practically and even scientifi- cally necessary untruth is to be found in the more or less dis- guised psychological idealism of all of these philosophers. According to their philosophical presuppositions, there is no independent reality; but we need, practically and scientifi- cally, to act as if there were. Hence, it is inferred, we need to believe what is not true. If, however, we refuse to accept psychological idealism — and we have seen no good reason for its acceptance — we are saved from the unpleasant dilemma in which these philosophers find themselves, and are at the same time able to retain the suggestions they give us as to a 1 Anti-pragmatism, Eng. Tr., p. 221. 2 /^.^ pp. 207, 250, 268. A CRITIQUE OF ANTI-INTELLECTUALISM 417 pragmatism that shall be scientific, i.e. intellectually justi- fiable, in its procedure. Essential pragmatism is not content to say, with semi- pragmatism, that all real live judgments which are true are in some sense useful to the person making them, although it would say that, with certain qualifications as to the kind of usefulness meant. (Judgments which serve to express im- mediate appreciation of ends would have to be recognized.) Nor is essential pragmatism satisfied, as is what we have called quasi-pragmatism, to have judgments constructed in the light of practical criteria, if these judgments are to be regarded as merely useful, or even practically necessary, but not true. It insists upon some sort of practical criterion of truth. But the attempt to state explicitly the essential nature of pragmatism has led to over-statements, in which much more is affirmed than can be easily or successfully defended. These over-statements may be divided into two groups, one of which, hyper-pragma- tism, although it goes beyond what is necessarily involved in essential pragmatism, is nevertheless a quite characteristic doctrine of current pragmatism ; while the other, pseudo-prag- matism, cannot be fairly regarded as a characteristic doctrine of current pragmatism, although, as we shall see, many lead- ing pragmatists occasionally allow themselves to lapse into forms of expression which, if taken literally, manifestly imply it. In the main, however, it goes not only beyond essential pragmatism, but beyond current pragmatism as well; and it may be regarded as existing for the most part in the imagination of the critics and in the minds of novices in the study of prag- matism. This pseudo-pragmatism is, or would be, as has been inti- mated, the doctrine that all judgments that happen in partic- ular cases to be useful in leading to the fulfilment of any kind of purpose, or even to the fulfilment of thoroughly worthy ulterior purposes, are true; or, in other words, that all satis- factory judgments are true, simply by virtue of their giving satisfaction to some particular desire. Now it is at once obvious that at least two varieties of this pseudo-pragmatism are theoretically possible, viz. the doctrine that what is useful for some particular purpose is true universally, and the doctrine 2e 418 THE PROBLEM OF KNOWLEDGE that what is useful for some particular purpose is true for that particular purpose. As might have been expected, it is the latter doctrine that eminent pragmatists have not always been completely successful in avoiding. Indeed it must be admitted that WiUiam James has been in this regard one of the worst of the offenders. Even the insist- ence that pragmatism is only a method, and is indifferent to particular results,^ gives some ground for suspicion. But we read further that ideas become true just in so far as they help us to get into satisfactory relations with other parts of our experience ; ^ that truth is a class-name for all sorts of definite working- values in experience ; ^ that true is the name of what- ever, in the way of behef, proves itself to be good, for definite, assignable reasons.^ As an illustration of what he means James declares that inasmuch as the Absolute affords reh- gious comfort to a class of minds, he unhesitatingly calls that Absolute true "in so far forth''; in giving people the benefit of a moral holiday, it is true.^ Logically, this leads James to the illogical doctrine of truths in mutual conflict; but here also he is '^ unhesitating " ; the greatest enemy of any one of our truths, he declares, may be the rest of our truths.^ Further on he assumes that when we make new appHcation of a '^cold- storage" truth, we can say of it either that it is useful because it is true, or that it is true because it is useful; and then he still more surprisingly adds, "Both of these phrases mean exactly the same thing." ^ Again he teaches that truth is only the expedient in the way of our thinking,^ and that on pragmaticprinciples we cannot reject any hypothesis if conse- quences useful to life flow from it.^ And yet, on the other hand, we must not overlook James's vigorous repudiation of the pseudo-pragmatic doctrine, at least in its cruder forms. Even in this same series of lectures he characterizes as an ''impudent slander" the charge that prag- matists say whatever they find it pleasant to say, and call it truth. 1^ In a later work he goes further. Not only does he 1 Pragmatism, pp. 45, 51, 54. 2 75.^ p. 53. 3 75.^ p. 68. " 16., p. 76. 5 76., pp. 73, 78. « 76., p. 78. ^ j^,., p. 204. « 76., p. 222. ^ 76., p. 273. 10 76., pp. 233-4 ; cf. The Meaning of Truth, pp. 70-1, where the charge is stigmatized as surprisingly shallow, and also p. 210, note. A CRITIQUE OF ANTI-INTELLECTUALISM 419 withdraw the saying that the Absolute is true in any sense ; ^ he assures Bertrand Russell that it is an "obvious absurdity" to suppose that any one who believes in a proposition must first have made out clearly that its consequences are good, and that his belief must primarily be in that fact,^ and in his reply to Marcel Hebert he disavows the doctrine ''that whatever proves subjectively expedient in the way of our thinking, is Hrue' in the absolute and unrestricted sense of the word, whether it corresponds to any objective state of things outside of our thought or not." ^ But we must not take this disavowal too uncritically. In the first place we would note that in his anxiety to repudiate the ''silly" doctrine Russell supposes the pragmatists to hold, James seems on the very point of disowning the most essential doctrine of pragmatism; good consequences, he says, are proposed by the pragmatist, not as the logical cue for his be- liefs, but as the cause or motive lying back of them.^ More- over, although his disclaimer as against the criticisms of Hebert shows that he honestly supposes himself to be free from at least the cruder form of pseudo-pragmatism (the doctrine that what is useful in a particular situation is true universally), we nevertheless find him defining the true as the expedient "in the long run" and "on the whole," ^ which qualifications plainly show that it is not simply "true for some particular purpose" that is intended. And even in the later work we read again that satisfactions grow pari passu as our ideas approach in- dependent reality, and that "the matter of the true is . . . absolutely identical with the matter of the satisfactory." ® He recognizes that the trouble Hes very largely in the ambiguity of the word practical ; ^ but he himself has done Kttle to reUeve that ambiguity. With Schiller the case is much the same as with James, in so far as pseudo-pragmatism is concerned. This doctrine he vigorously repudiates on occasion, but in his constructive statements on truth and its criterion he by no means always avoids it. He claims never to have been guilty of the simple 1 The Meaning of Truth, pp. viii-x ; 226-9. 2 j^,.^ p. 272. 3 /&., p. 231. * lb., p. 273. 5 Pragmatism, p. 222. 6 The Meaning of Truth, pp. 158-60. ' 76., p. 210. 420 THE PROBLEM OF KNOWLEDGE conversion from ''All truths work" to ''AH that works is true," ^ and says that not all that claims to be true, but only what has worked well, is to be accepted as true.^ But these distinctions and disclaimers are not very impressive when compared with such expressions as that whatever is relevant and conducive to our ends is true,^ that truth is the useful, eflB.cient, workable,^ that our truth is chosen,^ that it is unthinkable that any truth should fail to be satisfactory,^ that whatever works is true for the individual for whom it works,^ that different men are right in choosing different metaphysical systems,^ that if one enjoys his scepticism, or is satisfied to be inconsistent, he is at Hberty, from the pragmatic point of view, to be as sceptical or incon- sistent as he pleases,^ that since reUgion works, it is true, at least until superseded by something truer, ^^ and that our dis- carded ex-truths, although now error, really were truths in their day.^^ Similarly Alfred Sidgwick maintains, in the name of pragmatism, that so long as an assertion works, it is accepted as true, and is true for the purpose concerned, although next year's purpose may correct this year's truths.^^ Some other pragmatists, such as Papini, Le Roy, and others, have been perhaps even more guilty than James and Schiller in this connection, but the members of the "Chicago School" — at least the more responsible of them — have been fairly careful to avoid such pseudo-pragmatic utterances. Thus Dewey, in his important article, "What does Pragmatism Mean by Practical?" ^^ takes James to task for his careless manner of expression, saying that it seems unpragmatic for pragmatism to content itself with finding out the value of a conception whose own inherent intellectual significance prag- matism has not first determined by treating it not as a truth, but simply as a working hypothesis and method ; ^^ and that it is only consequences which are actually produced by the working ^Proceedings of the Aristotelian Society, 1910-11, pp. 163-4; Mind, N.S., XXI, 1912, pp. 532, 534. ^Studies in Humanism, p. 159; Proc. Aristot. Soc, 1910—11, p. 152. 3 Studies, etc., p. 152. * Humanism, p. 59. * Studies, p. 208. « Riddles of the Sphinx, 1910, p. 134. ^ Mind, N.S., XXI, 1912, p. 534. 8 Studies, etc., p. 18. » Journal of Philosophy, Vol. IV, 1907, p. 486. w Studies, p. 359. " 76., p. 212. 12 Mind, N.S., Vol. XXIII, 1914, p. 100. 13 Journal of Philosophy, V, 1908, pp. 85-99. " lb., p. 92. A CRITIQUE OF ANTI-INTELLECTUALISM 421 of the idea in cooperation with, or in appHcation to, prior reah- ties, that are good consequences in that specific sense of ''good" which is relevant to estabhshing the truth of an idea.^ More- over he disclaims having ever said that truth is what gives satis- faction, or having ever identified any satisfaction with the truth of an idea, save that satisfaction which arises when the idea as working hypothesis is applied to prior existences in such a way as to fulfil what it intends.^ A. W. Moore also insists that pragmatism is not, and must not be, a substitution of faith or will or feeling for thinking, ^ and that it is not enough to say that true ideas are the ideas which ''work"; they must meet the demand of the concrete situation in which they arise ; they must work in the way they set out to work.^ But even of the Chicago School it is true that pseudo-prag- matic ideas and forms of expression tend to creep in. It may be questioned whether an idea's "working in the way it sets out to work," ^ is a formula either unambiguous enough or otherwise adequate to be a criterion of truth ; it would apply in the case of errors acted upon and not yet discovered to be such, but still taken to be truths. Indeed, the distinction between "working" and "working as it set out to work" would almost seem to correspond to the distinction between the two types of pseudo-pragmatism noted above, the one holding that what works at all is true generally, the other that what works at all is true so long as it works. But it is nothing less than the quintessence of pseudo-pragmatism that we have in the doctrine of Bawden, that truth is that which works in rela- tion to a purpose or end, and that not opinions only, but truths also, are rightly subject to compromise and change.^ Some explanation of the too prevalent error of pseudo- pragmatism is afforded by the fact, which we have already noted, that pragmatists commonly represent truth or trueness as a quality of ideas, rather than of judgments. This being as- 1 lb., p. 93. 2 lb., p. 94. 3 75.^ VI, 1909, p. 294. * The Functional vs. Representational Theories in Locke's Essay, p. 67 ; Prag- matism and Its Critics, p. 87 ; cf. Dewey, Influence of Darwin, p. 150 ; Journal of Philosophy, V, 1908, p. 94. * See, besides references in the immediately preceding foot-note, H. H. Bawden, Principles of Pragmatism, p. 199. Cf. also Schiller, Riddles of the Sphinx, p. 133. ^ Principles of Pragmatism, pp. 202, 204. 422 THE PROBLEM OF KNOWLEDGE Slimed, it is indeed necessary to say with Mrs. Helen Thompson WooUey, one of Dewey's disciples, that a content may be true in one set of circumstances and false in another, because the truth is never in the content of an idea, but in its function.^ But what it is most important to remark is that it is primarily in the function of the idea in the judgment that its truth is to be looked for ; otherwise the statement is almost certain to be misleading. Truth is to be found not in the content of the idea, but in its function (in the judgment) ; but what pseudo- pragmatism forgets is that truth is not to be looked for pri- marily in the function of the judgment (in practical life) , but in its content. Failure at this point is what even Dewey is con- strained to charge against James. ''What Mr. James says about the value of truth when accomplished," writes Dewey, ''is likely to be employed by some as a criterion for ideas as ideas ; while, on the other hand, Mr. James himself is hkely to pass lightly from the consequences that determine the worth of a beUef to those which decide the worth of an idea." ^ But the other doctrine about truth that we have mentioned as going beyond essential pragmatism, viz. hyper-pragmatism, is much more characteristic of pragmatists, and may be taken as an essential element in the "wider" or more "radical" type of current pragmatism. In fact, we would hold that while the doctrine of the ultimately practical character of the criterion of truth is the good essence, this hyper-pragmatism is, in a peculiar sense, we would say, the bad essence of current prag- matism as a logical doctrine. Speaking broadly, while essential pragmatism finds the criterion of truth in its function, hyper- pragmatism identifies truth with its function. In this more extreme development of the movement William James has per- haps been the most outspoken leader. He adds to the prag- matic method, which he takes over from Peirce,^ "a genetic theory of what is meant by truth." ^ Stated with characteristic boldness, the teaching is as follows: "Truth happens to an idea. It becomes true, is made true by events. Its verity is in fact an event, a process : the process, namely, of its verifying itself, its Yeri-fication. Its validity is the process of its valid- 1 Journal of Philosophy, VI, 1909, p. 301. 2 j?,.^ y, 1908, p. 94. 3/6., Vol.. I, pp. 673-87; Pragmatism, pp. 46-7. ^ lb., pp. 65-6. A CRITIQUE OF ANTI-INTELLECTUALISM 423 ationJ' ^ ''Truth is simply a collective name for verification processes," ^ i.e. processes which guide conduct agreeably.^ Moreover, ''verification and validation themselves pragmati- cally mean . . . certain practical consequences of the verified and validated data," * so that the doctrine in its final form comes to be that "the truth of any statement consists in the con- sequences." ^ Similarly according to Schiller, truth means "successful operation on reality," ^ a "manipulation" of our objects which "turns out to be useful." ^ In short, "verity is verification." ^ Schiller's doctrine is reproduced by a disciple, J. W. Snellman, in the assertion that the meaning of truth is indistinguishable from its test.^ The same general position is taken by the members of the Chicago School. Indeed in this matter Dewey is especially pronounced. Truth, he declares, denotes "specific verifications";^"^ verification, or the effective working of the idea, and truth are one and the same thing — "this working being neither the cause nor the evidence of truth, but its nature." " Truth may be defined in terms of agreement, only in so far as the " agreement " is interpreted as not essentially different from success. ^^ A. W. Moore seems to be rather more successful than most pragmatist writers in guarding against hyper-pragmatic statements ; and he takes exception to Perry's interpretation of the pragmatist doctrine, as confusing the criterion of truth with its constitution. ^^ But Bawden, who is perhaps the enfant terrible of the Chicago School, insists that "if the truth be one thing, and the practical consequences a wholly different thing, then pragmatism is not true." ^^ And even Boodin, in that part of his logical theory where he is a pragmatist, asserts that there is no ultimate difference between truth and the test of truth. ^^ Very interesting in this connection is the story of the con- version of J. E. Russell from intellectualism to pragmatism. I lb., p. 201. 2 75,, p. 218. 3 76., p. 202. * 76., p. 201 ; cf. p. 205, and Meaning of Truth, p. xv. ^ 76., p. 52. * Studies in Humanism, p. 118. ''Humanism, p. 61. ^Journal of Philosophy, IV, 1907, p. 493. 9 Mind, N.S., XX, 1911, p. 241. i" Influence of Darwin, p. 109. II 76., pp. 139-40 ; Mind, N.S., XVI, 1907, p. 337. 12 Journal of Philosophy, IV, 1907, p. 202. is 76., p. 576. 1* Principles of Pragmatism, p. 203. i* Truth and Reality, pp. 196-7. 424 THE PROBLEM OF KNOWLEDGE In an article in the Journal of Philosophy for 1906 he contended that the pragmatist doctrine, that the truthfulness of the idea is not different from its success, would not bear the test of critical examination. While ready to admit that the true idea does not always possess practical value, he maintained that it was only because of its agreement with reality in some non-prag- matic sense that it could have this usefulness. What makes the idea which guides the traveller — or the traveller acting upon the idea — successful, he insisted, is that the idea is the right or true one; and what makes the idea right or true is its agreement with the traveller's actual environment. ^ In a later series of articles he contended that it was futile for the pragmatist to reason with one who is not a pragmatist. So long as the intellectualist adheres to his own original definition of truth, the arguments of the pragmatist are unavailing. In the intellectualist's sense of the term ''truth," pragmatism is not true ; it is true only in the pragmatisms sense of ''true." Thus pragmatism is unable to make one a pragmatist; it can save from doubt only one who happens to be or to become a prag- matist.2 Responses to this urgent "cry de profundis for salvation from doubt" came from Dewey, Schiller, and James. Dewey asked how the lost traveller could compare his idea with the environ- ment, except by acting upon it.^ Schiller confessed his inability to cure a patient who refused to take the prescribed remedy, and contended that no further recommendation for a theory should be expected than that it was internally consistent, and that, if accepted, it would be found satisfactory. If the doubter would be saved, he must choose pragmatism, and, doing so, he would find it the true way of salvation.^ James, insisting that pragmatism gives an intelligible, concrete account of meaning and agreement, challenged Russell to produce a similarly definite statement of what the intellectualist means by agreement.^ This challenge was seconded by Schiller.^ In reply to his would-be deliverers Russell admitted that for consistent pragmatism the verity of an idea is its verification, 1 Journal of Philosophy, III, 1906, pp. 599-601 ; cf. Philosophical Review, XV, 1906, pp. 406-13. 2 75., jy, 1907, pp. 57-64, 242-3, 292. 3 lb., p. 202. " lb., pp. 236-7, 486-7. ^ 7?,.^ pp. 295, 296. « 7^., p. 435. A CRITIQUE OF ANTI-INTELLECTUALISM 425 but insisted that while one could indeed make the venture of faith, and treat an hypothesis as true, it was possible, even while doing so, to remain in a state of theoretical doubt. ^ But the unanswerable challenge to give a satisfactory intellectualist definition of truth remained a source of disquietude; and it was by no means clear that one could consistently remain in theoretical doubt as to the truth of the pragmatist doctrine of truth, once he had accepted it as his working hj^pothesis ; for, if it worked satisfactorily to act upon the hypothesis that truth is satisfactory working, then truth must be satis- factory working. For two years nothing on the subject of truth appeared from Russell's pen in the philosophical journals, and then he made confession of his conversion to the pragmatist faith, announcing that his change of view had been mainly due to his own attempts to remain an anti-pragmatist, and as such to meet the attack of the pragmatist, and especially his chal- lenge to specify the element of meaning of truth which prag- matism does not contain. He confessed inability to show how an idea could be true prior to its verification.^ Later in the same year he told of his having come to the conviction that prag- matism is not only a tenable doctrine, but offers a more satis- factory solution of the problem of knowledge than the doctrine it displaces.^ On the same occasion Russell, with the zeal of a new convert, tried to bring out into the full light of pragmatism a writer (Oliver C. Quick) whom he found occupying much the same ground as he himself had formerly held. Repeating the chal- lenge that had been too much for himself, he inquired what other than a pragmatic meaning could be given to the terms ''agree with" and "correspond to." ^ Quick, however, was more ob- durate than he had been, saying in reply, ''I maintain that reality is other than value, though I cannot define clearly what it is." ^ Thereupon Russell reaffirmed his own pragmatic position, defining the truth of ideas in terms of their funda- mental value. ^ And, finally, in his book, A First Course in Philosophy, he not only identifies the truth of an idea with the 1 lb., pp. 489-90. 2 lb., Vol. VII, 1910, pp. 23-4. 3 Mind, N.S., XIX, 1910, p. 547. f/&., pp. 548-9. 5 lb., XX, 1911, p. 257. 6 lb., p. 539. 426 THE PROBLEM OF KNOWLEDGE successful discharge of its function ; he follows James in identify- ing this good working or trueness of an idea with its verification, or being made true, and this again with the good consequences or satisfactory experiences resulting from acting on the idea. "The good consequences of an idea . . . are its verification, not . . . that they merely prove that the idea was true, they are the trueness of the idea itself." ^ Thus we read the story of the conversion of the intellectualist, not onl}?- to essential prag- matism, but to the extremes of hyper-pragmatism. Now this hyper-pragmatism does afford, as its adherents maintain, what would be, if true, a "concrete account" of the nature of truth. But even so undoubted a good as concreteness may be purchased at too heavy a price. The pragmatic ref- utation of this extreme pragmatism is that it so confuses the idea of truth as to make it of very little practical value. In the first place, the pragmatist of this extreme type "cannot separate the truth of an idea from our knowledge of its truth" ; ^ and yet both in science and in common life, we are forced to make use of the idea of hypotheses which are true, though not yet verified; judgments which turn out to have been true, though when first made they were not known as yet to be true. James uses for such cases the terms "verifiable" and "virtually true," ^ and Schiller classifies such unverified hypotheses as truth-claims.^ Dewey also seems to think it sufficient to speak of hypotheses as "candidates for truth," ^ "true before- hand" being explained as meaning nothing but "ability to work";^ until tested practically, beliefs are mere dogmas, he avers, not truths.'' But this distinction of the pragmatists between actual truth and virtual truth, or mere truth-claim, or candidate for truth, does not coincide with the practically necessary distinction of science and common thought, between truth entertained but not yet verified and truth known to be such. It has no pigeon-hole wherein to classify correct guesses and all truths as yet unverified. If hyper-pragmatism were true, 1 A First Course in Philosophy, 1913, pp. 202-4. 2 Schiller, Journal of Philosophy, IV, 1907, p. 493. 3 Pragmatism, pp. 207-9 ; The Meaning of Truth, pp. 101, 164-5. ^Humanism, p. 98, note; Studies, pp. 147-8, 193; cf. D. L. Murray, Prag- matism, p. 42, 5 Influence of Darwin, p. 141. ^ lb., p. 163. ^ lb., p. 167. A CRITIQUE OF ANTI-INTELLECTUALISM 427 it ought never to be possible for us to make a true judgment without its being completely verified from the first. But we are constantly learning that judgments which were at first merely tentative were nevertheless true. This hyper-pragmatism also leads, in the second place, to the rather revolutionary doctrine of the essentially temporary and mutable character of truth. The unchangeabty true, James regards as an ''ideal vanishing-point, towards which we imagine that all our temporary truths will some day converge." ^ Schiller speaks of what we now know to be errors as "discarded ex-truths"; they ''were 'truths' in their day," but truth is a commodity which is of a perishable nature. ^ In fact, the doctrine is a common one in current pragmatism, and it is but a natural consequence of the hyper-pragmatist's fundamental confusion of the nature of truth. Alfred Sidgwick, himself a pragmatist, says that for the pragmatist "all truths are pro tern truths at best, and the duration of their validity is uncertain." ^ One is tempted to inquire whether it is an absolute and unchange- able truth that no human truth is unchangeably true, and to remark that it would be more in keeping with the supposedly empirical temper of pragmatism to wait for any particular belief to be refuted, instead of dogmatically assuming beforehand that it is certain to be outgrown. It surely will not be per- manently satisfactory to hold that no truth will permanently satisfy, or that all things else are in a flux, and only pragmatism has come to stay. But in any case enough has been said to show that, weighed in the balances of its own criterion of "working," this extreme pragmatism as a theory of the nature of truth is found wanting ; hyper-pragmatism fails to work, except in the direction of de- stroying our practically necessary conception of truth. More- over, if, as has been maintained above, and as we shall more fully justify in the sequel, the essence of pragmatism can be set forth without making use of this doubtful principle of hyper- pragmatism, on the pragmatic ground that no difference should be recognized unless it makes sl difference, such an extreme doctrine might well be rejected by the essential pragmatist 1 Pragmatism, pp. 222-3. 2 Studies, pp. 212-3. ' Journal of Philosophy, II, 1905, p. 269. 428 THE PROBLEM OF KNOWLEDGE himself. As we have just seen, the only practical difference it makes seems to be a difference for the worse. But if hyper-pragmatism is such an unnecessary and in- convenient doctrine, how did it come to find so large a place in the creed of pragmatists? What is its explanation? As a matter of fact there are several considerations, almost any one of which would suffice to explain psychologically the genesis of hyper-pragmatism, but none of which, singly or in combina- tion with each other, is adequate to give it logical justification. In the first place, and probably most potent of all, is the effect of assuming that truth must be a quality of ideas, rather than of judgments. On this assumption there can be no truth except as the idea is brought into relation to reality, for a bare logical idea, an abstract predicate can only be true if it has something to be true to. It is concluded, therefore, that truth cannot be a property of ideas antecedent to verification,^ that it is a property of ideas only in verification,^ and so on, through all the charac- teristic inferences of hyper-pragmatism.^ But even on the basis of the assumption that truth is a property of ideas, the hyper- pragmatist infers more than is warranted. The idea would have to be brought into relation to reality, as it is in the judg- ment, to be true; but not necessarily as it is in the verified judgment. And so, more adequate than James's expression, ''virtual truth," or Schiller's "truth-claim," or perhaps even Dewey's ''candidate for truth," would it be to say that an idea is hypothetically true ; it would be true if it w^re asserted of a certain reality in a certain situation for a certain purpose, or certain purposes, in such a way as to fulfil certain conditions — just what these conditions are being the exact matter of dis- pute in connection with the definition of truth. But since the idea might fulfil these conditions without the individual judging hav- ing at the time the experience in the light of which the judgment is known to be true, it cannot be said that there is even a one-to- one correspondence between instances of truth and instances of verification, much less a remainderless identity between them. ^ Dewey, Journal of Philosophy, VI, 1909, p. 433. » Schiller, Journal of Philosophy, IV, 1907, p. 493. 3 Compare, in this connection, J, E. Russell's earlier and later positions, Journal of Philosophy, IV, 1907, p. 290 ; VII, 1910, p. 24. A CRITIQUE OF ANTI-INTELLECTUALISM 429 A second explanation of hyper-pragmatism, and one of scarcely less importance, is to be found in the peculiar conse- quences of James's transition from Peirce's pragmatic doctrine of meaning to the pragmatic doctrine of the meaning of truth. If the meaning of anything is best discovered by examining its consequences, it follows that the meaning of truth is best discovered by examining the consequences of truth; so that, pragmatically speaking, any truth is, ultimately, the practical difference it makes. But it ought to be remembered that this is true only in the same way that it is true that what love, or hate, or peace, or war, or righteousness, or sin is, is its conse- quences, the difference it makes practically in human experience. Manifestly it will not do to take every special pragmatic meaning forthwith as a definition. A definition must be reversible, simply convertible ; but special pragmatic meanings are no more reversible than is the relation of cause and effect. Of course we commonty assume that a thing is what it means; but, strictly speaking, an important distinction obtains between the two. Even from the standpoint of an acceptance of the pragmatic method, the definition states a certain universal minimum of pragmatic meaning, viz. what the thing is, or means, for all purposes ; but in addition to this it has a multitude of special pragmatic meanings, viz. what it may mean (mediate, be a means to) for certain special purposes, or, in other words, what consequences it will lead to when used in a certain way, as means to a certain ultimate end. The former, viz. "what it is," is its most proximate meaning; the latter, i.e. "what it means," means what it is more ultimately, in special cases. Now it is the mistake of the intellectualist that he tends to confine meaning to that which can be expressed in a reversible proposition, or definition, ignoring the fact that all meaning, even definition, is relative to purpose. But it is the mistake of James and his hyper-pragmatist disciples that they tend to eliminate meaning in the sense that is expressible in a revers- ible proposition, or definition, and to confine it to the multitude of additional special pragmatic meanings. And even this would not be so confusing, if this sum of special pragmatic meanings were not forced in the case of truth to do service as a definition. It is this procedure that forms the basis for the charge that he 430 THE PROBLEM OF KNOWLEDGE takes his pragmatic meaning of truth as true "in the intel- lectuahst sense." ^ Still another very potent influence in the direction of hyper- pragmatism has been the failure of pure intellectuahsm to give a satisfactory account of the nature and criterion of truth, combined with the impression that the more radical type of current pragmatism is the only logical alternative, since the pragmatic method is vaUd. This is shown conspicuously in the story of Russell's controversy with the pragmatists and his final capitulation.2 But that hyper-pragmatism is not a neces- sary consequence of the pragmatic method has been indicated, and in the constructive part of our discussion it will be our task to set forth another and more satisfactory alternative to absolute intellectuahsm. A further motive, and one which has grown out of the one just mentioned, has been the determina- tion to carry through to the end a consistent anti-intellectuaUst programme. This motive seems to have been especially operative in the controversial writings of Schiller. Another root of hyper-pragmatism has been the pragmatic view of reality, as fluctuating in correspondence with the flux of human purposes, as being what it is for us because so deter- mined by human will, individual or social. This appears rather prominently in the writings of Schiller and Murray, as a pragmatic realism within the hmits of a pluralistic subjective idealism,^ and in the " experience philosophy " of the Chicago School,^ according to which psychology and logic between them are considered competent to deal philosophically with the nature of reality, without any further metaphysics. Reahty is practical, they hold ; ^ taking a content of immediate experience as real makes it real, logically speaking — and that is the ultimate way of speaking — and what is thus real in the beginning of the judging act may be quite different from what is real at the end of the 1 See J. E. Russell, Journal of Philosophy, IV, 1907, pp. 61-3 ; J. B. Pratt, What is Pragmatism ? p. 128 ; James, The Meaning of Truth, pp. 197-200. ^Journal of Philosophy, IV, 1907, pp. 202, 291, 295-6; VII, 1910, p. 23; Mind, N.S., XIX, 1910, pp. 547-9; XX, 1911, p. 539; Russell, A First Course in Philosophy, pp. 202-5. 3 SchUler, Mind, N.S., XVIII, 1909, pp. 182-3 ; Murray, ib., pp. 389-90. * See criticisms by W. Fite, Philosophical Review, XV, 1906, pp. 1-16. s Dewey, "Does Reality Possess Practical Character?" in Essays . . . in Honor of William James, 1908. A CRITIQUE OF ANTI-INTELLECTUALISM 431 process.^ Now where the judgment is thus regarded as a capital operation on reahty, the idea is naturally interpreted as the surgical instrument, and its truth as nothing more than its efficiency, or the success of the operation. But all this is valid only on the more than questionable supposition that a realistic view is untenable. Finally, a minor ground of hyper-prag- matism, or perhaps a mere encouragement on its way, is the appeal to the etymology of the term verification.^ But trans- mutations of meaning have been so extensive and frequent that the value of an argument from etymology is now generally recognized as being well-nigh infinitesimal. But even when it avoids the errors of pseudo-pragmatism and hyper-pragmatism, what we have called essential pragmatism has its own difficulties. No one is a genuine convert to prag- matism, we have maintained, unless he proposes to live, intel- lectually speaking, by the principle of measuring truth, however cautiously, hy the standard of practical value, of usefulness. But once safely converted, it remains for the pragmatist to show his still doubting friends that he is able to recognize just what sort and what degree of usefulness may be taken as a guarantee of truth. Obviously not every sort or degree of practical value can be taken as an indication of truth, if the notion of truth itself is to retain for us any practical value. In the application of the pragmatic criterion for the deter- mination of truth, several definite problems have been en- countered. Of these the chief have been the avoidance of *' crass utilitarianism," the overcoming of ultra-individualism, and a due recognition of the theoretical interest, system, and strictly scientific methods. One of the commonest charges levelled against the pragmatic method has been that it is upon *Hhe dead level of utilitarianism." ^ Now it is undoubtedly true, as even the anti-pragmatists would admit,^ that originally consciousness and, in man, the judging process were valuable 1 A. W. Moore, Journal of Philosophy, IV, 1907, p. 571. 2 James, Pragmatism, p. 201. 3 T. De Laguna, Dogmatism, and Evolution, p. 140 ; cf. G. A. Tawney, "Utili- tarianism in Epistemology," Journal of Philosophy, Vol. I, 1904, pp. 337 ff. ; W. Caldwell, Pragmatism and Idealism, p. 136. * E.g. Bradley, Principles of Logic, pp. 459-60 ; Essays on Truth and Reality, pp. 75, 141. 432 THE PROBLEM OF KNOWLEDGE chiefly as means of better adjusting the animal organism to its environment, so that the physical life might be preserved and propagated. In that primitive situation the biological function of judgments, i.e. the way in which they functioned in the service of the physical life of the individual and of the race, was, roughly speaking, an index of their truth.^ But if it should be assumed that not only then but now and always the only test of truth is its function in man's struggle for physical existence, we would have an animalistic pragmatism which could not be ade- quate as a theory of the test of truth employed by any being whose life was above the merely animal level. It is a fact, however, that in conscious life new interests are constantly developing, many of which are not centred in the fate of the physical organism at all. Moreover these new interests peculiar to man as a spiritual personality may lead to a transvaluation of all former values, so that instead of life's being interpreted in its lowest terms, as the physical existence of the individual and of the race, it is interpreted in its highest terms, as the spirit- 1 It was on the basis of this fact that Georg Simmel developed, twenty years ago, a species of biological pragmatism, anticipating not a few of the features of the Chicago instrumentalism, but tending to reduce the criterion and nature of even the highest human truth to the level of mere usefulness for the furthering of the animal life (" Ueber eine Beziehung der Selectionslehre zur Erkenntnistheorie," Archiv filr systematische Philosophie, I, 1895, pp. 34-45). Having felt obliged, in view of such facts as that of the dependence of our representations upon the specific energies of our "psychical organs," to con- clude that we cannot reach the reality of things in themselves, Simmel combines with this representational agnosticism the theory that since for the lower animals satisfactoriness for the furthering of life is the only basis for distinguishing between "representations," so it must be in the case of man. Among the innumerable "representations" which occur, those which prove themselves biologically useful become fixed according to the well-known pro- cess of natural selection, and thus come to be regarded as the " true " representation of the world. Even when truth is imagined to have some other meaning than usefulness in the natural struggle for the furtherance of life, it can have ultimately no other criterion. But as a matter of fact the trueness of any thought means the uniformly satisfactory biological conse- quences of using it — nothing more. But, as Simmel himself remarks, it is a serious question whether the concept of truth will endure, when denuded thus of the notion of objective validity. That truth is, even in man, nothing but the value of mental contents for the animal life, is not, there seems good ground to surmise, the theory of truth most valuable for man's moral character, and so, ultimately, even for his animal life; and if this be true, then Simmel's theory of truth, even by bis_pwn criterion, is untrue. A CRITIQUE OP ANTI-INTELLECTUALISM 433 ual development and efficiency of the individual and society. Eucken's accusation against pragmatism, that ''it does not sufficiently distinguish between the natural desires and the elevation of life, between the decoration of a given world and the struggle for a new one, between what is useful and what is good,^^ ^ is not unworthy of consideration. Truth can be measured by a higher standard than its function in the struggle for bare existence, viz. by its function in the struggle for a better existence. Pragmatism, as instrumentalism, must remember that instead of consciousness and judgments being regarded as mere means for the promotion of the physical life, the physi- cal life is now regarded, even by people of ordinary spirituality, as simply or chiefly instrumental in the promotion of the con- scious life in its spiritual aspects.^ The ideal interests no longer exist for the sake of the physical, but the physical for the sake of the ideal. ''Man began to think in order that he might eat : he has evolved to the point where he eats in order that he may think." ^ Animalistic pragmatism, then, gives place to a humanistic doctrine, in which it is proposed to test the truth of judgments by their utility in the service of that life in which all the peculiarly and legitimately human interests are recognized as being of fundamental importance. The ultimate end, by being useful toward which the truth must, as means, accredit itself, must include the "perfect harmony of our whole life." * Once the ends in view are thoroughly accredited as humanly and spiritually necessary, it may be assumed, according to this humanistic pragmatism, that those judgments are valid which are ultimately necessary for the achievement of these ends. Thus necessity, in the sense of what is humanly and spiritually necessary, remains the test of the truth of judgments. But sometimes even humanistic pragmatism presents itself in an unduly individualistic form. The individual man as a purposive active being is taken as the measure of all values, including the truth of judgments. "What works," it is insisted by Schiller, "is true for the individual for whom it works." ^ 1 Knowledge and Life, Eng. Tr., pp. 94-7 ; cf. Main Currents of Modern Thought, pp. 79-81. * For a definition of the term "spiritual " see p. 448, infra. 3 W. P. Montague, Journal of Philosophy, VI, 1909, p. 489. * Schiller, Humanism, p. 61. 6 Mind, N.S., XXI, 1912, p. 534. 2p 434 THE PROBLEM OF KNOWLEDGE *'Men with different fortunes, histories, and temperaments ought not to arrive at the same metaphysic," he claims, ^'nor can they do so honestly ; each should react individually on the food for thought which his personal life affords, and the resulting differences ought not to be set aside as void of ultimate signifi- cance." ^ But elsewhere Schiller seeks to correct this ultra- individualism, and to pay due respect to "the social character of truth." 2 ''Society," he says, '^exercises almost as severe a control over the intellectual as over the moral eccentricities and non-conformities of its members. . . . Whatever, there- fore, individuals may recognize and value as Hrue,' the 'truths' which de facto prevail and are recognized as objective will only be a selection from those we are subjectively tempted to recog- nize." ^ With the Chicago School, on the other hand, the safeguarding against extreme individualism is no mere afterthought. A. W. Moore protests that the variety of pragmatism with which he is acquainted thinks of the "private consciousness" not only as born of, but as growing up in, and therefore continuing all the while vitally and organically related to, its social matrix, so that not only in its origin, but in its continued development and operation this consciousness, with its judgments and truth, must always be a function of the whole social situation. The need for readjustment is not "the need of some one, lone, ma- rooned organism or mind only/' and the readjustment, in those instances in which it does occur, is " always in and of a 'social situation.'" ^ According, then, to this revised or ortho- dox pragmatism — whichever it may be — it would appear that not only are the judgments we make social products; their truth must be decided by their experienced value to society. But even this social pragmatism is not without its diffiiculties. Strictly interpreted, it would lead to some curious results. For instance, in the days of the undisputed supremacy and social satisfactoriness of the Ptolemaic astronomy the universe was geocentric, but in the days of Copernicus it began to change its fundamental constitution, until at length it settled down into a multitude of heliocentric solar systems. 1 Studies, p. 18. 2 76., p. 155. 3 lb., p. 153. ^ Pragmatism and Its Critics, pp. 230, 232. A CRITIQUE OF ANTI-INTELLECTUALISM 435 Thus, accused of ultra-utilitarianism and ultra-individualism, pragmatism has been led to suggest the measurement of truth by spiritual edification and social acceptance. But are even these tests quite adequate? Certainly what has apparently been spiritually edifying has not always been true, nor can the criterion of social acceptance be made to seem adequate except at the cost of giving up our common-sense doctrines of the per- manence of truth and the world's non-dependence upon human experience for its existence and fundamental nature. In short, the tests examined so far fall indubitably short of fulfilling the conditions of scientific verification and fail to do full justice to certain elements of truth in intellectualism. To the task of solving the problems presented by the need of consistency and system, by the existence of the "theoretical interest," and by the normative character of the methods of science, current pragmatism has addressed itself, and in some instances with a considerable degree of success. This is es- pecially true in the case of the matter of consistency and system. The verification of consistency has come to be regarded as an essential part of the verification of life, the interest in ''ration- ality" being regarded as the fundamentally and ultimately practical interest in bringing into harmony the various ''practi- cal interests" recognized as valid. ^ This pragmatic interpreta- tion of rationality enables the pragmatist, then, to feel "that what he now thinks goes with what he thinks on other oc- casions." 2 In dealing with the theoretical interest in so far as it is broader than the mere interest in consistency, current pragmatism has not been, perhaps, quite so successful. "Reflective need comprehends theoretic and aesthetic need as well as practical need" ; ^ and the problem of the pragmatist is to find some com- prehensive sense of the "practical" which will include the other two as well as the more obviously practical. James confesses, 1 Cf. A. K. Rogers, Religious Conception of the World, p. 71. 2 James, Meaning of Truth, p. 211; cf. Pragmatism, pp. 216-17; Schiller, Studies, p. 151. But, we are tempted to ask, does all that the pragmatist commonly permits to be called "true" (in the judgments of others and even in his own past judgments, when satisfactory for the purposes which prompted them) " go " with what he now judges to be true ? See p. 451, infra. 3 G. A. Tawney, Journal of Philosophy, Vol. I, 1904, p. 340, 436 THE PROBLEM OF KNOWLEDGE and Dewey charges against him especially, that the term practi- cal has been used too carelessly by pragmatists.^ But, in general, while it is insisted that theory is an outgrowth of practice and incapable of independent existence as mere in- tellection,^ we get little further information as to the nature of intellect beyond the reiterated assertion that it is a special movement or mode of practice.^ Science is not inaptly de- scribed by Dewey as ''just the forging and arranging of instru- mentalities for dealing with individual cases of experience";^ but what is to be said about the pursuit of science as something interesting apart from its further application ; what about the interest in truth for its own sake ? There is apparently a lack of candor at this point among pragmatists, due doubtless to their fear of conceding too much to the anti-pragmatist. And yet, as we have already said, the outstanding representatives of pragmatism have not intended to undermine scientific pro- cedure, but rather to establish that procedure as the model for all philosophy. And there are not wanting statements of the nature of the ''working" required by pragmatism as criterion of truth, which constitute fairly good accounts of the process of scientific verification. James says, "To 'agree' in the widest sense with a reality can only mean to be guided either straight up to it or into its surroundings, or to be put into such working touch with it as to handle either it or something connected with it better than if we disagreed." ^ Similarly, according to Dewey, "the objective reality which tests the truth of the idea is not one which externally antecedes or temporarily co-exists with the idea, but one which succeeds it, being its fulfilment as intent and method."® Again he says, "Some assumption about the possibiUty of a change in the state of things as experienced is the idea — and its test or criterion is whether this possible change can be effected when the idea is acted upon in good faith." ^ And again, "It seems unpragmatic for pragmatism to content itself with finding out the value of a conception whose 1 James, Meaning of Truth, p. 207 ; Dewey, "What Does Pragmatism Mean by Practical?" Journal of Philosophy, V, 1908, pp. 85-99. 2 Schiller, Studies, p. 128. ^ Dewey, Influence of Darwin, pp. 125-6. * The Logical Conditions of the Scientific Treatment of Morality, p. 8. ^ Pragmatism, pp. 212-13 ; cf. The Meaning of Truth, p. 157. 8 Journal of Philosophy, IV, 1907, p. 313. ^ Influence of Darwin, p. 135. A CRITIQUE OF ANTI-INTELLECTUALISM 437 own inherent intellectual significance pragmatism has not first determined by treating it not as a truth, but simply as a working hypothesis and method. ... I have never identified any satisfaction with the truth of an idea, save that satisfaction which arises when the idea as working hypothesis or tentative method is applied to prior existences in such a way as to fulfil what it intends."^ Finally, J. E. Russell has this to say, ''The truth of an idea consists in the value of that idea in so guiding and controlling experience as to bring us into direct experiential relations with the particular object or part of the real world we may be seeking to know and practically to possess. This functional value of an idea is what we mean by its truth." ^ Manifestly what each of these writers has in mind is the process of verification in the empirical sciences. Are we to understand, then, that the only novelty introduced by essential pragmatism is a biological language into which the methodology of science may be translated? Or is it a way of getting the appearance of scientific justification for practically valuable philosophical doctrines by bringing both the acknowl- edged science and the valuable philosophy under a common formula? This is a crucial point which current pragmatism has left altogether too obscure, giving occasion for the gibe quoted above : ''If it is new, it is nonsense; if it is old, it is obvious." And the failure here is simply the last remainder of that pseudo-pragmatism which — perhaps not altogether un- wisely — leaves always vague and somewhat undefined the con- sequences by which the truth is to be tested.^ Perhaps prag- matism may yet be useful, and thus in some indubitable sense pragmatically justified, in showing how some of the contents of a spiritually valuable philosophy may become genuinely scien- tific ; but current pragmatism has not yet gone so far, nor has it clearly seen, apparently, that such an event is within the bounds of possibihty. 1 Journal of Philosophy, V, 1908, pp. 92, 94. ^Mind, N.S., XX, 1911, p. 539. 3 Cf. W. Caldwell, Pragmatism and Idealism, p. 51. CHAPTER XIX Ckitical Monism in Logical Theory We have examined the various attempts of logical dualism and an absolute logical monism, both intellectualistic and anti-intel- lectualistic, to solve the problem of truth, and have not found any that leads to wholly satisfactory results. Absolute intellect- ualism insists that in truth there is some sort of identity between idea and reality, but just what sort or degree of identity, it seems unable to state. Moreover it has failed properly to assimilate what has been formulated as the "Law of Significant Assertion," the fact, namely, that the predicate must always be different from the subject. Absolute anti-intellectualism in its anti-conceptualist form in sceptical fashion gives up the problem, at least so far as thought is concerned. In the form in which it appears in current pragmatism, while it holds, in a way that gives promise of proving tenable, to "the theoretic value of practice," and seems therefore at best to have hit upon something which does contain the criterion of truth, the differentia of truth as a species of some higher genus, still in its common hyper-pragmatic form it has too much ignored and even lost sight of the higher genus of which this is the specific difference. These results of our critical examination of rival theories of truth suggest for our further consideration and con- structive elaboration the idea that the solution of the truth- problem lies in the direction of a synthesis of certain elements of intellectualism on the one hand, and pragmatism on the other. May it not perhaps turn out that we shall be able to derive the proximate genus for our definition of truth from the one side, and the differentia of the species from the other ? The position toward which we have been moving, not only in the present discussion of the problem of truth, but also in our former discussion of the problem of acquaintance, is that in judgment an idea, an abstraction from reality, is predicated of 438 CRITICAL MONISM IN LOGICAL THEORY 439 some reality, generally of a reality immediately experienced in the past or at present, either by one's self or others, or at least experienceable in the future. But in view of the fact that, at the moment of judging, the subject-matter of the judgment is not ordinarily — if, indeed, ever — completely presented ; and in view of the further fact that it would seem unnecessary for the person judging to represent to himself what is at the moment fully presented, it begins to appear that predication is such representation as is required to supplement the presentation of the reality which constitutes the subject-matter of thought; it is, or aims to be, representation of the reality under con- sideration in so far as it needs to be represented, in view of its being already only partially presented, or already only partially presented and represented, which latter it is by virtue of pre- vious judgments, or of similar mental acts. According to this view, then, the typical judgment would be analytic of its sub- ject, rather than synthetic, because its subject is not a mere idea or thought-construct, but an independent reality with its primary and secondary qualities and relations. Only as re- lated to tertiary qualities and relations would the judgment be synthetic of its subject. On the other hand it could be freely admitted that all real live judgment is synthetic of the concept or idea we are coming to have of the subject. It should be noted that this view does not involve the absolute dualism in epistemology which we have seen sufficient reason to reject. In all judging there is a duality of subject and predi- cate, of reality and idea, of represented and representing; but this necessary duality does not involve an absolute dualism. Representation does not exclude previous and further possible presentation ; on the contrary it can make good its claims only if there can be and is direct presentation. One who is an ab- solute intellectualist in logical theory, and an absolute monist, idealistic or realistic, in epistemology, can find no place for knowledge by representation, and consequently no place for the truth of judgments, which obviously undertake such a repre- sentation. On the other hand the absolute intellectualist who is also an absolute dualist in epistemology, while he would make all consciousness, like judgment, merely representative, can find no representation which amounts to knowledge, because without 440 THE PROBLEM OF KNOWLEDGE direct presentation there is no touchstone by which the sup- posed representation may be measured, and thus, if not rejected as untruth, vindicated as truth, instead of being left as either truth or a mere practical substitute for it, we know not which. But even when the problem of acquaintance has been solved in the way outlined in our constructive statement above,^ the strict intellectualist is nonplussed by the problem of truth. Sometimes, indeed, he adopts the coherence theory and main- tains that the agreement which constitutes the essence of truth is the agreement of the judgment or proposition with other judgments or propositions. The only approach to a plausible excuse for this confusion of truth with mere consistency is to be found in the idealistic doctrine that there is no essential dif- ference between things and ideas, or propositions — a fallacious doctrine with which we have already sufficiently dealt. Very commonly, however, the intellectualist recognizes that judgment is representational, and that there may be true judgments; but just what constitutes the truth of judgments he is unable to say. Some sort and degree of identity or representation is required ; but the question is. What sort or what degree of identity or of representation is sufficient to insure the truth of the judgment? If, in judgment, we represent what is not at the moment adequately presented, and do so because we need to do so, our need being simply our need of the judgment for some practical purpose, may it not be that when the represen- tation satisfies our practical need, the judgment is true? But to say so would be to cease to be a mere intellectualist ; it would be to have adopted what might be regarded, from that point of view, as the essential element (the good essence) of pragmatism. But pragmatism itself does not remain unchanged when it consents to the definition of truth in terms of identity with or representation of reality. If there is to be a permanent settle- ment of the controversy between the intellectualist and the pragmatist, the latter must concede to the former that the par- ticular practical purpose in the interest of which a judgment is made may be satisfied by the judgment in some instances, without the judgment being therefore necessarily true. For 1 Ch. XIV. CRITICAL MONISM IN LOGICAL THEORY 441 example, if a nation, A, is at war with two nations, B and C, it may adequately serve the practical purposes in the interests of which the judgment is made if a soldier of A mistakes a soldier of B for a soldier of C. Indeed must it not always be, as the intellectualist claims, the purpose to know, the purpose of the investigator, the truth-seeker, fulfilment of which is to constitute verification, and not necessarily the purpose to make some further use of the truth after it has been obtained ? But then, would not to concede this to the intellectuahst necessarily mean the capitulation of the essential pragmatist ? Not necessarily. It remains to ask, What sort of purpose is the purpose to know? And as we have seen, what makes one a pragmatist, essentially, is the insistence that, as in science, so in philosophy and all truth-seeking, the idea in question should be used as a working-hypothesis, and the truth of the resulting judgment tested by the way in which the idea works. An idea is constructed to serve, in the guidance of action, as a sub- stitute for a further immediate perception of the reality which is the subject of the judgment; and if, when the immediate perception does occur, it prompts to the same action as did the original idea, may it not be claimed, with much force, that the idea '^agrees" with, or is practically the same as, the percep- tion? Here we fall back upon the critical realistic monism of our previous discussion, according to which that which is im- mediately given in perception is in part an independent reality. It should be noted, however, that there is a difference between the explicit and the implicit subject of the judgment. The ex- plicit subject includes all that is given perceptually and furnished apperceptively, while the implicit subject includes all that can be truly predicated of the subject, with the exception of tertiary quahties and relations.^ The judgment represents in its predi- cate what is not presented, or what needs to be represented again. Thus while there is in all ordinary cases of true judg- ment an identity between the predicate and some phase of the implicit subject, there is always, in judgments that have any significance, a difference between the predicate and the explicit subject. This consideration throws light upon the relation of 1 See Ch. XIV supra. 442 THE PROBLEM OF KNOWLEDGE the new "Law of Significant Assertion" to the traditional "Law of Identity." May it not be, then, that the test of truth indeed is, as the intellectuahst has supposed, some sort of identity between the idea and reaHty, but that this identity is some sort of practical identity, i.e. identity sufficient for practical purposes, even if the question as to just what practical purposes these would need to be may have to be left as yet undetermined ? At any rate this much may be said, that the cognitive purpose, as distin- guished from the purpose to make use of truth, is the purpose to obtain or frame an idea which shall prove at least sufficiently identical with the reality for practical purposes. Or, to put the matter differently, every cognitive purpose is an employee, the right-hand man, as it were, of practical purposes, and the employee's ultimate satisfaction is in the satisfaction of the employers. Sometimes the employee may modify the em- ployers' demand, but in general the business of the employee, the cognitive purpose, is to secure an idea which is sufficiently identical with reality to suit the employers, the practical purposes. But, it may be asked, may not the original em- ployee, the cognitive interest and activity, become independ- ent of its old employers and set up business for itself? The answer is that it may indeed act independently of its old employers, and in relative independence of practical in- terests, but this is not to say that truth about reality — and all truth is about reality — can ever be determined in absolute independence of all practical demands. In industrial and com- mercial affairs, even when the former employee sets up business for himself, he is not yet absolutely independent ; he is the em- ployee of society, and is made at times to feel most acutely that his own satisfaction is to be obtained only in and through the satisfaction of the community. And, most obviously, the football coach — to refer to Royce's illustration — is an em- ployee, the worth of whose activities is to be measured entirely by their serviceableness in directing the activities of the team. And so it is in the case of the cognitive interest. While it may gain independence, so far as particular practical activities are concerned, it can never gain absolute independence of the de- mands of practical life in general. When Royce admitted that CRITICAL MONISM IN LOGICAL THEORY 443 every idea is a plan of action, he admitted the nose of the prag- matist camel into his intellectuaUst tent. But, in further insistence upon the important distinction between the cognitive purpose and the purpose to make use of truth when it is known, we must keep in mind the difference between the purpose of the original judgment and the purpose of the later statement. The judgment is always relatively sincere ; the statement, as we know, need not be so. It is what the re- porter takes as true when collecting his materials that we are concerned with here, and not with what he gives as true to the readers of the daily paper. Understanding the term ''practi- cal" in this sense, then, the hypothesis here suggested is that the mark of truth is some sort or degree- of practical identity of the idea with the reality, of the predicate with the subject. And so at the heart of the good essence of pragmatism we seem to find representationalism, the good essence of intellect- ualism. It is not a representationalism which contradicts prag- matism, but one which supplements the pragmatic criterion at the same time that it is supplemented by that criterion itself. For while current pragmatism may give, even if in too vague and general a way, the differentia of the species, viz. practical value, it does not bring out sufficiently, if at all, the proximate genus, viz. representation of reality. This defect of the one-sided current pragmatism it would be the aim of what we may call representational pragmatism to remedy as far as possible. In our attempt to state the prag- matic criterion we found that we had to make use of the in- tellectualist's favorite idea of identity, interpreting it, however, in a functional way. Similarly, in attempting to define truth in terms of representation, may it not be that one can suc- ceed only by recognizing the pragmatic criterion ? We seem, then, to be on the verge of a definition of truth which shall be a "higher synthesis" of intellectualism and pragmatic anti-intel- lectualism. In leading up to this definition, however, it is well to take account of the fact that representation belongs to things outside of explicit judgments, as well as existing in judgments which claim to be true. Our problem may therefore be re- garded as not only that of finding the specific difference between judgments which are true and other judgments ; it is also the 444 THE PROBLEM OF KNOWLEDGE problem of finding the specific difference between the represen- tation found in true judgments and the representation belonging to meanings, apart altogether from explicit judgment. The functional analysis of an idea's meaning shows it to be, primarily, potential mediation of purpose ; it is representation which can , mediate purpose. As we have seen, meaning is essentially mean-ing, mediation, being a means; what an idea means, ultimately, is what it is a means to ; ^proximately it is the means itself, viz. representative material, proxy experience, a product of thought with its practical function, either actual or potential. Thus, as was noted long since by C. S. Peirce, the purposes which the idea can mediate form the key to the meaning of the idea. What Peirce ought to have done was to have gone farther and used this key, the meaning of meaning, to unlock the meaning of truth, and not to have been frightened back by his glimpses of pseudo-pragmatism and hyper-pragmatism. Meaning, then, is representation which can mediate purpose; but in the case of every live judgment, some possible purpose has become actual, and in that judgment some meaning is actually employed to accomplish that purpose. We arrive, then, at the following tentative definition. What is taken as truth is representation (of subject by predicate, of reality by idea) sufficient to mediate satisfactorily the purpose with which the judgment is made} But what is really true must 1 An approach to this position is briefly indicated in the following sentences from E. D. Fawcett's The Individual and Reality, 1909, p. 38 : "Often the agree- ment [of propositions with outward fact] may seem inconsiderable, nay, tri- fling ; but provided that such agreement forwards a purpose, the proposition or arrangement of propositions is sufficiently true. Truth means propositions which, in view of our ends, can be taken as, and substituted for, the appearances with which they agree." Oliver C. Quick indorses the pragmatic criterion, while rejecting the current pragmatic definition of truth. He himself, how- ever, leaves truth undefined {Mind, N.S., XIX, 1910, pp. 218-30), and seems to consider the formulating of a satisfactory definition impossible {ih., XX, 1911, pp. 256-7). Quick's position is thus in almost complete antithesis to that of Bertrand Russell, who, as we have seen, claims to define the nature of truth, while regarding it as having no criterion that can be stated in universal terms. J. B. Pratt indorses the pragmatic test of truth, but reverts to a defi- nition of truth which sacrifices clearness to simplicity. "Truth," he says, "means that the object of which one is thinking is as one thinks it" {What is Pragmatism? 1909, p. 67). Pragmatically interpreted, this definition will serve ; but intellectualistically interpreted, as it is evidently intended to be, it is involved in all the old epistemological and logical difficulties. CRITICAL MONISM IN LOGICAL THEORY 445 he representation sufficient to mediate satisfactorily whatever pur- pose or purposes ought to he recognized in making the judgment. In other words, real truth is practical identity of idea with reality, of predicate with suhject, where the practice in question is ulti- mately satisfactory, as well as the mental instrument which serves it. Now this representational pragmatism is truer to the in- tellectualist ideal than intellectualism itself is able to be. So long, for example, as the subject is taken as if it were a (logical) idea, like the predicate, as it seems to be by the idealistic intellectualist, the equating of ''subject" and ''predicate," being really the equating of one idea or pos- sible predicate with another, would in some cases be pos- sible, although it could never amount, even here, in any real judgment, any judgment that expresses meaning, to an ahso- lute identity. But even so, such equation of two abstract predi- cates would give no information about reality, the subject- matter of which both are, or may be, separately predicated. Indeed, even from the practical point of view, two ideas cannot be identified save as both are thought of as predicated of the same reality, with no practical difference in the consequences ; and then the identity is of the practical sort. But when the sub- ject is a reality, and not a logical idea considered as if it were a reality, although on intellectualist grounds it becomes even more hopeless to try to identify subject and predicate, on grounds of representational pragmatism, even here, it would seem, a solution of the problem is possible. According to rep- resentational pragmatism, in true judgment the one essential relation of predicate to subject is that of functional equivalence in the control of the action required. The judgment is true when the idea will do practically as well at least as further experience of the thing in stimulating and controlling action in adjustment to that thing. If it should be objected that the subject of the judgment has been previously represented in various ways, and so is different, relatively to the thinking subject, from what it would have been if it had not been thus represented, this may be readily admitted by the representational pragmatist. And if a very precise, even if somewhat unwieldy, statement of representa- 446 THE PROBLEM OF KNOWLEDGE tional pragmatism is desired, its definition may be stated so as to allow for this fact. If the reality which the subject-term {taken as predicate) represents {sufficiently for all purposes which ought to be considered in making the judgment — except the particular purpose or purposes which call for this last judgment), is repre- sented by the predicate of this judgment sufficiently for all the purposes which ought to be considered in making the judgment, whether the purposes which originally called for the judgment, or others, then the judgment in question may be taken as true. Here we have, then, in contrast with current pragmatism, a view which explicitly recognizes the ideal element in truth. The practical failure of ordinary pragmatism at this point has been in large part the basis for the charge, to which we have already referred, that it leads to sordid utilitarianism. Truth is to be measured — so this view will have it — not simply by the idea's working ''in the way it sets out to work," ^ but also by the way in which it does set out to work. Ends, and not simply adjustment of means to ends, come in for critical examination. The moral quality of the purpose is often re- flected in the judgment itself, and learning the truth becomes — in its higher reaches almost always, it would seem — a moral achievement. But while recognizing the ideal character of truth, represen- tational pragmatism makes this ideal of truth essentially human. It substitutes for the insoluble, artificial problems of current epistemology and intellectualist logic, the soluble, real prob- lems of the functional psychology and logic of the processes of cognition ; and in so far as any practically valuable judgment falls short of ideal truth, there are norms by which it may be corrected. W. Caldwell's criticism,^ that the doctrine that truth should be tested by consequences is useless, seeing that omniscience alone could bring together in thought or in imagina- tion all the consequences of an assertion, loses much of its weight as against a representational pragmatism stated in terms of the purposes which ought to be recognized. For it often occurs that the consequences are knowable by the individual suffi- ^ Dewey, Influence of Darwin, p. 150 ; Moore, Pragmatism and Its Critics, p. 87 ; Bawden, Principles of Pragmatism, p. 199 ; Schiller, Riddles of the Sphinx, 1910, p. 133. 2 Pragmatism and Idealism, 1913, p. 127. CRITICAL MONISM IN LOGICAL THEORY 447 ciently for the purposes which ought to be considered in the situation ; and in such cases there is ordinarily no reason to suppose that the judgment made will not be permanently satisfactory. According to such a pragmatism, even telling "the whole truth," whenever it was a moral duty, would be- come at the same time a real possibility. It would be tell- ing what was, practically speaking, the whole truth, so far as all purposes which ought to be recognized were concerned ; and except where it was thus a moral duty, it would not be a human possibility, under any definition of truth. Truthfulness, simi- larly, would consist in the habitual care to make one's state- ments always approximate one's own judgments sufficiently for whatsoever purposes ought to be recognized in each particular situation. 1 Moreover, there may be degrees of approximation to the truth and degrees of verification of the truth ; but, given the purposes which ought to be recognized, the judgment which represents reality sufficiently for these purposes is, from the point of view of representational pragmatism, true. Mathematically worked out laws, as in physics and astronomy, are only hypothet- ically truths, except in so far as they have been verified. Many of them have been sufficiently verified empirically for practical purposes, and so may be taken as practically true of the actual world. If by more critical tests they should be verified more completely, this would not make them any truer than they were before. But if by means of the more critical tests a discrepancy should be found between the mathematically deduced law and the actual fact, then for the purposes which dominated these tests, the supposed law is not true. If these are purposes which ought to be recognized by humanity, then the more accurate empirical observation must be regarded as the truth, and not the mathematical anticipation. If, however, there is no valid human reason for recognizing such hyper-critical purposes, there is practically no difference between the two expressions ; the one statement is as true, practically, as the other. More- over, on this theory it would seem that representations of reality sufficient for all practical purposes are not to be rejected as untrue simply because of the possibility of making the rep- * This does not, of course, decide the case either for or against rigorism in ethics. 448 THE PROBLEM OF KNOWLEDGE resentation closer, were there any occasion to do so. For ex- ample, for the purposes which ought ordinarily to be recog- nized, the carrying out of the value of tt to a few decimal places gives a practically true judgment ; but in some situations a more extended determination is required, so that in the former situations this later judgment would have contained irrelevant representation, as well as truth. The contradiction between the two judgments, therefore, when each is viewed in situ, is easily seen to be merely formal and not real. But we begin to see that representational pragmatism must encounter some very serious difficulties. In the first place, what are these ''purposes which ought to be recognized"? The obvious preliminary answer is that they must be stated ultimately in terms of human welfare, interpreted from a point of view in which the distinctly spiritual interests are duly dominant. But let us see just what this may be taken to mean. The universal human interests are perhaps seven : the hygienic, the economic, the (narrowly) social (i.e. interest in others and in fellowship with them), the scientific, the aesthetic, the moral, and the religious.^ The ''distinctly spiritual interests," as those concerned with ultimate and permanently valid ideals, or values which transcend the demand of the merely animal life, individual and racial, are the scientific, the aesthetic, the moral, the religious, and the social — this last in so far as one's fellows are viewed as ends rather than as means. A spiritual life is one in which the spiritual interests are properly coordinated with each other as ultimate ends, and made duly dominant over the life. Ultimately, the hygienic and economic interests are to be regarded as means to the realization of the spiritual inter- ests as represented by the ideals of universal human well-being and brotherhood (social), knowledge of the truth (scientific), contemplation of the beautiful (aesthetic), perfection of char- acter and conduct (moral), and fellowship with God (religious). But now, interpreting in the light of this explanation the clause, "the purposes which ought to be recognized," we find that representational pragmatism, as defined, offers us, for the ascer- 1 For this classification I am indebted to Professors A. W. Small and C. R. Henderson. Professor Henderson's list differs from that of Professor Small in making the moral and the religious distinct interests. CRITICAL MONISM IN LOGICAL THEORY 449 taining of truth, a criterion within a criterion. The scientific interest is here represented as one of the interests in relation to which what claims to be true is ultimately to be tested ; but, on the other hand, is not the scientific interest the '^ cognitive interest" of which we have spoken, thorough satisfaction of which ought to be regarded as all that needs to be sought? Indeed, have we not pointed out that what essential pragma- tism — and so, representational pragmatism — has undertaken to do is to universalize the procedure of the experimental sciences? If, on the other hand, it be said that science itself, in its judging of the truth, can only seek to represent reality sufficiently for all purposes which ought to be recognized, why should the scientific interest be mentioned as a separate interest which truth must satisfy ? And yet, do we not seem to need to include the scientific interest, the disinterested interest in truth, in order to guard against too hasty generalization ? But there are further difficulties ahead, especially in connec- tion with the problem of the permanence of truth. To be sure, representational pragmatism would enable one to take a more conservative attitude toward this question than obtains in current pragmatism. It is of course obvious that, as we have seen, even pragmatism, as it is and has been, has often shown unnecessary haste in concluding that it must, in order to be consistent, maintain that all truths are of but temporary validity. But representational pragmatism comes nearer to a positive vindication of the permanence of truth. Every honest judgment intends to be of permanent validity, and if at any later time it is seen to need revision, this is commonly to be explained either as due to the fact that the purposes active in the original judgment were deficient with reference to the situation, or as due to a lack of will or ability for mental thoroughness, so that in either case the earher judgment was not really true, but only seemed to be so. If any judgment is really true, the pre- sumption is in favor of its predicate a ways remaining the idea which will represent the reality sufficiently for all purposes which ought to be recognized in making the judgment. That many of our judgments in practical life are of this character, no one really doubts. Nor need we say, with Royce, that such truths are accessible only in the realm of our knowledge of the forms 2g 450 THE PROBLEM OF KNOWLEDGE that predetermine all of our concrete activities.^ If that were so, we could have no real or permanent truth about anything which we are ordinarily practically concerned to know. The representational pragmatist can claim not merely hypothetical judgments, but many categorical judgments — judgments of historical fact for example — as absolutely and permanently true ; the hard and fast intellectualist, as we saw, and as Royce admits when he says, ''Absolute truth is not accessible to us in the empirical world, in so far as we deal with individual phenom- ena, '^ 2 is not logically entitled to claim even that much. But in connection with what we have just been saying the difficulty is just this. On the one hand, in now judging any past judgments, our own or those of others, we necessarily make use of the criterion of non-contradiction, according to which it must be maintained that what was once, strictly speak- ing, true is always true, that what we cannot now judge to be true, e.g. the Ptolemaic astronomy, never was in reality true. But, on the other hand, can we say that the Ptolemaic astron- omers did not fulfil representational pragmatism's conditions of arriving at the truth ? Did not the geocentric astronomy — although it contradicts our modern heliocentric view — repre- sent reality sufficiently for all the purposes the early Ptolemaic astronomers ought to have considered, in view of the limited data accessible at that time ? According to an unsupplemented representational pragmatism, when the representation of reality satisfies the absolute ''ought" of the moral imperative in the making of the judgment, it ought to be absolutely true. And yet, in the case of the Ptolemaic astronomy we seem to have come upon judgments which, although when made they satis- fied the moral imperative, must now be judged to have been contrary to fact, erroneous — in fine, absolutely untrue. Thus representational pragmatism which seemed to promise a solution of the problem of truth, runs into self-contradiction and begins itself to suffer disintegration ; it seems about to fall apart once more into its constituent elements, intellectualism and mere pragmatism. Or, to change the figure, this represen- tational pragmatism, which offered so fair a prospect of a via media between intellectualism and current pragmatism, seems ^ William James and Other Essays, p. 251. ^ lb., p. 249. CRITICAL MONISM IN LOGICAL THEORY 451 now, before we have travelled it far, to bear the sign. No thoroughfare. Must we then retrace our steps and return to either intellectualism or current pragmatism? Neither pros- pect is at all inviting. If we choose intellectualism, we must resign ourselves to the conclusion — in so far as we may allow ourselves to come to any conclusion — that no really true judg- ment has been or ever will be made. On the other hand, if we choose mere pragmatism, at the very best we shall have to face the following dilemma. On the one hand we may say — in spite of all that can be said in the name of rationality, con- sistency, system — that all judgments which, when made, satisfied the practical purposes for which they were made, are to be permanently regarded as having been true ; so that there are many instances of true judgments which nevertheless con- tradict each other — a conclusion which works utter havoc with our indispensable everyday notion of truth. Or, on the other hand, if we refuse to choose this horn of the pragmatist dilemma, we may deny the permanence of truth, as a consequence of which we should have to say, for instance, that two thousand years ago it was true that the sun revolved about the earth, but that nowadays the truth is that the earth revolves about the sun. In other words, while in our astronomy we should have to judge the theory of the Ptolemaic astronomers untrue, in our pragma- tism we should have to judge it true — again a self-contradic- tion which, unless corrected, would utterly destroy any usable notion of truth. Is there then no way of escape from the impasse into which, even with our representational pragmatism, we seem to have been led? Without going over into anti-conceptualism, which would mean the sceptical giving up of the problem of the truth of judgments altogether, can we find a unitary criterion and formulate a unitary definition of truth, without falling into the futilities of either intellectualism or current pragmatism? Before we follow any of these counsels of despair, let us see whether our representational pragmatism may not be so revised and developed as to meet the requirements of the situation in which we find ourselves. Manifestly we are entitled to say this much, that even when the data are insufficiently accessible for full knowledge of the truth — unless, as is not always the 452 THE PROBLEM OF KNOWLEDGE case, the situation is one in which no judgment ought to be made at all — a 'person has a moral right to believe that that judg- ment is true in which the idea (predicate) represents the reality judged about sufficiently for oil the purposes which ought to be considered in making the judgment. It would have to be ad- mitted, of course, that in some cases judgments which one has had a moral right to believe to be true have nevertheless been shown to have been untrue. For truth, according to our revised representational pragmatism, would have to be defined, to bring out its distinction from mere morally justified behef, in some such way as this : Representation of reality by idea, of subject by predicate, such that in all situations calling for decision between the judgment in question and its contradictory, it will be found satisfactory in view of all the purposes that ought to be considered. But how do we know that there are any such judgments ; or, since we do seem, as a matter of fact, to have the right to believe that many of our judgments are of this sort, what is the criterion of this absolute and permanent {i.e. real) truth? How do we know true judgments to be true, if there are instances in which we do know this? This is a question which will lead us over, ultimately, into a discussion of methodology, or the problem of proof ; but it must be considered here also. If we are to have a real definition of truth, we must discover its real criterion — a criterion that can be really used with permanent satisfaction, in view of all purposes which ought to be considered. Indeed, without such a criterion we should not even be able to main- tain that there are any judgments which we have the moral right to believe to be true. It has been said that we have this right in connection with judgments which represent reality sufficiently for the purposes which ought to be considered in making the judgment. But the question is always pertinent, Is it true that we have considered all the purposes which ought to be considered in making the judgment? How can we know that the purposes considered are the right ones? Is all that we can say simply that they are the right purposes, if it repre- sents reality sufficiently for the right purposes to say that they are the right purposes? How shall we avoid the suggested unending circular regress, and actually get any measuring done with our criterion? CRITICAL MONISM IN LOGICAL THEORY 453 The question in this latter instance has thus come to be, How do we recognize ultimate (as distinguished from merely instrumental) values? And to this, obviously, the answer can only be. By immediate experience and appreciation, or, in other words, by direct intuition. But may not much the same thing be said in answer to the other question as to how we can ever know that what we now judge to be true will be permanently satisfactory in view of all purposes that ever ought to be con- sidered. The answer suggested is that in immediate experience of reality we may verify, i.e. intuitively perceive the absolute and abiding truth of a judgment. In order to round out our revised representational-pragmatic definition of truth, we must have recourse, it would seem, not indeed to a Bergsonian anti- conceptualism, but to what in Bergson is the positive counter- part of that doctrine, viz. intuitionism, the doctrine that truth is to be found in an immediate experience of reality. Of course truth does not consist in an immediate experience of reality, for it is a quality of judgments, which are essentially mediating, representational. But the truth of a judgment is indeed ^' found," discovered in immediate experience, when what its predicate represented (i.e. presented virtually, or in proxy fashion) is actually presented in the immediate experience to which the purpose to verify it (by acting upon it as a working hypothesis) leads. All truths which are either ac- tually verified or verifiable are of this sort; and it is worth noting that for the definition of the truths of science, i.e. scientifically verified truths, we must take into account not only intellectualism and the pragmatic form of anti-intellectual- ism, but the intuitional form of anti-intellectualism as well. Indeed our position might well be termed scientific representa- tional pragmatism, not only a synthesis of intellectualism and current pragmatism, but a further synthesis of representational pragmatism and intuitionism. It is the procedure of science become conscious of its own fundamental nature.^ We see, 1 Our synthesis of intellectualism, pragmatism, and intuitionism includes, we believe, the valid elements in H. W, Wright's suggestion that intellectual con- sistency, technical eflSciency, and emotional harmony are criteria of truth (Philosophical Review, XXII, 1913, pp. 606-22. See p. 370, supra). Em- pirical intuition does not coincide with "emotional harmony," of course; in our opinion the latter must be viewed as a notoriously unreliable, but often very 454 THE PROBLEM OF KNOWLEDGE therefore, why it was necessary, in defining truth in repre- sentational-pragmatic terms as ''representation of reahty, suf- ficient for whatever purposes ought to be considered in making the judgment," to include, implicitly if not explicitly, the cog- nitive purpose of the scientist. But the query may be suggested. Is not this the adoption of a new criterion altogether, viz. that of immediate intuition instead of, and not merely in supplementation of, the pragmatic cri- terion? But to this the answer must be negative. Bergson to the contrary notwithstanding, ''intuition without concepts is blind," or practically so. For truth at any rate, there must be concepts, judgments, representation. And, since the idea is never identical with its subject, except for practical purposes, we can never have a satisfactory definition of truth (i.e. true representation) except in pragmatic terms. Nor can we safely take the spontaneous judgments which emerge out of imme- diate experience as infallibly true. Practice without intuition has often more truth than certainty ; but intuition without practice has quite as frequently more certainty than truth. And in order adequately to supplement mere intuitionism we need more than the bare "negative pragmatism" that Hocking has allowed; for, as we have already contended, unless some sort of positive pragmatism is justified, not even negative pragmatism is true. Moreover, to return to a point upon which we have already touched, we must never forget that the completely verifying perception is often either temporarily or permanently unattain- able by human beings, or else not important enough to be sought at the necessary expense of something else. Does this neces- sarily mean the total cessation of belief, the total absence of knowledge? Not according to everyday life, not according to science, and not according to a sufficiently critical theory of truth. It often occurs that acting upon the idea continues to work so uniformly well in connection with its most intimately associated practical purposes, that even in the absence of the immediately verifying perception, the idea is kept in action, and rightly so ; that is, we believe our originally hypothetical valuable, variety of the former. Cf. criticism of Hocking's mystical intuition- ism, Ch. VIII, supra. See Ch. XV, supra, also. CRITICAL MONISM IN LOGICAL THEORY 455 judgment to be true, and are morally justified in this belief. We are practically certain that the idea is practically identical with the reality, an immediate experience of which we either necessarily, or deliberately but justifiably, forego. And the only but all-sufficient justification of this is that the race has needed to postulate, and through long and successful experience has acquired the inveterate habit of postulating, that nature, or reality in general, is dependable.^ Here, again, therefore, our revised representational pragmatism is simply the logical theory of everyday scientific procedure. What, more particularly, a truly scientific procedure is, it will be our task to inquire in the following chapter ; but before our discussion of the problem of truth is brought to a close, it may be well briefiy to consider certain criticisms commonly passed upon current pragmatism, in order to see whether our revised representational pragmatism can successfully meet these incidental, and possibly minor, tests. A charge frequently made by absolutists is that pragmatism fails to do justice to the transcendent and superhuman character of truth. Our answer to this, so far as a scientific representational pragmatism is concerned, is to be found in large part in what was said of the ideal character of truth ; but in connection with the question as to whether there is not an actual superhuman truth, our answer would be. Doubtless there may be ; but what is generally meant by "absolute truth," or truth as it is for "the Absolute" (of ab- solute idealism) , is simply a regulative idea. If absolute truth be defined, with Schiller, simply as "truth adequate to every human purpose," 2 we are furnished with a standard sufficiently acces- sible for our most critical needs, and in this sense the contention that we are in possession of absolutely true propositions may be readily granted.^ But when the absolutist assumes, or, by whatever process, concludes, that there must of necessity actu- 1 "That things do work together and our needs are satisfied when a certain set of postulates are conformed to, is, in so far forth, evidence of the correctness of the postulates. . . . The theory is not true because it satisfies our needs, but the fact that it satisfies our needs is evidence that the theory fits into the organ- ism" which the whole universe seems to be, since "all parts of the universe . . . act together" and have apparently "grown to be what they are in organic unity of development." (C. L. Herrick, Journal of Philosophy, I, 1904, p. 596.) 2 Studies in Humanism, p. 213. 3 Cf. p. 389, supra. 456 THE PROBLEM OF KNOWLEDGE ally be an eternally existent sum-total and systematically unified experienced harmony of all possible true judgments, he indulges in speculative dogmatism. Superhuman truth, if we are going to speak of it at all, we would do well to call divine, rather than absolute ; for while, so far as our present discussion is concerned, even of God, existence and attributes are matters of speculation, the whole conception is less ambiguous than that of 'Hhe Absolute" of current metaphysics. And if we are going to speak of God's truth, there is no manifestly vaUd re- Hgious reason why it should not be regarded as essentially similar to man's. It is surely not a timeless, changeless, pur- poseless, absolutely complete representation, in one act of thought, of an eternally-complete reahty which is also content of an eternally-complete immediate experience; for why, in such a case, should there be representation at all? May it not more probably be representation, the content of which may vary from time to time, and yet which is sufficient always to mediate satisfactorily whatever purposes God may have in view. This is not to say that God is in every way anthropo- noetic, but that if there is such a thing as God's truth, it must be essentially similar, or even identical, with man's truth. Another important test to be applied to any theory of truth is the question whether it is ^'self-critical" or self-refuting. Is our scientific representational pragmatism true, according to its own definition of truth ? Both absolute intellectualism and the absolutely anti-intellectuahstic theory of current pragma- tism, are, as has been intimated, self-refuting. On the one hand, the idea of an absolute identity between idea and reahty, between predicate and subject, is not absolutely identical with what functions as truth in actual human experience. On the other hand, the idea of mere practical usefulness, or working value, of ideas will not always work as a substitute for what we mean by truth. But revised, or scientific, representational pragmatism is self-critical. To say that truth is representa- tion of reahty sufficient for whatever purposes ought to be con- sidered by any one who may ever have to decide between that judgment and its contradictory, is itself a representation of the reality in question (viz. truth), sufficient for whatever purposes ought to be considered by any one who may have to decide be- CRITICAL MONISM IN LOGICAL THEORY 457 tween it and its contradictory. Moreover, that any judgment (in a given situation in which judgment concerning a certain subject-matter is morally required) which represents the reality sufficiently for all the purposes which the person making it ought to consider, is a judgment which that person has the moral right to make and to believe to be true — this itself is a judgment which we have not only the moral right, but, we would claim, a fully verified scientific right to make. We see no reason, then, for rejecting scientific representational pragma- tism, the only definition of truth remaining unrefuted. While the traditional intellectualism gives the proximate genus of truth (representation of reality), but not its specific difference (suffi- ciency for all proper practical purposes), and while current prag- matism rightly, even if too vaguely, insists upon the specific difference, but wrongly ignores the proximate genus, scientific representational pragmatism combines the complementary par- tial truths of the two positions. Finally, there is a formal test of definitions, which may be appHed to our definition of truth. L. S. St ebbing has urged in criticism of pragmatism that unless the pragmatic dictum, "All truths work," is simply convertible, it fails to provide a criterion.! This is not quite accurate ; it is not the criterion, at least as stated thus, too broadly to be a real criterion^ that must be simply convertible, but the definition; but the spe- cific difference by means of which the definition is constructed is the criterion, the test, par excellence. Our definition of truth, however, unlike those of intellectualism and current pragmatism, will stand this test. To say. All judgments in which the predi- cate represents the subject sufficiently for all purposes which ought to be considered at any time when any one may have to choose between the judgment in question and its contradictory, are true, is as true as our definition of truth, of which it is the simple converse. It will appear, then, that once more, in our treatment of the chief problem of logical theory, we have been led to a position that may be characterized as critical monism. We have noted the evident unsatisfactoriness of an absolute logical dualism, ^Mind, N.S., XXI, 1912, p. 471 ; XXII, 1913, p. 250; cf. Pragmatism and French Voluntarism, 1915, pp. 154-6. 458 THE PROBLEM OF KNOWLEDGE and have seen the finallj^ disappointing character of absolute logical monism in its various forms, viz. on the one hand, ab- solute intellectualism (intellectualistic absolute logical monism) in its epistemologically duaHstic, idealistic, and reaHstic vari- eties; and on the other hand, absolute anti-intellectualism (anti-intellectuahstic absolute logical monism), whether anti- conceptualistic or pragmatic. We are left with but one theory which can be regarded as both tenable in the face of attack, and positively justifiable, viz. scientific representational prag- matism, or, to give other possible designations, critical logical monism, critical pragmatic monism, critical pragmatic logical monism, intuitional-pragmatic representationalism, or critical monism in logical theory.^ 1 In this and the two immediately preceding chapters I have included, with- out the use of quotation marks, some excerpts from my article entitled," Repre- sentational Pragmatism," in Mind, N.S., XXI, 1912, pp. 167-81. B. THE PROBLEM OF PROOF (METHOD- OLOGY) CHAPTER XX The Problem of Scientific Method The problem of mediate knowledge is the problem of proving the truth. In the immediately preceding chapters we have dealt with what, regarding logic as the normative science of the truth of judgments, we take to be the most fundamental problem of philosophical logic, or logical theory, viz. the problem of truth. We must now take up the remaining problem, viz. the problem of proof, which may be regarded as the central concern of methodology. But the problem of proof" is the problem of the production of certainty of the truth in a way that is logically satisfactory. We shall therefore have first to consider briefly the nature of certainty in general, and of logical certainty in particular. Now the problem of certainty is not, in the first instance, a logical problem at all, but a psychological one. And probably the best available criterion of psychological certainty is readiness to act upon the judgment, not tentatively and with a view to verification, but, finally, irrevocably. Certainty in this broad sense is the state of mind accompanying judgment or belief, in which there is such a readiness to act irrevocably, given the appropriate situation. (Our view would allow for a real dif- ference between knowing, and knowing that we know. And yet, if we do not know that we know, and dwell upon this nega- tive fact, it may destroy our readiness to act, and this would mean the destruction of our knowledge. On the other hand, to know that we know is a safeguard to our knowledge; it keeps it steady, free from unnecessary fluctuations. And prob- ably this is the chief value of a constructive epistemology. 459 460 THE PROBLEM OF KNOWLEDGE It makes for a certainty which is adequate, even after the most comprehensive sort of criticism.) But there are two main sorts of psychological certaintj^, viz. logical certainty, and certainty which falls short of logical certainty. Logical certainty may be defined, in preliminary fashion, as sufficiently critical psychological certainty, provided the term ^'sufficiently critical" be taken seriously enough; although it is perhaps quite as informing to say that psychologi- cal certainty is sufficiently critical when it is logical. But in any case it will readily appear, in view of our previous discus- sion, that there are two main varieties of logical certainty, viz. that in which the judgment has been fully verified in immediate perceptual experience, i.e. in perceptual intuition; and that in which such direct perceptual verification is, for some good and sufficient reason, unnecessary. It may be remarked in this connection — and the considera- tion is of great importance for epistemological theory — that on the basis of an absolute dualism in epistemology, according to which no perceptual intuition of reaHty would be possible, while there might perhaps be truth in human judgments, there could be no certain knowledge that it was the truth. Indeed truth would be indistinguishable from what seemed, as a matter of fact, to be a practical substitute for truth. It is only when we have had, or can in some way find access to, immediate experience of reality, with which we can compare our ideas, that we can know that what in any particular case functions satisfactorily is really true, and not a mere temporary substitute for the truth. It is one thing to know that we have either truth or an ap- parently satisfactory temporary practical substitute for it; it is quite another thing to know that we have representation which is true, because it is the functional equivalent of further immediate experience of the reality, so far as all purposes which ought to be considered are concerned. The former we might have on the basis of a dualistic epistemology; the latter re- quires epistemological monism. If there is to be knowledge of reality, representations must be comparable with presentations. x\nd yet, granted that there is someivhere for us a direct acquaint- ance with independent reality, it is perhaps not inconceivable that, in view of our general knowledge of the nature of this THE PROBLEM OF SCIENTIFIC METHOD 461 reality, there may have been produced, in certain instances, a sufficiently critical or logical certainty of the truth, on the basis of a prolonged and varied experience of the satisfactory working of the hypothesis, without there ever having been such an im- mediate experience of the reality in question as would at once have constituted its complete verification. But, granting that logical certainty is at least a sufficiently critical intellectual readiness to act irrevocably upon an idea or proposition, our discussion of the problem of mediate knowledge will not be complete until we shall have treated of the method of the production of this sufficiently critical intellectual readi- ness, or logical certainty, this certainty of truth resting upon ade- quate experiential grounds ; in other words, we must now take up the problem of the scientific method of proof, or of the pro- duction of logical certainty. And when we call our problem one of scientific methodology, we assume, of course, that the tradi- tionalistic method, of resting finally upon some external au- thority, is out of the question here. What we are seeking is a method fitted to be employed by all independent investigators and thinkers. The scientist is not satisfied simply to know, or even to know that he knows ; his ideal is to know how he knows, in order that he may proceed with sure and steady step to still further intellectual conquests. Indeed, science may be re- garded as including not only systems of verified judgments about reahty, but also an adequate system of verification. How, then, does the scientist, as such, come to know? As we confront this methodological problem, we find, as in the case of each of our previous investigations, that the points of view chiefly represented may be classified under an absolute dualism and the two corresponding one-sided absolute monisms. In this case over against the absolute methodological dualism we find on the one hand a rationalistic absolute methodological monism, and on the other hand an empirical absolute methodo- logical monism. The absolute dualism need not detain us long. What we have in mind here is simply the widespread doctrine that there are two methods of proof, radically different from each other and irreducible to any common denominator, other than that they are both methods of producing logical certainty, or proof. We refer, of course, to deduction and induction, and 462 THE PROBLEM OF KNOWLEDGE to the common tendency to interpret the former after the manner of pure rationaHsm and the latter in accordance with pure empiricism. But the question naturally arises as to why there should be two ultimately different ways of doing one thing. Unquestionably dualism is to be accepted either only tentatively, or only as a last resort, because of the failure to establish, on sufficiently critical grounds, some form of monism. We turn at once therefore to an examination of certain views which represent more or less completely a rationalistic ab- solute monism in methodology. The tendency of pre-Kantian rationalism toward this extreme is now fully recognized. Des- cartes, dissatisfied with all that claimed to be science in his day, with the sole exception of mathematics, undertook to follow the mathematical model in philosophy, proceeding by careful de- duction from whatever premises should be found, in spite of the most rigorous criticism, to admit of no reasonable doubt.^ Spinoza, an apt pupil, followed with his Renati des Cartes Prin- cipiorum philosophice pars I et II., more geometrico demonstratce, and his Ethica, ordine geometrico demonstrata. And according to Leibniz, in so far as we are empiricists, which we are in three- fourths of our actions, we simply "act in like manner as ani- mals'' ; it is only the knowledge of eternal and necessary truths {i.e. truths arrived at by deduction from definitions, axioms, postulates, and primary principles which have no need of proof) which distinguishes us from mere animals and gives us the sciences.^ During the past few decades there has appeared, largely under the influence of the mathematicians, a recrudescence of this highly rationalistic tendency in methodology. One of the frankest expressions of this tendency is to be found in the recent essay on " The Principles of Logic," by Louis Couturat. " Dem- onstration," this author insists, "consists in deducing from given premises or hypotheses the consequences or conclusions which they formally impty in virtue of the laws of Logic. From the algorithmical point of view it consists in passing from prem- ises to conclusions by means of transformations permitted by the laws of the calculus. There can be no logical and correct 1 Discourse on Method, Parts I and II. 2 Monadology, § § 28-35. THE PROBLEM OF SCIENTIFIC METHOD 463 demonstration except at this price ; we must not take a single step which is not justified by the logical laws : all recourse to * evidence' or to 'intuition' must be rigorously excluded."^ Similarly Bertrand Russell avers that what is called induction is either disguised deduction or a mere method of making plausible guesses.- ''In the final form of a perfected science, it would seem," he says, "that everything ought to be deductive." ^ Much more, then, from this point of view, is the simple arithmeti- cal process of "demonstration by recurrence," which Poincare insists is "mathematical induction," ^ to be regarded as nothing but deduction.^ Now this monistic methodological doctrine, that no method of proof is scientific except deduction, and that deduction is absolutely non-empirical, independent of intuition,^ is closely related to recent developments of formal logic. We refer to the new "symbolic logic," or "logistic," which is offered as a more exact "science of logical form" than the traditional syllo- gistic logic. This symbolic logic, it may be noted, is, like the syllogistic logic it is intended to displace, the logic of consistency simply ; it is the science of the logical form of abstract science. It is the science of hypothetical truth ; but, inasmuch as the hypoth- esis in question may be either unknown to be true or known to be untrue, it cannot be said to be the logic of truth. The sciences whose procedure it describes assert implications rather than facts. Ileal logic, the logic of real or categorical truth!' is the science of the logical form of descriptive, empirical science. In support of the statement that recent deductive or rational- 1 Encyclopedia of the Philosophical Sciences, edited by Windelband and Ruge, Vol. I, p. 184. ^Principles of Mathematics, 1903, p. 11. 3 Our Knowledge of the External World as a Field for Scientific Method in Phi- losophy, 1914, p. 34. ^ Science and Hypothesis, Eng. Tr., 1905, pp. 7-16. 5 See E. B. Holt, The Concept of Consciousness, pp. 11, 12. The view is common. 6 We would regard "intuition" not as absolutely a priori in the case of space and time, or anywhere else, but as relatively a priori and relatively empirical (see Ch. XVI, supra) ; so that, from our point of view, the appeal directly to intuition would be mediately to experience. ' Cf. F. Bacon, Novum Organum, Bk. I, Aphorisms 11-14; J. S. Mill, A Sys- tem of Logic, Bk. II, Ch. Ill, § 9 ; F. S. Schiller, Formal Logic, 1912, p. 8. 464 THE PROBLEM OF KNOWLEDGE istic monism in methodology is intimately bound up with the new formal logic, we would cite the opinion of Bertrand Russell that now, 'Hhanks to the progress of symbolic logic, especially as treated by Peano," the Kantian view that mathematical reasoning is not strictly formal, but always uses ''intuitions" (by which term, following Kant, he means the a ^priori knowl- edge of space and time, whereas Poincare speaks of the intui- tion of pure number), is "capable of a final and irrevocable refutation." '' By the help of the ten principles of deduction and ten other premises of a general logical nature," he contin- ues, " all mathematics can be strictly and formally deduced." ^ Now if, as has been intimated, abstract sciences cannot be said to assert more than implications, or hypothetical truths, and if, consequently, these ''hypothetical truths" {i.e. the apodosis, apart from the protasis) may be actually untrue, we must not accept without further question even the conciliatory statement of the mathematicians, that when mathematical entities (such as non-Euclidean space or numerically infinite collections) are said to exist, it is only mathematical or logical existence, i.e. freedom from contradiction, that is meant. So far as anything that purely formal logic can take account of is concerned, these entities may be said to be free from con- tradiction ; their existence follows logically from certain assump- tions. But this does not mean that they can be said to be "free from contradiction" so far as real logic, the logic of truth, is concerned. The descriptive sciences, the logical form of which real logic undertakes to describe, are always open toward reality. All relevant truth, even that which is still undiscov- ered, is potentially a part of such a science. Only that, then, which is in agreement with fact can be said to be, in relation to the other parts of the science, free from contradiction, or to have logical existence, so far as the logic of truth is concerned. For the sake of greater definiteness upon this important point, we shall deal in greater detail with the illustrations to which we have referred. In the geometry of Bolyai and Lobachevski the notion of non-Euclidean or "curved" space is introduced hy the assumption that more than one parallel ^ Principles of Mathematics, p. 4. Cf . Foundations of Geometry, 1897, where it is asserted that projective geometry is "wholly a priori." THE PROBLEM OF SCIENTIFIC METHOD 465 to a given straight line may be drawn through any given point. Assuming the possibiHty of such a pluraHty of parallels through a given point in real space, then it follows that real space is ''curved" (non-Euclidean) ; and so also if it be assumed that no parallel can be drawn. But the question remains : Is there any reason at all for making such assumptions in a science of real space, and so for supposing that real space is non-Euclid- ean? The case of numerically infinite collections is similar. A large degree of abstraction from the concrete and empirical was accomplished by means of Dedekind's theory of number, according to which the fundamental and original idea of num- ber is obtained by abstraction from all special characters, in- cluding quantity, of the group of numbered objects, with the single exception of the relation of order between those objects. Numbers would then be primarily ordinal, and the cardinal numbers derivative, the result of making an aggregate of num- bered objects, or of abstracting still further from order. ^ With this non-quantitative view of number there seemed to be less call for objection, on logical grounds, to the notion of series and collections numerically infinite. The new definition of infinite, anticipated by Bolzano,^ and worked out by Dede- kind ^ and Georg Cantor,^ as a collection which is similar to a proper part of itself, i.e. which is such that its elements can be set out in a relation of one-to-one correspondence with those of a proper part of itself, was made to seem a logically permissible concept; there could be no objection in the nature of any ''intuition" of the necessarily finite character of all quantity. It was regarded as quite demonstrable that there are, within the realm of consistent mathematical definitions, infinite sys- tems, as, for example, one's own system of possible thoughts, 1 R. Dedekind, Was Bind und Was Sollen die Zahlen ? Nos. 73, 161 ; cf. Couturat, L'infini mathematique, pp. 334-5 ; Royce, The World and the Individual, Vol. I, pp. 528-30. A similar development, in even greater detail and rigor, but without the emphasis upon the primacy of ordinal numbers, is to be found in Frege (Die Grundlagen der Arithmetik, 1884, etc. Cf . Whitehead and Russell's Principia Mathematica) . 2 Paradoxien des Unendlichen, 1851, § 20. 3 Op. cit., No. 64. * "Ein Beitrag zur Mannigfaltigkeitslehre," Crelle's Journal, Vol. LXXXIV, 1878, pp. 242-58. Cf. Frege, op. cit., §§ 84-6. 2h 466 THE PROBLEM OF KNOWLEDGE including, as it does, a thought about each thought in the system.^ Now this non-EucHdean geometry and mathematics of the infinite would be quite harmless philosophically, if it were always clearly understood that, properly speaking, as we have intimated, and as even Russell reminds us, nothing is affirmed therein except implications. ^ But besides asserting, as they are entitled to do, that the conclusions are consistent with the premises, "logisticians^' commonly assume that there was nothing logically objectionable in the initial assumption of two parallels to a given straight line through one and the same given point, or in Dedekind's and Cantor's definition of an infinite system. It is assumed that there can be no objection to the assertion that non-Euclidean or ''curved" space exists, or that there exist numerically infinite series and collections, provided it be understood that existence means logical exist- ence, freedom from contradiction.^ This accounts for the fact that logisticians seem commonly to regard it as more or less of a '' toss-up'' as to whether or not real space is non- Euclidean, and as to whether or not there exist infinite collec- tions. It has become the fashion, under the influence of logis- tic, to maintain that perhaps, even though the finest practical geometrical measurements fail to give us the slightest reason to suppose it, real space may nevertheless be ''curved" or, if Euclidean, of more than three dimensions. But let us take geometry as a science of real space. And let us assume the validity of the principle of parsimony, accord- ing to which, since there is no necessity for the assumption of either non-Euclidean or four-dimensional space, space must 1 Dedekind, op. ciL, No. 66 ; cf. Bolzano, op. cit., § 13 ; Cantor, Grundlagen einer allgemeinen Mannigfaltigkeitslehre ; Couturat, Uinfini mathematique ; Russell, Principles of Mathematics, I, pp. 143-4, 147, 194, 357-8 ; Philosophical Essays, pp. 77-8 ; Our Knowledge of the External World as a Field for Scientific Method in Philosophy, Chs. VI, VII. Royce, op. cit., I, Supplementary Essay, especially pp. 501-14. A further tendency to abstract from experienced reality is seen in the doctrine that the indemonstrable axioms with which pure mathe- matics begins are really disguised definitions. (Couturat, referred to by Poincar6, Science et Methode, pp. 161, 175-6. ) 2 Principles of Mathematics, p. 5 ; cf . Royce, William James and Other Essays, p. 239. 3 Royce, World and the Individual, 1, p. 511, note; cf. Poincar6, Science et Methode, p. 162. THE PROBLEM OF SCIENTIFIC METHOD 467 be regarded as Euclidean and simply tridimensional. On this basis it becomes clear that the hypothesis of more than one parallel to a given straight line through a given point leads to absurdity. It introduces into the intended science of real space an element of mutual contradiction between propositions. The hypothesis must therefore be rejected as imtrue. And so we would claim to be justified in contradicting the assertion that, so far as we can say, space is as likely as not either non- Euchdean or four-dimensional. For such an assertion there has been found not a single good reason. The theory has not even been shown to be a necessary postulate of '^ practical reason.'* As a scientific hypothesis it grossly violates, as we have seen, the principle of parsimony. It runs counter to practical need, to common sense, and to immediate intuition. The only thing that can be said for the new geometries is that they show that it is hypothetically true that space is either curved or of four dimensions ; i.e. space is of this sort, if some- thing is true, which, as a matter of fact, we are certain enough for all practical purposes, is untrue. And so also with regard to the notion of infinite collections. Not only do we not know that there are such collections ; the notion itself, we would say, is unscientific and — in our view of logic as being properly the logic of truth and not of mere consistency — illogical. In the first place, to assume such a collection violates the principle of parsimony; no one, so far as we know, has experienced such a collection, nor is there any scientific need of assuming it. Moreover, experience and *' rational intuition," when we are sufficiently critical, disallow the notion. An infinite collection, if there could be such a thing, would be a collection such that adding to it would not increase it, and subtracting from it would not diminish it. But we know, by intuition capable of enduring the severest criticism, that there can be no such collection. Other condi- tions remaining the same, "adding to" involves ''increasing"; and so the definition, when the meaning of its terms is considered, is seen to be simply self-contradictory. Expressing the defini- tion in ordinal terms does not help; an ordinal number, as defined by logisticians, is the whole series from the first up to that number taken in that order, so that it really involves 468 THE PROBLEM OF KNOWLEDGE a corresponding cardinal number. It is commonly supposed that the number of real points in any given Hne is infinite. But a real point must be a location which, under conditions conceivable without contradiction, is discriminahly different from any other location. A point that is not, under any really conceivable conditions, discriminahly different, is not a dif- ferent point. This never takes us beyond a finite number of real points in any real hne. Indeed, to suppose a numerical infinite leads to the ''Cantorian antinomies" and others, such as that there are both more and no more points in space than there are whole numbers in the series of numbers — an out- come which ought to be regarded as the redudio ad absurdum of the notion of the numerical actual infinite.^ We admit, to be sure, that it is sometimes convenient to use this fiction of an infinite collection, in spite of its self-contradiction. But the doctrine that all convenient fictions are true is a doctrine that ought not to be dignified with the name pragmatism ; it is the worst possible sort of what we have called pseudo-pragmatism. If, then, these notions are not even logically unobjectionable — as, from the point of view of real logic, they certainly are not — it surely cannot properly be maintained that the objects in- tended by them exist.^ Our position, then, is that since real science is fundamentally description of reahty, and real logic the science of the form of thought necessary for arriving at the truth about reahty, '^ abstract .science" is to be regarded, in proportion to its ab- stractness, as merely instrumental to real science, while systems of conclusions which are simply consistent with arbitrary and experience-contradicting assumptions are not, properly speaking, real science at all. We would agree with Alfred Sidgwick and F. C. S. Schiller ^ that our a priori laws and universal proposi- tions must be interpreted in the light of experience, apphed 1 Cf. Poincare, Science et Methode, pp. 152-5, 201-3, 212-13 ; Dernieres Pemees, pp. 132-7. 2 Sometimes the doctrine of an actual infinite is welcomed because of its agreement with other philosophical doctrines, which the philosopher is inter- ested in defending, such as the concrete objective idealism of Royce. See The World and the Individual, I, Supplementary Essay. 3 See Sidgwick, "Applied Axioms," Mind, N.S., XIV, 1905, pp. 42-57; The Application of Logic, 1910, etc. ; Schiller, Formal Logic, 1912 ; Humanism, pp. 85-94 ; Studies in Humanism, pp. 8, 9 . THE PROBLEM OF SCIENTIFIC METHOD 469 to reality as experienced, and verified, before they can be logically regarded as true of reality. We may also refer to the distinction, which goes back to Mill,^ between induction, as the scientific process of establishing general propositions, and deduction, as the exhibition and use of the product of that scientific process. This distinction may easily be pressed too far, however; no demonstration, to one's self or to others, of truth about reality is possible either inductively or deductively, it would seem, without reference to and dependence upon experience, either directly or through 'intuition." Logistic, though it claims to contain all that is good in meth- odology, turns out indeed, on closer scrutiny, to be not a method at all. It is a science of certain hypothetical objects, objects which, as we have just seen, are often so very hypothetical as scarcely to merit the appellation '^possible." But also from another side we find the claims of the logisticians extravagant. Russell seems to regard all sciences as merely appUed logistic, — logistic with the x's and y's replaced by hydrogen and carbon. All that is system in any sense is logistic. But on examination of the systematic connections used by logisticians, for instance ''implication," ^ we find reason to suspect that the logisticians' systematic connections have suffered an abstraction, and even transformation, of such an extreme sort that the notions of relevance and intimacy of connection actually necessary in dealing with concrete proofs have been generalized away. From our point of view, according to which logic is the science of the sort of processes of thought that must be employed in order to realize the ideal of truth in our judgments, such ab- straction is most objectionable, and the importance of logistic is much diminished thereby. But we would suggest a further criticism of rationalistic methodology. Can even logical consistency be shown without any appeal to ''intuition," and so, ultimately, to experience? It may be remarked that since the publication of Russell's Principles of Mathematics the list of twenty principles and fundamental premises, which were supposed to be adequate as a basis for mathematics, has been greatly revised by the 1 System of Logic, Bk. II, Ch. III. 2 My attention has been called to this point by Dr. H. T. Costello. 470 THE PROBLEM OF KNOWLEDGE author — a fact which would seem to indicate that the processes of science are not so purely deductive as was supposed. One is reminded, in this connection, of Poincare's criticism, that not o^ily do the indemonstrable propositions assumed at the outset involve ''a new act of intuition," but it is a mistake to suppose that these original appeals to intuition are the last that will be necessary for mathematics.^ Not only is ''the principle of com- plete induction,'' according to Poincare, " at once necessary to the mathematician and irreducible to logic " ; he claims that logic itself is sterile, until fertilized by intuition, except that in certain cases it is able to engender the antinomy.^ We would therefore have to reject as untenable what we have called rationalistic absolute methodological monism, i.e. the tendency to hold that pure deduction, without any dependence, however mediately, upon experience, and this alone, is adequate to give us scientific knowledge about reahty, or even about what is logically possible. Granted such dependence upon experience or ''appeal to intuition" as may be found necessary in deduction, we would admit that the deductive sciences, however abstract, do give us knowledge of impHcations, and thereby also indirectly knowledge about one concrete reality, viz. the human mind, or the necessities of human thought. But with reference to objective existence in general, no abstract deductive science can be known to lead to true results, unless all the indemonstrables assumed are known actually to exist; and even here, as we shall contend, prior to empirical verifica- tion, the results can, without dogmatism, be taken as valid of reality only in view of our genetic doctrine that the fundamental categories of human thought have been moulded upon an im- mediately experienced independent reality. Before closing this critique of absolute rationahstic monism as a methodological doctrine, some attention should be given to the employment in philosophy, especially recently, of the so- called dialectical method. In Hegel's system the dialectical method has a fundamental place, and is closely integrated with his metaphysical doctrine. The truth, in his theory, is the 1 Science et MSthode, pp. 175-6 ; cf. pp. 177-8, 192, 195, 207-8. 2 "Les math^matiques et la logique," Revue de metaphysique et de morale, XIII, 1905, p. 817 ; Science et Methode, pp. 211-12. THE PROBLEM OF SCIENTIFIC METHOD 471 concrete universal, i.e. the World or Reality as an organized whole ; and this truth, the Absolute Idea which is Absolute Reality, philosophy seeks to possess. The concepts of the understanding, useful as they are in the special sciences and in common life, give us at the best something less than the truth, mere abstractions instead of the whole. But there is even in ordinary finite thought a tendency to transcend itself and to reach out toward the Absolute Idea, and this tendency mani- fests itself in the contradictions which arise in common thought and demand some more adequate concept for their solution. Thus thought takes on a dialectical movement, tending from thesis through antithesis to a higher synthesis until finally the Absolute Idea is attained. Now it is true enough that rational reflection does tend to fall into this dialectical form, and where there is constant depend- ence upon further experience to furnish the contradictions to our earlier theses, there can be no vaHd objection from the standpoint of methodology. But the Hegelian dialectic is sometimes interpreted — as by McTaggart recently, for example — as dependent upon experience for its first concept only, the bare concept of being, and thereafter developing by positing over against this concept its contradictory, and then overcoming the opposition by means of a higher concept, which necessarily appears as involved in the synthesis of the antithetical concepts. For example, the first triad in Hegel's metaphysical dialectic might almost be put in the form of a conundrum. When is being not-being? to which the answer seems to be. When it is becoming. 1 Now this doctrine of McTaggart — whether it is also what Hegel meant to teach is more doubtful — may very well be called a rationalistic absolute methodological monism. With the exception of the first step, the whole process of dis- covering and proving the truth about reality is accomplished, it is claimed, by pure reason alone, without the aid of experience. But there are two criticisms to be made. On the one hand the dialectic, as McTaggart actually employs it, is by no means independent of empirical data; the higher synthesis is really taken in each case from the fund of general knowledge which has 1 See J. M. E. McTaggart, Studies in the Hegelian Dialectic; Studies in Hegelian Cosmology. 472 THE PROBLEM OF KNOWLEDGE grown out of past experience. Even ''becoming" refuses to appear as the mere result of compounding ''being" and "not- being" ; it is plucked fresh from the fields of direct intuition, of immediate experience. But, on the other hand, the dialectic, as McTaggart conceives it, if followed out faithfully, would soon lead the thinker, as may readily be imagined, into a barren desert of meaningless abstractions. Our conclusion, then, is that a rationalistic absolute monism in methodology, aiming, whether as deductive science or dialectical philosophy, to escape all dependence upon experience for either invention or verifica- tion, neither ought to be nor can be applied with consistency. We shall now turn to an examination of empirical absolute monism in methodology. The empiricist movement through- out the history of modern philosophy has been primarily method- ological in interest. The course of its history, however, may be viewed as exhibiting, twice over, the tendency to pass from the original methodical appeal to experience, with full confidence in the power of thought adequately to interpret this experience, to a position of scepticism with reference to the value of thought for knowledge, and an abandonment to the immediate data of experience. The first of these movements is that from Bacon to Hume ; the second, that from Comte to Bergson. Between Hume and Comte stands Kant, siding for the most part with the empiricists as against dogmatic rationalists, in so far as scientific method is concerned, but partaking of both the scepti- cism of Hume and the positivism to be developed by Comte. Francis Bacon, reacting from the experience-defying specu- lative dogmatism of the scholastics, set up as the ideal for the thinker an unbiassed and methodical investigation of the laws of nature, with a view to mastery over the forces of nature. Reasoning power he regarded as of very insignificant importance in this task, as compared with the possession of the ''novum organum," the new inductive method of investigation. The empirical method of the "interpretation of nature," he compared to the use of a compass and a ruler in drawing circles and straight lines, with which the unpractised man is able to obtain good results, whereas the rationalistic dogmatist, pretending to an "anticipation of nature" by means of thought alone, is to be likened to a draughtsman who may be more talented and expert, THE PROBLEM OF SCIENTIFIC METHOD 473 but who obtains inferior results, because he is compelled to operate without instruments. Indeed, in view of nature's great subtilty, a purely speculative philosophy of nature is but a kind of insanity. No great progress in learning is to be looked for through ''anticipations"; ''our only hope is in genuine Induction," ascending to axioms from particulars, and only from axioms so constructed descending again to particular effects.! Bacon's advocacy of empirical methods was very timely; but his methodology was too one-sided, in that he did not sufficiently recognize the fact which is so full of difficulty for the extreme empiricist, that there are "certain conceptual order- systems whose exactness of structure far transcends, in ideal, the grade of exactness that can ever be given to our physical observations themselves." ^ For such oversight, however, Bacon is less to be blamed than are many of his empiricist suc- cessors, for whom a rapidly increasing body of mathematically exact "natural laws" has been available. This fact must be fairly dealt with in our constructive undertaking. In the extreme empiricism and consequent scepticism of Hume, we see, by way of contrast, how indispensable to science is the con- structive activity of reason. Comte's positivism deliberately aimed to turn away from all dogmatic speculations concerning what is beyond the realm of experience, and to confine intellectual effort to a simple scien- tific description of the phenomenal and verifiable. Metaphys- ics he regarded as a vain attempt to support the fantastic structures of traditional theology, fast toppling under the attacks of the empirical sciences. With the progress of induc- tive science, metaphysics becomes more abstract, seeking to explain phenomena by abstract substances or essences, and events by final causes. The final stage of metaphysical thought is reached when events are explained as being caused by nature and natural causes; whereas in the final state of thought nothing will be attempted beyond a formulation of the laws of * Novum Organum, Preface, and Bk. I, aphorisms 1-3, 9-14, 19-21, 26-30, 103, etc. 2 Royce, "The Principles of Logic," in Windelband and Ruge's Encyclopedia of the Philosophical Sciences, Vol. I, p. 88. 474 THE PROBLEM OF KNOWLEDGE phenomena on the basis of a simple generahzing description of the facts of observation and experiment.^ John Stuart Mill made important contributions to the tech- nique of induction, to some of which we must presently refer; but he is important in this immediate connection as marking a rather close approximation to what we have styled an empirical absolute monism in methodology. Influenced by the English tradition of empiricism and by the positivism of Comte, as well as directly by the rapidly developing empirical sciences, he undertook to construct a logic in which induction, as the process of inference from particulars to particulars, should appear as the only scientific method. '^Deduction" is still recognized but only as the process of reading off the signs previously employed to register former inferences from particulars to particulars. Ratiocination is simply the interpretation and appHcation of inductions ; the syllogism can do no more than dole out driblets of old knowledge as they may be needed ; it always involves a begging of the question and can never lead to any really new knowledge. The '^ deductive sciences," such as arithmetic and geometry, are really inductive ; they differ from the obviously inductive sciences in that they confine themselves to interpret- ing old inductions, without needing to resort to new observa- tion and experiment. Their theorems are necessary truths only in the sense of necessarily following from hypotheses. The axioms employed, many of them surreptitiously or uncon- sciously, are all experimental truths, inductions from the evi- dence of the senses. The definitions are mere verbal proposi- tions, explanations of the meaning of a name, together with an implied assumption of the existence of things corresponding to them; in other words, they are axioms also, old inductions in disguise.^ There is much in Mill's doctrine that is suggestive and at least partially justified; but his general position undoubtedly calls for criticism. Familiar illustrations from coromon life, such as that of the startling discovery, in the well-known anecdote, that since the abbe's first penitent was a murderer, and since a certain nobleman was the abbe's first penitent, that 1 Cours d'une philosophic positive, passim. » A System of Logic, Bk. I, Ch. 8 ; Bk. II, Chs. 1-6. THE PROBLEM OF SCIENTIFIC METHOD 475 nobleman was to be regarded as a murderer, show that deduction can of itself lead at times to important new knowledge. More- over, the constantly increasing body of new knowledge concern- ing logical and mathematical relationships is further presump- tive evidence of the fecundity of deduction. And if indeed it should turn out to be a tenable position that there is throughout the deductive sciences a constant dependence upon empirical verification for certainty, still, in the light of the achievements of symbolic logic, that dependence must be much less obvious, essentially, than Mill supposed.^ The further question as to whether Mill is right in reducing definitions to axioms, or Couturat in reducing axioms to definitions, or whether both are wrong, we shall reserve for consideration in the constructive part of our discussion. ''I should maintain," saysRoyce, ''that the mystics are the only thoroughgoing empiricists in the history of philosophy." ^ The truth of this statement may not be very obvious in the case of religious mysticism, but the fundamental features of Bergson's philosophy would indicate that at the extreme of methodological empiricism we are bound to find a theoretical mysticism. Nor are we likely to find in recent philosophy a more pronounced example of empirical absolute methodological monism than in the methodological doctrine of Bergson and his disciple LeRoy, whatever we may think of their actual practice. Not only is deduction incompetent, according to Bergson, to give us genuine knowledge about that which is ultimately real; even induction is not sufficiently empirical. Both intellectual processes are dependent upon our intuition of space, as is also even our idea of number ; but, inasmuch as space and the spatial world are to some extent constructs of finite intelligence (although having also something to do with the development of intellectual forms), neither logical process can be depended upon for knowledge, because both tend in- evitably to spatialize the reality with which they deal. Only " pure perception " and other modes of intuition such as we 1 Russell, Principles of Mathematics, I, pp. 3, 4, 10, 106, 373-4, 457-8 ; Royce, Sources of Religious Insight, pp. 94-6, 98 ; William James and Other Essays, pp. 211-12, 246-9. 2 The World and the Individual, I, p. 81. 476 THE PROBLEM OF KNOWLEDGE approximate in our immediate awareness of the life of our own self which endures, can be truly cognitive in any ultimate or metaphysical sense. • Even induction should be used only as a stepping-stone to intuition. Even science at its best, according to LeRoy, is the instrument of action only.^ In this way, then, as happened before in the case of Hume, an extremely monistic empirical methodology undermines the very founda- tion upon which it originally proposed to stand. We are now ready to consider the possibility of combining the justifiable elements of both the rationalistic and the em- pirical form of methodological monism, without leaving an unexplained absolute dualism of reason and experience, of deduction and induction. Before attempting a constructive statement on our own account, it will be helpful to say something of the contributions made toward such a critical methodological monism by J. H. Poincare. First of all, however, we shall refer to the methodological position of Kant ; for while it would be vain to look within the limits of the Kantian system for any very satisfactory illustration of the critical monism which seems to be demanded in methodology, our exposition of the theory of Poincare will be facilitated if the main features of Kant's doctrine are borne in mind. Having retained and developed further his inherited rational- istic doctrine of the synthetic activity of the mind, Kant might have been expected to be found favorably disposed to the doc- trine of the possibility of extending our knowledge of reality by means of purely deductive processes, or by a non-empirical dialectic. As a matter of fact he accepted the knowledge-value of such synthetic activity — in so far as he can be said to have accepted it at all — only within the limits of ''possible human experience." Strictly speaking, for Kant not even within actual human experience can the synthetic activity of mind give genuine knowledge of reality. The forms of mental activ- ity are declared to be absolutely a priori; consequently the object of actual sense-experience must be regarded as a con- struct, essentially different in constitution even from the bare 1 Bergson, Time and Free Will, pp. 77 £f., 84, 225 ; Matter and Memory, pp. 26, 64, 69, 77, 84-5, 297, 303-4 ; Introduction to Metaphysics, passim; see Poincar6's The Value of Science, pp. 112 ff. THE PROBLEM OF SCIENTIFIC METHOD 477 sense-material which is its content, and of no Hkeness at all, so far as we can say, to the independent reality which is assumed to be causally responsible for that raw-material of human ex- perience. It becomes impossible, therefore, to speak of any known or any knowable reality, save that which is real only in and for human conscious experience, actual or possible, and through the constructive activity of human thought ; we have not knowledge, but merely a human makeshift for it. But if for even so little as this there must be, as something for the mental activity to work into shape, the raw materials of sense, or, at the very least, temporal and spatial ''intuition" (which, from our point of view, is also ultimately empirical), it must be evident that the creations of pure reason are twice removed from reality and from knowledge of it ; they are not even made out of materials produced by the reahty they would be and enable us to know. The upshot of all this, so far as methodology is concerned, is to lead to a sanctioning of induction and, under dejSnite limits, deduction, for the gaining of such ''knowledge" as is humanly available; but inasmuch as the best we can ever get is not knowledge, but something which, as the best thing available, goes by that name, obviously in the end all methodological considerations are of Httle importance; both induction and deduction are, in the last analysis, epistemologically futile. We may say, then, that while Kant intended to leave room for what we would call a critical empirical monism in methodology, and while this was quite in keeping with his critical perceptual monism/ it was obscured and handicapped by the absolute genetic dualism to which we have referred,^ and by his absolute epistemological dualism,^ or his wavering between that position and an idealistic epistemological monism. The greatest contributions that have been made to the phi- losophy of scientific method for many a day, as many would agree, are those of Poincare. His fundamental problem, cor- responding to the Kantian, How are synthetic judgments a priori possible ? has been how to account for the union of novelty and absolute certainty in pure mathematics. In 1894 he stated this problem as follows : "The very possibility of the science of » See Ch. XV, supra. 2 See Ch. XVI, supra. ^ Cf. Ch. II, supra. 478 THE PROBLEM OF KNOWLEDGE mathematics seems an insoluble contradiction. If this science is deductive only in appearance, where does it get that perfect rigor no one dreams of doubting? If, on the contrary, all the propositions it enunciates can be deduced from one another by the rules of formal logic, how is it that mathematics is not reduced to an immense tautology?" ^ Mathematical physics, too, with its synthetic judgments a priori, is a fact to be ex- plained,^ as is also geometry, in connection with which the prob- lem presents itself in a new aspect in view of the development of non-Euclidean systems.^ In attacking these problems Poincare adopts a fundamen- tally empirical attitude, and directs his attention toward the nature and function of hypotheses in the various mathematical and physical sciences. He finds that there are two main classes of hypotheses, viz. generalizations and ''neutral hypoth- eses," or conventions. The former are fruitful, because they are capable of definite verification or refutation. The latter cannot be completely verified, nor can they be refuted ; but, being more or less convenient as guides to action, they are to be regarded as thought-constructs which are merely useful, and concerning which all questions of truth or falsity are irrel- evant. He mentions another class, viz. ''natural hypotheses" ; but these do not figure largely in his discussions, and it would seem that they are simply hypotheses which would come within the terms of the definition of conventions, but which it seems necessary to interpret not as for that reason neither true nor false, but as generalizations not at present capable of complete empirical verification.^ Among the conventions he would include definitions, although of course definitions are not arbitrary conventions, but the most convenient constructions available. Mill's contention that the definition implies an axiom, a statement about existence, Poincar6 would admit only when "existence" is defined in the 1 "Sur la nature du raisonnement math^matique," Revue de MStaphysique et de Morale, 1894, p. 371 ; Science and Hypothesis, Eng. Tr., p. 5. * Science and Hypothesis, p. 101. » lb., pp. 3, 29, etc. * lb., Introduction (by Royce), pp. xxi, xxiv, and also pp. 109-10; cf. Vaihinger's distinction between "hypotheses" and useful or indispensable "fictions," Die Philosophie des Als Ob, passim. THE PROBLEM OF SCIENTIFIC METHOD 479 purely logical sense of freedom from contradiction.^ But what he is especially concerned to urge is that the first principles of geometry are mere conventions — not arbitrary, indeed, but simply convenient, and neither true nor false.^ Apart from such arithmetical processes as may be involved, those axioms which have nothing to do with space, but are purely analytical propositions,^ the axioms of Euclidean geometry, it is declared, are shown by the developments of non-Euclidean systems to be not synthetic a 'priori judgments, nor yet experimental facts, but simply the most convenient of many logically possible sets of conventions, or disguised definitions.^ Our space of three dimensions has simply been imposed by ourselves upon nature, because of its comparative convenience ; and the same thing is asserted of time.^ Essentially the same position is taken with reference to mathematical physics. The special principles of mechanics reduce in the last analysis to a mere convention, which we have the right to make, because we are certain before- hand that no experiment can ever contradict it. More explic- itly, the principle of the conservation of energy reduces to the proposition : There is something which remains constant ; and this, although forever unverifiable and irrefutable, we assume because of its practical convenience. Even the Copernican astronomy is to be preferred to the Ptolemaic, not as any truer, but simply as more convenient.^ There are very evident suggestions here, it may be remarked, of a certain type of pragmatism, or quasi-pragmatism,^ and of immediate empiricism, or some form of psychological idealism.^ Both the pragmatism and the idealism are avowed in the Dernier es Pensees.^ Poincare has been careful to insist that science is an end as well as a means, ^^ and that it must give genuine knowledge, foresight, as otherwise it would not be even useful.^^ But even with these safeguards the idealistic pragma- 1 Science et Methode, pp. 139, 161-2. 2 Science and Hypothesis, p. 3, Chs. III-V ; The Value of Science, Chs. Ill, IV ; Science et MMhode, Bk. II, Ch. I. 3 Science and Hypothesis, p. 29. ■* lb., pp. 38-9, 53, 65. 5 The Value of Science, p. 13 ; cf. Ch. II ; Science and Hypothesis, p. 67. « Science and Hypothesis, Chs. VI-VIII, especially pp. 85, 93, 98-100 ; The Value of Science, Chs. V-IX, especially p. 76. 7 See Ch. XVIII, supra. s See Ch, VI, supra. <» Pp. 146-8, 157-8. 10 The Value of Science, pp. 8, 9. ii lb., p. 115. 480 THE PROBLEM OF KNOWLEDGE tism which would reduce not only atoms and molecules but space, time, and energy to mere useful devices of human thought is hard put to it to give an intelligible meaning to the objectivity of science. What is maintained is that the relations which are found constantly to obtain in certain groups of sensations are the only elements of the experienced that are common to many minds, and that they are, as such, therefore, the only objective reality.^ The other hypotheses, the generalizations, are based upon and verified in particular experiences. They all rest upon the in- duction which makes us expect the repetition of a phenomenon when the circumstances under which it first happened are reproduced. "If all these circumstances could be reproduced at once, this principle could be applied without fear ; but that will never happen ; some of these circumstances will always be lacking. Are we absolutely sure they are unimportant? Evi- dently not. That may be probable; it cannot be rigorously certain. Hence the important role the notion of probability plays in the physical sciences." ^ It becomes highly important, in order to eliminate uncertainty as completely as possible from our generalizations, to choose for purposes of induction signifi- cant facts, facts which will serve many times and which thus reveal a law. Scientific method, in short, is the judicious choice of facts upon which one can safely build generalizations.^ But we have, as the result of induction, only probability thus far ; there has been nothing to account for the peculiar certainty attaching to the novelties which make their appearance in mathematics. To explain this Poincare maintains that while the procedure of mathematical science in its purity, i.e. in arithmetic, is not induction in the ordinary sense of generalizing on the basis of sense-experience, neither is it mere deduction. It is a true case of the synthetic judgment a priori, and is based ^ The Value of Science, pp. 135-8. 2 Science and Hypothesis, p. 4. 3 The Value of Science, pp. 4-9 ; Science et Methode, pp. 7-18, 307-11. C. S. Peirce points out that we can draw probable conclusions concerning a set of facts of determinate constitution, if we choose fair samples of the collection, observe their constitution and generalize carefully ; and that this is so because the possible samples which agree with the constitution of the whole are more numerous than those which disagree. See exposition by Royce, in Windel- band and Ruge's Encyclopedia, pp. 82-8. THE PROBLEM OF SCIENTIFIC METHOD 481 upon an intuition of a special sort, viz. the intuition of pure number, which can never lead us astray, however far we may carry the generalizations based upon it.^ The dependence upon Kant at this point is quite evident ; but Poincare's doctrine is far from being identical with the Kantian, with its a ^priori intuitions of space and time and its category of number, and we must go on to examine more closely what the French philosopher has to say about this intuition of pure number and about the mathematical induction based upon it. The syllogism, according to Poincare, leads to nothing essen- tially new ; it adds nothing to the data contained in the prem- ises. ^ In order to make any science, even arithmetic, more than pure logic is necessary. This something more is what he calls intuition.^ Now intuition may be broadly defined as immediate presentation of reality (or truth) through an inherent power of the mind. When used broadly by Poincare, the term is meant to include the data of sense and imagination ; but in the narrower sense in which he commonly refers to it as the basis of a priori arithmetical knowledge, it seems to refer to an active property of the mind itself. It is necessary for the learning of mathematics, and in the application of mathematical results. It is the instrument of invention in both mathematics and physics, and as such necessarily precedes demonstration, fore- seeing conclusions and suggesting arguments, sometimes by sudden illumination which is due to the influence of subliminal processes.* The '4ogisticians," Peano, Russell, and Couturat, claim to dispense with intuition in pure mathematics. ''Even if we admit," writes Poincare, ''that it has been proved that all the theorems could be deduced by purely analytical processes, by simple logical combinations of a finite number of axioms, and that these axioms are nothing but conventions, the philosopher would still retain the right to investigate the origin of these conventions, in order to see why they have been judged pref- erable to the contrary conventions."^ The "indemonstrable ^ Science and Hypothesis, Ch. I, especially pp. 7, 13, 14 ; The Value of Science, pp. 19, 20, 23. 2 Science and Hypothesis, p. 5. ^ j<]iq Vahie of Science, p. 19. * Science and Hypothesis, p. 14 ; The Value of Science, pp. 21, 23 ; Uenseigne- ment mathematique, I, 1899, pp. 157 ff. ; Science et Methode, pp. 56-9, 82, 309. ' Science et Methode, p. 158. 2i 482 THE PROBLEM OF KNOWLEDGE propositions" with which the logisticians begin, and which they would explain as mere conventions, disguised definitions, are, according to Poincare, ^'in each case a new act of intuition." Nor are these appeals to intuition the last that will be necessary in ''pure" mathematics, it is contended. Even Couturat, who would construct the new logic without the idea of number, is nevertheless obhged to introduce references to number over and over again.i Without intuition, logic is sterile, except that it may engender contradictions and so make progress by their elimination.^ But Poincare is far from agreeing with Bergson that we ought ever to try to dispense with logic in our pm-suit of knowledge. Bare intuition, unless it be the intuition of pure number itself, can never give us either precision or certainty. Precision is found only in our definitions and in what is logically deduced therefrom ; and, in general, the more precise our ideas become, the more we are forced to abstract from intuition and the less objective our knowledge is found to be. And as for certainty, that depends upon proof, and all proof is a logical process.^ It is not logic without intuition that can make any science, nor intuition without logic, but logic in combination with in- tuition. In other words science, even mathematics, is essen- tially inductive, rather than deductive. Mathematics is no exception to the rule that experiment is the sole source of truth, that it alone can teach us anything new, and it alone can give us certainty. Pure mathematics need not appeal to material objects for verification, but its propositions are arrived at in truly inductive fashion, by proceeding from the particular to the general; only, the particular in this case is found in the intuition of pure number, ''the only intuition which cannot deceive us." Mathematical induction also differs from all other induction in that it is absolutely rigorous, and its results are absolutely certain.^ The axiom of this mathematical induction, or of demonstration by recurrence, which gives rise to all purely mathematical reasoning, Poincare enunciates as 1 Science et Methode, pp. 156, 176-7. 2 75.^ pp. 159^ 2II. 3 The Value of Science, pp. 17, 18, 20, 23, 25, 79 ; Science et Methode, pp. 130-1. * Science and Hypothesis, pp. 2, 3, 101 ; The Value of Science, pp. 20, 23, 25 ; Science et MUhode, pp. 160, 162, 309. THE PROBLEM OF SCIENTIFIC METHOD 483 follows: "If a theorem is true for the number one, and if it has been proved that it is true for the number n + 1 provided it is true for n, it will be true for all the positive whole num- bers." 1 In much of this we can follow Poincare, but at some points we must choose another path. Let us indicate some of the most fundamental of our objections to his doctrine. In the first place it seems impossible to accept his positivistic psycho- logical idealism as epistemologically valid.^ To be sure, there is some justification for his identifying the Cantorians with the realists and the pragmatists with the idealists, because the realism which he has in mind consists in the anti-empirical attributing of independent reality to universal ideas, as the Cantorians tend to do ; while the idealism he is thinking of is that which regards these universals as conventions, depend- ing upon the mind of man for their being, structure, and function, having been devised and chosen with reference to practical human purposes, as the pragmatists contend. But we can also sympathize with the remark of Hermite, quoted by Poincare,^ ''I am an anti-Cantorian just because I am a realist." If one believes in an independently real world of physical energy in space and time, with which one has immedi- ate and mediate cognitive relations, while he cannot adopt Poincare's sweeping reduction of the entities of physical science to mere convenient products of human thought for practical purposes, he will be almost as strongly disinclined to accept the Cantorian doctrine that the nature of reality, even so far as it is set forth in mathematics, is discoverable without any dependence whatever upon experience or intuition. More- over there are indications in Poincare's essay in the recent volume entitled Le materialisme aduel,^ and in the Dernieres Pensees that his own thought latterly was moving in the realistic direction.^ But the line of distinction, so rigidly drawn, between gener- alizations and mere conventions, is calculated to excite sus- picions with reference to the fundamentals of Poincare's doc- 1 Science and Hypothesis, pp. 11-14, 38; The Value of Science, pp. 19, 20, 23. 2 See Ch. VI, supra. » Dernieres Pensees, p. 160. * Paris, 1913. 5 See H. C. Brown, Journal of Philosophy, XI, 1914, p. 231. 484 THE PROBLEM OF KNOWLEDGE trine. How comes it that there are two radically different types of mental instruments, both of which intend to represent an independent reality, but of which only one sort actually does represent any reality, and that not the present reality intended, but a past or future group of sensations, while the other, however useful in leading us, actually represents noth- ing at all, either in external reahty or in past or future sensa- tion? If, however, a realistic theory were adopted, this near-dualism of generalization and convention might be appre- ciably reduced. Then, it would appear that our more con- venient and therefore less arbitrary ''neutral hypotheses," as well as our ''generalizations," are framed to represent reality as it is, whether immediately experienced by us or not, and that these ideas tend to become inconvenient just in propor- tion as the facts of experience tend to refute them. Holding that reality is broader than the content of human experience, we would be entitled to say that even hypotheses which can neither be completely verified nor refuted, if the most con- venient hypotheses that can be devised, are probably either true or reasonably close to the truth. Convenience is a mark of rationality, and rationality, according to our realistic and empirical doctrine, has taken shape by being moulded upon reality. There are also difiiculties in the way of accepting, as a whole, Poincare's doctrine of intuition. He uses the term "intuition" more nearly in the Kantian than in the Bergsonian sense, as being not only an immediate awareness of reality, but also in some sense a contribution of mind, not through thought, but through the activity of some other innate faculty. In all three, Kant, Bergson, and Poincare, the term "intuition" refers to an at least immediate awareness through some original activ- ity of the mind ; but whereas Kant applies it primarily in con- nection with space and time, regarding quantity and nimiber not as intuitions, but as categories of thought, and whereas Bergson regards our awareness of time, as duration, as our most completely intuitive consciousness, space being only partially, if at all, an intuition, and in some apparently unde- termined measure a construct, with quantity and number as further derivatives, Poincare, contrary to both, explaining THE PROBLEM OF SCIENTIFIC METHOD 485 space and time as conventions, thought-constructs, is primarily interested in affirming the intuitive character of our conscious- ness of pure number. Our suggestion would be that in the case of both space and time our primary consciousness is intui- tive, a direct awareness of reality, much like what Bergson asserts with reference to duration. Our consciousness of num- ber, however, seems to be partly due to immediate presentation of the serial combinations and activities of external reality,^ but partly also to a recurrent activity of the mind itself. Ac- cording to this view the origin of the number-consciousness would be essentially realistic and empirical, as it is in the case of space and time; but the apparently more intuitive char- acter of the consciousness of number is due to its having a more subjective origin than our consciousness of space, what- ever Bergson may have to say of our awareness of time. Thus we would be able to say that our awareness of number is due to an activity of mind, as Kant held ; that, in part at least, it is also derivative from our at least partially intuitive conscious- ness of spatial and quantitative objects, as Bergson maintains ; and yet that it is itself, fundamentally at least, an intuitive consciousness, as Poincare has claimed. It is a creative men- tal activity by virtue of which a character of reality is directly revealed, even if it does come to be a category of thought, the particular mode of application of which does depend upon the purpose of the moment. From this point of view the sharp distinction between the forms of sensuous intuition and the categories of thought tends to disappear. Both are regarded as due to creative psychical activity, but at the same time as being moulded upon the independent reality presented in and through that activity. It is convenient, however, to use the term intuition in connection with the more original and im- mediate states of consciousness, and the term category in con- nection with those that are more mediate and derivative. This leads us to question whether Poincare is justified in setting up so strong a contrast as he does between intuition 1 E. Picard suggests, very plausibly, that we owe the idea of cardinal number (number pertaining to a group) to our sense of sight, and the idea of ordinal number to our sense of hearing {Der Wissen der Gegenwart in Mathematik und Naturwissenschaft. Wissenschaft und Hypothese, XVI, Leipzig, B. G. Teubner, 1913). See Journal of Philosophy, XI, 1914, pp. 556-7. 486 THE PROBLEM OF KNOWLEDGE and logic. On the one hand, is it true, as he holds in opposi- tion to the logisticians, that the syllogism can never teach us anything essentially new ? ^ And on the other hand, is it true, as he assumes, in agreement at this point with his oppo- nents, that deductive logic can dispense altogether with intui- tion ? 2 In our opinion the negative answer to both questions is quite defensible. We have seen that under certain cir- cumstances important new information may be disclosed by syllogistic reasoning. This may occur when the right premises are brought together, either by accident, as in the case of the old priest and his first penitent, or by a sagacious selection of the character represented by the middle term.^ And if ordinary deduction is not sterile, why should we not look for intuition in it as well as in what Poincare calls mathematical induction? Is not knowledge of implication, as truly as knowledge of num- ber, originally intuitive? Corresponding to Poincare's axiom of mathematical induction we have the axiom of the syllogism. Instead of the scholastic dictum de omni (that whatever can be affirmed of a class may be affirmed of everything included in the class) we would suggest the following formulation : What is true of any subject is also true of any subject with which (either individually or as the class of which this ob- ject is a member) it is (so far as all purposes which ought to be considered are concerned) numerically identifiable by means of some infallible mark.^ That this is an infallible 1 Science and Hypothesis, p. 5. 2 Science et MUhode, pp. 152-9. 3 See James, Principles of Psychology, Vol. II, pp. 342-5. 4 Bertrand Russell has a formulation which is similar, except that it fails to bring out so clearly the originally intuitional and empirical character of knowl- edge of implication. His statement is, "If anything has a certain property, and whatever has this property has a certain other property, then the thing in question has the other property" {Our Knowledge of the External World, etc., p. 57). For some empirical accounts of deductive reasoning, see J. S. Mill, A System of Logic, Bk. II, Ch. II, § 3, where the positive formula is " Things which coexist with the same thing coexist with one another," and the negative, "A thing which coexists with another thing, with which other a third thing does not coexist, is not coexistent with that third thing " ; and James, Psychology, Vol. II, p. 340, where " the two great points in reasoning " are thus stated : " First, an extracted character is taken as equivalent to the entire datum from which it comes; and second, the character thus taken suggests a certain conse- quence more obviously than it was suggested by the total datum as it originally came." THE PROBLEM OF SCIENTIFIC METHOD 487 mark, and that the numerical identification can be legitimately made, depend for their being known upon intuition, and ultimately upon experience. Moreover, whereas Poincare has classed such axioms as that equals to the same thing are equals to one another, as ''analytic judgments a priori/'^ we would ask, Are these not generalizations (and as such, instruments of thought) on the basis of an intuition, in this particular case, of the transitivity of the relation of equal- ity ? It would seem, then, that intuition may be held to accom- pany even our logical processes, at least in so far as they are not mere processes of unintelligent routine. (This is very evidently suggested by the strict limits to the new knowledge that may be inferred from the premises.) On this view all significant deduction is virtually induction. This is not a one- sided or absolute empirical monism in methodology, for deduc- tion, the syllogism and the a priori have been given their due. Rather is it to be called a critical empirical methodological monism. We must now briefly outline the course of the scientific method of proof, as seen from the point of view of our critical empirical methodological monism. The sciences are generally classified as abstract, descriptive, and normative. It should be noted, however, that in the normative sciences we simply have certain materials which have been drawn from the de- scriptive sciences, selected and organized with reference to the realization of some universal ideal, as that of truth in logic, beauty in aesthetics, and morality in ethics. They can be suffi- ciently dealt with for our present purposes, therefore, in our discussion of the method of descriptive science. The method of the abstract sciences demands some attention, however. It should be noted at the outset that the abstractness of these sciences is relative. It is quite evident, for instance, that, our consciousness of three-dimension space being interpreted as essentially intuitional, not conventional, the Euclidean geome- try is abstract with reference to the physical world, and con- crete with reference to space. Even pure arithmetic is con- crete with reference to number ; but there seem to be important grounds for maintaining that mathematical physics is quite ^ Science and Hypothesis, p. 29. 488 THE PROBLEM OF KNOWLEDGE abstract with reference to the actual world.^ In speaking of abstract sciences, however, we shall have in mind chiefly the mathematical sciences, which are more or less abstract with reference to the physical world. Doubt as to the abstractness or concreteness of a science may be overcome by raising the question whether the science is true of the reahty concerned categorically, or only hypothetically. If it is true categorically, the science is concrete ; if only hypotheticaUy, it is abstract. The definitions with which the abstract sciences begin, while not necessarily arbitrary, are nevertheless conventions, which are to remain constant throughout the whole process. They are not necessarily disguised axioms in the sense of generaliza- tions concerning the real experienced world. Poincare is cor- rect, as against Mill, in maintaining that the existence assumed is nothing more than mathematical existence, i.e. freedom from contradiction, from the point of view of the logic of con- sistency. In view of the non-EucHdean geometries and the science of infinite aggregates, it will be seen that even freedom from contradiction-producing characteristics is not a necessary assumption in the definitions fundamental to an abstract science. The explicit assumptions (axioms and postulates) of an abstract science may be anytliing even approximately conceiv- able. In some cases the assumptions accord with experience; in other cases, while not verifiable, they do not contradict experience, and are not arbitrary, but the most convenient conventions that can be devised ; in still other cases, however, the assumption may be quite arbitrary and not especially con- venient, but practically contradicting experience, and even running foul of the best efforts of the imagination. It may be conceded to Couturat and his feUow-logisticians that the inde- monstrable axioms of mathematics and symbolic logic are dis- guised definitions ; but this concession has no great significance. As axiom or postulate, the ''that," in the mathematical sense of permissibihty, i.e. supposed possibihty, or freedom from contradiction, is emphasized; as definition, the ''what" re- ceives the emphasis. Commonly the most fundamental ^ See E. Boutroux, Natural Law in Science and Philosophy ; Poincare, Science and Hypothesis, pp. 94-8 ; H. Driesch, History and Theory of Vitalism, pp. 223- 31. THE PROBLEM OF SCIENTIFIC METHOD 489 assumptions of ordinary thought are taken over; but among the assumptions sometimes introduced we find the following : two parallels to a given straight line, both passing through the same point — or, in other words, non-Euclidean space (Lobachevski) ; motion without friction ; a stationary earth and the heavenly bodies moving in perfect circles (the Ptolemaic astronomy) ; an actual infinite, or a whole such that there is a one to one correspondence between its elements and the elements of one of its parts (Dedekind, Cantor, Royce, Russell, Couturat). The point which it is of the greatest im- portance to remember here is that the abstractness of the abstract sciences is of two possible sorts, viz. the abstractness which comes by the subtraction of empirical elements, and the abstractness which comes by the imaginary substitution of arbitrary for genuinely empirical elements. Euclidean geome- try will serve to illustrate the one, and the non-Euclidean systems the other. Broadly speaking, the former abstract- ness is in the interest of practice, while the latter is productive of mere curiosities, which serve only to stimulate speculative wonder. The procedure in the abstract sciences is largely deductive, ''analytic" ; but, as we have tried to show, this does not mean that it is not at all inductive. It appeals to intuition, though it refrains from appealing to all possible sorts of intuition, be- cause it has abstracted from experienced reality in its full con- creteness, and is interested only in certain phases of the whole. The main question for scientific methodology is that which has to do with the ^^ novum organum," the method of the descriptive, or empirical,^ or overtly inductive sciences. In such sciences the preliminary definitions are largely formal, with just enough of the content indicated to enable the investi- gator to identify the object to be studied. The content of the definition constantly grows as the processes of investiga- tion are successful. A full definition, one adequate for all possible purposes, may be regarded as the goal of empirical research. 1 By "empirical science" we do not mean a scientific description of experi- ence as such, but a scientific description of reality as it is known immediately in, or mediately through, experience. 490 THE PROBLEM OF KNOWLEDGE Among the postulates and other assumptions fundamental to any empirical science are properly included, besides the assumptions most indispensable to common sense, the axioms and postulates of science in general, and the relevant results of the other sciences. The existence of the subject-matter to be investigated is also presupposed, either as already practi- cally certain, on grounds of experience, or else simply as a fundamental working hypothesis to be tested by empirical methods. As the investigation proceeds there accumulates a body of empirical data, or what may be regarded thenceforth as pre- suppositions. In ordinary perceptual experience there is a certain measure of discrimination and choice, only those prod- ucts of sense-activity being selected, ordinarily, which have some relation to some subjective interest; but in scientific observation and experiment this discrimination and choice are much more pronounced. In any scientific observation the intention is to accumulate simply such data as are relevant and may conceivably be made the basis of inductive inference. With this hmitation, the collection of data aims to be the choice of what Peirce calls ''fair samples"; within the field of the relevant the data must be such that we have no reason to sup- pose they have not been chosen at random. All inductive inference, or generahzation, is based upon one fundamental principle, sometimes called the uniformity of nature. Less dogmatically put, it is the principle of the de- pendableness of nature. Viewed as an hypothesis, it is the last to be fully verified, and yet, in the undogmatic form we have suggested, it must always be the last to be given up. What Mayer is reported to have said with reference to the theory of the conservation of energy, ''I discovered the new theory for the sufficient reason that I vividly felt the need of it," ^ is still more emphatically true of every scientist in relation to this fundamental principle. We must now take up definitely the question of the method or methods of induction, i.e. of discovery of ''laws," which may be regarded, for our present purposes, simply as generali- zations, stating what the thing under investigation does under 1 E. Mach, Popular Scientific Lectures, p. 184. THE PROBLEM OF SCIENTIFIC METHOD 491 certain conditions. As might be expected from this definition, the main Hnes of procedure are simple enough. ''By indirec- tion" we proceed to "find direction out." From the above principle of the dependableness of nature, and from the theoret- ical constructions to which we shall presently refer, there may be deduced certain major hypotheses, from which in turn may be deduced minor hypotheses. By minor hypotheses are meant such as are capable of being refuted or completely verified in single crucial experiments, where acting upon the hypothesis leads to an experience in which there is immediate awareness either of the reality or of the unreality of what was supposed in the hypothesis. (In some cases actual experiment may not be necessary, the appeal, perhaps even in the framing of the hypothesis, to the known results of past experiences, or to ''intuition," being sufficient for verification.) Refutation of a minor hypothesis involves, it should be noted, refutation of the major hypothesis from which it was deduced, and also of the logical theory concerned; but verification of the minor hypothesis does not mean complete verification of the major hypothesis or general theory. To assume the opposite would be to commit the fallacy of affirming the consequent. Practi- cally complete verification of major hypotheses may be ob- tained, however, by the employment of Mill's well-known methods of experimental inquiry, the Method of Agreement, the Method of Difference, the Joint Method of Agreement and Difference, the Method of Residues and the Method of Con- comitant Variations. The canons of these methods are stated by Mill, as is required by his phenomenological philosophy, in terms of "unconditional, invariable antecedent" phenomena as causes ; ^ but, translated into reahstic terms, as required by our epistemological theory, they would run somewhat as follows : An indication of something causally related to a phenomenon may be found either in some circumstance in which all the instances of the phenomenon agree, or in some circum- stance in which alone the two instances of its occurrence differ, or in such circumstances as vary whenever the phenomenon to be explained varies; and when part of a phenomenon has been accounted for, an indication of the cause of the remainder 1 System of Logic, Bk. Ill, Ch. VIII. 492 THE PROBLEM OF KNOWLEDGE may possibly be found in the circumstances which have not served to guide to the already discovered causes. These indications are sometimes of great service in leading to the framing of a successful hypothesis as to the causation of the facts under consideration. Once a law has been discovered, it makes possible both prediction and the partial control of future experience, on the assumption that under the same condi- tions the thing will act as before. The further experience result- ing furnishes further data for induction. But over and above all these laws or generalizations as to the nature of observed facts or the course of observed events, there is a place for scientific theory, which is essentially further a ^posteriori definition of the subject-matter under investigation. What it is beyond immediate experience, is capable of being learned to some extent in the light of what it is and does within immediate experience. Such theory, again, as has been noted, suggests further hypotheses which may be empirically tested. And obviously, from our realistic point of view, it is not at all necessary to regard entities with which theoretical construc- tion deals as being mere conventions, so long as no hypotheses deduced from the theory are refuted. Moreover, once such refutation has occurred, the entity, as conceived, can no longer be legitimately assumed, even as a convention. All sciences, then, it would appear, are descriptive, either categorically or hypothetically. They are hypothetical when some condition or conditions need to be expHcitly stated or kept in mind, in order to avoid misunderstanding and practi- cal error. They are categorical when all conditions are so in accord with experience and intuition that they can be taken for granted, without explicit statement. But on the other hand, not only the original definitions, but the empirical data and the generalized descriptions or laws, become assumptions forthwith, from which deduction may proceed. And inasmuch as all these assumptions may also be viewed as fragmentary or real definition, categorical or hypothetical, of a reality, or reali- ties, all science is thus, it would seem, broadly speaking, deduc- tive. But the assumptions are either directly derived from experience or ''intuition," or else they are made on certain conditions, the meaning of which is also empirically derived, THE PROBLEM OF SCIENTIFIC METHOD 493 immediately or ultimately ; and so, it would appear, all science is also, broadly .speaking, inductive. Even scientific theory is description, as well as assumption. It has to do not with an absolutely, totally unexperienced reality back of experienced processes. It is anticipatory or divinatory further description of reahty or processes, some of which may not be, in the more direct sense, humanly experienceable — description, more- over, on the basis of what is and has been thus directly experi- enced. This view of theory as further partial description of a not yet fully experienced and perhaps not fully experienceable reality, partly on the basis of what is or has been experienced of that particular reahty, and partly on the basis of what has been experienced of reahties in general, may be likened to the completing of given arcs of a circle or ellipse, by means of a knowledge of the general nature of circles and elHpses, as derived from experience or "intuition." All science, we may then say, is deductive, and yet all is inductive, just as we saw that all science is descriptive, although obviously it is always necessarily to some extent abstract. This discovery that the method of all really scientific proof, i.e. of all demonstration of the truth about reahty, is one and the same, being both inductive and deductive, enables one the better to decide between the rival claims of abstract, descrip- tive and normative science, respectively, to be the one funda- mental form of scientific procedure, to which the other forms may be reduced. According to some (e.g. J. S. Mill) all real science is descriptive; definitions are disguised axioms, which in turn are interpreted as empirical generahzations, tentative or final. From this point of view, truth is correspondence. According to others (e.g. B. Russell, Couturat) all real science is abstract ; axioms are definitions in disguise ; truth is cohe- rence (according to Russell, a multiple relation). According to still others (e.g. F. C. S. Schiller, and in some points, J. Dewey and Wm. James) all science is essentially normative (a system of judgments which it is good to believe for practical purposes) ; all axioms are postulates ; truth is identical with consequences that are good. Now it is true enough, on the one hand, that science is always, of necessity, more or less abstract, and on the other hand that the various sciences may each be viewed as 494 THE PROBLEM OF KNOWLEDGE organized about some fundamental practical interest, so that they take on the aspect of organized instrumentalities as related to some ideal, or norm. But the fundamental character of science, we would claim, is that of being a description of reality. Axioms are essentially akin to empirical generalizations, and must stand the test of comparison with the facts of experience. Abstractness is to be permitted only for the sake of convenience ; otherwise it is to be reduced as far as possible in the interests of knowledge of reality. Definitions are especially to be watched, as possible sources of abstractness, and are to be constantly revised and given more concrete content in the light of further experience of reality. Normative sciences are to be regarded as resulting from a process of selection from the results of descriptive sciences for some relatively constant special purpose or organized group of purposes; and while all science may be said to partake, fundamentally, of this characteristic, it must not be supposed that the sciences are mere expressions of purpose, requiring no verification beyond practical useful- ness in a general way. Every proposition must be brought into comparison with reality as experienced. Postulates are to be taken as hypotheses and examined with a view to empiri- cal justification. The above-described, really unitary scientific method, avoid- ing, as it does, an absolute dualism of deduction and induction and the two one-sided absolute monisms (the deductive, or rationalistic, and the inductive, or empirical) may well be called, as we anticipated, a critical empirical methodological monism. This scientific method is the method of proof, i.e. the method of producing logical (sufficiently critical, or intellectually adequate) certainty {i.e. intellectual readiness for definitive action) with reference to the truth about reality. Thus our conclusions in the various separate investigations which we have been obliged to undertake are seen to converge toward what is, in general, one and the same philosophical position. In epistemology proper we were led to a critical realistic monism. Obliged, for the completion of our solution of the problem of acquaintance, to make excursions into the morphology of knowledge and genetic logic, we found ourselves THE PROBLEM OF SCIENTIFIC METHOD 495 with a critical perceptual monism in the former field, and a critical empirical monism in the latter. In logical theory, again, we arrived at a critical pragmatic monism, and finally, in methodology, at a critical empirical monism. Our result is thus critical monism, epistemological, morphological, genetic, logical, and methodological. This critical monism has this much in common with the point of view occupied by Hoeffding, and to which he applies the same name, that it ''strives to maintain the thought of unity without dogmatizing." It seeks to avoid absolute dualism, but does not insist upon arriving at an absolute monism. Hoeffding' s interest, however, is almost entirely in maintaining his critical monism in connec- tion with metaphysical problems, particularly the problems of ''substance" and of "the one and the many." Moreover, his "critical" principle is based upon Kantian presuppositions, and is not without its suggestions of agnosticism.^ Our own criti- cal monism, on the contrary, which in the present volume has been applied only to problems included within the general field of epistemology, departs fundamentally from the Kantian point of view, and looks directly to the sciences, in which, with their carrying of the unifying process as far, but only as far, as the facts will allow, the pace is set for all philosophical under- takings. One may surmise that this principle of critical mon- ism, with its union of the attitudes of faith and scepticism, would prove no less fruitful in metaphysics than in the sciences and epistemology ; but to anticipate further such results would not agree well with the critical ideal of proceeding "with- out dogmatizing." Finally, to forestall one not improbable even if, as we think, superficial objection, our critical monism, we would say, can- not be dismissed as "eclecticism." It would hope, to be sure, to do justice to those valid elements and approaches to the truth which are to be found in most of the systems rejected. But while it aims, definitely and persistently, to avoid the fundamental errors of other philosophies, it has not been inter- ested in any process of culling out whatever attractive and 1 B.. KoeMing, Philosophy of Religion, Eng. Tr., pp. 33-4, 57, 65-9 ; Problems of Philosophy, pp. 135-7, 144; "A Philosophical Confession," Journal of Phi- losophy, Vol. II, 1905, pp. 85-92. 496 THE PROBLEM OF KNOWLEDGE separately plausible doctrines there may be in the different theories considered. On the contrary, it has been primarily concerned to maintain both internal consistency and fidelity to fact. Indeed, one of the author's principal fears is that in his actual procedure he may not have been, in a possible sense of the term, eclectic enough. The critical portions of the work are designed to show the need of a new system, and for that reason they have had to ignore many things for which the philosophies examined are undoubtedly worthy of universal appreciation. INDEX OF AUTHORS [Italic figures refer to the more important passages.] Albertus Magnus, 79. Alexander, H. B., 255. Alexander, S., 217, 222, 284-236, 244, 251, 261, 262-264, 266, 271, 290, 294, 303-304, 306, 307, 308, 315, 363, 396. Aliotta, A., 146. Ameseder, R., 204. AngeU, J. R., 272, 273, 275, 317, 319. Apelt, E. F., 41. Aquinas, T., 391. Aristotle, 86, 88, 89, 317, 391. Aster, E. von, 196. Avenarius, R., 63, 111-112, 113, 160, 219, 220, 230, 251. Bacon, F., 463, 472-473. Bain, A., 381. Bakewell, CM., 206-208. Baldwin, J. M., 351, 4II-412. Balfour, A. J., 136. BardiH, C. G., 38. Baumann, J., 64. Bawden, H. H., 119-120, 408, 421, 423, 446. Bergson, H., 109, 120-125, 162, 220, 299, 312, 313, 314, 317, 322, 337, 339, 340, 343, 347, 357, 360, 361, 362- 363, 4OO-4O6, 415, 416, 453, 472, 475, 476, 482, 484-485. Berkeley, G., 16, 45, 82, 93, 96, 97-98, 116, 133, 161, 223, 325. Bolyai, W., 464. Bolzano, B., 465, 466. Boodin, J. E., 227-229, 247-248, 277, 279, 395-396, 412-413, 423. Bosanquet, B., 146, 154-159, 163, 314, 381-382, 383. Boutroux, E., 317, 488. Bradley, F. H., 30, 130, 136, 146-153, 154, 155, 157, 159, 160, 163, 176, 181, 218, 375-381, 382, 383, 388, 407, 431. Brown, H. C, 109, 483. Bush, W. T., 220, 224, 267-268. Caird, E., 22, 133, 136, 137. Caird, J., 96, 134-135. CaldweU, W., 431, 437, 446. Cantor, G., 303, 465, 466, 468, 483, 489. Carr, H. W., 222. Cassirer, E., 195-198. Clifford, W. K., 55, 100-102, 103. Coe, G. A., 75, 77. Cohen, H., 83, 193, 194, 195-198, 317, 339 353 Cohen, M. R., 206, 264-265, 304. Comte, A., 18, 473-474. Copernicus, N., 20, 23, 414, 434, 479. Cornelius, H., 66-68. Costello, H. T., 469. Couturat, L., 462, 466, 475, 481, 482, 488, 489, 493. Darwin, C. 356. Dedekind, R., 465, 466, 489. Delacroix, H., 75. De Laguna, T., 431. Descartes, R., 93, 246, 352, 462. Deussen, 60, 75, 76, 77,. 78. De Wette, W. M. L., 41. Dewey, J., 92, 114, 117-118, 119, 170, 219, 221, 224-227, 228, 229, 238, 273, 274-275, 283, 299, 320, 323, 408, 409, 420-421, 422, 423, 424, 426, 428, 430, 436, 446, 493. DUthey, W., 25, 33-34. Dionysius, 88. Drake, D., 53, 54. Driesch, H., 299, 317, 344, 488. Dunlap, K., 270, 271-272. Eckhart, 79. Eddy, M. B. G., 50. EUiot, H., 360. EUisen, 194. Eucken, R., 208-209, 317, 433. EucHd, 109, 254, 414, 464, 465, 466, 467, 478, 479, 486, 488, 489. Ewald, O., 197, 201. Fawcett, E. D., 444. Fechner, G. T., 165, 182, 317. Fichte, J. G., 25, 57, 85, 96, 104-105, 106, 141, 162, 195, 231, 317, 338, 339, 353. Fischer, K., 20, 21, 57, 104. Fite, W., 430. FouHlee, A., 105-106. Frege, G., 465. 2k 497 498 INDEX OF AUTHORS Fries, J. F., 37, S8-4I, 42, 347. Frischeisen-Kohler, M., 34. Fritzsch, P., 183. Frost, E. P., 272, 274, 275-276, 291. Fullerton, G. S., 114-115, 192-193, 217, 221, Gibson, W. R. Boyce, 208, 209, 317, 373-374. Gomperz, T., 86, 88. Green, T. H., 95, 134-136, 155, 317, 353, 375. Guyon, Mme., 77. Haldane, R. B., 136. HamHton, W., 26-27, 28, 216. Harris, W. T., 131. Hartmann, E. von, 36, 57, 60-63. Hebert, M., 419. Hegel, G. W. F., 25, 62, 63, 127, 128, 130-131, 132, 133, 134, 135, 136, 137, 138, 140, 155, 157, 162, 175, 176, 181, 195, 196, 197, 198, 206, 218, 238, 317, 338-339, 347, 353, 374, 375, 376, 381, 470, 471. Henderson, C. R., 448. Herbart, J. F., 36, 4^-44, 45, 46, 48, 195, 317, 339. Hermite, C, 483. Herrick, C. L., 455. Hobhouse, L. T., 217, 218, 244-246, 260, 261, 360. Hocking, W. E., 121, 123, 161-180, 347, 403-404, 413-414, 454. Hodgson, S. H., 30-31, 114, 115, 219, 221-222, 223, 224, 228, 259-260, 261. Hoeffding, H., 3, 4, 20, 27, 194, 496. Hoffman, F. S., 80. Holt, E. B., 217, 223, 251-254, 279, 280-281, 286-287, 291, 296, 300, 303, 305, 392, 394, 463. Hopkins, E. W., 76. Howison, G. H., 189-190, 206, 317. Hume, D., 19, 20, 23, 27, 63, 98-99, 100, 133, 182, 183, 213, 316, 363, 354, 472, 476. Husserl, E., 202-203, 206. Jacobi, F. H., 37, 38, 42. James, W., 10, 52, 99, 114, 116-117, 122, 219, 221, 222-224, 227, 228, 238, 251, 266-267, 270, 271, 291, 317, 319, 340-341, 342, 344, 345, 356-366, 357- 369, 361, 363, 4O6, 408, 411, 4I8-419, 420, 4^2-423, 424, 426, 427, 428, 429, 431, 435, 436, 486, 493. Jevons, W. S., 377. Joachim, H, H., 383-384. Jones, A. H., 51. Jones, E. E. C., 398-400. Jones, R. M., 79. Jordan, D. S., 360. Judd, C. H., 221. Kant, I., 14, 17-24, 25, 26, 28, 31, 32, 36, 37, 38, 39, 40, 41, 42, 47, 49, 57, 58, 59, 60, 61, 69, 78, 83, 84, 85, 96, 103, 106. 112, 123, 128, 132, 133, 136, 182, 193, 194, 195, 196, 197, 198, 199, 206, 230, 231, 299, 317, 331-332, 338, 339, 346, 348, 361-362, 353, 362, 364, 365, 370, 372, 402, 462, 464, 472, 476- 477, 481, 4S4-485, 495. KeUogg, V. L., 360. Knox, H. v., 413. Kroner, R., 203. Kuelpe, O., 36, 61, 68-70, 114, 230. Ladd, G. T., 8^9, 60-61, 73, 317. Lamarck, J. B., 35Q. Lange, F. A., 193-196. Lanz, H., 203. Leibniz, G .W. von, 55, 93, 96, 182, I84- 185, 317, 362-363, 371-372, 462. Le Roy, E., 415, 420, 475, 476. Liebmann, O., 193, 196. Lipps, T., 106-107, 108. Lobachevski, N. J., 464, 489. Locke, J., I4-I6, 17, 19, 21, 93, 96, 213, 247, 317, 322, 353, 371, 372, 373. Lotze, R. H., 36, 44-50, 135, 182, 183, 184, 185, 313, 317, 322, 372. Lovejoy, A. O., 66, 240, 246, 310, 311, 372, 408. Luther, M., 195. McCosh, J., 216. McDougall, W., 260-261, 290, 315, 316, 317, 319, 360. McGilvary, E. B., 217, 227, 238-240, 277-279, 304, 306, 307, 392-393. MacGregor, D. H., 95. Mach, E., 110-111, 112, 219, 220, 249, 414, 416, 490. McTaggart, J. M. E., 190-192, 339, 471-472. Mansel, H., 27, 28. MarshaU, H. R., 103-104, 361. Marvin, W. T., 217, 248-249, 279, 280- 281, 296-297, 300, 302, 303, 309, 392, 394. Mayer, J. R. von, 490. Mead, G. H., 118-119, 220. Meinong, A., 203-206, 231, 264, 266. Meyer, M., 408. Mill, Jas., 354. 1 INDEX OF AUTHORS 499 Mill, J. S., 27, 63, 100, 116, 182, 223, 316, 354, 463, 469, 4^4-^75, 478, 486, 488, 491, 493. Miller, D. S., 52, 213, 277. Montague, W. P., 217, 223, 227, 248, 251, 254-258, 259, 264, 279, 287-280, 295-296, 302, 303, 305, 306, 307-308, 392, 398-394, 433. Moore, A. W., 119, 408, 409, 412, 421, 423, 431, 434, 446. Moore, G. E., 217, 222, 230, 236-238, 239, 261-262, 266-267, 290, 306, 315, 319. Moore, G. F., 76. Morgan, C. Lloyd, 361. Muirhead, J. H., 136. Miinch, F., 201. Miinsterberg, H., 198, 201. Murray, D. L., 408, 426, 430. Natorp, P., 81, 83, 86, 88, 195-198. Nelson, L., 39, 4I-42. Nunn, T. P., 217, 240-24^, 244, 251, 261, 262, 293-294. Occam, W. of, 243. Oldenberg, H., 76. Ostwald, W., 317. Papini, G., 420. Parmelee, M., 360. Parmenides, 77. Pater, W., 314. Paulhan, F., 317, 410. Paulsen, F., 22, 60, 182-183, 193. Peano, G., 464, 481. Pearson, K., 102-103. Peirce, C. S., 4IO-4II, 422, 429, 444, 480, 490. Perry, R. B., 95, 217, 220, 223, 248, 250-251, 279, 281, 283-285, 286, 291, 297, 300-302, 303, 306, 308, 392, 393, 423. Petzoldt, J., 112. Pfeiffer, F., 79. Philo, 86, 88. Picard, E., 485. Pitkin, W. B., 217, 227, 251, 279, 281, 281-283, 297, 298-300, 304. Plato, 8, 22, 63, 78, 81-89, 133, 197, 202, 206, 302, 303. 304, 336-337, 339, 341, 391, 402. Plotinus, 86, 87, 88, 89-91, 162. Poincare, J. H., 108-109, 303, 414-415, 416, 463, 464, 466, 468, 470, 476, 477- 487, 488. Porphyrius, 89. Porter, N., 50, 216. Pratt, J. B., 430, 444. Pringle-Pattison, A. S., 26, 51-52, 185. Ptolemy, 414, 434, 450, 451, 479, 489. Pythagoras, 1. Quick, O. C., 425, 444. 191. 305, 316. Rashdall, H., 187-188, Reid, T., 25, 26, 50, 213-21', Reinhold, K. L., 37-38, 42. Renouvier, C., 188-189, 317. Rickert, H., 198, 200-201. Riehl, A., 22, 25, 31-33. Rignano, E., 360. Ritchie, D. G., 88, 136. Rogers, A. K., 146, 435. Romanes, G. H., 360. Royce, J., 131, 136, I4I-I46, 154, 158, 159, 160, 163, 166, 253, 839-340, 341, 350, 384-390, 413, 442-443, 449-450, 465, 475, 478, 481, 489. Russell, B., 70-71, 217, 218, 222, 231, 234, 240, 242-244, 261, 262, 264, 266, 285, 293, 294-295, 302-303, 304, 305- 306, 309, 396-398, 403, 419, 444, 463, 465,' 466, 469, 475, 481, 489, 493. Russell, J. E., 423-426, 428, 430, 437. Schaub, E. L., 105. Schelling, F. W. J., 57-58, 61, 62, 162, 195 317 338. Schiller, F.'c. S., 50, 150, 188, 218, 317, 383-384, 407, 408, 409, 410, 419-420, 421, 423, 424, 426, 428, 430, 433, 434, 435, 446, 455, 463, 468, 493. Schinz, A., 416. Schleiden, M., 41. Schopenhauer, A., 36, 57, 58-60, 62, 63, 78, 142, 182, 195, 317, 323. Schulze, G. E., 24. Schuppe, W., 56. Sidgwick, A., 420, 427, 468. Simmel, G., 432. Singer, E. A., 272, 276-277, 291. Slosson, E. E., 407. Small, A. W., 448. Smith, N. K., Ill, 220-221. Sneath, E. H., 216. Snellman, J. W., 423. Socrates, 1, 87. Spaulding, E. G., 150, 217, 248, 249- 250, 279, 281, 282, 297-298, 303. Spencer, H., 25, 27-30, 354-355. Spinoza, B. de, 63, 93, 182, 462. Stebbing, L. S., 457. Stewart, J. A., 81, 83, 84, 88, 328. Stirling, J. H., 131, 132. 500 INDEX OF AUTHORS Stout, G. F., 217, 218, 222, 264-266, 290, 294, 305, 361. Strong, C. A., 62-56, 102, 219, 255. Tawney, G. A., 431, 435. Taylor, A. E., 81, 82, 84, 86, 87, 146, 154, 159-161, 337. Thorndike, E. L., 271, 272, 273-27 4^ Ueberweg, F., 79. Vaihinger, H., 18, 19, 107-108, 415, 416, 478. Volkelt, J., 63-66, 347. WaUace, W., 105, 131. Walsh, C. M., 21. Ward, J., 185-187, 218. Watson, J., 8, 133, 136, 137, 138-141, 375. Watson, J. B., 272, 273, 274-275, 291. Whitehead, A. N., 465. Windelband, W., 3, 86, 198-200. Wolf, A., 56, 217, 245, 246, 265, 266, 290, 315. Woodbridge, F. J. E., 7, 8, 217, 219, 224, 233-234, 244, 269-270, 272, 276- 277, 282, 291. Woodworth, R. S., 270-271, 272, 317. Woolley, H. Thompson, 422. Wright, H. W., 370, 453. Wright, W. K., 363. Wundt, W., 36, 68, 69, 112-114, 219, 220, 221, 230, 317. Yajnavalkhya, 75. ZeUer, E., 86. INDEX OF SUBJECTS This index is at all detailed only in the case of the author's constructive statements. For purposes of reference to the expositions and criticisms of the views of others, the Analytical Table of Contents and the Index of Authors should be used. Absolute, 30, 333. Absolutism, 93, 180, 338. Abstraction, 85 n. ; see Idealism, ab- stract. Activism, 312-19, 363 ; see Creative activity. -(Esthetics, 2, 4. Agnosticism, Chs. II-IV. fallacy of, 23^ ; cf. 33. Anti-conceptualism, 401-6. Anti-inteUectualism, Ch. XVIII; 371. Apperception, 330, 339, 342. Apriori, Ch. XVI. absolute, 41 n., 365. relative, 41, 330, 364-5. Assumptions, 488-9, 490. Axioms ; see Assumptions. Behaviorism, 272-7, 285-7, 319. Categories, 361-4. Causality, 316. Certainty, 331, 369, 459-61. Color-blindness, 321. Conception, 341, 350. Consciousness, 3-4; Ch. XII; 313, 314-22. Consistency, in pragmatism, 435, 451. logic of, 369-70 ; see Logic, real and formal. Cosmolog5% 5. Creative activity, 55, 214, 312-14, 315- 22, 329, 363. Deduction, Ch. XX. principle of, 486. Definitions, 488, 489, 492. Dialectic, 127-8, 166-7, 190-2, 339, 470-2. Dogmatism, with epistemological dualism, 51, 71. of absolute idealism, 127, 135, 137. of the new realism, 309, 348. Dreams, 322, 342. Dualism, epistemological, Chs. II-IV. definition, 13, 14. involves agnosticism, Chs. II-IV; 14, 44, 52, 70. morphological, 336. genetic, 351-2. logical, 370. methodological, 461-2. Eclecticism, 495-6. Empiricism, activistic, 19 n., 41, 357-65. in genetic logic, 353-65. in methodology, 372-94. immediate; see Idealism, psycho- logical, disguised, radical ; see Idealism, psychological, disguised. Empirio criticism ; see Idealism, psy- chological, disguised. Epistemology, 3, 4-5, 6-7, 333-4. relation to metaphysics, 7-10. subdivisions of, 10. Ethics, 2, 4. Experience, 20, 21, 44, 52, 70 ; see Em- piricism, pure, philosophy of; see Idealism, psychological, disguised. Freedom, 317-18, 381. Geometry, Euclidean and non-Euclid- ean, 464-5, 466-7. Hallucination, 321. History, philosophy of, 4, 5. Hypotheses, 490-2. Idealism, epistemological, Chs. V-IX; 13, 14. definition of, 72-3. subdivisions of, 74-5. practica,! and theoretical, 73. 501 502 INDEX OF SUBJECTS Idealism, relative, 73-4. mystical, 75-80. logical, 81-9, 201-6. disguised, 85, 230-1. mystical-logical, 89-91. psychological, 92-120. fallacies of, 94-6. activistic, 96, 104-6. disguised, 96-7, 109-120, 219-30. mystical-psychological, 120-5. absolute, Chs. VII, VIII ; 128-9. faUacies of, 129-30, 133-4, 140-1, 143-4. intellectualistic, 130-41, 154-9. voluntaristic, 130, 141-6, 159-61. mystical ; see Idealism, mystical- logical-psychological, logical-psychological, 126-61. objective, 128. subjective ; see IdeaKsm, psycho- logical, concrete, 128. abstract, 128, 192-208. personal, 129, 181-92, 206-8. fallacies of, 183-4. mystical-logical-psychological, 130, 161-80. theistic, 181-90. pluralistic ; see Idealism, personal, rehgious, 208-9. spiritual ; see Idealism, rehgious. Ideas, as constructs, 84. Identity (and truth), Ch. XVII; 438- 42, 445. lUusion, 321, 324, 325-6. Images, positive after-, 320. negative after-, 320. IndividuaHsm, 92, 433-5. Induction, Ch. XX. mathematical, 463, 482-3. methods of, 483, 491-2. principle of, 490. Infinite, 465, 467-8. Inheritance, 354, 359, 360, 364. Instinct, 360-3. InteUectuaHsm, Ch. XVII ; 371. Interaction, 50, 93, 320. Interpretation, 339-40, 342. Introspection, 271-2, 344-6. Intuition, 60, 170-1, 346-9, 401-6, 453- 4, 469-70, 484-7. Judgment, 81, 341, 343. analytic and synthetic, 332. Knowledge, immediate, Chs. II-XVI ; 327 mediate! Chs. XVII-XX. dependent on immediate, 311-12; cf. Chs. II-IV. definition of, 311, modes of, Ch. XV. morphology of, Ch. XV. theory of ; see Epistemology. Logic, scientific, 2, 370. philosophical, 4 ; Chs. XVII-XIX. genetic, Ch. XVI. real and formal, 369-70, 463, 467, 468-9. symbolic ; see Logistic. Logistic, 462-70. Meaning, 410-11. Memory, 342. Metaphysics, 3-4, 5, 6. relation to epistemology, 7-10. relation to the sciences, 69. Methodology, Ch. XX, especially 487- 94. Mind, 94, 313, 316. Monadism, 93. Monism, critical, 494-6. in epistemology proper, 13, 19 n., 56, 309 ; Ch. XIV ; 334-5. in morphology of knowledge, 338- 50. in genetic logic, 355-65. in logical theory, Ch. XIX. in methodology, 476-94. (absolute) epistemological, Chs. V- XIII ; 13, 14. ideahstic, Chs. V-IX. reahstic, Chs. X-XIII. (absolute) morphological, 336-50. conceptual, 336-7. perceptual, 337. (absolute) genetic, 352-56. rationalistic, 352-3. empirical, 353-56. (absolute) logical, Chs. XVII, XVIII. intellectualistic, Ch. XVII. anti-inteUectuahstic, Ch. XVIII. (absolute) methodological, 462-76. rationahstic, 462-72. empirical, 472-76. numerical metaphysical (singular- ism) ; see IdeaHsm, absolute. Mysticism ; see Idealism, mystical ; mystical-logical ; mystical- psychological ; mystical-log- ical-psychological ; also 87, 160. and psychology, 92. and empiricism, 475-6. Ontology, 5. INDEX OF SUBJECTS 503 Parallelism, 320. Perception, 330, 336-50. in a complex, 343-4. Philosophy, definition of, 1, relation to sciences, 1-3. its field, 3. its problems, 3. Postulates ; see Assumptions. Pragmatism, 407-58. current, 407-37. scientific, 408, 414-17, 436-7; see Pragmatism, scientific represen- tational. semi-, 410-14. quasi-, 410, 414-17. pseudo-, 410, 417-22. hyper-, 410, 422-31. essential, 417,, 431, 441. "absolute," 413. negative, 413-14. animalistic, 432-3. representational, Ch. XIX. scientific, 449, 453-5, 456-7. Proof, 331, 369 ; Ch. XX. Psychology, metaphysical, 5. relation to idealism, 92. definition of, 318-19. Purpose, and relations, 332-3. and truth, 380, 390, 440 f., 443, 444- 5, 446-7, 452, 454. Qualities, primary, 15; Ch. XI; 322-8, 329, 331. secondary, 15 ; Ch. XI ; 313 £f., 322- 8, 329, 331. tertiary, 327-8, 329. Ratiocination, 341, 343, 486. Rationalism, in genetic logic, 352-3. in methodology, 462-72. Realism, epistemological, Chs. X-XIV ; 13, 14. logical, 83, 85, 86-9, 202-6, 230-1, 302-6. neo-, Chs. X-XIII. naive, 212-13. natural, 213-17. activistic, 316. Reflex arc, 320. Relations, internality and enternality of, 49, 293-302, 332-3. primary, secondary, and tertiary, 328. knowledge of, 348. Relativity, 30. Religion, philosophy of, 5. Representation, 329-30, 439-41. Sciences, relation to philosophy, 1-3, 69. abstract, 2, 487-9, 493-4. descriptive, 2, 489-94. normative, 2-3, 487, 493-4. Sensation, 313-27. Solipsism, 103-4, 119-20, 129, 139. Soul ; Spirit ; see Mind. Subconsciousness, 318-19. Subjectivism ; see Idealism, psycho- logical. Theism, 170. Theology, metaphysical, 5. Theory, scientific, 492. Thing-in-itself, knowable, 67, 327. Truth, 81, 145, 331 ; Chs. XVII-XIX ; see Logic, real and formal. logic of, 369-70. definition of, 444-5, 446, 452, 457. ideal element in, 446, 448. human, 446-7. permanence of, 449, 451. absolute, 449-50. superhuman, 455-6. Utilitarianism, in logical theory, 431- 3, 435. Values, 3, 306-8, 328-9, 348, 349. Variations, spontaneous, 355 ff., 359. Wisdom, philosophy as, 1, 3. Printed in the United States of America. *HE following pages contain advertisements of books on kindred subjedls. A Student's History of Philosophy By ARTHUR KENYON ROGERS Professor of Philosophy in Yale University. New York, 1901. New edition, 1907. ^11 pages, 8vo, $2.00 A feasible and practical introduction to the history of philosophy suited to the mental development of the college student. It gives an accurate account of philosophical de- velopment which includes all the material that can fairly be placed before the student beginning the study of this subject. The chief aim is to gain simplicity of statement without los- ing sight of the real meaning of philosophical problems. Throughout, the author has borne in mind the viewpoint of the student, a feature which has contributed greatly to the success of the book. Wherever possible, the thought of the writers has been given in their own words, thus combining a practical source book and a fully developed narrative history in one volume. Although the author recognizes that the his- tory of philosophy centers about the lives of individual men, he has made his book distinctly a history of philosophy and not of philosophers. THE MACMILLAN COMPANY Publishers 64-66 Fifth Avenue New York A Brief History of Modern Philosophy By Dr. HARALD HOFFDING Professor in the University of Copenhagen Translated with the author's permission by C. F. Sanders Professor of Philosophy in Pennsylvania College Clothy i2moy $i._50 In a concise and interesting manner the author discusses the following subjects, which constitute the parts or books into which the volume is divided : The Philosophy of the Renaissance, The Great Systems, English Empirical Phi- losophy, Philosophy of the Enlightenment in France and Germany, Emanuel Kant and Critical Philosophy, The Phi- losophy of Romanticism, Positivism, New Theories of the Problem of Being upon a ReaHstic Basis, New Theories of the Problems of Knowledge and of Theories. "There is no saner or wiser guide to the study of philos- ophy now living than Professor Hoffding. Particularly valu- able are his appreciations of contemporary thinkers, including Boutroux, Bradley, Bergson, and Eucken." — Educational Review. "The work constitutes at once a masterly analysis and a valuable guide for sincere seekers after truth." — Review of Reviews, " It would be hard to find anything better adapted to the needs of the student or one that will more surely yield what- ever is asked of it in the way of clear, concise, and compact information." — The Argonaut. THE MACMILLAN COMPANY Publishers 64-66 Fifth Avenue New York A History of Philosophy With Especial Reference to the Formation and Development of Its Problems and Conceptions By Dr. W. WINDELBAND Professor of Philosophy in the University of Strassburg. Authorized translation by James H. Tufts, Ph.D., Professor of Philosophy in the University of Chicago. New York, 1893. Second edition revised and enlarged. y 2b pages ^ 8vo, $4.00 A portrayal of the evolution of the ideas of European philosophy, with the aim of showing through what motives the principles, by which we to-day scientifically conceive and judge the universe and human life, have been brought to consciousness and developed in the course of the movements of history. The choice of material has fallen everywhere on what individual thinkers have produced that was new and fruitful. The material is apportioned as follows : The Phi- losophy of the Greeks (23-154); The Hellenistic-Roman Philosophy (155-262); The Philosophy of the Middle Ages (263-347); The Philosophy of the Renaissance (348-436); The Philosophy of the Enlightenment (437-528); The Ger- man Philosophy (529-622); The Philosophy of the Nine- teenth Century (623-682). THE MACMILLAN COMPANY Publishers 64-66 Fifth Avenue New York An Introduction to Kant's Critical Philosophy By GEORGE TAPLEY WHITNEY AND PHILIP HOWARD FOGEL Assistant Professors of Philosophy in Princeton University 226 pages, 121710, $1.00 An excellent text for courses devoted exclusively to the study of Kant, and one which may also be used to distinct advantage in more general courses in which a considerable time is devoted to the study of the Critique of Pure Reason. Too often Kant has been taught as merely a part in a scheme of philosophy, and his ideas warped to suit the general scheme ; but in this book is brought out the many-sidedness of Kant's system in itself and for itself. The conflicting tend- encies in Kant's thought have not been ignored for the sake of a unified interpretation, but in the observations which are made at various points in the discussion, those aspects of the diverging tendencies which seem to be insisted upon by Kant and which seem to be involved in his fundamental position, are emphasized. Through the selections from Kant and the observations upon them, an attempt has been made to give a true state- ment of him, but no pretence is made of giving a complete interpretation of Kant. The aim is to give a statement of him which brings out the continuity of the thought, which emphasizes the problems he considered, and how they arose — in short, a statement which meets the needs of the ordinary student. THE MACMILLAN COMPANY Publishers 64-66 Fifth Avenue New York ^ ^ -^ ^ 9^ V ^^4- ^^^ ^ .-^^ ^ v-^v' * #"-^. '. -ft. ^^•^ ^^^% V * ^ "o. A^ ^ -^ * « /• ^ A^ " ^ * « .<^ A. ^, o .0^ V ^A , "^ ^ , X -^ A ^. ^. % "'-:. C5^ > ^— :>^^ " uV— "/. o V ^>^— ^>^^ ->/^^ 2 *• xO^ V xO^ V 'V^ .V^ ^■...ov^ V s- A-A-' _^_filf«tefc. ^^ AV^r LIBRARY OF CONGRESS 019 953 512 2 I .. !• i:;