MASTER NEGATIVE 
 
 NO. 93-81469- 
 
M I C R O F 1 L M E L) 1993 
 COLUMBIA UNIVERSITY LIBRARIES/NEW YORK 
 
 •»? 
 
 as pail of the 
 Foundations of Western Civilization Preservation Project 
 
 ?? 
 
 Funded bv tlie 
 NATIONAL END0WK4ENT FOR THE HUMANITIES 
 
 Reprodueiions inav not be made without permission from 
 
 Columl^a Unnersity Library 
 
COPYRIGHT STATEMENT 
 
 The copyright law of the United States - Title 17, United 
 States Code - concerns the making of photocopies or 
 other reproductions of copyrighted material. 
 
 Under certain conditions specified in the law, libraries and 
 archives are authorized to furnish a photocopy or other 
 reproduction. One of these specified conditions is that the 
 photocopy or other reproduction is not to be "used for any 
 purpose other than private study, scholarship, or 
 research." If a user makes a request for, or later uses, a 
 photocopy or reproduction for purposes in excess of "fair 
 use," that user may be liable for copyright infringement. 
 
 This institution reserves the right to refuse to accept a 
 copy order if, in its judgement, fulfillment of the order 
 would involve violation of the copyright law. 
 
A UTHOR: 
 
 THALHEIMER, ALVIN 
 
 TITLE: 
 
 THE MEANING OF THE 
 TERMS: 'EXISTENCE' .. 
 
 PLACE' 
 
 PRINCETON. N. J. 
 
 DA TE: 
 
 [1920] 
 
COLUMiUA UNIVERSITi' LIBRARIES 
 PRESERVATION DEPARTMENT 
 
 Master Negative // 
 
 BiliHOGRAPliiC MICROFORM i A R G ET 
 
 Original Material as Filmed - Existing Bibliographic Record 
 
 Kestrictioiis on Use: 
 
 ► 1 T •> 
 
 ■ MKl iJk -f*,^ 
 
 ' T 'X *"> 1 
 
 A, O tt I 
 
 I I mm 
 
 ^■Miqp 
 
 Thalhoimf>r, Alvln, 1894- 
 
 The meaning of the tarns 'exiutenoe' and 'reali- 
 ty - A dlB3nrtation ... by Alvln Ihalheimor ... ! 
 rrincoton, N.J., 1 rinoeton Univorolty press, jl9S0,i 
 
 1 1 -^"^ T 
 
 ^ P« 23 cm. 
 
 TheelB, Johns Hopkinn Univ. 1918. 
 
 
 • ) 
 
 TECHNICAL MICROFORM DATA 
 
 FILM SIZE: _3if^___ REDUCTION RATIO: ///V^ 
 
 IMAGE PLACEMENT: lA ^ JB^ IIB 
 
 DATE FILMED: 6^7/f3 INITIALS Pjf 
 
 FILMED 1!Y: RESEARCH PUBLICATIONS, INC WOODBRIDGF. CT 
 
c 
 
 Association for information and Image IManagement 
 
 1100 Wayne Avenue, Suite 1100 
 Silver Spring, Maryland 20910 
 
 301/587-8202 
 
 Centimeter 
 
 1 2 3 
 
 iiiiliiiiliiiiliiiili 
 
 111 
 
 rrr 
 
 IN 
 
 Inches 
 
 I 
 1 
 
 I 1 1 
 
 iliiiiliiiili 
 
 TTT 
 
 7 8 
 
 iliiiiliiiil 
 
 1.0 
 
 9 10 n 
 
 iiiliiiiliiiiliiiiliiiili 
 
 r\ I 
 
 45 
 
 2.8 
 
 50 
 
 
 
 3.2 
 
 
 3.6 
 
 2.5 
 
 2.2 
 
 I.I 
 
 1.25 
 
 **- III 4.0 
 
 1.4 
 
 2.0 
 
 1.8 
 
 1.6 
 
 12 13 14 15 mm 
 
 liiiiliiiiliiiiliiiiliiiiliiiiliiii 
 
 TTT 
 
 I i I 
 
 I 
 
 / 
 
 MRNUFPCTURED TO PIIM STflNDfiRDS 
 BY fiPPLIED IMRGE, INC. 
 
 / 
 

 4^- .r.-»...l: 
 
 Uniyrfcxchang«. 
 
 52618^ iO" 
 
 ii 
 
 THE MEANING OF the TERMS: 
 ^EXISTENCE' AND ^REALITY^ 
 
 A DISSERTATION 
 
 Submitted to the Board of University Studies of 
 
 THE Johns Hopkins University in conformity 
 
 WITH THE Requirements of the Degree 
 
 OF Doctor of Philosophy 
 
 BY 
 
 ALVIN THALHEIMER 
 
 '■ •~i«s? 
 
 BALTIMORE 
 1918 
 
 PRINCETON UNIVERSITY PRESS 
 PRINCETON. N. J. 
 
h 
 
 I 
 
 » 
 
 \ 
 
 vs t * 
 
 Columbia ^Hnibfrsitp 
 
 intljeCitpof^etogorb 
 
 LIBRARY 
 
 f 
 
 
 
THE MEANING OF the TERMS : 
 'EXISTENCE' AND 'REALITY' 
 
 
 %, 
 
 A DISSERTATION 
 
 Submitted to the Board of University Studies of 
 
 THE Johns Hopkins University in conformity 
 
 WITH THE Requirements of the Degree 
 
 of Doctor of Philosophy 
 
 BY 
 
 ALVIN THALHEIMER 
 
 BALTIMORE 
 1918 
 
 PRINCETON university PRESS 
 PRINCETON. N. J. 
 
f 
 t 
 
 (5 
 
 Published igj* 
 IVnUea in the United States of Amtr 
 
 Ka 
 
 .>j. 
 s 
 
 
 TABLE OF CONTENTS 
 
 Chapter I page. 
 The Nec<i for a Definition of ''Existence" and Reality" 5 
 
 Chapter II 
 The Reai ?« the Permanent '9 
 
 Chapter III 
 The Real as the Sensible ^^ 
 
 Chapter IV 
 The Real as that which is Related ^' 
 
 Chapter V 
 The Problem of the "Real Centaur ;" Belief and Existence 85 
 
 Chapter VI 
 The Authors Definition of "Existence" and "Reality" 99 
 
 sj 
 
 ^t 
 
 I 
 
.^> 
 
 CHAPTER I 
 
 The Need for a Definition of ''Existence'' and 
 
 ^^Reauty." 
 
 Not long- ago it was characteristic of philosophers to 
 deal with the world as a whole. To determine the funda- 
 mental nature of the heterogeneous experience with which 
 wc are confronted, to draw unity out of diversity, was 
 supposed to be the business of philosophy. All the world 
 was found to be "will" or ^'intellect" or '^spirit," and a way 
 was thereby opened by which the individual was enabled to 
 feel at home in his environment and in sympathy with it. 
 The aim of philosophy was to discover the fundamental 
 character of ''reality in general," and at the same time to 
 offer consolation to the individual by showing that the world 
 fundamentally is akin to himself. 
 
 But now, if we cannot say that all this has been changed, 
 we can at least note a strong movement in a different direc- 
 tion. We can note a tendency to consider the parts before 
 the whole, to use James's phraseology ; a tendency to busy 
 ourselves with some specific part of the world and to let the 
 world as a whole take care of itself. Specific problems, we 
 find, have become interesting and worthy of attention for 
 their own sake. Less value is being put upon the broad 
 vision necessary for synthesis and generalization, and more 
 value upon the exercise of keen and concentrated intellectual 
 effort of the sort that is required for analysis. To this 
 renewed interest in analysis the movement known as neo- 
 realism has offered the most organized support. But the 
 method of analysis is by no means tied up with one epistemo- 
 logical doctrine or with one theory of consciousness. It is 
 the method of all those, and their number seems to be grow- 
 ing, who are interested in specific problems for their own 
 sake, who do not consider all philosophizing merely a means 
 to some generalization as to the nature of the universe, and 
 who rate high among the charms of philosophy the alertness 
 required to dissect intricate concepts and to avoid lurking 
 confusions. 
 
 It is as an attempt at such analysis that the present dis- 
 sertation is conceived. We are frequently asserting of some 
 
 5 
 
 
entity that it exists or that it does not exist. What then 
 does it mean to "exist'' ? That is the question with which 
 we are to deal. And we are to deal with it without concern 
 for the future of our souls. We are to deal with it by con- 
 centrating our attention upon the concept "existence/' by 
 analyzing out various senses in which the term has been 
 used, and by making each of these senses of the word 
 definite, precise, and meaningful. 
 
 Indeed, if we are to come to close quarters with any con- 
 cept, "existence" is the one that requires our attention. For 
 we have here a concept that is particularly important in our 
 judgments. In nearly everything we say we are implicitly 
 saying something about existence. Except for a few classes v 
 of judgments, each judgment we make is an assertion of 
 the existence or of the non-existence of some entity ; it is '' 
 an assertion, that is to say, that can be turned into an exist- 
 ential proposition. It has sometimes been said that when- 
 ever we make a judgment we imply the existence of the 
 subject of that judgment. But that this is not the case it 
 has not been hard to show. We need only think of an 
 hypothetical proposition of the form : If A is, B is. Here 
 neither A nor B is said to exist, but the existence of B is 
 said to be contingent upon the existence of A. Exactly the 
 same interpretation holds when the proposition is thrown 
 into the categorical form: A implies B. And since per- 
 haps the most important judgments we have are judgments 
 of this latter form, judgments in which the existence of the 
 subject is not asserted, we cannot accept the thesis that in 
 all judgments the existence of the subject is assumed. But 
 though it cannot be maintained that a judgment assumes the 
 existence of its subject, it does not follow that judgments 
 as a rule have no existential import. Indeed, nearly all 
 propositions, it seems to me, can be turned into propositions 
 having a reference to existence. And it may be worth while 
 to show this in some detail in order that the importance of 
 the concept "existence" may be realized. 
 
 Let us start with the proposition: "Some men are bald.'* 
 Here I am asserting the existence, not only of men, but of 
 bald men. "Bald men exist" is exactly equivalent to "Some 
 men are bald." There is the same assertion of existence 
 when the particular proposition is negative. "Some men 
 are not patriotic" means that some unpatriotic men exist. 
 
 6 
 
 ij In most particular propositions, consequently, we are 
 \ making assertions that are equivalent to assertions of 
 h existence. Singular judgments, on the other hand, are, 
 ■ if they are negative, equivalent to assertions of non-exist- 
 ence. "George is not at home today" means that a George 
 I who has the quality of being home today is a non-existent 
 entity. When I think of such a George I am thinking of 
 something that doesn't exist. And when I assert that 
 George is not at home today I am asserting the non-exist- 
 ence of this object. In the case of affirmative singular 
 propositions there is a two-fold reference to existence. 
 Suppose, for example, I make the judgment : "John Smith is 
 now eating his dinner." I mean that a John Smith who is 
 now eating his dinner is a real object: when I think of such 
 a John Smith I am thinking of an existing entity. And a 
 John Smith who is not now eating his dinner is a myth, a 
 non-existent entity. When I lay down the affirmative sin- 
 gular proposition : A is B, I am asserting that the A which 
 is B is an existing entity, and that the A which is not B is a 
 non-existing entity. And I am saying nothing more than 
 this. It is in this way, then, that we find the existential 
 import of most particular and singular propositions. Par- 
 ticular propositions are exactly equivalent to assertions of 
 existence. Negative singular propositions are exactly 
 equivalent to assertions of non-existence. And affirmative 
 singular propositions are equivalent to propositions that 
 attribute existence to one entity and non-existence to 
 another. 
 
 With universal propositions the transformation is slightly 
 more involved. Suppose my judgment is that all men are 
 mortal. Then what my judgment amounts to is the asser- 
 tion that immortal men do not exist. To attribute mor- 
 tality to all men is to attribute non-existence to the class of 
 immortal men. When, on the other hand, I say : "No stone 
 is alive," my assertion means that living stones do not 
 exist. With these examples of Brentano's^ I quite agree. 
 I believe Brentano is correct in holding that universal cate- 
 gorical propositions, and hypothetical propositions that can 
 be thrown into categorical form, may be transformed into 
 assertions of non-existence. When we say "2 and 2 are 4," 
 
 ' Brentano — Psychologic vom cmpirischcn Standpwnkie, 1874, vol. i, 
 P- 283. 
 
I think we mean that a collection of two entities together 
 with two other entities such that the collection does not con- 
 tain four entities is a non-existent object. And I think that 
 there is nothing more that the proposition in question does 
 mean A proposition of this kind does not assert that the 
 subject exists ; but it does, nevertheless, make some assertion 
 about existence, or, rather, alwut non-existence. It may 1^, 
 as Holt says, that most propositions are about entities that do 
 not physically exist.-' But it does not follow that the dis- 
 tinction between existence and non-existence— in some sense 
 of "existence"— makes no difference to these proiwsitions. 
 It may be. as G. E. Moore says, that 2 and 2 are 4 "whether 
 there exist two things or not."''' But it does not follow that 
 mathematics is a non-existential science, if we mean by a 
 non-existential science one that has no use for the concepts 
 "existence" and "non-existence" in any of their senses. 
 Indeed if "non-existemial" is defined in this manner, an 
 existential science is precisely what mathematics is. And 
 not only mathematics but almost all scientific propositions, 
 and indeed almost all judgments, are existential. 
 
 Of course, it may be said that these categorical proposi- 
 tions are not equivalent to propositions asserting non-exist- 
 ence in general, but to propositions asserting non-existence 
 in some definite realm, as, for example, in the realm of 
 mathematics. But at least they assert non-existence of 
 some sort. Let us then l>e content with this for the present^ 
 We have shown that almost all propositions are equivalent 
 to propositions asserting existence or non-existence of some 
 sort We have shown that existence is a vitally necessary 
 concept, that no science can hope to escape a reference to it. 
 The distinction between various realms of existence we 
 
 shall consider later on. • , *. .^ 
 
 Our thesis is not that all propositions are eqmva ent to 
 propositions asserting existence or non-existence, for th re 
 are some propositions of which this is not true. When the 
 proposition with which we are deahng defines a word th 
 meaning of this proposition will not be exhausted f we 
 transform it into the existential proposition into which we 
 should normally transform it. If a pentagon is a plane 
 
 (1899). P- 1^- g 
 
 I 
 
 figure bounded by five straight lines, no doubt a pentagon 
 that is not a plane figure bounded by five straight lines does 
 not exist. But the proposition in which the word "|)enta- 
 gon" is first introduced means more than this. Something 
 similar might no doubt be said of such a proposition as : 
 ''this is a rose," where the meaning of the word "rose" is 
 given by pointing out one of the objects to which it refers. 
 A more important limitation, however, that must be put 
 upon the theory that has preceded, comes when we are deal- 
 ing with predications of existence and of the concepts that 
 are prior to existence. When existence is explicitly predi- 
 cated of an entity, there is no reference to existence other 
 than that which is explicit. If we were to transform ''All 
 A's exist'' as we have transformed ''All A's are B," we 
 should find ourselves saying : "A's that do not exist do not 
 exist" ; and we should be involved in a hopeless regress of 
 tautological propositions. Similarly when we predicate 
 "non-existence", or a term equivalent to "non-existence", 
 there is no reference to existence other than that which is 
 explicit. Again, in dealing with the predication of concepts 
 prior to the distinction between existence and non-existence, 
 this reference to existence cannot be found. Let us call 
 "subsistence" that which may be predicated of any entity. 
 Then subsistence and the qualities in terms of which exist- 
 tence is defined are prior to existence. When I attribute 
 one of these qualities to an entity, there can be no reference 
 to existence; otherwise we could never open our mouths 
 without being presumed already to know the meaning of 
 existence. 
 
 Before leaving this subject perhaps I ought to show how 
 I should meet an objection that might be raised against the 
 theory that has been put forth. I have said that particular 
 propositions and affirmative singular propositions imply the 
 assertion of existence. It may be objected, consequently, 
 that propositions such as: " Some Greek Gods acted im- 
 morally" and "Ivanhoe lived under the Plantagenets" make 
 no assertion of existence. It seems to me, however, that 
 such propositions, taken in the usual sense of their terms, 
 do make this assertion of existence, and, since what they 
 assert to have existed did not exist, these propositions are 
 false. If they are meant to be true, they are worded incor- 
 rectly. The one about Ivanhoe should read : "Ivanhoe is a 
 
 9 
 
fictitious character who is supposed to have lived under the 
 Plantagenets." In this proposition, however, there is no 
 assertion of existence according to the theory I am defend- 
 ing; for we are dealing with one of the cases that have 
 been expressly excepted. We are dealing with a case in 
 which non-existence is explicitly l^eing predicated ; for to 
 call Ivanhoe a fictitious character is to call him non-existent. 
 Apart from the classes of propositions that have been ex- 
 cepted, all categorical propositions can be transformed, and 
 transformed without loss of meaning, into assertions of 
 existence or of non-existence. And since most of the judg- 
 ments we make are of this form or can be put into this 
 form, we see that we are engaged for the most part in 
 making assertions that are equivalent to existential judg- 
 ments. In science, in mathematics, in daily life, the propo- 
 sitions with which we deal are equivalent to propositions 
 asserting existence or non-existence of some kind. Our 
 knowledge is built up of propositions that are equivalent to 
 existential propositions. And so, though we do not always 
 mention the word "existence," the assertion or denial of 
 existence is implied in nearly all of the propositions with 
 which we usually concern ourselves. 
 
 ''Existence" consequently is a most important word. The 
 term we may avoid mentioning, but the concept will be lurk- 
 ing in the background. We may attempt to develop a doc- 
 trine that shall be purely a Gegenstandstlieone, but as soon 
 as we advance beyond the most fundamental concepts, our 
 assertions will be equivalent to assertions predicating an 
 existence or non-existence of some sort. A concept so im- 
 portant in our thinking certainly merits our attention. And 
 It merits our attention all the more because its meaning is 
 so vague and chaotic. When we first ask ourselves what 
 the difference is between an entity that exists and one that 
 does not, no clear and unambiguous answer comes to our 
 minds. We seem to be given no additional information 
 about an entity when we are told that it exists. Indeed there 
 have been many writers who have said that to think of an 
 entity and to think of it as existing is one and the same 
 thing. They have been able to put forward such a thesis 
 because the word as we first meet it seems to be almost com- 
 pletely without meaning. If we look to the ordinary usage 
 of words to solve our problem, we look in vain. We can 
 
 10 
 
 find little meaning in the word simply -by observing how it is 
 used in our every-day language, for in common speech the 
 word "existence" is applied in a most haphazard manner. 
 We look through a microscope and we say that what we see 
 exists; we press our eyeballs and we say that what we see 
 does not exist. In both cases we are supplementing our 
 native powers of observation and gaining an experience that 
 contradicts the experience we gain through exercising our 
 unaided vision. The two operations have nearly all char- 
 acteristics in common. And yet we say that the one opera- 
 tion puts us in contact with existing entities and the other 
 with non-existing entities. As to the existence of other en- 
 tities, such as God, we find the greatest diversity of opinion. 
 And we find little more than the dogmatic assertion of exist- 
 ence, or the equally dogmatic assertion of non-existence. 
 We find no simple definition of existence by the application 
 of which we might tell whether God does, or does not, exist. 
 
 The use of the term is in the highest degree chaotic. Two 
 entities may stand in practically the same position with 
 regard to all the marks that w^e might take to be marks of 
 existence; and yet one may customarily be called existent 
 and the other non-existent. Furthermore, the term at first 
 sight lacks meaning to such an extent that the proposition : 
 "the Pope exists" is only slightly more informative than the 
 proposition : "the Pope is an Abracadabra" ; that is to say, 
 "existence" is almost as meaningless a predicate as the 
 typically meaningless term Abracadabra. It is evident, con- 
 sequently, that if the term is to be used at all, it must be 
 defined. The concept which it represents must be given 
 some more definite meaning; it must be rendered richer in 
 content. It is this that we shall attempt to do. We shall try 
 to find out what some of the more important philosophers 
 have meant by "existence" when they have used the term. 
 And as a result of such historical inquiries we shall see what 
 may be said as to the way in which the term may or should 
 be defined. 
 
 When we come to consider the term "existence," there are 
 a whole host of related terms that come to mind. There are 
 "essence," "being," ''subsistence," and "reaHty" in English; 
 "Sein;' "Dasein;' 'Wirklichkeit/' and ''Realitaf in Ger- 
 man; and a similar array of terms in other languages. 
 
 II 
 
"Essence/' I suppose, may be described as that part of an 
 entity which remains when we remove existence from the 
 concept of it. It is a term that should be quite easily dis- 
 tinguished from "existence," and so will cause us little con- 
 cern. To be sure, in the writings of some philosophers, 
 notably in the case of Spinoza, the relation between essence 
 and existence offers difficulty. But the source of confusion 
 is the failure to give "existence" a definite and limited mean- 
 ing: when "existence" is made meaningful, it is not difficult 
 to see what is meant by "essence." 
 
 "Subsistence" is a term that has recently become popular. 
 And deservedly ; for it is a term that is quite serviceable. 
 It is that which may be predicated of any entity, of an exist- 
 ing entity as well as of a non-existing one. "Being" is 
 frequently used in the same sense. For example, Russell* 
 says : " 'Being' is that which belongs to every conceivable 
 term, to every possible object of thought — in short, to 
 everything that can possibly occur in any proposition, 
 true or false, and to all such propositions themselves . 
 Numbers, the Homeric Gods, relations, chimeras and four 
 dimensional spaces all have being, for if they were not 
 entities of a kind, we could make no propositions about 
 them. Thus being is a general attribute of everything, and 
 to mention anything is to show that it is." It is in this sense 
 that I shall use "being;" I shall treat it as a synonym of 
 "subsistence." And when I find it given a more limited 
 scope, I shall assume that the writer with whom I am dealing 
 is talking about what I am calling "existence." 
 
 Existence is not the same thing as subsistence or being. 
 If it were, non-existence would, of course, be a self-contra- 
 diction ; it would be as impossible to think : "Jupiter does 
 not exist" as it is to think : "Jupiter does not subsist." And 
 yet we frequently find assertions that imply just this identi- 
 fication of existence and subsistence. According to Hume, 
 "there is no impression nor idea of any kind of which we 
 have any consciousness or memory, that is not conceived as 
 existent."^ Existence, accordingly, is a predicate that be- 
 longs to any object of which we may think. It belongs to 
 the centaur just as subsistence does. And just as a non- 
 subsisting object is a self-contradiction, so, it would seem, 
 
 * Russell: Principles of Mathematics, vol. i, p. 449. 
 'Hume: Treatise of Human Nature; Bk. i, Pt. 2, Sec. 6. 
 
 12 
 
 is a non-existing one. "Existence," in short, comes to be 
 equivalent to "subsistence." Now, when we are engaged 
 in introspection, we seem to find some basis for making 
 "existence" as broad in extension as "subsistence," and as 
 devoid of intention as the latter is. "Existence" is a rather 
 meaningless word as we first meet it, and adds almost noth- 
 ing to the concept to which it is joined. But it is not to be 
 assumed that "existence" is to be left in this condition. 
 We have enough words to act as universal predicates ; "ex- 
 istence" should be redeemed and rendered definite. 
 
 Indeed, it is usually recognized that "existence" is not to 
 be attributed to everything. And so we are only following 
 customary modes of speech when we say that some entities 
 exist and others do not exist. However, not only is it cus- 
 tomary for "existence" to be given a limited application; 
 it is absolutely necessary that "existence," or some term 
 like it, have a limited application if we are to have any 
 knowledge at all. For nearly all of the propositions we use 
 can, as we have seen, be transformed into existential propo- 
 sitions. If now to attribute existence to an entity is to 
 attribute nothing, or nothing specific, to it, all of these 
 existential propositions are meaningless. And some of 
 them become worse, for those in which non-existence is 
 predicated of some entity come to be self -contradictory. 
 Knowledge depends, consequently, on the use of a concept 
 like "existence" that shall be a quality of only some entities 
 and not of all. We must, therefore, reject the view that 
 identifies "existence" with "subsistence." "Subsistence" or 
 "being" is one thing, and "existence" is another. 
 
 Of the other terms I have mentioned it is hard to dis- 
 tinguish one from another. There have been attempts to 
 distinguish "existence" from "reality," but the majority of 
 writers I have examined make no clear distinction between 
 the two. In German, where there are the four terms : '^Ex- 
 istenz:' ''Dasein;' ''Wirklichkeit;' and "Realitat:' there 
 have ]yttn similar attempts.^ But the result has been merely 
 a collection of verbal distinctions that finds no justification 
 in the previous history of these words. I shall, therefore, 
 group them all together and shall use "existence" and 
 "reality" interchangeably. I shall be dealing with the con- 
 
 " Friedrichs : Bcitrdf^c 211 einer Geschichtc iind Thcorie des Hxis- 
 tential-urteils, p. 13. 
 
 13 
 
cepts that are called concepts of ''existence," those that are 
 called concepts of ''reality," and those that are called con- 
 cepts of "being," where "being" does not mean mere sub- 
 sistence. And I shall call them, now concepts of "exist- 
 ence," and now concepts of "reality." If, in the absence of 
 any generally recognized distinction between these terms, 
 I were to be especially careful which term each individual 
 author preferred to use, I should be laying too much stress 
 on peculiarities of terminology. And I should end by 
 drawing up a list of senses in which "reality" has been used 
 that would almost exactly duplicate a list of senses in which 
 "existence" has been used. For me, then, "existence" 
 means "reality," and "reality" means "existence." And 
 the concepts which I shall principally be engaged in dis- 
 cussing, I shall call now by one name and now by the 
 other. 
 
 It is settled, then, that we shall talk about "existence" or 
 "reality," and that what we shall mean by these terms will 
 not be mere subsistence. We may be asked, however, with 
 what sort of existence we are concerned. There is existence 
 in the realm of mathematics, existence in the realm of 
 scientific theories, existence in the pliysical world, existence 
 merely in thought. We shall be told, in short, that "exist- 
 ence** is a term that by itself is quite incomplete. When- 
 ^ever we predicate existence, we must predicate existence in 
 some realm, if our assertion is to be useful. All entities 
 exist, but they do not exist in the same realm. Therefore, 
 it is nothing to predicate existence, but it is useftil to predi- 
 cate existence in some specific realm. Zeus does not exist 
 in the physical world : he does exist in Greek mythology. 
 Consequently, wlien we read "Zeus exists," we do not 
 know whether the proposition is true or false, for we do 
 not know what realm of existence is understood. If the 
 world of Greek mythology is referred lo. the proposition is 
 true: if it is the physical Vvorld that is meant, it is false. 
 At times some of these realms of existence are graded 
 according to their importance. Fritz Medicus,"^ in dealing 
 with mathematical objects, distinguishes between the reahn 
 of concrete objects, the realm of abstract objects, and the 
 realm of pure extension. To say that an entity does not 
 
 ' Mediais: Bcmcrkumjcn cum Problem dcr Exisieuz mathcmatischer 
 Gegenstande ; Kantstiidien, 1914, p. i. 
 
 14 
 
 i 
 
 , 
 
 - 
 
 exist is to assert that it exists only in a realm lower down on 
 the scale than that realm about which we are talking. If 
 we say the number two does not exist, we mean that it does 
 not exist in the realm of concrete entities, but has only an 
 abstract existence. If we say four-dimensional space does 
 not exist, we mean that it does not exist in the realm of ab- 
 stract entities, but belongs only in the realm of pure exten- 
 sion. At other times the various realms of existence are 
 not graded. It is insisted that existence is always relative 
 to .some realm of entities, to some "universe of discourse" 
 or "situation," but there is no attempt to compare these 
 various "universes of discourse" and these various "situa- 
 tions." 
 
 There is for F. H. Bradley this multiplicity of worlds. 
 There is for him "my present actual world, and the am- 
 biguous existence of what has been and is about to be. 
 There are the worlds of duty and of religious truth, which 
 on the one side penetrate and on the other side transcend 
 the common visible facts. . . . Above the sensible sphere 
 rises the intellectual province of truth and science, and, 
 more or less apart from this, the whole realm of the higher 
 imagination. Both in poetry and in general fiction, and 
 throughout the entire region of the arts and of artistic per- 
 ception, we encounter reality. . . . Because there are many 
 worlds, the idea which floats suspended above one of them 
 is attached to another. . . . And hence when an idea floats 
 above, or is even repelled by, one region of the world, there 
 is available always another region in which it inheres and 
 to which as an adjective it is attached."^ There is, in short, 
 always some world, some universe of discoiu'se, in which 
 a given entity is real. 
 
 It must be admitted that there is something attractive in 
 this view. It enables us to go behind any question as to 
 an entity's existence by making a distinction. We can say : 
 "In one realm it exists; in the other it does not exist." 
 And this usage finds some support in everyday langtiage. 
 We do sav of a lunatic that his million dollars exist in his 
 head. But in our ordinary speech we also recognize an 
 existence that is absolute existence. If we ask the man in 
 the street whether the lunatic's million dollars exist, he will 
 answer immediatelv that thev do not exist. He will not ask 
 
 8 i:* 
 
 F. H. Bradley: Essays on Truth and Reality, pp. 31-32. 
 
 15 
 
us to specify which reahn of existence we are discussinP- 
 
 ■^ Co;r;:ahfo"; '^-t ^^^"- °*' -^-:~s 
 
 course th^f ic ft. • always one universe of dis- 
 
 course that is the universe of real obferfQ q.. j . 
 
 wuh Bradley. There is. he reco.; IS^ a s en^ tti ch 
 e tsTa™,:,'"'^^- 'fr'^ ^"" ^'^^''""- ™^"e?s and d:;.' i' 
 
 extet Tetruro„r:' i:i;:.f ;,e:.j^„^-^: r ^ t' 
 
 worlds of unreal, o^ only part.anJC, enm£ 1^ ilt: 
 h.m a reahn of reality in xvlnch there is absolute real tv A 
 a ..atter of fact, we only obfuscate the c^ stionfto a, 
 uuuy s e.x.stence. and n.ake a solution more remote when 
 ve talk- about various realms of existence. VVl en .'-e a k 
 uhether an entUy exists, we are asking, u hether i exSL n 
 
 ence ni thought or ui pure extension ,loe. not concern „■ 
 And It ,s to be noticed that when we insist > ,T 
 account various reahns of existence, tet^fa To" 'e" 
 meanmg of existence is no, solved, but is r'epla e" Uy hos 
 of new questions. We have now to ask what k neam bv 
 
 ■n the r::t.Mn;r:r:! .::;--,-- ,;^ -'-.e 
 
 reahn of .ciencc It seem- , T ■ ^•^'■''^■"'-'-' '" t''^ 
 
 concentr-.,/ 't;^c^"'^ to be u'lse.st. consequentlv, to 
 
 iiitans to exist in the one r<")lTii f]-..,t • 
 
 , . ^ ^^'^^ uajin that ]> most iniuort'inr in 
 
 ;;;': "h": ^^""' ■"' '"^' ""■'^•^■'^ ^■•^'- "■ - ^'" - er 
 
 e es^vl ■^'"^' °^";.'"^'"'-' ^"'' -^- l'l^-<-'l.v to content our- 
 ^ehes u th si^ccous d.stmctions that do not otter a solution 
 but merely cover up the difticultv. --oiution 
 
 I shall be e.igaged in the main, therefore, in considering 
 « hat .t means to extst absolutely, to exist ,n general Whe,, 
 
 ent t es hat ,s not a co lecfon of four entities does not exist 
 and th s ,s what 1 have said the as.sertion that two and 
 t vo are four cotnes to-it is about this .sort of existence 
 It scents to me. that we are talk.ng. We are .savm^ t"u t' 
 -ch a cotnb.nat.on does «., exist absolutely, .n'the way 
 "' -'>^cn the world of Platonic tdeas or the Inminiferoul 
 
 9 r f • ■% 
 
 ^ Ibid, p. 31. 
 
 "F H. P>radley: Appearance and Reality, ch. 24. 
 
 16 
 
 ,< 
 
 ir% 
 
 th 
 
 ether has been asserted to exist absolutely. It is possible, 
 of course, to say that what is asserted of such a combination 
 is non-existence in the realm of mathematical entities. But 
 it seems to me unnecessary always to be bringing in a refer- 
 ence to some particular universe of discourse. Existence 
 in an a'bsolute sense can, we shall see, be made meaningful. 
 And when this has been done we can determine how many 
 realms of existence are to be recognized and what each of 
 these types of existence is to mean. After all, thev are all 
 species of one existence that is absolute existence. And if 
 we insist on always being mindful of these various species, 
 we complicate our subject unnecessarily. I shall, conse- 
 quently, consider all assertions of existence merely as asser- 
 tions of an absolute existence. Existence of some kind, and, 
 consequently, absolute existence, is implicitly being predi- 
 cated in nearly all judgments. Notwithstanding this fact, 
 existence as generally used is quite deficient in meaning. It 
 follows that this state of affairs must be remedied, that we 
 must learn what absolute existence is, in order that these 
 judgments may be meaningful. 
 
 In order, too, that we may solve those questions that are 
 explicitly questions of existence, we must know what 
 absolute existence is. Let me suppose that as I walk along 
 a road I see something lying on the ground ahead of me 
 that I take to be a snake. As I approach nearer, I conclude 
 that I was mistaken and that what lies on the ground is 
 really a piece of cord. I am interested in determining what 
 really is there. Is it a snake or a piece of cord? When, 
 now, I ask myself whether or not a snake is really there and 
 whether or not a piece of cord is really there, I am dealing 
 with the same question. I am dealing with the same alterna- 
 tive: snake or cord? For to decide in favor of the cord 
 is to judge that the snake that seems to be there is unreal 
 and the cord that seems to be there real. Whenever we ask 
 7vhat is real, we are asking whether some seeming entity is 
 real. And whenever we ask whether an entity is real, we 
 are asking what is real. For when the question is answered 
 in the affirmative the content of reality is different from 
 what it is when the question is answered in the negative. 
 To ask whether an entity is real is consequently no less prac- 
 
 17 
 
tical and no more artificial than to ask what is real.^' To 
 answer either we need to know w^hat it means to Ixi real. 
 We cannot tell whether an entity deserves the predicate 
 "real" unless we know the meaning of this predicate. And 
 we cannot determine the content of reality without reality 
 being a meaningful term. It is quite necessary, then, that 
 the term he made meaningful. And this task of making 
 explicit itF meaning — or rather its meanings — is the one to 
 which we are to address ourselves. 
 
 *^ Dewey: Essays in Expcrimenlul Lo^ic, p. 8, note. 
 
 
 i8 
 
 CHAPTER II 
 
 The Real as the Permanent 
 
 No question in Occidental philosophy, so far as we know, 
 is older than this question: What is it to be real? It was 
 recognized from the start that in it lies the beginning of 
 philosophical wisdom. For when the Milesians looked out 
 and found themselves confronted by a world of infinite 
 variety and ceaseless change, they asked themselves what the 
 meaning, the ''nature," of all this is. That is to say, they 
 asked what is fundamental and 'Veal" in all this hetero- 
 geneity. They desired to separate out the w^heat from the 
 chaff, to find out what truly exists, so that they might cling 
 to this and disregard the rest. For, as long as we occupy 
 ourselves with illusory objects we have no knowledge; it 
 is only when we limit our attention to objects that are "real" 
 that we can build up knowledge and science. This funda- 
 mental truth was recognized in the early stages of Greek 
 philosophy ; and so these early Greeks set about finding out 
 wliich one or more of the objects confronting them w^as 
 "real." 
 
 The real for them is that w^hich is permanent and abiding. \ 
 "As Anaximandros and most of the physicists say," it is 
 that which "is immortal and indestructible."^- It is that "of 
 which all things consist, the antecedent from w^hich they 
 have sprung and into which they are finally resolved. "^^ 
 That which is permanent and enduring alone is real; the 
 rest is mere unintelligible becoming. This is the interpre- 
 tation that has been put upon these Ionics by Aristotle and 
 most subseqtient philosophy. There have been some,^^ how- 
 ever, who have asserted that what these Milesians wanted 
 to know when they inquired into the "nature of things*' 
 was not the permanent entity underlying all the changing 
 ones, but was the origin of, and process implicit in, the 
 changing world we usually have before us. They are said 
 
 '* Aristotle : Phys. iii, 4 ; 203b 7. 
 "Aristotle: Met. i, 3; 983b 6. 
 
 " Woodbridge : The Dominant Conception of the Earliest Greek 
 Philosophy; Philosophical Review, vol. 10 (1901), p. 359- 
 
 19 
 
 " 
 
to have wanted to know how the worlfl works, how and 
 whence becoming becomes. Rut they are more usually held 
 to have been eni^aged in in([uiring into the nature of the 
 permanent fact amid so much chani^e. in attempting to 
 discover the underlying reality which — as real — must he 
 permanent. However that may be, when we come to Par- 
 menides we meet the clear and unequivocal assertion that 
 that which is "real" is permanent and abiding. His muse 
 tells him: "There is left but this single path to tell thee of: 
 namely, that being is. And on this path there are many 
 proofs that being is without beginning and indestructible; 
 it is universal, existing alone, immovable and without end; 
 nor ever was it, nor will it be, since it now is, altogether, 
 one and continuous. "^^ Other entities appear and then 
 vanish; that which is real persists forever. It is in this 
 sense that "reality" is used, not only by Parmcnides, but 
 by almost all the Pre-Socratic philosophers. From Par- ^ 
 menides to Anaxagoras the real is that which persists 
 unchanged, unaffected by the lapse of time. There is dis- 
 agreement as to the number of such permanent entities and 
 the qualities that they possess, but there is no one to deny 
 that whenever such permanent entities are found they are 
 to be called "real." W'e have here in the concept of per- 
 manence, consequently, what so far as we can determine is 
 the original meaning of the term "reality." All other mean- 
 ings of "real" have supervened upon it and are probably in 
 some way derived from it. ^ 
 
 But why should stich unanimity give way to subsequent 
 diversity? Primarily it is because "permanence" and its 
 antithesis "change," like most concepts, are stibject to analy- 
 sis, and on analysis they each yield new meanings which 
 are not equivalent to one another. Some particular kind of 
 permanence is hit upon, and reality is made to consist of 
 entities that have as an attribute this sort of permanence; 
 or some particular kind of change is hit upon, and unreality 
 is made an attribute of entities that exhil)it this sort of 
 change. One way in which an entity can change is by 
 appearing in one form to one person and then appearing in 
 a changed form to another. And so entities that in this 
 way change their appearance from person to person, or 
 
 '' Parmcnides : On Nature, vv. 57-63. 
 
 20 
 
 I 
 
 i 
 
 change tlie way in which they appear to the same person 
 from time to time, come to l)e called unreal. They are not 
 permanent; they vary from subject to subject and from 
 tmie to tmie. Consequently they are unreal. Quite early 
 does relativity thus come to mean unreality. "By use there 
 is sweet." says Democritus, "by use there is bitter: by use 
 there is^warm. and by use there is cold; bv use there is 
 C(jlor. But m sooth there are atoms and the void.'"* Now 
 when permanence has come to mean freedom from this sort 
 of relativity its meaning has changed. What we usually 
 thmk of as permanence is a qualitv in the object itself and 
 has no relation to the number of ways in which the object 
 IS perceived. And so when reality is limited to objects with 
 this new sort of permanence, the term is being used in a 
 new sense. 
 
 Or, again, permanence is made to consist in a sort of 
 logical permanence, in immutabilitv of attributes. An entitv 
 cannot be permanent if it has a quality that varies according 
 to the entity with which it is compared. A tall man, to be • 
 permanent and real, must always be tall; he cannot be tall 
 as compared with one man and short as compared with 
 another. -Slx." to be real, must always be "more by a 
 half; ,t cannot be "more by a half" as compared with 
 tour and less by a half as compared with twelve '^ Such 
 an opinion is sometimes attributed to Protagoras He and 
 other Sophists are sometimes said to have held all entities 
 unreal whose qualities are in this sense relative. For them 
 that IS to say, there were included among unreal entities 
 those with this .sort of relativity. 
 
 Cleavage along another line" is more fundamental Two 
 very different senses of "real" are the natural result of two 
 very different senses of permanence. In one sense an entitv 
 IS permanent that persists through time, that exhibits a per- 
 durance that knows no end. But using the word in another 
 sense, entities have been called permanent that are entirely 
 outside the realm of time. Being entities to which time is ' 
 not applicable, they too are free from change and in this 
 sense permanent. During the greater part of Greek phi- 
 losophy permanence means perdurance. The permanent 
 and real entities are those that were at the start and will 
 
 " Democritus, fr. 125. 
 
 "Plato: Theaetatiis. 154, 157-160. 
 
 21 
 
continue forever. But the "Bein^" of Parmenides abides 
 in the sense that it is out of time entirely. From the 
 "Timaeus" also we can derive the notion of a Being- to 
 whom time is not applicable. And it is this timeless sort 
 of permanence that, partly as a result of the ''Timaeus." 
 was conceived by some of the Schoolmen to be an attribute 
 of real Being. The essence of reality, however, cannot be 
 both perdurance and timelessness. The two concepts are 
 at war with one another. We are face to face with two 
 contradictory notions of what it is to be permanent and 
 real. "Real" is applied, now exclusively to those entities 
 that perdure, now exclusively to those that are timeless. 
 When it is used to point to perduring entities alone, the 
 timeless ones are unreal, for they manifestly do not per- 
 dure. This is one of the points Plato makes against the 
 timeless Being of Parmenides. It is unreal because it is 
 not in time as an entity must be to be real.^^ When, on the 
 other hand, ''real" points to timeless entities alone, the 
 perduring ones are unreal. The two senses quite evidently 
 are mutually exclusive. 
 
 And so ''reality" has come to mean very different things. 
 Used in one sense it is equivalent to timelessness ; used in 
 another it is equivalent to perdurance through time. Fur- 
 thermore, the meaning of permanence is often modified so 
 that what we have in mind is not the unchanging character 
 of the object in itself, but the stability and permanence of 
 its relations to the sentient organism or to other entities. 
 As we have seen, at times entities that appear differently 
 from subject to subject are called unreal and at times 
 entities with qualities that vary, entities with relative quali- 
 ties, are unreal. But while some have identified reality 
 with permanence when its meaning is modified in one of 
 these two ways, others have identified reality with perma- 
 nence taken in its normal sense. That is, they have meant 
 by a perm.anent entity one that, regardless of the changes 
 in its relations, is itself unaffected by the lapse of time. 
 And using "real" in this sense, the relativity of sense-per- 
 ceptions is no sign of unreality. Everything depends upon 
 the sense in which the word is being used. 
 
 Let us suppose that by permanence we mean this intrinsic 
 permanence, and that it' is with this that we are identifying 
 
 '* Plato: Parmenides, 141, 152. 
 
 22 
 
 • 
 
 reality.^ Which, then, are the entities that we are calling 
 ''real"? It is possible that our antithesis between the per- 
 manent and the impermanent refers to a distinction among 
 the objects of our perception. Mountains, rocks and tower- 
 mg trees seem to last indefinitely. They are there to greet 
 our vision whenever w^e turn our eyes their way. But frail 
 flowers and beautiful sunsets must be enjoved when they 
 are given, for they are short-lived and soon pass awav. 
 It is possible, accordingly, that we are using "real" to point 
 to the mountains and "unreal" to point to the sunsets. But 
 if w^e use these words in this way, we are, it seems to me, 
 diverging quite far from the usage of common speech! 
 There is to-day, so far as I can judge, no tendency in com- 
 mon si>eech to call mountains real rather than sunsets, and 
 Gothic cathedrals real rather than soap bubbles. It is not 
 only, however, that such a definition of reality would give 
 the term an unusual meaning. The more serious objection 
 is that such a definition is as yet by no means precise. For 
 there is no sharp distinction between short-lived entities 
 and long-lived entities : on the contrary, there is a gradual 
 transition from the lightning flashes and soap bubbles to 
 the mountains and stars. Consequently, permanence is, f 
 after all, relative: and to say merely that an entity to be 
 real must be permanent is not to tell us just how long an 
 entity must last to be real and just how^ evanescent it must 
 be to be unreal. It is a definition, in short, that cannot be 
 applied with any definiteness and precision. 
 
 But when we contrast the permanent with the imperma- 
 nent, we may not be pointing to the qualities that are given 
 us in perception. Instead we may be referring to a "sub- 
 stance" that stands behind these qualities. This substance 
 may be conceived to be permanent, and, since permanence 
 means reality, this substance may be conceived to be real. 
 With a view of this kind we can associate the name of ' 
 Herbert Spencer. For Spencer objects consist of qualities 
 that change and an unsensed substratum that endures, v 
 "The most conspicuous contrast presented in the vivid 
 aggregate as a whole, as well as in each of its parts," he 
 says, "is the contrast between that which perpetually changes 
 and that which does not change, between each ever-varying 
 cluster of vivid states and their unvarying nexus. This 
 transcendent distinction needs a name. I must use some 
 
 23 
 
mark to imply this duration as disting-uished troni this 
 transitoriness — this permanence in the midst of that which 
 has no permanence. And the word existence, as apphed 
 to the unknown nexus, has no other meanini;-. It expresses 
 nothing beyond this primorchal fact in my experience."^ ^ 
 It is to be questioned, however, whether "existence" does 
 ofenerallv refer to this nexus. No dou])t "existence" can 
 be used to refer to substances and "non-existence" to refer 
 to qualities, but certainly when we use these words in this 
 sense we are giving them an unusual meaning. When we 
 say that an entity does not exist, we certainly do not usually 
 mean that this entity is merely a quality and not a substance. 
 Furthermore, such a definition does not enable us to dis- 
 tinguish between such substances as phlogiston and the 
 gods and goddesses of mythology on the one hand, and 
 such substances as trees and men on the other. We can 
 think of a persisting substratum called Diana that supports 
 an ever-changing group of qualities. Now she is hunting 
 with bow and arrow, now she is driving the silver chariot 
 of the moon. She is conceived as a substance just as this 
 desk is. And so a definition of this kind does not assist us 
 in makme a distinction between the two. To be sure, a 
 similar objection can be made to any definition of existence. 
 No definition will be able to determine the content of exist- 
 ence with absolute precision."^ But this definition leaves 
 the content undetermined just where we are looking for 
 help. What we want is a definition that wdll make a dis- 
 tinction between dififerent kinds of substances, one that will 
 definitely rule out such entities as Diana and Zeus and 
 fairies and centaurs, as they are usually conceived. We 
 want to be told what kinds of substances are real ; we do 
 not want to be told simply that an entity to exist must be a 
 substance of some sort. 
 
 One of the most important of all the senses of reality has 
 arisen, most probably, directly out of this notion of i)er- 
 manence that we have just been discussing. To be real an 
 entity must persist. But to persist means to i)erdure through 
 all time. Consequently, a real entity must persist at the 
 times when I am not thinking of it. It must, that is to say, 
 be independent of my consciousness of it. And so reality 
 
 '" Herbert Spencer : Principles of Psychology, 2d ed-, 1877. §467- Cf. 
 
 also § 59- 
 
 --' Infra, cb. 5- 
 
 24 
 
 \ 
 
 comes to mean independence of consciousness and inde- 
 pendence of consciousness comes to mean reality. This 
 meaning of reality is already peeping out of a passage in 
 Democritus. "By the senses," he says, ''we in truth know 
 nothing sure, but only something that changes according 
 to the disposition of the body and of the things that enter 
 into it or resist it,"-^ If we could get at something that con- 
 tinues on its way regardless of the percipient, w^e should be 
 getting at reality. But we cannot arrive at such real enti- 
 ties through the senses ; through them we can only arrive 
 at entities that depend upon the body. And it is the inde- 
 pendent ones alone that are real. This linking of reality 
 with independence has persisted. At the present time we 
 often find reality described in terms of independence. For 
 there is a widespread conviction that unreal entities do not 
 perdure through those instants at which they are not objects 
 of thought, while real entities perdure through those in- 
 stants at which they are not being thought of as well as 
 through the instants at which they are objects of con- 
 sciousness. 
 
 Now, when a real entity is described as one that is not 
 dependent on my thought of it, or on any thought of it, 
 there is one class of entities that is definitely being called 
 unreal. These are the ''ideas" that have been said to be 
 the objects of our consciousness. When I think, I think an 
 idea that may or may not have an entity beyond to which it 
 means to correspond. But an idea of this kind does not 
 persist when I am not thinking of it: or, in any case, it 
 does not persist when no consciousness is thinking of it. It 
 is conceived to be dependent on consciousness, whether on 
 my individual consciousness or on consciousness in general. 
 If, then, to exist is to be independent of consciousness, these 
 "ideas" do not exist. It follows, consequently, that one 
 cannot be an epistemological dualist proclaiming the exist- 
 ence of ideas and at the same time be using "existence" to 
 mean independence. 
 
 This definition, however, while it determines the onto- 
 logical status of "ideas" in the epistemological sense, is 
 open to the same objection as that which we made against 
 the identification of existence with permanence. After all, 
 it is in order to distinguish between centaurs and fairies 
 
 \ 
 
 »\ 
 
 Democritus : fr. 9. 
 
 25 
 
and horses and men that we feel called upon to use the 
 terms "existence" and ''non-existence." A definition of 
 existence m terms of independence, however, does not help 
 us in the least to make distinctions of this kind. Ivanhoe 
 and Walter Scott pretend to be in the same position so far 
 as independence of my consciousness is concerned. Diana 
 and all the other deities of Greek mythology were taken 
 by the orthodox to be independent. They were ''immortal" 
 and were conceived to be entities that would persist when 
 the consciousness of mortal man is no more. Phlogiston, 
 again, pretended to be independent of consciousness. It was 
 conceived as an element that persists through those instants 
 at which no one is thinking of it, as well as through those 
 instants at which it is an object of thought. In fact, un- 
 less we have been very much affected by idealism, all entities 
 come to us pretending to be independent. And if the thesis 
 of idealism is ever before our minds, they all alike come 
 seeming to be dependent on consciousness, — the island 
 Madagascar as well as the island Atlantis. A definition in 
 terms of independence, consequently, does not separate out 
 the entities we are primarily interested in excluding from 
 existence. Indeed, some of those who most loudly insist 
 on the independence of real entities recognize that a defini- 
 tion in terms of independence would be unsatisfactory. 
 Perry, for example, in order to support the view that real 
 entities are independent of consciousness, deals at length 
 with the notion of independence. And yet in developing a 
 theory of independence he disclaims any intention to define 
 reality in terms of independence.^- 
 
 The meanings of "real" we have thus far been led to 
 consider have all been due to ambiguities in the notion 
 of permanence. There have been various meanings attached 
 to the word "real" as there have been various senses in 
 which permanence has been understood. But the senses of 
 the word that we are now to consider are connected in a 
 different fashion with the concept of permanence. The 
 different senses to which we are now to attend are due, 
 not to ambiguities in the term itself, but to differences 
 of opinion as to w^hat is concomitant with permanence. 
 Permanence is held to be co-extensive now with one quality, 
 now with another. And, since permanence is equivalent 
 
 "The New Realism, p. 117. 
 
 26 
 
 \ 
 
 
 a 
 
 to reality, reality will, of course, be held to be co-extensive 
 now with one quality and now wath another. But if it is 
 co-extensive with these qualities, it may be defined in terms 
 of these qualities. And so the notion of permanence drops 
 out of sight and reality is held to be equivalent, directly, 
 to these qualities newdy come upon the stage. 
 
 It is in this way, I think, that we must explain the genesis 
 of the notion that entities known by "reason" are alone real. 
 These entities were probably called "real" in the first in- 
 stance, not because they were known by reason, but because 
 they were permanent. And probably before "unreal" came 
 to mean "known through sense perception," objects per- 
 ceived by the senses were called unreal because they were 
 subject to change. The entities that change continually, 
 among which nothing abides, were identified with the en- 
 tities of which we become conscious through sense percep- 
 tion; and those that remain unchanged were identified with 
 those that we know through reason. And so the real came 
 to mean the intelligible and the unreal the sensible. As we 
 have found Democritus saying, "By the senses we in truth 
 know nothing sure, but only something that changes accord- 
 ing to the disposition of the body and of the things that 
 enter into it or resist it."^"^ Intelligible entities, it is implied, 
 do not vary with the subject; they remain unaltered and 
 persisting, and consequently they alone are real. The Pla- 
 tonic dialogues are the great source of inspiration for this 
 identification of the real with the intelligible. There we 
 find in abundance passages in which the objects of the 
 intellect, the Ideas, are eulogized and called "real." and 
 in which the objects of the senses are called "unreal. "^^ A 
 realism or rationalism of this kind comes to the Middle 
 Ages with Saint Augustine. The author of the "City of 
 God" show^s the same partiality for these intelligible entities. 
 The mind, says he, "is disabled by besotting and inveterate 
 vices not merely from delighting and abiding in. but even 
 from tolerating His unchangeable light, until it has been 
 gradually healed, and renewed, and made capable of such 
 felicity."-^ Did we, how^ever, have a mind that was entirely 
 
 M 
 
 Democritus : f r. 9. 
 '^Cf. esp.— Phaedo : 65, 74; Republic: 525. 
 St. Augustine— City of God; Bk. 11, Sec. 2. 
 
 27 
 
 » 
 
purified, then we should have a mind capable of contem- 
 plating the immutable spiritual things that are stipremely 
 real and abidmg. 'Jliere are. in fact, two ways in which 
 materia] things may be known. They "are known in one 
 way by the angels in the Word of (^orl, in which are seen 
 the eternally abiding causes and reasons according to which 
 these things were made, and in another way in which these 
 things are seen as they are in themselves/ In the former 
 way, they are known with a clearer knowledge ; in the latter 
 they are known with a dimmer knowledge, a knowledge 
 rather of the bare works than of the design. ''2« xt is 
 knov/ledge of the spirit that, implicitly, is true knowledge, 
 knowledge of the spirit that alone contemplates true reality' 
 Scattered through the Middle Ages we find marks of this 
 Platonic doctrine. That "in which there is any mutable 
 element," says Saint Anselm,^' '*is not altogether what it 
 IS. . . . And what has a past existence which is no longer, 
 or a future existence which is not yet,— this does not 
 properly and absolutely exist." It is God who truly is, God 
 who is pure spirit and immutable. Similarlv in the seven- 
 teenth century we find the distinction made between the 
 spiritual things that are truly real, and the sensible things 
 that stand on a lower plane. We find the distinction made 
 by Descartes, especially in the small "Search after Truth" ; 
 we find it made by Malebranche and the more pious Car- 
 tesians, by Spinoza and by Leibniz. 
 
 Thus, the word "real" is used to point to intelligible 
 entities. The only real entities are those that are known 
 by^ reason. Sensation and imagination are not distin- 
 guished; the entities that are known by either of these 
 faculties are unreal. The same seventeenth century phi- 
 losophers who used the word "real" in this sense frequently 
 employed the phrase: "clear and distinct ideas." To be 
 sure, when this phrase was upon their lips, they usually 
 were talking about judgments. But often we may regard 
 the object held in view to be an entity that is intuited 
 rather than a judgment that is asserted. And when this 
 IS the case, it is the entities that are ''clear and distinct" 
 that are real and those that are obscure or confused that 
 
 n 
 
 Ibid: Bk. ii, Sec. 29. 
 
 St. Anselm: Proslogium, ch. 22. 
 
 28 
 
 k m€ 
 
 
 are unreal. This, however, is not a different use of the 
 word "real" from that which makes "real" and "intelligible" 
 synonymous. For the entities that are referred to as "clear 
 and distinct" are generally precisely those that a/e in- 
 telligible. Mathematical concepts, God, and moral ideals 
 are intelligible entities and they are at the same time objects 
 that are "clear and distinct." They are not clear and dis- 
 tinct in the sense that we can form images of them that 
 have sharp outlines, for these are entities of which we can 
 form no images at all ; but they are clear and distinct in the 
 sense that they can be thought through without contradic- 
 tion. "Intelligible," "clear and distinct," "self-consistent," 
 and "real" — all mean one and the same thing. The very 
 fact that so many terms were used to point to the same 
 class of entities shows how important these entities were 
 held to be. Tliey were the real entities; all others were 
 unreal and at the same time unimportant. 
 
 On the other hand there is a tendency to identify the real 
 and permanent with the stable and reliable. And entities 
 that are present to the senses are then identified with those 
 that are stable. Other entities than these sense-data are 
 mere fancies, cobwebs of the brain that can be swept aside. 
 But these things that are seen and felt are not ephemeral. 
 They cannot be wished away; they are enduring. And so 
 the term "real" comes to mean these sense-data. This, at 
 least, seems to be the easiest way in which to explain the 
 genesis of the view that sense-data alone are real. Whether 
 this view arises in this way or not, however, we find this 
 sense of the term coming into use rather early. In 
 Epicureanism "real" is used to mean sense-data quite evi- 
 dently. To be real means to be present to the senses. It is 
 a sense of "real", moreover, that has remained. In modern 
 philosophy nothing is more common than the use of the 
 term in this sense. From Bacon and Hobbes to James and 
 Bergson there has been a line of philosophers to whom this 
 has been the meaning of "real." And in the course of 
 the extensive history of this sense of reality, new meanings 
 of the term have l)een introduced. Just as in the first place 
 different kinds of permanence were discovered and reality 
 identified with each kind separately, so here there are found 
 to be various ways in which an entity can be present to the 
 
 29 
 
senses and realil)- is tlien identified with each kind of 
 sense-data in tnrn. But the consideration of these develop- 
 ments we shall reserve for another chapter. (Chap. III.) 
 
 Development alont^ another line leads to a rather para- 
 doxical conclusion. We meet no success in looking for that 
 which is permanent. We decide tliat all objects change, even 
 those most nearly stable. But in the depths of our despair 
 we light upon (^ne thing that remains forever, namely, 
 change itself. For if all objects are continually changing, 
 then change itself never ceases. Of course the reference 
 cannot l>e to any specific case of becoming. It would be 
 absurd to hold that a given example of change is an example 
 of permanence, that of the entity under consideration the 
 part that changes remains. It is not the changing entity 
 that is permanent, but the fact of change. What remains 
 unaltered is the laze that all entities change. We find our 
 permanence, consequently, not in phenomena themselves, 
 but in the laws to which phenomena are sul)ject. And so 
 we come to identify permanence and realilv with law and 
 with subjection to law. In Hermann Lotze the meaning of - 
 the term "real" undergoes (piite explicitly a development of 
 this sort. In summarizing one of his discussions, he says: 
 *'\\'e gave up seeking the permanent element of things in 
 a state of facts always identical with itself, and credited our- 
 selves with finding it in the very heart of change, as the 
 uniform import of a law which connects a multiplicity of 
 states into one rounded whole."-'' In this way "real" comes 
 to mean the law to which changing phenomena are subject, 
 the system to which changing phenomena belong. It comes 
 to mean the system of interrelated entities and the inter- 
 related entities that belong to this system. But if the mean- 
 ing of "real" is to be precise and definite, we must elaborate 
 this idea and make further distinctions. However, this too 
 will be left for another chapter. (Chap. IV.) 
 
 There can in this general way be found a common origin 
 for most of the meanings the term "real" has been given. 
 But from this common origin there develops the greatest 
 diversity. Senses of "reality" come into use that are utterly 
 opposed to one another. All of the entities that are "real" 
 if the term is used in one sense are "unreal" when the term 
 
 P-78. 
 
 Lotzc: Metaphysic, Bk. i, ch. 4, § 38. Trans, by Bosanquet. 1884, 
 
 30 
 
 is used with another meaning. We come to the pass where 
 we cannot in the least tell whether an entity is real or not 
 unless we know in which of these many senses "real" is being 
 used. Now^ of these many conflicting senses, which is to 
 be chosen? The sense in which reality is identified wdth 
 permanence has a certain claim due to its priority. But we 
 have found permanence to be an ambiguous concept; and 
 we have found that a definition of reality in terms of per- 
 manence does not assist us in making the distinctions we are 
 interested in making. Furthermore, mere priority certainly 
 establishes no convincing claim. W^e cannot say, because 
 "real" was said to mean the permanent before it was said 
 to mean that which is present to the senses, that "real" is 
 not to be given this latter meaning. All senses of the term, 
 both original and derivative, stand on a par as possible 
 senses in which the term may be used. And when we choose 
 to use the term in one of these senses rather than in another, 
 it is by an arbitrary choice that we decide that the term is 
 to have for us this one meaning rather than some other. 
 
 It is rare, however, to find a writer who realizes that there 
 is something arbitrary in giving the term "real" the meaning 
 he chooses to give it. For the most part, a writer will use 
 the term in one of its senses without noticing the fact that it 
 has often been used in very different senses. "Real," as 
 we have seen, has often been used to mean that which is 
 "intelligible," that which is given to the intellect alone. And 
 this use of "real" has often been accompanied by the employ- 
 ment of many pious and edifying expressions. A noble life 
 is identified with one conversant about the intelligible 
 entities, an ignoble life with one whose attention is limited 
 to sensible entities. Statements of this kind are so fre- 
 quently associated with, and apparently based upon, state- 
 ments that intelligible entities alone are real, that one would 
 think these edifying sentiments could flow out of such a 
 meaning of the term "real." And yet the identification of 
 the "real" with the "intelligible" is merely a matter of 
 definition. These intelligible entities, that by definition are 
 given the predicate "real" also, may of course be held to 
 be important and praiseworthy on other grounds. But to 
 the extent to which they have been held important because 
 they have the predicate "real," we have a right to enter an 
 
 31 
 
objection. For. as we have seen, it i.> only by an arbitrary 
 definition that the intenio:ible entities are called 'Veal." 
 While those whom 1 may c\ill F^atonists are applying the 
 term to these entities and are withholding it from the objects 
 of the senses, other.> whom we shall consider in the next 
 chapter are applyini,- it to the sensible entities, and are not 
 applying it — some of them, at least, are not — to any of the 
 intelligible entities. Since now one has no more right to 
 the term than the other, an author is free to elect that the 
 term shall carry for him one meaning or that it shall carry 
 for him the other meaning. One exercises choice in giving 
 the word one meaning or another; and in this sense there 
 is something arbitrary about the way in which "real," as 
 the result of such a choice, is implicitly or explicitly defined. 
 There are certain methods of approach that are especially 
 likely to make one lose sight of this element of choice. This 
 is particularly the case when an author fixes his attention on 
 the concept tliat "real" symbolizes for him and disregards 
 the word itself that does the symbolizing. These concepts 
 or notions or ideas, we may admit to Ix- fixed and objective 
 — in the n^fi(re of things. Reality in the sense of mtelli- 
 gibility may l)e as objective as you please, and so may reality 
 in the sense in which it is synonymous with sensibility. But 
 it is arbitrary to make the icrni "reality" refer to one of these 
 objective concepts rather than to the other. Accordingly, 
 the meaning of the tenn is not objectively given. The term 
 has been used to refer to any number of different concepts. 
 And so when we make use of it, it will not be clear which 
 of these man\- concepts is meant unless we make it clear. 
 It behooves us, therefore, in employing the term "real," to 
 state the meaning the term has for us. We must lay down 
 a definition of reality, a statement of what it means to us to 
 be real. It will not be sufficient to say of "reality" or ''exist- 
 ence" as Descartes--' does, and as Kanf'^" and others do after 
 him, that it is simple and clear. Nor will it be sufficient 
 to let the matter go with a vague stateinent such as Wolff's : 
 "Existence is the completing of possibility,"'"*^ a statement 
 
 -"•* Descartes ; Reply to the Sixth Set of C^bjections. Oeuvres, cd. by 
 Adam and Tannery, vol. 9. p. 225: Principles of Philosophy. Part i, 
 Prin. 10. 
 
 ■"'Kant: Der einci^ mogliche Beweisyrund zu eincr Demonstrahun 
 des Daseins Gottcs. ll'crke, ed. by Hartenstein, vol. 2. p. 115. 
 ''Wolff: Ontulogia. § 174. 
 
 32 
 
 
 that does not tell us at all which of the many possible senses 
 of the word is being used. We need a definition, either an 
 explicit definition or one implicit in the context, that will 
 show us which sense of the word is intended. Without such 
 a definition we cannot judge of the truth of those proposi- 
 tions of the author's in which the term ''real" appears. With 
 such a definition, on the other hand, not only do such 
 propositions become meaningful, but we have a criterion 
 that enables us to tell which entities are "real" in the sense 
 of the word that is chosen and which are unreal. 
 
 33 
 
CHAFI^ER III. 
 
 The Real as the Sensible. 
 "Real" and "existing" are terms that we are using' all the 
 time to mark off that which is substantial and in a certain 
 sense important from that which is a mere phantasm, that 
 which is a mere subsistent or mere idea. We use these 
 terms with perfect freedom, as if there were a iKtitctly 
 clear and well recognized distinction between the real and 
 the unreal. But. as a matter of fact, there is. as we have 
 seen, by no means such unanimity of opinion as to what 
 objects shall be called "real" and what "unreal." One 
 philosopher assumes that one set of objects are primarily 
 deserving of the predicate "real." another that the term 
 "real" points to quite another set of objects. In the present 
 chapter we shall deal with the opinion that assumes that the 
 term "real" denotes primarily the objects present to the 
 senses. This is but one out of many meanings the term in 
 question has been given ; some have used ''real" to mean 
 ideas that are clear and distinct, others to mean that which 
 IS in time and space, others to mean that which has power 
 or force, others to mean that which we ha\e on divine 
 authority. 
 
 But the view of "reality" that we .>hall deal with in diis 
 chapter applies the term to those entities which it is per- 
 haps most frequently used to denote. And since it is in this 
 sense that the term is used most commonly, when we use it 
 so. we are more than usually likely to forget that in other 
 contexts the term has other meanings. "Real" is said 
 primarily to denote sense-data without anv reason being 
 stated by the author why this rather than some other mean- 
 ing is given the word, without the author giving any evi- 
 dence to show that he is even conscious the word is some- 
 times used in a different sense. And (juite frequently "real" 
 IS used to mean sense-data without any explicit statement 
 in which the term is definitely given this meaning. 
 
 We shall deal in this chapter with those who apply the 
 term "real" primarily to sense-data. Things that we hear 
 and taste and feel and see are real, it is assumed. A table. 
 
 34 
 
 if it is something we see before us, something present to the 
 senses, is real. The reality of God, or of an historical 
 event alleged to have taken place centuries ago, is open to 
 dispute. But this table before me I see ; it is a sense-datum, 
 and that is sufficient to establish its reality on incontestable 
 grounds. Such a doctrine as to the meaning of "real" is 
 usually accompanied by another opinion from which it 
 cannot be easily distinguished. I refer to the opinion that 
 we are first aware of sense-data and that investigation 
 should start out with these simple data and gradually lead 
 up to conceptional generalizations. One is an opinion on 
 the nature of the real, the other an opinion on the order in 
 which the various kinds of objects of thought do and 
 should come before the mind ; one answers a question in 
 ontology, the other a question in psychologv and what may 
 be called methodology. And yet since one opinion often 
 leads one unwittingly to accept the other, and since the 
 opinions with which we are dealing are not always ex- 
 plicitly stated, but must be extricated from statements in 
 which some indefinite sort of priority is attributed to sense- 
 data, it is difficult to distinguish between these two ideas. 
 
 in the Epicurean philosophy, as we find it summarized 
 by Diogenes Laertius, these two principles are both en- 
 dorsed. We read : "Again, the reason cannot pronounce 
 on the senses ; for we have already seen that all reasoning 
 has the senses for its foundation."^- Here we have stated 
 the psychological priority of perceptual objects, and the 
 ontological priority of these objects based upon that fact. 
 In the next sentence the "real" is identified with sense-data 
 in as explicit a manner as we can expect to find. "Reality 
 and the evidence of sensation establish the certainty of the 
 senses; for the impressions of the sight and hearing are 
 just as real, just as evident, as pain." The reality of sense- 
 data, that is to say, is assumed. Sense-data are real because 
 they just are. This may seem puerile, but it is not; there 
 can be no better reason given. For "real" is one term and 
 "sense-data" another. And to use them together in this 
 manner can only mean that they are being used to point to 
 the same objects. The term "real" is being used in the 
 sense in which it means sense-data. "Real" is being de- 
 
 "Diogenes Laertius: Lives of the Philosophers. Translation by 
 Yonge, p. 4^5. 
 
 35 
 
.1 
 
 fined to mean sense-clata : that is the onJv and sttfficicnt 
 reason why sense-data arc real. 
 
 For how can the psychological prioiitv of perceptual 
 objects show their reality? Assuming that I am first aware 
 of percepts and later of concepts, whv should I apply the 
 term ''real" to these entities that first become objects of my 
 consciousness? Others who do not denv the psychological 
 priority of percepts call the concepts "real." It can only be 
 that ] am using the term "real" in a different sense from 
 that in which they are using it. 1 am using it to denote 
 those objects that are psychologically prior, or, more con- 
 cretely, sense-data. Or. rather, Epicurus is using it so, 
 since, if we can rely on Diogenes Laertius, he asserts both 
 the psychological and the ontological priority of sense- 
 data. 
 
 But as we come to close quarters with this notion of 
 sense-data we notice two sets of entities to which it may 
 apply. We may be naming *'sense-data" those objects that 
 excite the sense organs and cause the observer to be con- 
 scious of sights or sounds. Or we may be referring to the 
 peculiarly vivid objects of consciousness that we assume to 
 be due to such excitation of the sense organs. I am not 
 concerned in differentiating the cause of consciousness 
 from the object of consciousness in the case of veridical 
 perception. For my present purpose it is a matter of indif- 
 ference whether, when we perceive correctly, the object of 
 consciousness and the cause of consciousness be taken as 
 one or as two. But since there are times when we err in 
 our perceptions, since our senses deceive us, which group 
 of entities do we mean by *'sense-data" ? Do we mean all 
 entities that excite the sense organs and result in conscious- 
 ness, whether the resulting experience be valid or illusory? 
 Or do we mean vivid objects of consciousness presumably 
 due to stimulation of the sense-organs, whether or not they 
 correspond to— or, if you will have it so, are identical 
 with — the causes of this consciousness? In the system of 
 Epicurus this distinction between the causes of conscious- 
 ness and the objects of consciousness is not made. 
 And the distinction does not have to l>e made, since 
 Epicurus does not admit the possibility of error in 
 sense-perception. That is, there is for him no vivid 
 object of consciousness that is not identical in content with 
 
 the cause of that consciousness. There is nothing that we 
 see that has not acted upon us. ^The visions of insanity 
 and of sleep have a real object," writes Laertius,^^ -for 
 they act upon us ; and that which has no reality can pro- 
 duce no action." Since, then, the causes of consciousness 
 are identical in content with the objects of consciousness it 
 makes no difference under which name we assert the reality 
 Of this group ot entities. But to us who do recognize the 
 possibility ot error in sense-perceptions, and for whom in 
 consequence, the distinction between the causes of conscious- 
 ness and the objects of consciousness does have meaning 
 It is the latter group, the objects of consciousness, that 
 hpicurus seems to have called "real." The vivid sun that 
 seems to be two hundred feet away Epicurus said was a 
 real sun; to the visions of insanity and of sleep he at- 
 tributed, too, a real object. What we should call the objects 
 of perception but not the causes of perception he called 
 "real." But for Epicurus these objects of perception were 
 identical in content with the causes of perception. Epicurus 
 meant by "real" what he meant by ''cause of perception" or 
 by ''object of perception," but he meant by "real" what we 
 mean by "object of perception" and not vvhat we mean by 
 "cause of perception." But in making these distinctions 
 between Epicurus's point of view and our own and between 
 causes of perception and objects of perception, let us not 
 forget the larger statement on which these distinctions are 
 refinements. In any case, when Epicurus used the term 
 "real" he referred to what we may indefinitely call "sense- 
 data." "Real" is used in the sense which is of present 
 interest to us; it is used in the sense in which it points 
 primarily to the entities present to the senses and not to 
 concepts of reason or to the mystical "One" that is All, or 
 vvhat not. 
 
 In modern times the empiricism that has now become so 
 widespread found its first important exponent in Francis 
 Bacon. Bacon was more interested in the methodological 
 principle that investigations should start with sense-percep- 
 tions and gradually advance to broad generalizations than 
 he was in applying the term "real" to sense-data. He was 
 convinced that all opinions have their basis in sense-per- 
 ceptions. "Man," says the first aphorism of the Novum 
 
 "" Ibid, p. 435. 
 
 2>7 
 
Organuni. "as the minister and interpreter of nature, does 
 and tmderstands as much as his observations on the 
 order of nature, either with regard to things or the mind, 
 permit him, and neither knows nor is capable of more." 
 All of us, then, get our opinions from our sense-data. 
 Some, however, build big and inspiring, though hasty and 
 unstable, structures on this foundation, while others ad- 
 vance cautiously and leave reluctantly the secure ground 
 from which they start. ''There are and can exist but two 
 ways of investigating and discovering truth," says Bacon, 
 and both start with sense-perceptions. Even he who is most 
 enamored of the flights of reason must start with the senses. 
 But ''the one hurries on rapidly from the senses and par- 
 ticulars to the most general axioms, and from them as 
 principles and their supposed indisputable truth derives and 
 discovers the intermediate axioms," while "the other con- 
 structs its axioms from the senses and partictdars, by 
 ascending continually and gradually, till it finally arrives at 
 the most general axioms." The former, Bacon says, is the 
 method his contemporaries use. while the latter is **the true 
 but unattempted way."^"* 
 
 Bacon is the high priest of sense. His new method is a 
 method in which sense-perceptions have the most important 
 part. And when he comes to use the word "real," he uses 
 it, too, to mean that which is present to the senses. The 
 sense-data, according to him, are "real"; the products of 
 the mind with which they become entangled are false. But 
 in contrast to the position of Epicurus, it is the causes of 
 perception and not the objects of perception that Bacon 
 makes real. The entities that present themselves to the 
 sense-organs and result in consciousness, these are the real 
 entities. Unfortunately there is often a certain manipula- 
 tion, a distortion, and a contamination in the process of 
 getting known, so that what we are finally conscious of is 
 not identical in content with the entity first presenting itself 
 to the sense-organ. It is the crude sense-datum, not yet 
 worked over by the mind, that he calls "real"; and a 
 process of "getting known" that shall not affect the con- 
 tent, that he holds desirable. The cause of perception is 
 real; we can be conscious of this "real" by neutralizing the 
 
 *^ Bacon: Novum Organutn, Aphorism 19. 
 
 38 
 
 A 
 
 damage done by the mind in altering the content of the 
 
 Bacon let men please themselves as they will in admiri,i<. 
 and a most adoring the human mind, thi's is cenam 2 
 
 to its T T """'"^ ^"'°"^ '^'' '^y °*" objects according 
 to ,ts own figure and section, so the mind, when it receives 
 .mpress,ons of objects through the sense, cannot be tru ted 
 to report thent truly, but in forming its notions mixefup 
 Its own nature with the nature of things." 
 
 And not only is Reason the criminal. Sense also is at 
 fault Besides the distortions of reason or mind, we must 
 guard against the deceptions of the senses themselves 
 Bacon ,s not one of those who say the senses are always 
 correct, who say that what we call deceptions of the senses 
 are mistakes of judgment in interpreting the data of the 
 senses^ According to him the objects of sense-perception 
 otten fail to tally with the causes of perception. And when 
 there IS this failure, it is the causes of perception and not 
 the objects of perception that are real ; the latter are illusory 
 and unreal. And so even the objects of sense with which 
 we start must be submitted to a process of scrutiny and 
 correction. The cause of perception, to which the term 
 real is restricted, is the objective of knowledge In 
 order to be aware of the content of this cause of perception 
 we must correct the errors to be found in the objects of 
 sense-perception and neutralize as far as possible the dis- 
 tortions mtroduced by mmd. Or. as Bacon summarizes 
 h.s position, the evidence of the sense, helped and guarded 
 by a certain process of correction. I retain. But the mental 
 
 pTrTrele. -"■' '°"°^'^ ^'^ ''' °*' ^^"^^ ' '^ '^^ -- 
 
 o{^nl!^"'rl "U°^" ^'''^' '' '""^h "^^^ ^i"^P'e than that 
 of Bacon. The "Essay on the Human Understanding" is 
 
 somewhat encyclopedic in character; the manv subjects that 
 are discussed in it are not treated with the singleness of 
 purpose we find in Bacon's philosophical writings Con 
 sequently we do not find in Locke this whole-hearted 
 advocacy ol the priority of sense-data. We find passages 
 "Plan of the Instauration ; in the edition by J. M. Robertson Crcx)5), 
 
 p. 250 
 
 "Ibid. Ed. by J. M. Robertson, p. 256, 
 
 39 
 
evidencing an empiricism that is quite marked, but we 
 should be slow to accept these passages as typical. Yet 
 traditionally Ivocke is considered an empiricist, and this 
 characterization is not without some justilication. On the 
 psychological side, his advocacy of the priority of sense- 
 data can be asserted without ciualification. Again and again 
 he tells us that all ideas come from experience.-'" The mind. 
 Locke says, cannot originate a single new simple idea.'*^ 
 These simple ideas, since they cannot he derived from the 
 mind itself, must all be the result of entities directly present 
 to the senses. Consequently the psychological priority of 
 sense-data is well established. Simple ideas are identified 
 with sense-data, and complex ideas are derived from, and 
 consequently subsequent to, these simple ideas. 
 
 In many passages, too, Locke asserts the ontological 
 priority of sense-data. The o1)jects of sense-perception 
 are associated with a reality of a sort that other objects 
 have not. An entity I see or have seen I know to l>e real, 
 hut the reality of an entity that is not in connection with 
 my senses is quite unreliable. "For if I saw," says Locke, 
 "such a collection of simple ideas as is wont to be called 
 'man' existing together one minute since, and am now alone, 
 I cannot be certain that the same man exists now."-^^ That 
 is to say, we can assert the reality of the man one minute 
 ago on the ground that he was a sense-datum, but we have 
 no such basis for attributing reality to the man of the 
 present moment. Those entities that are sense-data are real 
 in contrast to the entities that are not ; and those entities 
 that have been sense-data are also given the name "real." 
 ''Real" is used in a sense in which it points out entities that 
 are or have been present to the senses. 
 
 In other places the reality of sense-data is contrasted 
 with the possible unreality of facts we learn from authority. 
 I may learn on authority that gold is malleable, but the 
 malleability of gold will still be a fact of doubtful reality. 
 VV^hen I have before me gold that is hammered out, then, 
 l:)eing a sense-datum, malleability is a fact."*^* We know 
 
 ''Essay on the Human Understanding, especially Book 2, chapter i. 
 ** Book 2, chapter 2, section 2, and Book 4, chapter 4, section 4. 
 '"Ibid; Book 4, chapter ir, section 9. 
 
 ^ Ibid; Book 4, chapter 12, section 9. Cf. also Book i, chapter 4, sec- 
 tion 2Z- 
 
 40 
 
 ^ ^f 
 
 what we experience, but do not necessarily know what we 
 have on authority. It is true that we are here dealing with 
 the certainty of our knowledge and are not here dealing 
 directly with the ontological status of the objects of that 
 knowledge. We are dealing in the first instance with the 
 category of modality and not with the category of reality. 
 There is, however, an intimate connection between the two. 
 An object that I do not know to be real belongs to a class 
 containing both real and unreal entities. If objects that we 
 have on authority were all real, I should know an entity 
 to be real as soon as it is described as such an object. 
 And since, according to Locke, we do not know entities to 
 be real that are given on authority, some entities that we 
 have on authority may be real and some may be unreal. 
 "Object learned on authority" does not imply "real object" 
 as "object present to the senses" does. 
 
 The meaning of reality that I have been bringing out is 
 that in which the word is applied to sense-data exclusively 
 and is not applied exclusively to facts learned on authority, 
 nor, indeed, to any entities that are not sense-data. It is a 
 sense in which the distinction between real and unreal 
 holds of the mere content, regardless of whether the entity 
 under consideration be "thing" or ''idea." But at other 
 times it is the "thing" that Locke calls real in contrast to 
 the ''idea," the "thing" with its primary qualities and not 
 the various ideas in our minds representing the "thing." 
 It is not that Locke often explicitly identifies the real with 
 the "thing" and the unreal with the "idea." But there is, 
 through much of his book, a bias in favor of the "thing" 
 as substantial, reliable and real. Now, in so far as Locke 
 calls the "thing" real, he is again using the "real" to sig- 
 nify a sense-datum. For the "thing" is the cause of per- 
 ception, and the sense-datum, it will be remembered, can be 
 taken to mean either the cause of perception or the object of 
 perception. In this connection, where "real" means the 
 "thing" and not the "idea," it obviously means the sense- 
 datum as cause of perception and not as object of percep- 
 tion. But in the cases with which we had previously to 
 deal, the cases in which "real" means the sense-datum and 
 not the fact that we learn on authority or by conjecture, 
 it is hard to tell whether the sense-datum that is real is the 
 
 41 
 
cause of perception or the object of perception. Probably 
 there, too, it is the cause of perception that is real. The 
 hammered gold that is a sense-datum probably is 'Veal" in 
 that it affects my sense-organs and not in that it is a vivid 
 object of consciousness. But there are passages in which 
 Locke seems to imply that when it is the content that we are 
 dividmg into real and unreal, the distinction between die 
 content of the cause of perception and that of the object 
 ot perception is unnecessary. He seems to feel that 'the 
 vivid objects of perception never point to the wrong causes 
 of perception, that objects of sense-perception have internal 
 marks ^distinguishing them from dream objects and illu- 
 sions.^^ And if we conclude from these passages that Locke 
 does not. in these passages, recognize the possibility of error 
 in sense-impressions, we must admit that it is no more 
 necessary tor him than it was for Epicurus to distinguish 
 between the content of the cause of perception ancrthat 
 of the object of perception. 
 
 The Locke whom we have been considering has applied 
 the term "real" to sense-data exclusivelv. Xow he has 
 applied it to the thing outside the mind in contrast to the 
 ideas we have of this thing, now to the object of experience 
 in contrast to the fact learned on authority or by conjecture 
 In the one case he means the cause of perception; in the 
 other It IS doubtlul whether he means the cause of percep- 
 tion or whether he fails to find it necessarv in practice to 
 distinguish between the content of the cause of perception 
 and that of the object of perception. But in anv case the 
 real has meant a sense-datum. We were prepared in the 
 beginning, however, to find that Locke was not so thorough- 
 going a sensationalist as Bacon. There are many passa-es 
 in which he asserts the reality of entities other than sen^'se- 
 data, where he gives the term "real" other meanings For 
 example, substance, he says, is "real." Why? It is real 
 because it is impossible for us to conceive of attributes 
 without a substratum, not l)ecause substance is a sense- 
 datum.^^ -Real" here is used in a sense in which it means 
 that the opposite of wdiich is inconceivable, not in the sense 
 in which it means that which is a sense-datum. Again, 
 
 yind; Book 4, chapter 2. section 14, and Book 4, chapter 11. 
 Ibid; Book 2. chapter 23, section 4. 
 
 42 
 
 > !• 
 
 Locke at times uses "real" in the sense of "possible." Ob- 
 jects of thought that are possible he calls "real." though 
 he does not say they "exist." It is a sense of "real." that 
 is to say, in which the term ceases to be synonymous with 
 "existence." It is in this sense that mathematical entities 
 and moral concepts such as "absolute justice" are "real."^^ 
 They are not sense-data, they in some cases do not "exist," 
 and yet they are called "real." All objects of thought that 
 are possible, the component parts of which are compossible, 
 are real. We have knowledge of reality when we perceive 
 the agreement between our ideas, for the agreeing ideas 
 constitute a "real" And so in many places "reality," if 
 not "existence," is used in a sense in which it by no means 
 points to sense-data. "Real" is used by Locke in more 
 than one sense. In some passages the term "real" points 
 to a sense-datum of one sort or another. In other passages 
 it means something other than that which is a sense-datum. 
 And so if we call sensationalism the assertion that to be 
 real means to be a sense-datum of some sort, Locke may 
 be called a sensationalist in his definition of "real," but 
 certainly not a consistent one. His "real" in many con- 
 texts refers to the sense-datum and to nothing else. But 
 in other passages his use of "real" is very similar to that 
 traditionally attributed to Leibniz. And it is a sensational- 
 ism that at its best is expressed haltingly and without 
 enthusiasm. 
 
 Bishop Berkeley, insofar as he asserts the reality of 
 sense-data, asserts it in a more clear-cut and a simpler man- 
 ner than does Locke. W^ith him as with Locke all of our 
 ideas arise from experience. Some objects are present to 
 our external senses, some objects are present to our internal 
 sense, and the rest of our ideas are complexes founded 
 upon these two forms of sense experience. "The objects of 
 knowledge," he writes, "are (i) ideas actually imprinted 
 on the senses, (2) ideas perceived by attending to the 
 passions and operations of the mind, or, lastly, (3) ideas 
 form.ed by help of memory and imagination — either com- 
 pounding, dividing, or barely representing those originally 
 perceived in the aforesaid ways."^"* Psychologically, then, 
 
 ^^ Ibid; Book 4, chapter 4. sections 6-8. 
 
 ** Principles of Human Knowledge, part i, section i. 
 
 43 
 
the data of the external and uiternal senses are prior-w.th 
 Berkeley as n>th Locke-to the objects formed by help of 
 memory and inm.^ination. But when it conies to restricting 
 
 he term real to these entities psychologically prior! 
 Berkeley .s more defnnte than f.ocke. The objects of the 
 senses as contrasted with those of the imagination are 
 
 real. And ue do not have to wonder wliether it is the 
 
 mZ- "' 'f^'^li''"" T '^'' "'^■'■'" °f perception that he 
 makes real. Sense-data are real as oh/ccts of perception 
 They are real in that they are vivid, brilliant, compelling 
 objects, and the objects of the imagination are unreal in 
 that they are weak, faint, and subject to change at will 
 1 he Ideas imprmted on the senses by the Author of nature 
 are ca, ed real , kings, an.l those excited in the imagination 
 being less vivid and constant are more properly tenned 
 Ideas or ,magcs of things, which thev copv and repre- 
 sent. ■ And again we read : "The ideas of sense are 
 allowed to have more reality in them than the creatures of 
 the mind. 
 
 It is accordingly the vivid objects of percq.tion that 
 Berkeley uses the term "real" to point out. He cannot 
 admit, if he adheres to this meaning of reality, the possi- 
 bility of error in sense-perceptions. For to do 'so would be 
 to run counter to his identification of realitv with vivid 
 objects ot i>erception by making some of these 'vivid objects 
 of perception unreal. And Berkeley does seem to call all 
 these vivid objects "real," though there are some to which 
 the term as commonly used would not apply, some which 
 m the denotation given the term more usually, would be 
 called unreal. "For my part," savs Berkeley,"" "I can as 
 well doubt of my own being as of the being of those things 
 mhKh I actually perceive by sense: it being a manifest con- 
 tradiction that any sensible object should be immediately 
 perceived by sight or touch, and at the same time have no 
 existence m nature, since the very existence of an unthink- 
 ing being consists in being perceived." 
 
 In all this the term "real" most clearly and explicitly 
 refers to sense-data. But in other connections the term 
 quite as evidently does not refer to sense-data. I have in 
 mind those passages where Descartes's phrase "clear and 
 
 '''Ibid: section 33. Cf. also sections 36, go 
 "Ibid: section 88. 
 
 44 
 
 4 
 
 1 
 
 distinct ideas" shows itself. In the introduction to the 
 "Principles of Human Knowledge/' for example, general 
 abstract ideas are declared unreal. Ideas such as "triangle" 
 are declared unreal, not on the ground that they \re 
 creatures of the imagination, lacking the vividness of ex- 
 perience, but on the ground that they are objects of thought 
 that are not clear and distinct. The two criteria must be 
 distinguished. For many objects with the compelling 
 power and vividness of experience are hazy and inchoate in 
 form and possibly inconsistent on analysis, while others 
 that are quite definite and self-consistent may be but 
 creatures of the imagination. General abstract ideas, then, 
 are unreal because they are not clear and distinct. Such a 
 statement points to a use of real that, following Descartes, 
 makes objects real that are clear and distinct, and objects 
 unreal that lack these qualifications. Descartes had an 
 influence on the line of British philosophers that is not 
 always recognized. Locke, Berkeley and Hume by no 
 means developed a system of thought indigenous to the 
 British Isles, insulated from their contemporaries and 
 predecessors on the Continent. But the evidence of this 
 Continental influence, especially on the part of Descartes, is 
 SO very abundant that the point is hardly w^orth mentioning. 
 
 In the latter part of his life Berkeley may be assumed to 
 have again used 'Veal" in a sense in which it does not point 
 to sense-data. In ''Siris" he show^s himself particularly in- 
 terested, not in the entities that are psychologically prior, but 
 in the entities that are inferred from these sense-data. It 
 is the knowledge of spirits and of the Platonic Ideas that 
 he stresses, entities which I presume he would call real, 
 though they are evidently not sense-data. In so far, then, 
 as he calls spirits and the Platonic Ideas 'Veal," he is using 
 "real" in a sense in which it means, not sense-data, but 
 entities that we are aware of mediately and inferentially. 
 
 Hume's use of "real" is very similar to that characteristic 
 of Berkeley. He begins, as do Locke and Berkeley, with 
 the assertion that, psychologically, sense-data are the first 
 objects of our consciousness, and that it is from them that 
 our other objects are derived. The vivid sense-data are 
 called impressions, the derivative objects ideas. There is 
 some shifting, however, from the psychological thesis that 
 all ideas must be derived from sense-data, to the thesis that 
 
 45 
 
w^^^ 
 
 all valid objects of thought— though not all possible objects 
 of thought — must be derived from sense-data. It becomes, 
 not impossible, but inralid, to have objects of thought that 
 are not derived from sense-data. And so we get the onto- 
 logical principle that sense-impressions and ideas derived 
 from sense-impressions are ''real," while other objects of 
 consciousness are unreal. There are only impressions and 
 ideas derived from them, according to Hume: the rest is 
 arguing about words. But. contrary to his psychological 
 thesis, he admits a "rest" ; it is possible to have as an object 
 of consciousness what he would call a "mere word." For 
 example, some people talk about "power." But ''power" is 
 an object of consciousness that is a 'mere word"; it is a 
 pseudo-idea, unreal in that there is no impression from 
 which it is derived. 
 
 At times Hume restricts the meaning of reality further. 
 He uses the term to mean, not both sense-impressions and 
 the ideas derived from them, but only the \i\-id sense-im- 
 pressions. Sense-data arc real. The objects I am aware 
 of when I look at entities are real : l)ut when I turn 
 away and am aware of these entities as "still there," the 
 objects of my consciousness are unreal. The fiction of the 
 continued existence of these entities, Hume says, "is really 
 false," for — I suppose this is his meaning — when I am not 
 looking at these entities they are not sense-data, and are, 
 consequently, unreal."*^ 
 
 At other times Hume, like Berkeley, uses "real" to mean 
 the ''clear and distinct." Universals are unreal, not so 
 much because they are not sense-data or derived from 
 sense-data, but rather because they are not clear and dis- 
 tinct.^^ And power is unreal, not only in that it is not 
 derived from a sense-datum, but also because we have no 
 distinct idea of either some particular power or of power in 
 general.-*'^ Then we can find passages in Hume, and in 
 Locke and Berkeley as well, where all objects of conscious- 
 ness seem to be real. In the main, however, Hume uses 
 "real" to mean sense-data, and. with some wavering, ob- 
 jects derived from sense-data. Sense-data are real as 
 *'vivid objects" and not in the sense of "causes of percep- 
 
 *' Hume: Treatise of Human Nature, Book i, pt. 4, sec. 2; Cf. also 
 Book I, pt. 3, sec. 5. 
 
 *^ Ibid: Book i, j)t. i, sec. 7. 
 "^ Ibid: Book i, pt. 3, sec 14. 
 
 46 
 
 % 
 
 tion," though to a certain extent Hume seems to regard 
 the class of vivid objects as identical in content with what 
 we are accustomed to call the causes of perception. That 
 is to say, the result of his identification of "real" with vivid 
 objects of perception is not to make the objects of dreams 
 real. For he allows our ideas "in sleep, in a fever, in mad- 
 ness, or in any very violent emotions of soul" only to "ap- 
 proach" the vividness of impressions.^*^ All vivid objects 
 of consciousness are real then, though these vivid objects 
 never happen to be what common sense would call illusions. 
 In the other case in which Hume uses "real" to point 
 primarily to sense-data, in the case in which the denotation 
 of "real" is so widened that those ideas that are derived 
 from sense-impressions are also called real, Hume is using 
 "real" very looselv. For with merely this description of 
 "real," it is quite impossible to determine which objects of 
 thought are the ones that are "derived" from sense-data, 
 and which are the objects that are not so derived, that are, 
 consequently, "unreal." Be that as it may, what little defi- 
 niteness this definition of "reality" has forces it to be recog- 
 nized as belonging to the group of definitions in which the 
 "real" is primarily identified with sense-data. And so we 
 come to Kant. 
 
 We have previously met quite frequently with the asser- 
 tion that all our ideas begin with sense-experience. In 
 Kant we find this assertion repeated. "That all our knowl- 
 edge begins with experience," we read, "there can be no 
 doubt." "For," Kant continues, "how should the faculty of 
 knowledge be called into activity, if not by objects which 
 affect our senses, and which either produce representations 
 by themselves, or rouse the activity of our understanding 
 to compare, to connect, or to separate them."^^ And if 
 more passages to the same effect are desired, we can turn 
 to the beginning of the "Aesthetic" where we read : "What- 
 ever the process and the means may be by which knowledge 
 reaches its objects, there is one that reaches them directly 
 and forms the ultimate material of all thought, viz. : intui- 
 tion {Anschauung). . . . All thought, therefore, must, 
 directly or indirectly, go back to intuitions {Anschauun- 
 gen), i. e., to our sensibility, because in no other way can 
 
 "^ Ibid; Book i, part i, sec. i. 
 
 "Kant: Critique of Pure Reason, 2d edition, p. i. 
 
 47 
 
objects be given to us."''^ The effect of this, to be sun- 
 is to ffive sense-<lata a certain fundamental i)osition in the 
 genesis of the objects of o.;r consciousness, but it in no 
 uay involves the tern, "real.- It is important for our pur- 
 poses, however, because, with Kant as with the writers we 
 have just exannne.l. the assertion of this psychological 
 priority of sense-data lea,ls to the assertion of the onto- 
 logical priority of these entities. Now, in considering 
 Kant s position, we shall have to distinguish Ix^tween the 
 crude material of knowledge and the entities that result 
 from the addition to this crude m.iterial of the mental 
 factors that play so large a part in knowledge, between the 
 unsynthetized manifold present to the senses and the syn- 
 thetized. co-ordinated manifold that is the world as we see 
 it^ Kant implicitly denotes by "real" now one of these sets 
 of entities. n<nv the other. When he seems to imply that 
 the crude, unsynthetized manifold is -'real." it is the mental 
 factors that, in contrast to this unsvnthetized manifold 
 seem to be called "ideal." The distinction between these 
 two elements, the crude bare data of sensation and the 
 mental factors that make knowledge possible, Kant makes 
 quite plainly. Connection, he .savs,^'^ "does never lie in 
 the objects, and cannot be borrowed from them by percei)- 
 tion and thus be taken into the understanding but it is 
 always an act of the understanding, which itself is nothing 
 but a faculty of connecting a priori, and of bringing the 
 mamfold of given representations under the unity of apper- 
 ception." .And looked at from one point of view, this con- 
 nection, not lying in the objects, does not form a real part 
 of the world of objects: it is a man-made addition superim- 
 posed upon nature. Though the connection is an important 
 and valuable addition, it is the unsvnthetized manifold that 
 IS the "real." There are passages, ] sav, in which Kant 
 seems to take this position, in which he applies the term 
 "real" exclusively to the crude objects not vet taken up and 
 worked over by the understanding For example, in the 
 Anticipations of Perception" in the Second Edition of the 
 Kritik dcr rcincn Vcrmmft. he says that i)henomena con- 
 tain "over and above the intuition, the material for some 
 one object in general (through which something existing 
 Ibid: i.st edition, p. 19. 
 "Ibid: 2(1 edition, Deduction of the Categories § 10. 
 
 48 
 
 
 
 ■■4 
 
 4* 
 
 
 in space and time is represented) ; that is, they contain the 
 real of sensation, as a merely subjective representation, 
 which gives us only the consciousness that the subject is 
 affected, and which is referred to some object in general. 
 Now, there is a gradual transition possible from empirical 
 to pure consciousness, till the real of it vanishes completely 
 and there remains a merely formal consciousness (a priori) 
 of the manifold in space and time." The crude data, while 
 not the entities that Kant desires to emphasize, are called 
 "real." And things-in-themselves. which, in so far as 
 they have any content at all, may be taken to have the con- 
 tent of these crude data, are also given the attribute "ex- 
 istence." "For that it existed by itself," Kant says,^* 
 "without any reference to ourselves and possible experi- 
 ence, might no doubt be said when wt speak of the thing by 
 itself." 
 
 Now these crude data, in so far as they are called "real," 
 are they real as causes of sensation or as objects of sensa- 
 tion? That is to say, are these crude entities real because 
 they affect the sense-organs and bring about the perception 
 of objects as we know them ? Or are they real as element- 
 ary, embryonic objects of consciousness, as the inchoate 
 part of the entities that are objects for nsF There is little 
 basis for attributing to Kant one of these positions rather 
 than the other. He does seem, however, to make no dis- 
 tinction between the content of the crude cause of percep- 
 tion and the content of the crude object — or, rather, poten- 
 tial object — of perception. That is to say, he seems to 
 deny the possibility of error where we are dealing with 
 these very elementary objects. All error is due to the 
 mental factors in knowledge. The crude object of percep- 
 tion is never unreal, never differs in content from the crude 
 thing-in-itself that affects the sense-organs. But the evi- 
 dence for this interpretation is very scanty.^^ It is scanty 
 because Kant for the most part denies any cognizable con- 
 tent to either the crude cause of perception or this very 
 elementary o-bject of perception. For the most part he 
 admits content only to the object of experience, the entity 
 that results from the assimilation of the crude material to 
 the mental factors that make experience possible. These 
 
 ""Ibid: 
 * Ibid: 
 
 First Edition, p. 493. 
 First Edition, p. 303. 
 
 49 
 
 ,.ii 
 
entities and not the crude data have content, and, in a 
 different sense of the word "real" from that which we have 
 been considering, these entities and not the crude data are 
 'Veal." 
 
 In fact, the sense in which 'Veal" means crude data is 
 the exception ; it is much more frequently that Kant em- 
 ploys the term to denote the object of experience. For the 
 crude data are really nothing when taken by themselves; 
 the mental factors are the sine quibiis non of knowledge 
 and even of perception. It is the combination of crude 
 data and mental factors that makes the world we know, 
 the world that Kant calls "real"— in this second sense of 
 the word. It is not the unsynthetized manifold by itself 
 that is "real" in this sense, but the unsynthetized manifold 
 affected by a process of connection and so turned into ob- 
 jects of experience. These objects of experience are the 
 entities that "real" most fre([uently denotes for Kant. 
 These entitie> are the important entitie.v; they are the only 
 objects that are objects- for-us. 
 
 Now when "real" denotes crude data, it is evidently 
 denoting a species of sense-data. When it denotes objects- 
 for-us, objects of experience, it is not being used so appar- 
 ently to denote sense-data. For these objects of experience 
 are objects that are permeated through and through with 
 ideal elements, while we generally think of sense-data as 
 simple entities that are to be contrasted with the products 
 of mind. But what are Kant's objects of experience? 
 For the most part, it seems to me. Kant's "objects of ex- 
 perience" refer to the objects that we see and hear, the 
 objects at hand, objects that we have been calling sense-data 
 quite as much as we have been calling the crude material 
 sense-data. That is to say, we must make a distinction 
 between two sorts of sense-data. On the one hand there 
 IS the crude material, the elementary object of sense-per- 
 ception; on the other hand there are the objects of sense- 
 perception as we know them, the ol)jects-for-us that we 
 see and feel and hear. Yet both groups of entities may 
 be called sense-data. The crude material has been held, 
 by those whose opinions we have considered, to be present 
 to the senses either as original cause or as elementary object 
 of perception. And the objects of sense-perception, in the 
 
 50 
 
 M 
 
 form they have for us, are sense-data, in contrast to the 
 objects of memory, of reason, and of imagination. Let us 
 turn our attention, then, to these objects of sense-perception, 
 which, with all the connection that is given them by mind, 
 are still sense-data in contrast to the objects of memory, 
 imagination, and reason. 
 
 It is to these synthetized objects of sense-perception, I 
 have suggested, that Kant in large part attributes reality. 
 Objects of experience rather than crude data are real, 
 and sense-data are the objects of experience par excellence. 
 Still it is not the synthetized objects of sense-perception 
 alone that are real, in the sense of "real" with which we are 
 concerned. These sense-data transmit their reality, so to 
 speak, to the entities that are derived from them and con- 
 nected with them. And so we have a use of "real" identical 
 with that which w^e noticed in discussing Hume: — "Only 
 sense-data and their derivatives are real." As a matter 
 of fact, while such a principle is upheld by Hume, it is 
 much more characteristic of Kant. "What is real in 
 external phenomena." we bud asserted in the "Critique," 
 "is real in perception only, and cannot be given in any other 
 w\'iy." "From such perceptions, whether by mere play of 
 fancy or by experience, knowledge of objects can be pro- 
 duced, and here no doubt deceptive representations may 
 arise, without truly corresponding objects, the deception 
 being due, either to illusions of imagination (in dreams), 
 or to a fault of judgment (the so-called deceptions of the 
 senses). In order to escape from these false appearances, 
 one has to follow the rule that whatever is connected accord- 
 ing to empirical lazvs zvith a perception is real/'^^^ And 
 again, in another passage, he writes r"^' "That there may 
 be inhabitants in the moon, though no man has ever seen 
 them, must be admitted but it means no more than that, 
 in the possible progress of our experience, we may meet 
 with them; for everything is real that hangs together wdth 
 a perception, according to the laws of empirical progress." 
 Passage after passage comes to the same thing. ''The 
 postulate concerning our knowledge of the reality of things 
 requires perception, therefore sensation and the conscious- 
 
 67 
 
 Ibid, 1st Edition, p. 376. 
 Ibid, 1st Edition, p. 493. 
 
 51 
 
 ''^ 
 
ness of it. not, indeed, immediately of the object itself, 
 the existence of wliich is to be known, but yet of a con- 
 nection between it and some real perception according to 
 the analogies of experience which determine in general 
 all real combinations in experience. . . . But if we do not 
 begin with experience or do not proceed according to the 
 lazi's of the eiiipirical connection of phenomena, we are only 
 making a vain display as if we could guess and discover 
 the existence of anything. ''^^ 
 
 Sense-data and the objects that are connected witii them 
 are the objects of possible experience, and them alone is 
 the word "real" — in the sense of the word we have been 
 discussing — used to denote. In contrast to these objects 
 of possible experience that are real, there are other entities 
 that are unreal. "It is possible experience alone that can 
 impart reality to our concepts; without this a concept is only 
 an idea without truth, and without any reference to an 
 object."'*^ That is to say, leaving out of account the crude, 
 inchoate data, there are two sorts of entities — the objects 
 of possible experience and the mere ideas. An unreal entity 
 is a ''mere idea, the objective reality of which can never 
 be shown in any possible experience," while the real world 
 is the world of sense w^hich "must be looked upon as the 
 sum total of all possible experience."^^ There are unreals, 
 then, though the content of these "mere ideas" is, as with 
 Hume, left quite indefinite. 
 
 Now, when it is the entities in the wu3rld of possible ex- 
 perience that are real, these sense-data and their derivatives 
 seem to be real in so far as they are vivid objects or con- 
 nected with vivid objects, not in so far as they are causes 
 of our thought. For, says Kant,^'^ "wnth reference to the 
 reality of external objects, I need as little trust to inference 
 as with reference to the reality of the objects of my internal 
 sense (my thoughts), both being nothing but representa- 
 tions, the immediate perception (consciousness) of which 
 is at the same time a sufficient proof of their reality." This 
 seems to imply that the objects of experience that are vivid 
 and those objects that are connected with such vivid objects 
 
 I 
 
 Ibid: 1st Edition, p. 225. 
 Ibid; 1st Edition, p. 489. 
 Ibid; I St Edition, p. 437. 
 *^ Ibid; 1st Edition, p. 371. 
 
 
 Be 
 
 eo 
 
 52 
 
 of experience are unmistakably real. Consequently, with 
 Kant, as with all who make sense-data real as object of 
 perception and not as cause of perception, there is ruled 
 out the possibility of error in the perception of vivid objects 
 of experience. Perceptions and objects connected with 
 them by a rule are real. But how recognize a perception? 
 A perception is a vivid object of experience; consequently 
 no vivid object of experience is unreal. 
 
 There are other definitions of reality implicit in Kant 
 than those I have brought out. Some are implied in pas- 
 sages I have not mentioned, w^hile others may be inferred 
 from passages I have given, provided a different interpre- 
 tation be given the term "experience." But the senses in 
 which Kant can be taken to use 'Veal" to mean sense-data 
 are, roughly speaking, two. He calls "real" the crude 
 material, taken now as mere elementary data, now as thing- 
 in-itself. And he calls "real" the complete, informed objects 
 of sense-perception and the entities connected with them by 
 a rule. The meaning of real when used in the first of these 
 two senses is not ambiguous, though the entities denoted by 
 it are practically without content. In so far as vivid objects 
 of experience and entities connected with them are real, we 
 lack any good criterion to distinguish the real from that 
 which is a "mere idea." "Reality" in this sense is rather 
 indefinite; the reality of vivid objects of perception is well 
 assured, but which objects are connected with these vivid 
 objects, and which are not, is not sufficiently determined.^^ 
 Since Kant these varying definitions of "reality" have 
 gone on repeating themselves. And the men in whose writ- 
 ings these definitions are implied have been at even less 
 pains, if possible, to make their definitions explicit and to 
 justify or to acknowledge the lack of justification for the 
 various senses in which they have used the term. In par- 
 ticular, the two senses in w^hich Kant used "real" to mean 
 sense-data have become commonplaces. The physicist 
 Mach, for example, uses "real" in almost exactly the first 
 of the two senses in which Kant used "real" to mean sense- 
 data. For Mach, as in part for Kant, it is the crude ma- 
 terial, the data without the infiltration of any mental ele- 
 ments, that alone are "real." Nature is composed of "sensa- 
 
 *" Cf. however: infra., pp. 71-73. 
 
 53 
 
I 
 
 tions" as its elements/''^ These crude entities, he imphes, 
 are alone ''there." By way of contrast, all mental con- 
 structs, such as hypotheses and abstractions, are unreal. 
 *' Atoms cannot he perceived by the senses," he finds; there- 
 fore "like all substances they are things of thought" and 
 implicitly unreal/'* The crude data alone are real. Sub- 
 stances are ideal abstractions and hy])otheses ideal con- 
 structions for the sake of economy. The crude data or 
 what Ayenarius calls "pure experience" — for Avenarius 
 holds a similar position — are the only objects of which we 
 can be conscious without addini; a mental construct. "Real" 
 conseciuently means them alone. 
 
 Another physicist, Karl Pearson, makes sense-data and 
 entities connected with sense-data "real." He uses "real" 
 in practically the same sense as that in which Kant 
 uses it when he calls sense-data and entities connected with 
 sense-data "real." "The reality of a thing." Pearson says,^^ 
 "depends upon the possibility of its occurring in whole or 
 part as a group of immediate sense-impressions." And 
 "a sine qua non of the existence of an actual black-board 
 is some immediate sense-impression to start with."^® Not 
 merely actual sense-data, however, but possible sense-data 
 inferred from actual data, are real. "I have heard of the 
 Capitol at Washington," says Pearson,^^ ''and although 1 
 have never been to America, I am convinced of the reality 
 of America and the Capitol-— that is, T believe certain sense- 
 impressions would be experienced by me if I put myself in 
 the proper circumstances." On the other hand, entities that 
 are neither actual nor possible sense-data are unreal. Atoms, 
 molecules, electrons, mathematical points, lines, and sur- 
 faces, the ether, matter — all are unreal, in so far as they 
 have not become objects of perception.^^ At times when 
 Pearson wants to give to some of these unreal entities a 
 name that for him has a particularly disagreeable connota- 
 tion, he calls them "metaphysical." It is in such an invidious 
 manner that he talks about ether and matter, about "force" 
 
 " Mach. : Science of Mechanics (Chicago, 1907), p. 482. 
 
 ^ Ibid; p. 492. 
 
 "Karl Pearson; Grammar of Science, 3d Edition, Part i, p. 41. 
 
 Ibid: p. 40. 
 'Ibid: p. 40. * 
 
 Ibid: pp. 193. 286. 
 
 54 
 
 and about an hypostatized "will." Apropos of the last two 
 he says:^^ "Both carry us into the region beyond our sense- 
 impressions; both are therefore metaphysical." Sense-data 
 and entities connected with sense-data are real. Some other 
 entities in the course of time reveal themselves as actual 
 sense-data or as entities connected with actual sense-data, 
 that is, as possible sense-data. Thereupon it becomes possi- 
 ble for us to attribute reality to them. But of other entities 
 that have not shown themselves to be possible sense-data 
 we cannot assert existence. And those entities that can 
 never be objects of possible experience — such as mathe- 
 matical points— are for Pearson definitely labelled "unreal." 
 The vivid o1)jects of sense-perception are real. The objects 
 connected with sense-data, that are also real, are the 
 objects that would be sense-data "if I put myself in the 
 proper circumstances." The objects connected with sense- 
 data are consequently delimited to a certain extent, but cer- 
 tainly not to an extent sufficient to enable us to tell whether 
 any given object is to be called "real" or "unreal." So far 
 as the entities connected with sense-data are concerned, the 
 definition, that is to say, suffers from the same indefiniteness 
 that we found in Hume's meaning of "real" and in Kant's — 
 when Kant uses "real" to refer to the objects of experience 
 and the entities related to them. 
 
 Another recent writer to use "real" to denote sense-data 
 is William James. His "real" points now to a hazy thing- 
 in-itself, now to the objects of sense-perception as we know 
 them. In his accounts of the genesis of thought he falls 
 back upon a thing-in-itself that is objective and real and 
 that sets the processes of consciousness going. We start 
 with a simple, unanalyzed "that," he tells us, something quite 
 lacking in definiteness and hence unattainable after early 
 childhood. This is sensation, the function of which **is that 
 of mere acquaintance with a fact."'*^ With its pristine 
 innocence, however, it, and it alone, puts us in contact with 
 reality. Or, as James says, it is only consciousness with 
 the "sensational tang'' that "directly encounters (to use a 
 word of Mr. Bradley's) a reality outside itself. "^^ It is an 
 unattainable thing-in-itself, consequently, a thing-in-itself 
 
 «e 
 
 as 
 
 es 
 
 Ibid: p. 119. 
 
 '''James: Principles of Psychology, 1890, vol. 2, p. 2. 
 " Ibid: vol. 2, p. 6. 
 
 55 
 
 
Mm 
 
 to which the original "blooming, buzzing confusion" most 
 nearly corresponds, that James is using the term "real" to 
 point out. 
 
 But at other times, and indeed more often, it is things 
 as we know them that are real. The vivid sense-perceptions 
 of a jumping, barking, hairy body ^'are the real dog. the 
 dog's full presence for my common sense."''"' It is the vivid 
 objects of sensation succeeding one another in a stream of 
 consciousness that give us the content of reality. "Dive 
 back into the flux itself, then." James says^"^ in paraphrasing 
 Bergson, whose thought he endorses, "if you wish to know 
 reality, that flux which Platonism, in its strange belief that 
 only the immutable is excellent, has always spurned; turn 
 your face toward sensation, that flesh-hound thing which 
 rationalism has always loaded with abuse." All content that 
 is given by sensation is real, all content that is not so given 
 is unreal.'^'* "Every examiner of the sensible life in concreto 
 must see that relations of ever>' sort, of time, space, differ- 
 ence, likeness, change, rate, cause, or what not, are just as 
 integral members of the sensational flux as terms are."'^''^ 
 Therefore relations are real. However, not merely actual 
 sense-data are real. As with Flume, with Kant, and with 
 Pearson, entities connected with sense-data are real. As 
 with all of these men, however, the distinction between those 
 entities that are real by reason of their connection with sense- 
 data and those entities that are unreal l>ecause they lack 
 such connection is not drawn so clearly as might be desired. 
 Events remembered to have taken place in the past seem to 
 be called "real," if there is a continuity from them to the 
 sense-data present."^' Another statement implies that entities 
 not sense-data now and for me are real if they "can be 
 experienced at some definite time by some experient."^"^ 
 And another criterion is the one that makes entities not 
 sense-data real if the thought of them "leads" me into the 
 presence of them.'^ But with any or all of these criteria. 
 
 " James : Essays in Radical Empiricism, p. 198. 
 
 "James: A Pluralistic Universe, p. 252. 
 
 '* James: Essays in Radical Empiricism, p. 42 
 
 " A Pluralistic Universe, p. 279. 
 
 '• Essays in Radical Empiricism, p. 213. 
 
 "'Ibid: p. 160. 
 
 ""Ibid: p. 55- 
 
 56 
 
 the distinction, I say, between those entities not sense-data 
 that are real and those that are unreal is quite vague. 
 
 Besides these meanings of "real," there are times when 
 James uses the term in still other senses, times when "real" 
 means neitlier the elemental ihing-in-itself nor sense-data as 
 we know them and their derivatives. James tends in some 
 passages to use the term in a sense that makes all things 
 "real," and so makes the term quite valueless. He does not 
 deny that concepts and the relations between them are just 
 as real in their "eternal" way as percepts are in their tem- 
 poral way."^^ In his broad-minded desire to recognize all 
 elements in the universe, he seems to me to give "reality" 
 at times so broad a meaning that it loses all connotation and 
 becomes altogether unserviceable. 
 
 I quoted above a passage in which James shows his 
 agreement with Bergson with regard to the ontological 
 status of sense-data. Bergson, that is to say. is another 
 who at times uses "real" to mean sense-data. Like so 
 many of the recent writers who imply this definition of 
 reality, however, he apparently regards this meaning as 
 so entirely a matter of common-sense that he nowhere 
 comes near making it explicit. Through a large part 
 of his writings the implication is apparent that, to get 
 at the world as it is, we must take from our objects the part 
 that is added by the intellect and so get at the crude data. 
 "We must appeal to experience — an experience purified, or, 
 in other words, released, where necessary, from the molds 
 that our intellect has formed in the degree and proportion 
 of the progress of our action on things. "^^ Time and states 
 of consciousness have really no magnitude. As we are 
 aware of them they have a magnitude, it is true, but this 
 magnitude is read into them by intellect. Real time, states 
 of consciousness as they really are, are without any such 
 magnitude. Spatial entities come through the process of 
 getting known unscathed; their content is not distorted. 
 That is to say, the mind is well equipped to handle such 
 entities. But temporal entities have their content altered in 
 the process of getting known. And so do psychological 
 data. "Now just as, in order to ascertain the real relations 
 
 ^ James: Some Problems of Philosophy, p. loi. Cf. also: Some 
 Problems of Philosophy, p. 7S, and Essays in Radical Empiricism, p. 17. 
 * Bergson : Creative Evolution ; translation by Mitchell, p. 363. 
 
 57 
 
of physical phenomena to one another, we abstract what- 
 ever obviously clashes with them in our way of perceiving 
 and thinking, so, in order to view the self in its original 
 purity, psychology ought to eliminate or correct certain 
 forms which bear the obvious mark of the external world. "^^ 
 It is the crude material, then, that is real. That part of the 
 objects of our consciousness that is due to intellect is unreal. 
 ''Real" is being used once more in the sense in which it 
 means crude data. 
 
 And there are many other arguments, so common that 
 they need not be identified with any one man. that imply the 
 reality of sense-data. Where we find the word "assump- 
 tion" there is nearly always this meaning of reality lurk- 
 ing in the background. The reality of sense-data is seldom 
 ''assumed"; it is undoubted, [kit the reality of entities that 
 are not sense-data is not quite certain; such entities are 
 hypothetical, they are assumptions. 
 
 Then, there are the apologists for religion who say that 
 there are many things we do and must believe on faith, 
 and that God belongs in this class. The implication is that 
 sense-data are nn(|uestionably real."- Other entities are 
 probably real, are to be held as real though we have not 
 seen them but have them only on authorit\- or hearsay; and 
 it is argued that there should be attributed' to God this same 
 inferred reality that we give to earthly entities we do not 
 ourselves directly experience. Those who deny God because 
 He is not a sense-datum are, of course, using "real" in this 
 same sense. God is unreal because "real" is being used to 
 denote sense-data exclusively. 
 
 And finally there are such questions as: How are uni- 
 versal judgments possible? We can know the particular 
 cases that we or others have met with in experience, but 
 how can we know those particulars that are subsumed under 
 the universal which we have not yet experienced?^ We can 
 undoubtedly make true judgments about objects of experi- 
 ence, but a knowledge of other objects is questionable. 
 Sense-data are real ; the knowledge of them raises no ques- 
 tion. But other objects, in what sense are they real, or, at 
 least, knowable:^ In all of these cases "real" is used to 
 
 4 
 
 81 
 82 
 
 Bergson: Time and Free Will ; translation hy Pogson. pp 223,224 
 St. Augustine: Confessions, VI, 5. 
 
 =;8 
 
 denote with certainty all sense-data, while the application 
 of the term to other entities is fraught w^ith some doubt, 
 calls for some hesitation and requires explanation. 
 
 One man uses "real" to mean the crude data, another 
 to mean objects of "experience" ; one to mean causes of per- 
 ception, another to mean objects of perception; one to mean 
 sense-data alone, another to mean objects of memory also, 
 and still another to mean, in addition, any entity vaguely 
 connected with sense-data. All these are shades of meaning 
 given to "real" while that term, generally speaking, still 
 points especially towards sense-data. Now, the account I 
 have given of these "reals" is like an elementary history 
 of philosophy that would content itself with; "Thales said: 
 'All is water' ; Anaximander said ; 'All is the indefinite' " ; 
 that is, it is like a syllabus that catalogs the opinions men 
 have held without stating the reasons for these opinions. 
 I have mentioned Epicurus, Bacon, Locke, Berkeley, Hume, 
 Kant, Mach, Pearson, James and Bergson, and to each I 
 have attributed one or more meanings of "real''; but I have 
 not explained why, for each of these men, "real" has the 
 meaning that it has. The fault, however, is not mine, but 
 theirs. Notwithstanding the fact that "real" is used in so 
 many dififerent senses, a writer will clearly imply that the 
 term has, for him, one of these meanings in particular; and 
 still he will give no reason for his choice, nor even call 
 attention to the fact that he might have chosen dififerently. 
 
 In fact, no one seems to be aware that he might have 
 chosen differently. To each "Reality" seems to consist, 
 fixedly and objectively, of those entities that he has denoted 
 "real." He seems not to know that reality is not "there" 
 before the term "real" is given a meaning; nor is he aware 
 that a different reality is there if "real" is defined differ- 
 ently. And so he confidently makes assertions that follow 
 directly from the definition of reality he has assumed, but 
 which would by no means be true were "real" interpreted 
 in a different sense. Perhaps he says wdth some feeling : 
 "To know reality we must appeal to experience." This is 
 quite true if "real" has the sort of meaning we have chiefly 
 considered in this chapter, if it means sense-data, experi- 
 enced entities. But if "real" denotes the sort of entities 
 that are not experienced — Platonic ideas, mathematical 
 
 59 
 
formulcT, the Infinite, the miracles of the Old and New 
 Testaments — this statement is quite as evidently untrue. 
 Sensationalism, that is to say, may be (juite close to one's 
 heart, and yet he based entirely on the definition of "reality" 
 that is implied. The question: "How are universal judg- 
 ments possible?" would not excite one who did not start 
 out with a bias towards the ontological priority of sense- 
 data. And so. too, with anti-intellectualism. The dis- 
 satisfaction both James and Bergson feel with the intellect 
 results from the fact that the objects of intellectualized 
 awareness are not 'Veal"— "real" being taken in the special 
 sense in which it means sense-data. So much that we hold 
 important is apt to depend on the definition of 'Veality" 
 implied, that we ought at least to make this definition ex- 
 plicit and to state the reason we employ the definition we 
 do rather than some other. 
 
 t 
 
 • t p 
 
 CHAPTER IV 
 
 Thk Rkal as That which is Related. 
 
 A great number of the events that pass before our con- 
 sciousness seem to accord with one another quite well ; 
 they form, so to speak, a coherent group of phenomena. 
 All of the members of this group are, for example, in space, 
 each one being at a certain distance from each other one. 
 Or they are all in time; any one is simultaneous with, or 
 before, or after, some other one. Perhaps we consider 
 them as all being subject to the same laws ; there are strik- 
 ing similarities, we say. in the way all of them behave. If 
 we have near the earth an unsupported body whose density 
 is greater than that of the air around it, it always falls to 
 the ground. Whenever a body is tilted so that its center 
 of mass falls beyond its base, the body always, if left to 
 itself, tumbles over. All of these entities form, in short, a 
 system of coherent, related phenomena. But there are 
 other objects that we are sometimes aware of that do not 
 fit into this system. In our dreams, in books, or even in our 
 waking life, we are aware of objects that behave in ways 
 that do not accord with our usual experience. We meet 
 with dogs that speak, with fairy godmothers, with giants 
 and ogres and centaurs and unicorns. Now, it is not un- 
 usual to call these objects that do not agree with our ordi- 
 nary experience unreal and to call those that fit into a 
 coherent system real. Using 'Veal" in this sense, entities 
 that form part of an interrelated whole are real, and those 
 outside of this system are unreal. The real entity is the 
 one that coheres in an ordered universe of related phenom- 
 ena. And the unreal entity is the one that is eccentric, 
 peculiar, unrelated to the world with which we are accus- 
 tomed to deal. 
 
 Basing his interpretation on certain passages in the 
 Saggiatore,^^ Cassirer*^"* says that Galileo holds that a large 
 part of science is to be concerned with the derivation of 
 
 83 
 
 M 
 
 60 
 
 390. 
 
 Galileo: Opcre, vol. 4, pp. 174, 258 (Edition by Alberi). 
 
 Cassirer : Das Erkenntnisprohlem, Second Edition, vol- i, pp. 389, 
 
 61 
 
plienomena from one another, and that "as long as this 
 derivation is not reached we possess no guarantee that we 
 move in the reahn of true Being and not in an invented 
 world of fable." That is to say, it is the entities that form 
 an interrelated system that are real, and tlie isolated, un- 
 related ones that are unreal; and it is only when we are 
 aware of its interrelations that we can he sure a phenomenon 
 before us is real. An entity related to the general system 
 of entities, accordingly, is real. The term "real" is used 
 to point to these entities; it uicans these entities. We have 
 before us. consequently, one of these cases w^here "real" 
 points to the component parts of a coherent system, and 
 where "unreal" points to the unrelated entity that stands 
 outside of this system. An entity to he "real" in this 
 sense of the term must sustain certain relations to other 
 parts of the universe. 
 
 But, we may ask. what sort of relations must an entity 
 have in order thus to he called "real?" It is not, accord- 
 ingly to Cassirer. on the basis of its co-ordination with other 
 individual phenomena that (lalileo calls an entity "real." 
 It IS only when the object in question agrees with the 
 generalizations that are called scientific laws that it is real. 
 When a phenomenon is consistent with the.-e laws it is real, 
 when it is inconsistent with thetn it is unreal, and when its 
 relation to these laws of nature is not known we ix>ssess 
 no assurance of its reality. Can we say that such a notion 
 of reality, whether it be really Galileo's or l)e incorrectly 
 attributed to him by Cassirer. is true or false? We can, 
 indeed, ask whether, by an application of this definition 
 to individual entities, the same entities would be called 
 "real" that are so called in ordinary parlance. But it would 
 be extremely hard to find out except in the most unenlight- 
 ening terms w^hat entities are called "real" in ordinary par- 
 lance. For the present, then, let us be content simply to 
 regard this notion of reality as involving a conceivable 
 definition of the term "real." It is not to 1)e described as 
 true or false, but merely as possible. 
 
 In our dreaming as in our waking life we are aware of 
 objects. In the one as well as in the other there passes 
 before our consciousness a series of events of the greatest 
 variety — battles, conversations, murders, loves. Jjut the one 
 
 62 
 
 
 M 
 
 i 
 
 >j 
 
 >» 
 
 series of events is usually called "unreal" and the other for 
 the most part called "real." Now when we start out with a 
 determination to call dream entities "unreal," we are usuallv 
 led to a use of the term "real" according to which this term 
 is applied to phenomena belonging to a system of inter- 
 related entities, and is not applied to unrelated entities. For 
 when we attempt to find the marks that distinguish dream 
 objects from others, the distinction that is most frequently 
 hit upon is the incoherence and absurdity of the former, the 
 impossibility of fitting them into our experience as a whole, 
 and the coherence and unbroken connections betw^een the 
 |)ai-ts of the latter, ft is not the lack of vividness that dis- 
 tinguishes dream objects, for, as Hobbes says,^' "they are 
 clearer than the imaginations of waking men, except such 
 as are made by sense itself, to which they are equal in clear- 
 ness." No, the distinguishing characteristic of dream 
 objects is their unrelatedness; it is this unrelatedness that 
 makes them dreams and makes them unreal. Consequently, 
 when we call dreams "unreal," we are usually using "real 
 in the sense in which it means the related entity and "unreal 
 in the sense in which it means the unrelated one. "Real," 
 accordingly, is being given a meaning similar to that which 
 we have just discussed in connection with Galileo. And 
 "unreal" again means the strange incoherent object that 
 stands outside of the system of entities that are related to 
 (^ne another and that behave in similar ways. 
 
 Descartes comes to use "real" in this sense when he con- 
 siders the characteristics of dream entities. "And I ottght," 
 he says at the end of his "Meditations," "to set aside all the 
 doubts of these past few days as hyperbolical and ridiculous, 
 especially that very cotnmon uncertainty respecting sleep, 
 which 1 could not distinguish from the waking state; for 
 at present 1 find a very notable difference between the two, 
 inasmuch as our memory can never connect our dreams one 
 with the other, or with the whole course of our lives, as it 
 unites events which happen to us while we are awake. And, 
 as a matter of fact, if some one. while I was awake, quite 
 suddenly appeared to me and disappeared as fast as do the 
 images which I see in sleep, so that I could not know from 
 whence the form came nor whither it went, it would not 
 
 '^' Hobbes : Dc Corporc, part 4, chapter 25, section 9. 
 
 63 
 
 4 
 
 ti»' 
 
be without reason that I should deem it a spectre or a 
 phantom formed by my brain [and similar to those which 
 T form in sleepl, rather than a real man." 'Phat is to say, 
 dreams are unconnected and wakino life is connected. And 
 dream objects, it is implied, are unreal, and the connected 
 objects of wakin- life real. "Real" accordin-ly. is bein- 
 used to mean that which is related and "unrear^ to mean 
 that which is unrelated. And the interrelation that ^.s in- 
 sisted upon with respect to real objects is not merely a con- 
 sistency with, and deducibility from, the general laws of 
 science. A phenomenon to be real, as Descartes seems to say 
 in the same paragraph, must be one such that no evidence 
 can be brought forward by the senses, the memory, or the 
 understanding, that is repugnant to it. 
 
 Let us stop here for a while to examine this notion of 
 repugnance. Let us see what sort of evidence might l)e 
 brought forward that would be repugnant to a given object. 
 There are undoubtedly propositions that Galileo and 
 Descartes would have 'considered intelligible laws. We 
 may mention as examples the rule that every event has a 
 cause and the law that all bodies fall to the earth with a con- 
 stant acceleration. Accordingly, a phenomenon that had no 
 cause, or one whose motion towards the earth took place in 
 a different fashion. w(nil(l be a phenomenon repugnant to 
 these intelligible laws. Since, moreover, consistency with 
 these laws is a requisite of reality, such an entity would be 
 an unreal entity. Now such a rule seems to me to give a 
 perfectly self-consistent method of separating out certain 
 phenomena as unreal, though the reach of phenomena to 
 which it may be applied is limited by the number of mtelh- 
 gible laws that are definitely laid down. One cannot, for 
 example, applv this rule to exclude from reality my riding 
 in an aeroplane last Tuesday unless there is some intelligible 
 law with which >uch a phenomenon is inconsistent. And 
 these laws must, 1 say, be laid down with some definiteness ; 
 otherwise it would be impossible to apply them. It may be 
 mentioned that, if reality is defined in this way, the particular 
 laws that are laid down 1)ec()me immune from overthrow 
 by what are known as negative instances, for the negative 
 instances are thrown out by definition as unreal entities. 
 But it is none the less possible to lay down definite proposi- 
 
 64 
 
 tion^ as intelligible laws and to call those entities that are 
 consistent with them real and those that are inconsistent 
 with them unreal. We have here, consequently, another 
 definition of reality, or rather many definitions, since the 
 content of reality will varv' as the specific propositions that 
 are laid down as intelligible laws vary. It is definitions of 
 this sf>ecies that Galileo and Descartes imply, I think, insofar 
 as they use the term "real" to point to entities that are con- 
 sistent with the eternal verities that may be called intelligible 
 laws and the term "unreal" to point to entities that are incon- 
 sistent with them. 
 
 But Descartes also requires that a phenomenon that is to 
 be called "real" be consistent with the evidence of the senses 
 and the memory, with, in short, "the whole course of our 
 lives." That men should be turned into stones, or should 
 come to life again when once dead, does not accord, we say, 
 with our experience. But there is a difference to be noticed 
 between the sort of accordance spoken of here and that 
 spoken of above. In the absence of the sort of consistency 
 that was demanded above there is real contradiction. That 
 event A should occur without a cause is contradictory to 
 the law that all events have causes. But the assertion that 
 a man has turned into a stone does not assert a phenomenon 
 that, precisely speaking, is contradicted by our experience. 
 Such a phenomenon is one that our experience has not 
 accustomed us to expect; but our experience being limited 
 to particulars cannot contradict it, cannot render it impossi- 
 ble. Such phenomena are sometimes experienced ; visions 
 of the Virgin Mary are by no means unknown. But they 
 are phenomena that we are not accustomed to experience. 
 And so a phenomenon that does not accord with our ex- 
 perience is one that is of a kind that we do not usually 
 experience. It is a rara azns that is dissimilar to w^hat w-e 
 usually meet with in the course of our lives. Such a char- 
 acterization of the unreal, however, is as yet, I am afraid, 
 too vague. It is not sufficiently definite to afford us a means 
 of picking out any group of specific events as unreal. For 
 what is the usual? The experiencing of devils w^as not 
 at all unusual in the Middle Ages, nor w^as the experiencing 
 of witches unusual in the sixteenth century. To determine 
 what is usual, we are forced to engage in a counting of 
 
 65 
 
numbers with the range of the field of entities that are to be 
 counted quite undetermined. If we Hmit ourselves to the 
 experience of the sixteenth century wdtches are not unusual, 
 though if we take into consideration the experience of sub- 
 sequent centuries they are. Again, substances that give off 
 emanations are unusual in our experience, though pieces of 
 radium that give off such emanations are not rare. Witches 
 that can fiy through the air are not rare ; most witches have 
 that power. But women in general who go flying through 
 the air on broomsticks are met with quite seldom. What 
 is usual, in short, depends entirely on the range of phe- 
 nomena we take into consideration. It depends on what 
 subjects we pick out whose experiences are to be investi- 
 gated, and on how narrowly we circumscribe the phenomena 
 whose reality is to be considered. 
 
 And so when we simply say that an entity to be real must 
 accord with our experience and do not lay down more 
 specifically the sort of consistency that is to l)e demanded, 
 we are without a rule by which we can exclude entities from 
 reality with any definiteness. We must have some specific 
 sort of consistency with experience demanded, so that we 
 can tell by applyitig this rule to an individual phenomenon 
 whether that phenomenon is meant to be called "real" or 
 not. More concrete demands of this kind are sometimes 
 laid down. Sometimes it is said that an entity to \yQ real 
 must be experienced not only by ourselves but by those 
 around us : A ghost that I see but that no one in the room 
 with me sees is unreal. Such a rule seems to me to give a 
 serviceable definition of reality. If we make such a demand 
 of reality, only that is "real" which is experienced by more 
 than one, while the realm of the unreal includes all those 
 entities which are experienced by but a single consciousness. 
 Or it is demanded that for an entity to be 'Veal" one must 
 l)e able to repeat it and to predict it. The phenomenon must 
 be a recurring phen(^menon and we must be able to predict 
 its behavior when it does recur. Here too we have a rather 
 specific definition of reality. It enables us to call radium 
 emanations "real" and Jacob's ladder up to heaven "unreal." 
 I believe that the class of entities marked out as unreal by 
 each of these definitions excludes some entities that are 
 commonlv called "real." 1 believe common sense, for 
 
 66 
 
 i 
 
 1 
 
 example, would call many phenomena that occur but once 
 "real" and would call the creation of some entrancing color 
 a real event, though perhaps it could not be repeated. But 
 I do not want to enter into this very deeply, for I find it 
 quite impossible in many cases to tell w^hat common sense 
 w'ould call "real." The naive man's use of the term "real" 
 is as inchoate as the contradictions betw^een the many senses 
 in which philosophers have used the term would have led 
 one to expect it to be; and his use of it has varied from 
 generation to generation as the philosopher's has, from the 
 Middle Ages when devils were "real" to now when they are, 
 I suppose, "unreal." But these two definitions are, none the 
 less, ways in which "real" may be defined. 
 
 In mentioning these two more specific sorts of consist- 
 ency with experience that may be demanded of real enti- 
 ties, I have anticipated somewhat. For Leibniz is the first 
 philosopher in whose writings I have found them laid down. 
 
 Leibniz uses the "real" to mean "that which is related" 
 to a much greater extent than does an}- previous philoso- 
 pher. With him the notion is by no means thrown out in 
 stray passages ; it is a notion that is expressed f requendy 
 and with considerable elaboration. The world of real enti- '^ 
 ties is conceived as a system of interrelated compossible 
 entities. And on the other hand, an unreal entity is one 
 that finds no place open for it in this system, one that lacks 
 the manifold relations that characterize real entities. For y 
 the world of real entities is chocked full of relations; each 
 one of these entities enters into relations with the others so 
 that all together they form an organic system in w'hich each 
 bit is essential. We cannot tamper with the slightest part 
 without having the whole universe, so to speak, come tumb- 
 ling down on our hands. And while all the entities within 
 the system are inextricably bound to one another, they are 
 jointly and severally free from relations w'ith entities out- 
 side it. And so we have this notion that the real is that 
 which is related quite unmistakably before us. xAn entity 
 is "real" if it belongs in the system, if it sustains the sort 
 of relations that all real entities do sustain towards one 
 another. And it is "unreal" if it comes without antecedents 
 and goes without consequents, a stranger that has no con- 
 nection wath the interrelated world we know. 
 
 67 
 
But just what sort of relations are we talking about? 
 What sort of relations does an entity that is within the 
 system and is to be called "real" sustain ?' And what sort 
 of relations does an "unreal" entity that is outside the 
 system lack? As with Descartes, what are demanded of 
 real entities are relations with the intelligible laws and with 
 the course of our experience. "The basis of the truth of 
 contingent and singular things is in the succession which 
 causes these phenomena of the senses to be rightly united 
 as the intelligible truths demand."''^' And a phenomenon to 
 be real must agree with experience. ''Undoubtedly/' says 
 Leibniz,'' • "the strongest proof" (of the reality of phe- 
 nomena, he means) "is the agreement with the whole course 
 of life." Entities that do so agree with the whole course 
 of life he uses the term "real" to denote. And entities that 
 lack this consistency wnth experience he calls "unreal." 
 "We can neither know nor ought we to desire anything of 
 sensible things than that they harmonize as well among 
 themselves as with indubitable reasons, and in such a way 
 that future things may in a certain degree be foreseen from 
 past things. Any other truth or reality will be sought in 
 them in vain than that which this vouches for. nor ought 
 sceptics ask anything else nor the dogmatics promise it.«^ 
 For it is these interrelated entities, as I look at it, that the 
 term "real" means. They are "real" because Leibniz is 
 calling them "real." And men moving through the air, 
 "sitting upon the hyppogryphs of Ariosto,""-' are "unreal" 
 because they do not accord with experience and therefore 
 belong to the class of entities that Leibniz designates by the 
 term "unreal." That which is related is the real and that 
 which is unrelated the unreal. This is the meaning Leibniz 
 gives these words. At times, indeed, there is a more definite 
 description of the characteristics of the related entities that 
 are real. Real entities are now those that appear in the 
 experience of others and now — and this, indeed, is insisted 
 
 ^Leibniz: New Essays on the Human Understanding, Bk, 4, ch. 4. 
 section 4. 
 
 *'Langlcy: Leibniz's New Essays, etc. (1916), p. 718; Gerhardt. vol. 
 
 7, p. 320. 
 
 ** Duncan : Philosophical Works of Leibniz, 2nd ed., p. d8— Anim- 
 adversions on Descartes's Principles of Philosophy"— On Article 4. 
 
 **Langley: New Essays, etc. (1916), p. 718; Gerhardt, vol. 7. P- 320- 
 
 68 
 
 I 
 
 
 on more frequently — they are entities that can be foretold. 
 But it is not their vividness that makes these objects real 
 and not their clearness and distinctness; it is their related- 
 ness. 
 
 To a considerable extent, however, even entities that are 
 related are not surely real. A *Veal" entity becomes quite 
 mysterious and its properties quite undiscoverable. There 
 is hardly any entity that one can put one's finger on and say 
 with assurance that it is denoted by the term "real." Even 
 if an entity has all of the relations that have been mentioned, 
 we still cannot be quite sure that it is meant by the term 
 "real." "By no argument can it be absolutely demonstrated 
 that there are bodies, nor anything keep certain well-ordered 
 dreams from being objects to our mind w^hich are considered 
 by us as true, and on account of the agreement among them- 
 selves with respect to use are equivalent to truths. "^'^ Such 
 well-ordered dreams would still be dreams and as such 
 unreal. Most related entities are, indeed, real, but some, in 
 spite of their relations, may be unreal. We never can tell 
 wdiether an entitv before us is real or not, and that not 
 because we are unable to discover the qualities that go with 
 reality, but because there are no qualities — such as w^e have 
 taken relatedness to be — that invariably go wath reality — 
 because there is no definite class of entities that "real" is 
 being used to point out — because, in short, the term "real'* 
 is being used without meaning. If, no matter how much 
 of relatedness or of whatever other quality you please w^e 
 admit in bodies, we still can not say that bodies exist, then 
 ^'existence" is being used without meaning. Agnosticism 
 is a conceivable theory if reality is given some rather definite 
 content and that content declared undiscoverable. But it is 
 meaningless if the term "reality" that points to what is for- 
 ever hidden points to nothing that can be conceived. And 
 that is the sort of agnosticism we have here. Reality has 
 no definite characteristics; related entities as w^ell as un- 
 related ones may be unreal; the term "real" points to noth- 
 ing. 
 
 Yet in spite of his wavering it was Leibniz who brought 
 into fashion this notion that the "fictitious" phenomenon 
 is the strange event out of accord with experience and the 
 
 »0 
 
 Langley : New Essays, etc., p. 719; Gerhardt; vol. 7. p. 320. 
 
 69 
 
^Veal" event the one that belongs to an interrelated system. 
 The romance of Astraea is unreal because it does not fit in 
 with our world; it could only be real if all our world were 
 different.^^ Dream objects, hyppogryphs, and all such en- 
 tities are unreal because they are strange phenomena that 
 stand outside of the connections that relate the various parts 
 of our ordinary experience and are not subject to the uni- 
 formities that hold there. Christian Wolff takes up the 
 same theme ; he uses the term '^unreal" to point to the same 
 class of entities. In a dream ''while you look at someone, he 
 suddenly changes into someone else or he vanishes straight- 
 way and no one comes back to take his place. "^^ Things 
 hai)pen in a strange, haphazard, and unreasonable manner. 
 And it is this that distinguishes them from real entities and 
 makes them dreams. In the interrelations of real entities 
 there is order, and in dreams there is none. Consequently, 
 "order distinguishes truth from fable."^^ The "real'' is 
 the ordered, the related; and, since all the relations that 
 hold between the ordinary objects of our experience are 
 summed up in the principle of sufficient reason, a real entity 
 is one that is subject to the principle of sufficient reason. 
 "If the principle of sufficient reason is removed, the real 
 world turns into a fictitious world in which we must look 
 to the will of man rather than to reason for the explanation 
 of the things that happen. ^^ It is order, then, subjection 
 to the principle of sufficient reason, that "real" means. An 
 ens veruni, a real entity, is defined as one in which "order is 
 given in the qualities that meet in it."^^^ But what sort of 
 order is required is not said; the meaning of the term is left 
 more formal and indefinite than by Leibniz. 
 
 According to Wolff, if we take away the principle of 
 sufficient reason the real world turns into the world of fic- 
 tion. We can have no reality at all without it, or, in other 
 words, order, relatedness, is a presupposition of reality. 
 Of course, relatedness is a presupposition of reality taken 
 in the sense in which Wolff uses that term. For. for him 
 " Letter to Bonrgct. Latta : Monadology, etc., p. 64. note 2 ; Gerhardt, 
 
 vol. 3, P- 572- 
 
 *= Wolff: Ontologia, § 493- 
 
 ''Ibid: §494- 
 
 ^ Ibid: § 77. Cf. also: Verniinftiges Gedanken von Gott, dcr Welt, 
 
 auch alien Dingen iibcrhaupt, § 142, ct scq. 
 Ontologia, § 4«>6. 
 
 70 
 
 i 
 
 i 
 
 I 
 
 60 
 
 f 
 
 reality means relatedness, and naturally, if there is no 
 relatedness there can be no reality. It is just as correct, 
 however, to say that reality is a presupposition of related- 
 ness. For if all reality were done away with, there could 
 be none of the relatedness that is equivalent to it. The 
 assertion that relatedness is the presupposition of reality 
 comes to no more and to no less than the definition of the 
 real as the related. It is defining the term "real;" though, 
 in view of the indefinite character of the demand that an 
 entity be consistent with the whole course of our lives, it is 
 giving a definition that is not very serviceable. And it is 
 one definition out of many. The "real" may be defined as 
 that which is clear and distinct or as the vivid object of 
 consciousness as easily as it may be defined as the related. 
 And so when Wolff and Kant talk about the presupposi- 
 tions of reality, they are not coming to the subject with a 
 more fundamental consideration than anyone else. They 
 are bringing forward their own notion of reality, a notion 
 that requires further elaboration if it is to be at all service- 
 able in distinguishing the real from the unreal, but one 
 that at least potentially is a definition of reality. 
 
 In Kant the notion that the real is that which is related 
 IS rather common. In our discussion of the view that identi- 
 fies the real primarily with sense-data, we have already 
 considered those passages in the "Critique of Pure Reason" 
 that say that "whatever is connected according to empirical 
 laws with a perception is real."^^ In those passages per- 
 ceptions are called "real," certainly. But so is the whole 
 group of entities connected with them, the entities that 
 with them form an ordered system of interrelated entities. 
 From one point of view "real" in these passages means 
 perceptions and the entities connected with them. But from 
 another point of view it means the elements in an ordered 
 concatenated whole of which perceptions are a part. How, 
 now, are we to tell an entity that is connected with a per- 
 ception according to empirical laws from one that is not? 
 In short, what sort of relatedness to a perception must an 
 entity show in order to be "real"? We have noticed above 
 two sorts of relations that may be demanded of entities 
 that are to be called "real." Either it is required that such 
 real entities be in accord with the whole course of experi- 
 
 "Kant: Critique of Pure Reason, ist edition, p. ^7^, Cf. also p. 22S- 
 
 71 
 
ence. Or it is required that they do not contradict definite 
 intelligil)le laws. In Kant's case it is the latter requirement 
 rather than the former that is emphasized. A ''real" entity 
 is one that is connected with a perception in the way that 
 the concepts of the understanding require. And more par- 
 ticularly, a "real" entity is one such that its relation to a 
 perception does not contradict the a priori laws laid down 
 in the "Analogies of Experience." P'^or a phenomenon to 
 be called "real" then, it mubt not contradict the law that the 
 quantity of substance is permanent, the law that every event 
 has a cause, or the law that there is dynanucal interaction 
 between contemporaneous entities. I describe these laws as 
 being laid down so that existence may be defined in terms of 
 them; Kant calls them i)resuppositions of existence. It is 
 through the concepts of the understanding alone "that 
 knowledge and determination of an object ])ecome possi- 
 ble. "^^ With respect to the second of these three laws he 
 makes this assertion more frequently. It is the relation of 
 cause and effect, he says, "which forms the condition of the 
 objective validity of our empirical judgments with regard 
 to the series of perceptions, and, therefore, also the condi- 
 tion of the empirical truth of them and of experience. "^^ 
 And again :^^ "The law of nature that everything which 
 happens has a cause, .... this law, through which alone 
 phenomena become nature and objects of experience, is a 
 law of the understanding which can on no account be sur- 
 rendered and from which no single phenomenon can be 
 exempted ; because in doing this we should place it outside 
 all possible experience, separate from all objects of possi- 
 ble experience, and change it into a mere fiction of the mind 
 or a cobweb of the brain." There are two ways in which 
 these laws can become the conditions of existence. Either 
 existence has first a definite meaning, and then this meaning 
 is found to be such that whatever is real is consistent with 
 these laws ; or else these laws are first laid down, and then 
 the "unreal" is defined as what is inconsistent with them 
 and "real" given a definition to correspond. The former 
 is undoubtedly the more usual way of looking at the rela- 
 tion between these laws and reality. But if reality is such 
 a vague concept, meaningless until defined, we cannot make 
 
 ""Kant: Critique of Pure Reason, ist edition, p. 310. 
 ^ Ibid: p. 202. 
 
 ^fbid: p. 542. 
 
 72 
 
 it a point of departure in our argumentation without first 
 defining it. We may say that Kant means by reality those 
 partieiiLir entities that are consistent with these three laws 
 and that thence he concludes that these three laws are con- 
 ditions of existence. But, assuming now that reality has 
 been used in many different senses, a certain amount of 
 arbitrariness must come in at some point. And so it is 
 quite as correct to say that Kant lays down these three laws 
 and means by the "real" that which is consistent with them 
 and by the "unreal" that which contradicts them. 
 
 Inconsistency here means real contradiction. We are not 
 troubled with the demand for a vague "accordance with 
 experience" that depends on whose experience and on how 
 much experience we consider; the "unreal" is that which 
 actually contradicts one of three specific laws. An appari- 
 tion that comes and goes again, that causes for a few 
 minutes a sudden increase in the quantity of substance in 
 the universe, is "unreal." So is a ghost that has no cause, 
 that springs up from no place; and so is an isolated phe- 
 nomenon that seems neither to affect nor to be affected by 
 the entities around it. These laws, as I remarked in dis- 
 cussing Galileo, are immune from overthrow by negative 
 instances. If I am aware of an apparent increase in the 
 amount of substance in the world, the increase of substance 
 is unreal. The term "unreal" means all phenomena that 
 increase the total quantity of substance. If we find an 
 event without a cause, we do not have to investigate fur- 
 ther to determine its reality; the term "unreal" by defini- 
 tion points to just such an object. We have here a per- 
 fectly possible way in which the term "real" may be used. 
 The difficultv, of course, is that these laws are far from 
 definite. If we were told just what a cause is and just what 
 constitutes dynamical interaction, our definition of reality 
 would be much more serviceable. If dynamical interaction 
 were something quite clear and specific, there would be cer- 
 tain entities that we should immediately be able to call 
 "unreal." But with "dynamical interaction" a vague and 
 unexplained term, no entities immediately appear to be 
 unreal. No entities are such that they cannot be conceived 
 to fit in with experience in some way. 
 
 The real is sometimes said to be that which is active, that 
 which has power to do and capacity to suffer, that which 
 
 73 
 
can act and be acted upon. Plato considers this sense of the 
 word ''real" in the ^'Sophist." He says, "My notion would 
 be that anvthing which possesses any sort of power to affect 
 another, or to be affected by another, if only for a single 
 moment, however trifling the cause and however slight the 
 effect, has real existence; and I hold that the definition of 
 being is simply power."^^^ But what does it mean to be 
 active? How can we pick out those entities that have power 
 from those that haven't it? Surely no entities come to us 
 pretending to be inactive. All seem able to do and to suffer. 
 Dragons can slay, and, with courage and good fortune, can 
 be slain; devils arc fearful only because of their power for 
 evil. Therefore in order to get a definite content for the 
 inactive and a limited content for the active, the notion of 
 activity must be defined and made specific. Often, however, 
 we find it said that the real is the active and find no further 
 explanation. Aristotle distinguishes between the potential 
 and the actual. He too calls the real the actual, that which 
 possesses activity, that which has an entelechy. For Leibniz 
 again the real is that which possesses force. And several 
 Germans since his day, perhaps on account of the common 
 ancestry of the noun "IVirklichkcif and the verb "unrken'' 
 have described reality in similar terms. For example 
 Platner, whom we shall mention later in another connection, 
 says that to exist is to act and to act is to exist. ^^^ But all 
 of these definitions of reality tell us very little. We are 
 almost as unable to distinguish the real from the unreal wdth 
 them as we should be without them. What we need is to 
 have this concept of activity defined in such a way that the 
 inactive will have some content. To say that the real is the 
 active is as valueless as it is to sav that the real is that which 
 Stands in dynamical communion with other entities. Indeed 
 the definition of the real as the active is exactly equivalent 
 to Kant's assertion that for an entity to be real it must be 
 subject to the third law laid down in the "Analogies of 
 Experience." That which can act and be acted upon is that 
 which stands in a dynamical commtmion with the objects 
 contemporaneous with it. And just as the principle of 
 dynamical communion is, as we have seen, indefinite, so the 
 
 Ml 
 
 I 
 
 \ 
 
 IW 
 
 Plato : Sophist, 2466-2476. 
 ^" Platner: Philosophische Aphorismen, § 739, Note 
 
 74 
 
 notion of activity is indefinite. With neither concept can 
 we point to any entities and say with assurance that they 
 are real or that they are unreal. 
 
 The real, it has been said, is that which is related. More 
 specifically, reality has been limited to those entities that are 
 in accord with our experience or to those that obey certain 
 intelligible laws. More specific yet, however, is the defini- 
 tion of reality that defines real entities as those that are in 
 space and in time. For when we are told that all real 
 objects are related to one another by being in the same space 
 and the same time, we have a definition that gives the unreal 
 some content. Many objects appear not to be in space and 
 many objects appear not to be in time. If, consequently, 
 we define reality in terms of time and space, these objects 
 are forthwith unreal. Here then we have indeed come upon 
 something that may be used to mark out the real from the 
 unreal, something that enables us to say forthwUh that cer- 
 tain definite entities are unreal. 
 
 We have already met with a passage in Plato that lays 
 down the thesis that whatever is not in time is unreal. An 
 object that does not participate in time does not participate 
 in being. ^^^^ When we come down to Hobbes. we find a 
 similar attitude taken with respect to space. ''If the triangle 
 exists nowhere at all," he writes, ''I do not understand how 
 it can have any nature; for that which exists nowhere does 
 not exist. "^^^ Sometimes it is required of a real entity only 
 that it be in time, sometimes only that it be in space. But 
 more often the two requirements are joined. Reality is made 
 conditional upon the possession of both a position in time 
 and a location in space. As Crusius. one of the philoso- 
 phers who wrote shortly before Kant, puts it, to give an 
 entity that is merely thought — that is, merely possible — a 
 position in time and a place in space is to give it existence. ^^"^ 
 "If a substance is to exist, it must exist immediately in some 
 place and at some time."^^^ 
 
 *" Plato; Parmenides, 141, 152. 
 
 '"** Hobbes: The Third Set of Objections to Descartes' Meditations- 
 Objection Fourteenth; Descartes: Oeuvres, ed. by Adam and Tannery, 
 vol. 9, p. 150. 
 
 ^** Crusius: Bntwurf der Notwendigen Verminft-Wahrheiten, 1753, 
 Sec. 46. 
 
 ^" Ihid; Sec. 57. See also Sec. 59. 
 
 75 
 
 -3 
 
Here we have a serviceable definition of reality, one that 
 enables us to pronounce with assurance upon the ontological 
 status of many objects. It does indeed make many objects 
 unreal that are often called real. It rules out of existence a 
 supra-spatial God, and the '^eternal verities," if they are 
 indeed eternal. The meaning of real, if the term is used in 
 this sense, is not exactly that which it has in common speech ; 
 but it is none the less rather definite. 
 
 There are one or two subjects, however, on which 
 this definition is as yet not quite specific. After Crusius had 
 described the real as that which is in time and in space, 
 Platner, in commenting upon this definition, raised an 
 objection. He objected because such a definition, he said, 
 presupposes the existence of time and space themselves. ^^® 
 In our own day much the same objection has been raised by 
 Marvin. ^^"^ Notwithstanding these probably independent 
 expressions of dissatisfaction with the definition we are con- 
 sidering, I find it difficult to tell what lies at the basis of 
 them. If by ''time" is meant the entities in time considered 
 collectively, I see no legitimate reason for dissatisfaction. 
 For if these entities, taken individually, are real by the defi- 
 nition of *'real," then they are real taken collectively. If by 
 *'time," in short, we mean the system of entities in time, then 
 ''time," on the definition we are considering, must likewise 
 be real. To be sure, the definition we are considering is a 
 definition that involves an ontology. But in this respect, we 
 shall find,^*^^ the definition w'e are considering is like any 
 other definition of reality that is precise and specific. There 
 may, however, be another meaning behind the statement 
 that a definition of reality in terms of time and space pre- 
 supposes the existence of time and space themselves. By 
 time itself there may be meant a time continuum that is 
 distinct from the system of temporal entities. And if this 
 be the meaning of "time," I can see how the definition as we 
 have it may be — not invalid — but ambiguous. The defini- 
 tion as we have it does not specify whether the continuum, 
 as contrasted with the entities in it, is, or is not, to be called 
 "real." This indeterminateness, however, does not seem to 
 
 loe 
 
 Plainer: Aphorismen, Sec. 739, Note. 
 "■"' Marvin : The Existential Proposition; Journal of Fhilosopk^, etc., 
 vol. 8 (191 1 ), p. 490. 
 ^'^'^ Infra, page 1 01. 
 
 m 
 
 \ 
 
 76 
 
 me to be unavoidable. We need only frame our definition 
 so that it becomes evident that time and space themselves are 
 meant to be called real or that they are meant to be called 
 unreal. Our definition needs only to 'be made specific on this 
 point; that is to say, it needs to be framed so as to permit 
 only one of these two possible interpretations. 
 
 Another ambiguity resulting from the identification of 
 the real with that which is in time and in space is more im- 
 portant. If reality refers to those entities that form one 
 system, to those entities each of which stands in certain 
 spatial and temporal relations to all the other real entities, 
 then there is but one space and one time. But if an entity 
 is called real provided it belong in any sort of space, then 
 there may be more than one space. If our definition is 
 taken in one sense, all entities that do not belong in the one 
 system selected are unreal. But if our definition is taken 
 in the other sense, some of these entities — namely, those 
 that are characterized by a spatial position of some sort — 
 may be real. The distinction is important in dealing with 
 the "ideas" of the so-called epistemological dualist. These 
 *'ideas," as sometimes conceived, are not in the one objective 
 space, but they are not devoid of all spatial qualities. The 
 house that is my idea is not in objective space ; it is not in 
 the spatial system that contains the real house. But it is to 
 the left of a roadw^ay that is also my idea; both the house 
 and the roadway have positions in a subjective space. Con- 
 sequently, if to be real is to be in a space, to have a spatial 
 position of some sort, this house that is my idea may be real. 
 But if only those entities that form a single system of inter- 
 related spatial entities are real, this house that is my idea 
 is unreal. 
 
 Kant, too, at times makes location in space and position 
 in time conditions of reality. While time and space are 
 transcendentally ideal, as he puts it, they are empirically 
 real. And every entity that is empirically real — that is to 
 say, real as a phenomenon — must be in time and in space. 
 At least all real objects except thoughts and feelings must 
 be in space. And all real objects without exception must be 
 in time. This condition, that all real objects must be in 
 time and all real objects except thoughts and feelings in 
 space, is given a priori. In this respect the situation is 
 
 77 
 
similar to that when reahty is defined in terms of intelhgible 
 laws that are laid down a priori. When that is said to be 
 real which does not violate the law that the quantity of 
 matter is always constant, this conservation of matter can- 
 not be disproved by experience. For any phenomenon that 
 seems to disprove the conservation of matter is thrown out 
 as unreal and disregarded. So here no experience can show 
 that reality is not conditioned upon position in time. For 
 any entity that seems not to be in time is by that very fact 
 to be called unreal. Time and space are given a priori as 
 conditions of reality in the sense that, having once been 
 laid down as conditions of reality, no subsequent experience 
 can show that they do not deserve that status. In a sense, 
 of course, Kant does not get them out of his head and lay 
 them down as conditions of reality. He makes several 
 psychological observations such as that we cannot form the 
 image of an entity not in time and space.^^^ But unless 
 reality is to be limited to those entities that can be imaged, 
 and this would be a notion of reality entirely distinct from' 
 the one we are considering, these psychological observations 
 have nothing to do with the terms "real" and "unreal." If 
 we declare that only those objects of which we can form 
 nnages are real, we may come out at the same i)oint. We 
 may arrive at the same conclusion, namelv. that only entities 
 with a position in time and space are real. But if 'reality is 
 (Mmud directly in terms of time and space, time and space 
 are a priori only because they are laid down as conditions 
 of reality. 
 
 Among our contemporaries position in time and location 
 m space are quite frequently made conditions of existence. 
 'T£xistent processes." says Montague.^"* "are those that 
 occur somewhere, non-existent are those that occur no- 
 where. The existent is in short that to which the Aristo- 
 telian category 'pan has a positive application/' But there 
 IS a tendency not to limit realitx so precisely. In the ''New 
 Realism" we find Montague saying: "The real universe con- 
 sists or the time-space system of existents together with all 
 that is presupposed by that system."'^ ^ We have, con- 
 
 109 1,'" /' ^ • • - T-. 
 
 Kam: Critique ot Pure Reason, ist ed.. p 24. 
 
 "VMontague: "Are Mental Processes in ' Space ^"-.Uo«^> toI 18 
 1908, p. 21. . • , 
 
 '''The New Realism, p. 255. 
 
 78 
 
 1 
 
 -<l 
 
 'I 
 
 I 
 
 sequently, the added task of finding out w^hat these entities 
 in time and space presuppose. It will be remembered that 
 in talking about sense-data, we came upon the thesis that 
 reality includes not only sense-data but all entities that are 
 implied by sense-data. And in the absence of a proposition 
 making the term ''implication" meaningful, w^e found it 
 impossible to distinguish betW'Cen those entities that are not 
 implied by sense-data and are unreal and those that are 
 implied by sense-data and are real. In the present case 
 we have the same dit^culty with the concept "presupposi- 
 tion." Of those entities that are not in time or in space, 
 which are presupposed by the time-space system and which 
 are not? Are extra-spatial ideas, of either the epistemo- 
 logical or the Platonic variety, so presupposed or are they 
 not? In order to decide this question, in order to give the 
 unreal some specific content, we must know wdiat "presup- 
 position" is. "Presupposition" and ''implication" — if "im- 
 plication" is equivalent to it — cannot be left undefined or 
 be defined in terms that assume we already know the mean- 
 ing of "real" and "unreal." 
 
 If reality is defined in terms of time and space we ought 
 to know whether time and space themselves are real. We 
 ought to be told whether an entity in any spatial system is 
 real, or only those entities that belong in one specific spatial 
 system that is pointed out. And if not only entities in time 
 and space but also entities that these spatial and temporal 
 entities presuppose are real, we ought to be told w^hat "pre- 
 supposition" means. There is one other point upon which 
 we rnay demand further information. Must an entity to be 
 real be at some definite place in space and time and not 
 vaguely located in all places, or is it only necessary that it 
 be not outside of space and time? For example, "Nature" 
 or the "cosmos" is not outside of space, but neither is it in 
 one place rather than in another. It has what the Scholas- 
 tics sometimes called "circumscriptive ubeity" but not "de- 
 finitive ubeity." It remains for us to specify in our defini- 
 tion^ consequently, whether entities of this kind may be real 
 or whether only entities with "definitive ubeity" are to be 
 accepted into the chosen circle. When these am])iguitie5 are 
 cleared up, time and space seem to be terms that can very 
 nell be used in the definition of reality. For manv of the 
 
 79 
 
entities with which we deal will lack the qualities that will 
 hcAt been made conditions of reality. There will be a con- 
 siderable number of entities that can immediately be seen 
 to be unreal. 
 
 The real is that which fits into a determinate order or 
 system of relations: this is the notion of reality we are 
 engaged in bringing out. '^Reality," when used in this 
 sense, has to do with a system of interrelated entities, a 
 system where experience is a whole, subject to certain 
 uniformities, held together by manifold interrelations. This 
 interrelatedness of experience, its essential unitv, is insisted 
 on pre-eminently by Spinoza. For in his metaphysics all 
 things are related to one another, all ha\ e in common their 
 derivation from God. And so we may expect to find in his 
 writings passages in which the term 'Veal" is used to refer 
 to entities that belong to the system of interrelated eruities 
 and the term "unreal" to entities ontside it. And there is 
 there this notion that the 'Veal" entity is the one that is 
 connected with experience and the unreal or non-existent 
 one the one that is not in accord with the laws of nature, 
 that violates the usual order of things. When we have a 
 definite phenomenon clearly before us. it exists, of course, 
 if its nature is stich that it must exist, and it does not exist 
 It its nature is such that it cannot exist. But. leaving aside 
 the help that this pronouncement affords us in determining 
 what entities are real, a clearly comprehended phenomenon 
 is real if it agrees with the order of nature and is not real 
 if It violates that order. *l.et us conclude again briefly,'* 
 says Spinoza, ''and see how it need in no wise be feared that 
 fiction will be confused with true ideas. As for the first 
 fiction of which we have spoken, where the thing is clearly 
 
 conceived we must only take care that its existence 
 
 be compared with its essence, and that attention is paid at 
 the same time to the order of nature."^ ^- Here is a notion 
 of reality similar to those we have been investigating ii 
 is not carried out in so much detail as we have found it 
 carried out by others. We have no definite statement of the 
 laws with which a phenomenon must be consistent in order 
 to be real ; and we are not told just w hat <nrt of consistency 
 with experience is demanded. But none the less "reality," 
 
 ^"^ Spinoza : On the Emendation of the Intellect, § 65. 
 
 80 
 
 m 
 
 or rather "existence," is being used to point to entities that 
 cohere in a system. 
 
 That all parts of the universe are connected with one 
 another is a frequently recurring theme in Spinoza. But 
 along with this doctrine there come two quite different 
 notions of reality. One is the notion we have examined, 
 the notion that an entity is real if it forms a part of this 
 interconnected whole. The other is the notion that only 
 the interconnected whole is real, and that all parts are un- 
 real Both may be said to describe the real as that which is 
 related. For one, "real" means the entity that sustains 
 certain relations and belongs to a system ; for the other, 
 "real" means a single object, and that is the whole inter- 
 related system. Indeed, the two senses of "real" do not 
 at all agree; all the individual phenomena that are "real" 
 in the sense that they belong to the system are "unreal" in 
 that they individually do not constitute the system. And 
 whereas, using "real" in one sense, unreality is made up 
 of entities that do not belong to the system, to many of 
 those using "real" in the other sense such unrelated entities 
 appear absolutely inconceivable. There are for them no 
 unrelated entities ; those entities that are called "unreal" 
 by those who use "real" in the first of these tw^o senses 
 belong in the system — in their opinion — quite as much as 
 do the entities that these others call "real." 
 
 Now, this notion that the whole organic thing called the 
 universe is real and that the unreal is the partial, fragmen- 
 tary bit of it is well recognized as Spinoza's. "To call 
 anything finite," he says, "is in reality a denial in part, and 
 to call it infinite is the absolute assertion of the existence 
 of its nature."^ ^^ When we are aware of the world in its 
 entirety we are aware of reality; when our attention is 
 limited to a small portion of the universe we are out of 
 contact with reality. "Falsity," reads one of the proposi- 
 tions of the Ethics, "consists in privation of knowledge 
 which is involved by inadequate or confused and mutilated 
 ideas."^^^ The universe as a related whole is real; its 
 parts are unreal. And the parts that are unreal are the 
 very parts that are real when Spinoza uses "real" in the 
 other sense. Generally speaking, "real" means for him the 
 
 U3 
 
 Spinoza: Ethics, Book i, Prop. 8, Note i. 
 
 '**Jbid: Book 2, Prop. 55- 
 
 81 
 
individual that is part of a system, when his attention is on 
 concrete phenomena. And he uses ''real" to refer to the 
 system as a whole, when he is dealing with entities at the 
 other end of the scale, with God and the eternal truths. 
 Since, therefore, Spinoza's interest lies so completely in 
 the realm of eternal truths, the term is used much more 
 frequently in this second sense, in the sense, that is, in 
 w^hich it points to the organized whole and not to the parts 
 of that whole. 
 
 In Hegel it is the organized whole that is real and not at 
 all the parts of that whole. The fragmentary parts of 
 experience are, indeed, said to have ''being*' ; but there is 
 no attempt, so far as I know, to distinguish between "being" 
 and "non-being" by saying that entities that belong to a 
 system have "being" and that those that do not liave "non- 
 being." "Being," that is to say, does not mean those enti- 
 ties that sustain certain relations and "non-being" does not 
 mean those entities that do not sustain such relations. Ap- 
 parently "being" means sense-data and "non-being" means 
 entities that are not sense-data. Consequently, reality — in 
 the sense of "being" — does not mean the part of a related 
 whole. The "real" is not used to mean that which is 
 related in this sense. And so the only sense in which the 
 "real" can mean the related is the sense in which it means 
 the relational whole itself. 
 
 it does, however, ilirou.c:h'>ut a large part of Hegel's 
 writings, certainly mean this related, organized whole, this 
 organized whole that may now be called Nature and now 
 the Absolute Idea. This unity "in which all characteris- 
 tics have coalesced"^ ^^' is the real; all more partial and 
 less inclusive entities are to some deqrec unreal. In this 
 Whole that is real are included entities w ithout "being" as 
 well as those with it; nothing is left out of the Absolute. 
 And so all conceivable entities are part of an interrelated 
 system; none are left out for those to call "unreal" who 
 would use the term to point t'^ entities that are not included 
 in the system of experience. 
 
 Among Hegel's English followers "real" again means 
 the organized whole in its entirety. But for some of them 
 it is also used in the sense in which it means the part of this 
 
 i 
 
 whole and excludes the entity that is not a part of it. Let 
 us take Joachim as an example. Reality is for him the 
 unified co-ordinated whole, the organized system of ex- 
 perience. "Truth in its essential nature," he says/^^ "is 
 that systematic coherence which is the character of a sig- 
 nificant whole." "A 'significant' whole," he adds, "is an 
 organized individual experience, self-fulfilling and self- 
 fulfilled." We do not get an entity that is able to stand on 
 its own legs, and so is self-fulfilling, until we get an entity 
 that is all-inclusive. And so "real" is being used to point 
 to this unique all-inclusive whole. But "real" is also used 
 to point to the individuals that cohere in the system and 
 "unreal" to point to individuals that do not belong in this 
 system. Or at least we can infer that "real" w^ould be so 
 used were Joachim talking about reality instead of about 
 truth. Immediate experiences, for example, are said to be 
 "real," or — as Joachim says (for he is talking about the 
 experiencing of those objects rather than about the objects 
 themselves) — are said to be "true." "Their 'truth' means 
 for us that a whole system of knowledge stands and falls 
 with them, and thai in that system they surv'ive."^^'^ This 
 is like Leibniz's assertion that for the romance of Astraea 
 to be real the whole world would have to be different. The 
 romance of Astraea was unreal because it did not fit in with 
 the system of things that is. It was an unrelated entity 
 and as such called "unreal." So, too, with Joachim it is 
 the unrelated phenomena, the parts that do not fit in, that 
 are "unreal." The immediate experience of the worshipper 
 of Baal is false because "the moral and religious experi- 
 ences of the past and present reveal themselves, when criti- 
 cally analyzed and reconstructed, as a texture into which 
 this immediate intuition can in no sense be woven ; they 
 form a system in which this would-be truth cannot as such 
 survive." That is to say, the real is that entity that fits 
 into experience, and the phenomenon that he sees who 
 sees Baal hurlins: thunderbolts is unreal in that it does not 
 accord with experience. The question is not as to whether 
 such an experience is self-fulfilling; it is as to whether it 
 "fits." 
 
 lie 
 
 Hegel: Encyclopedia of the Philosophical Sciences. Fart i 
 
 (Logic), § 236. 
 
 82 
 
 uo 
 
 117 
 
 Joachin : The Nature of Truth, § 26. 
 Joachim : The Nature of Truth, § 19. 
 
 83 
 
''Real" consequently is used in both senses, frequently 
 by the same author. It means the individual phenomenon 
 that is part of a co-ordinated interrelated experience; and 
 it means the unified organized whole. A criticism of the 
 first of these two notions would be a repetition of what 
 has gone before. It would reduce itself to the objection 
 that we have yet to be told precisely what sort of related- 
 ness is demanded of real entities. The other notion of 
 reality is a possible definition of the term; though here 
 reality is identified with an entity which, being one "in 
 which all characteristics have coalesced," is necessarily in- 
 definite. What I want to call attention to here, however, 
 is the absolute incompatibility of these two definitions of 
 reality with each other. \i "real" means the organized 
 whole, then the parts of that whole are "unreal." Using 
 "real" in one sense, the parts of experience, considered 
 separately, are unreal; using it in the other sense, it is 
 just these parts that are real. We liave come upon a con- 
 trast that shows the utter lack of uniformity in the use 
 of the term "real," for here are a grotip of entities that 
 it is being used now to point to, now to exchide. We see 
 the word's utter meaninglessness apart from a definition 
 of it, be it explicit or he it imi)licit in the context in which 
 the term appears. W^e see, in short, that we can do nothing 
 with the term "real" before we define it. 
 
 I. 
 ■ 
 
 CHAPTER V 
 
 The Problem of the "Real Centaur/' BeliEE 
 
 AND Existence. 
 
 Up to this point we have been moving slowly but we 
 have not encountered any difficulties that have threatened 
 to upset our plans. We have discovered several definitions 
 of reality, painfully detached them one from another, and 
 set them, forth one by one. But here it occurs to me that 
 perhaps our whole search is in vain. Perhaps it is im- 
 possible to define reality at all. necessary as such a definition 
 is if we are to make any use of the term. 
 
 For, suppose that I am thinking of a centaur or a 
 mountain of gold. These objects are fictions, objects of 
 the imagination if you will, but they are none the less 
 objects of my consciousness. Let me now suppose that the 
 entities of which I am thinking have in addition the quality 
 "existence" or "reality." I am thinking of an existing 
 centaur; the object with which my consciousness is toying 
 is a real mountain of gold. It is these existing centaurs 
 and real golden mountains that make us pause. It is they 
 that can lead us to think that perhaps the definitions we 
 seek are, from the nature of the case, impossible. For they 
 bring us face to face with a difficulty that is extremely 
 hard to overcome. 
 
 Among the subsistents, the objects that may be subsumed 
 under the all-inclusive category of "being," we find such 
 entities as existing centaurs and real golden mountains. 
 Apparently these entities belong in the class of real beings, 
 since they are in their definition explicitly given reality as a 
 quality. And yet if they are considered real, reality becomes 
 quite an un wieldly domain. Any object becomes real that 
 has this quality added to the concept of it. A Pegasus^^® be- 
 comes real provided that it is called a real Pegasus; an 
 island floating in the Atlantic becomes real provided it is 
 
 lis 
 
 84 
 
 Gassendi : The Fifth Set of Objections to Descartes' "Meditations"; 
 See, also, the discussion of the "existing lion" in the First Set of 
 Objections, by Caterus ; Descartes : Oeuvres, ed. by Adam and Tan- 
 nery, vol 9, p. 79. 
 
 85 
 
thought of as a perfect island and hence as a real island. ^^® 
 The category' of reality, if less inclusive than the category 
 of being, becomes quite as heterogeneous and unwieldly. 
 
 And so we must exclude the real centaurs and real winged 
 horses and real golden mountains from existence. We must 
 exclude them, that is to say, if reality is not to be as in- 
 coherent and unlimited as being itself. This exclusion is 
 usually brought about by saying that reality is not a 
 predicate at all. So say Gaunilon, and Gassendi, and, most 
 influential of all, Immanuel Kant.^^'" We can make our 
 centaur that is merely a concept anything that we desire 
 except that we cannot make it real. We can make our horse 
 winged, we can make fire come out of his nostrils, we can 
 make him roar like a lion; but we cannot make him a real 
 horse. Existence, in short, is not a quality that can be^ 
 attributed to the object of my thought. We think merely of 
 the entity; its existence is no proper part of the concept 
 of it. A hundred real thalers contain not a penny more 
 than a hundred thalers that are merely imagined. What we 
 can have as objects of thought are merely centaurs or 
 thalers. They are not to be described as real centaurs or 
 imaginary centaurs or real thalers or imaginary thalers. 
 And consequently we avoid the problem of the real centaur 
 by denying that such a combination of terms is an object 
 of thought at all. 
 
 But if existence is no attribute of the entities of which 
 I am conscious, I can never talk about existence at all. 
 The only entities about which I can think are entities that 
 are in some sense objects of my consciousness. Conse- 
 quently, if existence is never a part of them, 1 am cut off 
 from all commerce with existence. If what I think about 
 are merely centaurs and thalers, essences that never carry 
 within themselves either existence or a reference to exist- 
 ence, then existence is something entirely foreign and mean- 
 ingless. If none of the objects of my thought can breathe 
 a word about existence, there is no way in which I can 
 come to distinguish between the existent and the merely 
 subsistent. Accordingly, existence must in some way be 
 given in some of the objects of my thought. That is. some 
 
 319 
 
 120 
 
 Gaunilon : In Behalf of the Fool, sec, 6. 
 
 Kant : Critique of Pure Reason, ist eel., p. 597, et seq. 
 
 86 
 
 of these essences must contain within themselves as objects 
 of thought either existence or reference to existence. They 
 cannot be subsistents pure and simple that breathe no word 
 about existence. On the other hand, this existence or this 
 reference to existence cannot be given in them at will. 
 For in that case we have these real centaurs and real golden 
 mountains that are real merely because reality or a reference 
 to reality is put into the concept of them. Reality or 
 reference to reality must be given in some concepts, but 
 it cannot be given in any of them to which we please to 
 attribute such a predicate. 
 
 Here is where the difficulty arises when we attempt to 
 define reality. For, suppose that we define the real as the 
 permanent. Then, since permanence is equivalent to reality, 
 the status of permanence is the same as the status of reality. 
 Permanence, too, must be a predicate of some entities but 
 not of any entity to which we please to attribute it. On the 
 one hand, permanence or a reference to permanence must 
 be given as a quality of some of the objects of my thought. 
 Permanence, like reality, must be conceivable; it must have 
 meaning. But on the other hand, we cannot allow perma- 
 nence to be made a quality of any entity that we please to 
 call permanent. For in that case, as soon as I conceive 
 permanent centaurs or permanent golden mountains, such 
 centaurs and golden mountains are real forthwith. It is 
 not enough, consequently, to take the category of being 
 or subsistence and then restrict ourselves to those subsist- 
 ents that are conceived to be permanent. For any sub- 
 sistent will be conceived to be permanent if w^e put perma- 
 nence into the concept of it. No matter what the quality 
 be in terms of which we choose to define reality, we face 
 the same difficulty. Any entity can be conceived to be 
 independent, to be a sense-datum, to be a member of a 
 system of interrelated entities. So long as we restrict our- 
 selves merely to all entities that pretend to have one quality 
 or another, the real world we mark out will suffer from 
 instability. It will be a world that can be populated at 
 pleasure. Reality, accordingly, cannot be defined in terms 
 of a quality that can be made a quality of any entity. And 
 yet any quality that we may select will apparently be one 
 that can be attributed to any entity to which we please to 
 
 ST 
 
attribute it. "Any character you please may be imagined 
 and may remain merely imaginary."^^^ Entities that seem 
 to be permanent, or that pretend to be independent, or that 
 may be conceived to be sense-data, are legion. 
 
 And so reality must consist of entities that are really 
 permanent, or really independent or really sense data. If 
 reality is defined in terms of permanence, real entities are 
 those to which pennanence belongs, not those that seem to 
 be permanent. Of the entities that seon to be permanent, 
 some, we must say, are permanent and real and some are 
 not. A centaur or a golden mountain may Ije conceived to 
 be permanent; whether it is permanent and real is another 
 matter. The whole case is thrown o|>en once more. For 
 in order to determine the content of reality we must now 
 distinguish between the seeming permanent and the really 
 permanent, between that which is ruerely conceived as per- 
 manent and that to which permanence belongs. How, now, 
 are we to answer this new question? How are we to tell 
 with which entities permanence belongs and with which it 
 does noti^ If we define the really pernianeiU as that which 
 in addition to the quality permanence has the quality Q and 
 the seeming permanent as that which lacks the quality Q, 
 we only postpone the question. For the quality O can be 
 added to the concept of any entity; we consequently must 
 undertake to distinguish the entities to which the quality 
 O belongs from those to which it does not belong. We are 
 •caught in an infinite regress. For no criterion will be able 
 to rule out those entities that merely pretend to satisfy it. 
 And so it seems impossible ever to mark out the real from 
 the merely subsistent, from the entities that are merely 
 objects of thought, a conclusion that is fatal not only to 
 this dissertation but to all science as well. 
 
 There is a first step that must be taken if this outcome is 
 to be avoided. We must refuse to treat the concepts ^Veal" 
 and "unreal" in the same way. While not everything that 
 is conceived to be real is real, we must grant that every- 
 thing that is conceived to be unreal is unreal. My real cen- 
 taur is not real because I think of it as real. But a desk 
 that I think as unreal is by that very fact unreal. This 
 does not mean that the desk before me ceases to be a real 
 
 121 
 
 W. H. Sheldon: The Demolition of Unreality; Journal of Phi- 
 losophy, etc., vol. 13, 1916, p. 319. 
 
 desk when I form the concept of an unreal desk before me. 
 In this case there are two desks to be considered, one with- 
 out the predicate ''unreal" and one with it. And of these 
 two subsistents, only the one conceived with the predicate 
 ''unreal" is necessarily ''merely" a subsistent or ''merely" an 
 object of thought. When that which is in question is the 
 predicate "real," the situation is not similar. Here too we 
 may consider two entities, one centaur conceived without 
 the predicate "real" and one conceived with it. But in this 
 case not even the centaur conceived with this predicate is 
 necessarily real. 
 
 All entities conceived as unreal, then, are unreal. If, 
 consecjuently, reality is defined as permanence, all entities 
 that appear to be impermanent are unreal. Similarly, if 
 reality is defined as independence, all entities that pretend 
 to be dependent are unreal. If, then, a definition of reality 
 does nothing else, it does rule out a certain class of entities 
 as "unreal." This service, we must admit, is of considerable 
 importance. For it becomes ix>ssible as a result to de- 
 termine the status of a large number of important entities. 
 If to be real means to be present to the senses, then the 
 Platonic Ideas that are conceived as not present to the 
 senses are unreal. And the greater the number of qualities 
 than an entity must have if it is to be real, the greater will 
 be the number of entities that can definitely be classified as 
 unreal. If an entity to be real must have the qualities A, 
 B and C, any entity that pretends to be not A or not B or 
 not C will be unreal. Just which entities are those that 
 pretend to be not A or not B or not C will depend on how 
 reality is defined. On this account it is necessary to make 
 explicit the various definitions of reality. For each of these 
 definitions determines in a different fashion the content, or, 
 rather, part of the content, of the unreal. 
 
 If we define the real as the permanent, our definition de- 
 termines part of the content of the unreal. For then entities 
 that do not pretend to be permanent, entities that tnean to 
 be evanescent, are unreal. A similar result is accomplished 
 when we define the real as that which is related, when we 
 define it as that which is a sense-datum, when we define it 
 as that which is independent. In one case an entity that 
 does not mean to be a sense-datum will be unreal, in another 
 case an entity that does not mean to be independent will be 
 
 89 
 
 N 
 
 *# 
 
unreal. Yet one definition will mark out a larger number 
 of entities as unreal than will another. If we say that an 
 entity to be real must be a vivid object of sense-perception, 
 then a number of entities are immediately known to be un- 
 real. For there are a number of entities that are never held 
 to be vivid objects of sense-perception. If, however, we 
 define the real as the independent, we are still ignorant of 
 any large group of entities that are obviously unreal. For 
 there are not many entities that do not pretend to be inde- 
 pendent of a percipient. It is useless to define reality in 
 terms of independence because such a definition does not 
 separate out any important group of entities as obviously 
 unreal. And the fault that is here found with a definition 
 of reality in terms' of independence is the fault that in the 
 preceding chapters has been found with many other defi- 
 nitions of reality. It is useless to say merely that the real 
 is that which is "connected" with a perception, for unless we 
 know the specific character of the connection demanded, no 
 entity will come to us as obviously lacking in the connection 
 that is required and so no entity will come to us as obviously 
 unreal. We have in the preceding chapters objected to 
 certain definitions of reality in that they do not thus de- 
 termine part of the content of the unreal. The problem 
 that has been raised in the present chapter, however, applies 
 equally to all of the definitions that we have considered. 
 For while some of them determine quite definitely part of 
 the content of the unreal, none of them can determine with 
 any definiteness the content of the real. 
 
 A rather precise definition of reality will mark out cer- 
 tain entities as obviously unreal. But when all is said and 
 done, we are apparently left with an irresolvable element, 
 a surd, in the concept of reality. No matter how many 
 qualities we require an entity to show in order for it to be 
 called "real," and no matter how preciselv we describe 
 these qtialities in terms of which we are defining reality, 
 our goal is still some distance away from us. For after all 
 we have only succeeded in demanding of the candidate for 
 reality that it pretend to l>e A. and pretend to be B, and 
 pretend to be C. But if reality is to be kept free of real 
 centaurs and real golden mountains, it must be only some 
 of these entities that pretend to be A and B and C that are 
 real. If the content of realitv is not to be added to at will, 
 
 90 
 
 no number of limitations imposed on subsistence will be 
 sufficient to mark out the real. For, as we have seen, no 
 matter how many requirements we make, the number of 
 entities that can be thought to fulfill all these conditions 
 will be legion. Our purpose is to mark out the real. We 
 shall not succeed in our purpose by adding one more 
 quality, possession of which is to be made a requisite for 
 reaHty. On the other hand, if we leave the matter where 
 it stands, we cannot tell concerning the entities we have 
 not yet definitely labelled ''unreal'' which are to be called 
 ''real" and which are to be called "unreal." We have sepa- 
 rated out a class of unreal entities, but what we have left 
 contains both real and unreal entities. My centaur pre- 
 tends to be permanent and a sense-datum and independent 
 and a member of a system of interrelated entities. Your 
 desk makes the same pretensions. What reason, now, can 
 we find for calling your desk real and my centaur unreal? 
 And how am I to know when I am confronted by ether or 
 phlogiston that also makes the saine pretensions whether it 
 is to be called real or unreal ? One way out, theoretically, 
 at least, is to start with all of those subsistents not pre- 
 viously excluded, and to enumerate those that are real and 
 those that are unreal. Such a remedy, however, is just as 
 fatal as none at all. We get started on the road to knowl- 
 edge only to come to the end of the road at the same 
 moment.' Just as the concept ''reality" becomes meaning- 
 ful, it becomes useless, for we already know which entities 
 are real and which are unreal. 
 
 We must know what it means to be real ; otherwise the 
 process by which we call some entities real and others unreal 
 is without rhyme or reason. Besides, reality must be defined 
 in general terms, in tenns of a criterion that has yet to be 
 applied. And yet such a definition seems impossible on 
 account of the real centaurs and real golden mountains that, 
 it would seem, cannot be got rid of. Evidently reality can- 
 not be serviceably defined in terms merely of some objec- 
 tive quality such as permanence or independence. Let us, 
 consequently, have recourse to the concept: "belief." If 
 we say that an entity to be real must not only be conceived 
 to have the qualities A, B and C, but that this entity so 
 conceived must be believed by me, perhaps then we shall 
 have found a solution. Suppose I define the real as that 
 
 91 
 
which when conceived as a sense-datum is believed by me. 
 Let us see what hai)i)ens in this case to the real centaurs 
 and real g-olden mountains. We may take, as an example 
 of an entity that is conceived to be a sense-datum, a g-host 
 that I pretend is before me. If, now, all entities that are 
 conceived to be sense-data are real, this ghost is real. Rut 
 we have revised our definition; only those entities are real 
 that when conceived as sense-data are entities in which I 
 believe. And so I think of a ghost that pretends to be a 
 sense-datum in which T believe. But as soon as I consider 
 this ghost, I am aware that I am considering a ghost that 
 seems not to be an entity in which I believe. This entity 
 that is an object of my thought is given as a sense-datum 
 in which I believe, and at the same time is given as a sense- 
 datum in which I do not l>elieve. But since it is given as a 
 sense-datum in which I do not believe, it is unreal. For 
 only those entities are real that can be given as entities in 
 which I believe without being given at the same time with 
 the contradictory quality. 
 
 I hope that in this way I have avoided the real centaurs 
 and real golden mountains. An entity to be real, we have 
 said, must be given w'ith the quality A, and must not at the 
 same time be given with the contradictory quality non-A. 
 If, now, the quality A in terms of which we define reality 
 is some objective quality such as relatedness or [>erma- 
 nence, we have no way in which to eliminate the real cen- 
 taurs and real golden mountains. For we can have before 
 us a related centaur or a permanent golden mountain with- 
 out at the same time having before us a centaur that is not 
 related or a golden mountain that is not permanent. I do 
 not necessaril}- think of a golden mountain that is evan- 
 escent at the same time that T think of a golden mountain 
 that is permanent. And so a golden mountain that is 
 merely thotight of as permanent will ])e real provided that 
 reality is defined in terms of permanence alone. When, 
 however, we make belief a condition of existence, the status 
 of the golden mountain is qtiitc different. For at the satne 
 time that I think of a golden mountain in which I believe, 
 I think of a golden mountain in which I do not l>elicve. 
 At the same time that the golden mountain is given with 
 the quality A, it is given with the contradictory quality 
 non-A. And so, with reality defined partly in terms of 
 
 92 
 
 belief, the real world cannot be populated at will. An 
 entity does not become real merely because I pretend that it 
 is an object in which I believe. For certain entities in which 
 1 pretend that I believe are at the same time given to me as 
 entities in which I do not believe. These entities that are 
 given with the quality A are at the same time given with 
 the contradictory quality non-A; and so they are entities 
 that are forthwith unreal. 
 
 We seem to eliminate the real centaurs and real golden 
 mountains, it may be admitted, when we introduce the con- 
 cept of belief into our definition of existence. But, it may 
 be said, our apparent success is due to the fact that the 
 concept "belief" is subsequent to the concept "existence." 
 To believe in an entity, it may be said, is to believe in the 
 existence of that entity. The entity in which I believe is 
 the entity in whose existence I believe. Consequently, 
 when we' talk about belief, it may be held, we illicitly pre- 
 suppose a knowledge of the meaning of "existence." To 
 define reality in terms of belief, in short, ^is to define reality 
 in terms of' a concept that presupposes "reality." It is, it 
 may l)€ held, a circular definition, one that implicitly makes 
 use of the verv term that is to be defined. ^ ^ 
 
 Yet it is not, I think, in the nature of things that "exist- 
 ence" i^ prior to "belief." According to representative 
 psychologists, belief is for the naive man the natural and 
 normal condition. Only after certain disconcerting expert- 
 ences does he come to doubt the entities that are given to 
 him and to question their existence. Normally the feeling 
 or state of belief precedes the raising of any questions about 
 existence. The term "l>elief," it would thus seem, may be 
 used to point to a simple and non-cognitive psychological 
 condition, a psychological condition that does not imply 
 any cognitive distinguishing between unreality and reality. 
 When V^ertain ol)jects are presented to us, we are charac- 
 terized bv this feeling or state of belief. And there is this 
 belief it'mav be said, whether there is any consciousness of 
 existence or not. Belief may thus be described without any 
 presupposing of existence. It may be described as a psycho- 
 logical condition that is to be distinguished by its intrinsic 
 psychological characteristics. "Belief," in one sense, may 
 be defined in terms of existence, in terms of the entity to 
 which it refers. But "belief," in what is perhaps a different 
 
 93 
 
sense of the word, may be described in terms of its psycho- 
 logical characteristics. And it is a ''belief" to be described 
 in this latter manner that we are meaning to use. "Belief/* 
 as we use it, is not a concept that presupposes existence. 
 And so to define existence in terms of such a belief is not 
 to give a circular definition of the term that is to be defined. 
 ^y "an entity in which T believe" I do not mean an entity 
 in whose exisfence I believe. By ''an entity in which I be- 
 lieve" I mean an entity such that in being conscious of it I 
 
 am characterized by a specific psvchological condition called 
 belief." 
 
 When belief is descriljed in such psychological terms, the 
 connection between "belief" and "existence" is still to be 
 determined. "Existence" is a term that has not vet been 
 brought upon the stage. And when it is brouglit upon the 
 stage, it is a predicate that may be given to objects in which 
 I do not beHeve as well as it mav be given to objects in 
 which T do believe. "Belief" has not been defined in terms 
 ot "reality," and so "entity in which I believe" does not 
 imply "real entity." "Belief" is one term and "existence" 
 IS another. And so an entity in which J believe need not 
 exist unless either "belief" is defined in terms of "existence" 
 or "existence" is defined in terms of "belief." 
 
 \A'hen Descartes attempted to doubt every entity that was 
 presented to him. he found that there was one entity in 
 which he could not l)ut l)elieve. His own doubting, his own 
 mental functioning, was something that he found he could 
 not doubt, something in which he could not help but believe. 
 Here vvas an entity that was perforce an object of belief. 
 Yet if belief is understood to be a purel}' psychological con- 
 dition, if belief is understood to be something that is to be 
 described l\v its intrinsic qualities, then an entity that is 
 necessarily an object of lielief is not ipso facto an entity that 
 is real. If belief is not described in terms of existence, then 
 an entity in which I must believe is not necessarily a real 
 entity. "Reality" is in such a case a concept that has yet to 
 be introduced. And when it is subsequently introduced, it 
 need not be given as a predicate to these entities in which 
 I cannot but believe. Descartes does, to be sure, pass from 
 "entity in which 1 cannot but believe" to "entity that is real." 
 Because he can not doubt his own thinking, he concludes 
 
 94 
 
 V'/i 
 
 / 
 
 that that thinking must be real. Such a transition, it seems 
 to me, is implicitly a definition of reahty. It implies a 
 definition of reality to the effect that entities that cannot be 
 doubted must be real. It is not a transition to which in the 
 nature of things we must agree. It is a transition that is 
 valid only when existence is defined in terms of belief, or 
 belief in terms of existence. 
 
 A definition wholly in terms of belief is not the only possi- 
 ble definition of existence. Entities that cannot be doubted 
 are not entities that we must perforce call "real." And yet 
 a definition that makes some use of the concept "belief" is, 
 we have found, necessary, if we are to avoid the real cen- 
 taurs and real golden mountains. Only by having recourse 
 to "belief" can we eliminate these troublesome entities. 
 They cannot be eliminated if we define reality merely in 
 terms of the objective qualities— A, B and C. Of the enti- 
 ties given with the objective qualities A, B and C, only those 
 are real, we must say, that are entities in which T believe. 
 We must make some use of belief, though we may at the 
 same time define reality partly in terms of the objective 
 
 qualities A. B and C. 
 
 Accordinglv, of the entities that are conceived with the 
 qualities A, B and C, those, we shall say, are real that are 
 entities m which I believe. Now such a definition as this 
 will do verv well so long as I think of objects merely as 
 objects of my thought. Suppose, however, a second in- 
 dividual is doing the thinking. Then when he thinks of a 
 ghost present to his senses that pretends to be an entity in 
 which he believes, he will at the same time be thinking of a 
 ghost that seems not to be an entity in which he l)elieves. 
 In order that the real centaur shall not be admitted among 
 the objects of his thought, the real for him must be defined 
 as that which, \vhen conceived with the qualities A, B and C, 
 IS an entitv in which he believes. But the entities conceived 
 with the qualities A. B and C, in which he believes are differ- 
 ent from the entities so conceived in which / believe. Conse- 
 quently, the content of reality varies with the individual who 
 is doing the thinking. That is to say. the content of reality 
 is relative. For there is one group of entities that are real 
 for me and another group of entities that are real for you. 
 It is of course not a new thing for "real" to be used in 
 such a wa}' that the content of reality must var>' with the 
 
 95 
 
individiia] who is doing the thinking. The term is said to 
 have been used in just this way by Protagoras. ^^^ And it 
 has been used in just this way by James and some of his 
 followers. ^^'^ Now it has been said that the real is that 
 which appears, that w liich is believed. Now it has been said 
 that it is that which is interesting. But that which appears 
 to one man is different from that which appears to another. 
 That which is believed by me is different from that which 
 is believed by you. And that which is uninteresting today 
 was interesting yesterday. With such delinitions of reality, 
 consequently, to know whether an entity is real, we must 
 know whose thought, and what instant of that thought, is 
 in question. Now in so far as they put a certain amount of 
 relativity into the concept of reality, 1 believe definitions of 
 this sort to be correct. Reality, it seems to me, must be 
 relative to the individual whose thought is under considera- 
 tion if the real centaur is to be avoided. 
 
 When, however, we define reality in such a way that it 
 is relative to the thinker, we encounter a serious disad- 
 vantage. Let us suppose, for example, that we define the 
 real simply as that which is believed. Then, a^ we have 
 seen, the entities that are real for me are quite diflFerent 
 from those that are real for you. All sorts of entities are 
 real for some one or other. Consequently, to call an entity 
 "real" is to assert almost nothing about it ; it is to assert 
 merely that it belongs in the large and ill-assorted group 
 of entities that are believed to have been believed by some 
 one. We might, however, be willing to allow the term 
 ''real" to become vague and meaningless, if the term "real 
 for me" could be substituted for it. But the latter can not 
 be substituted for it without doing great violence to com- 
 mon speech. For when I say that the earth moves, I am 
 asserting that the earth's motion is real absolutely, not that 
 it is real for me. And so with all of the propositions that 
 can be transformed into existential propositions. In all of 
 them WQ are predicating a (luality that is meant not to be 
 relative to the individual making the judgment. 
 
 Consequently it seems wise to keep within as narrow 
 bounds as possible the relativity of realitw We can not 
 keep out the real centaurs and at the same time prevent tiie 
 
 •^ ' 
 
 I 
 
 in 
 
 Plato: Theactetus, 168-170. 
 '"James: Principles of Psychology (1890), Vol. 2, Ch. 21 
 
 96 
 
 concept from being to some extent relative to the thinker. 
 But we need not go the whole way. There is, as we have 
 seen, a surd element in the concept of reality,^ and to deal 
 with this we must have a recourse to "belief." But it is 
 not necessary to treat the entire concept as a surd. And 
 it is unwise to do so in view of the prevalent usage to the 
 contrarv If the whole meaning of "real" is dependent 
 upon the individuality of the thinker, "real" becomes quite 
 meaningless. But if the surd element is reduced to a mini- 
 mum, there is only a relatively small sphere within which 
 the ineaning of the term is undetermined. It is to reduce 
 this surd element that we must use some objective qualities, 
 such as permanence or independence, in the definition ot 
 reality. For the only meaning the term gets, it gets from 
 the place these qualities occupy in the definition of it. If 
 reality is to be meaningful, consequently, we must define 
 it not merely in terms of belief but in terms of certain 
 qualities A, B and C. The real must be that which is con- 
 ceived to have the qualities A, B and C, and at the same 
 time is believed. The latter requirement w^e can not avoid; 
 the former we neglect at the peril of leaving our concept 
 
 void of meaning. 
 
 The unreal must be defined, then, 1 think, as that which 
 appears not to be believed by me when conceived with the 
 qualities A, B and C. And of the remaining entities those 
 are real which appear to be believed by me when conceived 
 with the qualities A, B and C. This form, that I believe 
 the definition of reality must take, is rather cumbersome. 
 And perhaps its very cumbersomeness enables me to hope 
 that the real centaurs and real golden mountains have been 
 avoided. I think, however, that with this formula these 
 entities have truly been avoided. And I see no other way 
 in which they can be escaped and knowledge made possible. 
 
 The real, "^then, is not that which is conceived to have 
 the quality A, but that to which the quality A belongs. 
 And that\o which the quality A belongs is that which, 
 when conceived with the quality A, is believed. When, 
 consequently, in the preceding pages the real has been said 
 to be equivalent to that which is permanent or that which 
 is a sense-datum, we can not hold this to mean that the 
 real is equivalent to that which is conceived to be perma- 
 nent or conceived to be a sense-datum. At least we can 
 
 97 
 
not hold this if we are anxious to find definitions that will 
 not allow reality to be added to at will. We must hold 
 these definitions to mean that the real is that to which 
 permanence belong-s or that which is really a sense-datum. 
 And that which is really a sense-datum is that which, when 
 conceived as a sense-datum, is believed. It is only by some 
 such transformation as this that the surd element in the 
 definition of reality can be explained. Any definition of 
 reality, it seems to me, must be modified in some such way 
 as this if the "real centaur" is to be avoided. Consequent- 
 ly, when in the preceding pages I have said that the real 
 may be defined as that to which permanence belongs, I have 
 meant that it is possible to define the real as that entity 
 which is an object of belief when it is conceived to be 
 l>ermanent. 
 
 The real, then, is that which is l)elievecl bv me when it is 
 conceived to have the qualities A, B and C, and is not at 
 the same time disbelieved by me. This form an accept- 
 able definition of reality must take. But the qualities that 
 are to be written in where I have |)ut the symbols A and 
 B and C are not determined. For A and B and C we may 
 read "permanence" or "inde})en(lence" or "presence in sense- 
 perception," or the like. When, consequently, I have said 
 that it is possible to define the real as the permanent, 1 have 
 meant that it is possible to put "permanence'' where 1 have 
 put these symbols. There are a great variety of qualities 
 that may be used to take the place of these syml)ols A, B 
 and C ; but some qualities may ])e said to belong to real 
 entities much more meaning full}' tlian mav '^hers. Some 
 qualities, that is to say, can be used more successfully than 
 others to lill in these blanks represented by the letters A. 
 B and C. h'or some qualities, while a])])arentl\- they fill in 
 these blanks, are in reality quite as meaningless as the letters 
 whose places they are taking. If we are told, for example, 
 that a real entity is one that is a member of a system of 
 interrelated entities and are not told what constitutes this 
 relatedness, w^e are quite as much in the dark as if we are 
 told that a real entity is one that has the quality A. A 
 quality that is to be used as a substitute for these symbols 
 must be meaningful and specific ; it must not be a quality 
 such that when it is attributed to an entity we feel the 
 entity to be no more specifically described than it was before. 
 
 98 
 
 / 
 
 CHAPTER VI 
 
 The Author's Definition of "Existence" 
 
 AND "Reality" 
 
 The historical inquiries we have made have shown what 
 a large measure of disagreement there is as to the meamng 
 of the terns "existence" and "reality." Now only sense- 
 data are said to exist, now only entities that are not sense- 
 data are said to exist, Now only entities that have a posi- 
 tion in time are called real, now only entities that are out 
 of time are called real. "Existence" and "reality have 
 been defined in terms of permanence, in terms ot time and 
 space in terms of a svstem of interrelated entities, in terms 
 of activity, in terms of belief. The emities that are real 
 if the term is used in one sense are unreal if the term is 
 used in another sense. And if we have recourse to the 
 meaning of the term as it is used in common speech, either 
 we can not find out what that meaning is or we fand a 
 meaning different from any of those already discovered. 
 Wherever we look we merely add to the variety of mean- 
 ings before us. We find ourselves in the end confronted 
 by a whole host of meanings the term has had, no one ot 
 which stands out as the meaning the term ought to have. 
 
 And vet in anv one writer, or at least in any one con- 
 text there must be some one sense of "existence that is 
 being used For there are sure to be propositions laid 
 down there that are equivalent to propositions asserting 
 existence or non-existence. And these propositions are 
 meaningful onlv if existence has there some specific definite 
 meaning Existence can not be used in such a way that 
 we are left in doubt as to which of these various and con- 
 tradictory senses is intended. It must be used in one defi- 
 nite sense and not used vaguely. Each writer that is to 
 say must be clear as to which meaning the term existence 
 has for him The question can not be left undetermined 
 if propositions with an existential import are to be sig- 
 
 "' Iif view of the great variety of meanings the term has 
 had, however, there will be something arbitrary about this 
 
 99 
 
process of selection. Whichever sense of the term we 
 choose to use, there will be other senses of the term that 
 we might have chosen. There will always be some one for 
 whom ''existence" does not mean what it means for us. 
 And whichever sense of the term we choose to use, there 
 \A\\ always, I think, be some particulars in which the de- 
 notation of the term as we use it will not be identical with 
 the denotation of the term as it is used in common speech. 
 There will always be some entities that we shall be calling 
 "real" that in ordinary parlance are called "unreal," or 
 some entities that we shall be calling ''unreal" that in 
 ordinary parlance are called "real.'' The meaning "exist- 
 ence" will have will be the meaning it has for us, not the 
 meaning the word has for all philosophers nor the meaning 
 it has in ordinary language. And what is true of the term 
 "existence" is true of the term "reality." The sense in 
 which we use the term "reality" will show what we mean 
 by "reality" and not what "reality" must mean. We may 
 choose to define "existence" in one set of terms and "reality" 
 in another set of terms. Or we may choose to call "exist- 
 ence" and "reality" synonymous. But however we define 
 "existence" and however we define "reality," whether we 
 define the two terms in the same way or in different ways, 
 our definitions of both terms will be arbitrary. They will 
 show what "existence" means for us and what "reality" 
 means for us and not what "existence" and "reality" mean 
 for the world. 
 
 When we have once decided what "existence" means, 
 we shall have decided what entities are to be called existent 
 entities. A definition of existence, in other words, will 
 have determined the content of the world of existence. 
 Consequently, if the definition of existence must be arbi- 
 trary, the content of the world of existent entities must be 
 arbitrary. There will be one group of entities that exists 
 in the sense of the term that I am employing, another group 
 that exists in the sense in which some one else understands 
 that term, and still another group that exists as existence 
 is used in our ordinary speech. And so with reality. Just 
 which entities constitute the real world depends on the 
 meaning that is given the term "real." And since the 
 meaning given "real" must be a result of an act of choice, 
 
 100 
 
 the content of the real world will be the result of an act of 
 choice. In one sense, consequently, we make the real world. 
 For the content of that world depends on a choice of ours. 
 We may not make the world in the way in which the Ego 
 is conceived by Fichte to make the world, but we do make 
 the world in this other sense. Indeed, we must make the 
 real world if there is to be any real world at all; otherwise 
 the concept of a "real world" is vague and meaningless. 
 
 Any definition of "reality" or "existence" must, accord- 
 ingly, involve an ontology. It must determine in an arbi- 
 trary fashion the content of reality or of existence. This, 
 however, is just what is denied by Marvin. He asks : "How 
 ought we, then, to define the word 'to exist?'" And he 
 replies: "With the minimum of ontological assumption; 
 for the definition of existence ought not itself to be an 
 ontology."^^-* But, natural as the desire is to maintain a 
 judicial impartiality when we define existence, the facts 
 will not allow us to entertain such hopes. We cannot pre- 
 vent the definition of existence from involving an ontology. 
 For in whichever sense we choose to use the term "exist- 
 ence," we are selecting and constructing the content of 
 existence so as to excltide some entities that have been said 
 to exist. We cannot define the term in such a way that all 
 of the senses in which the term has been used will be found 
 acceptable. And to define it in such a way that nothing is 
 told us respecting the content of existence is to give no 
 definition at all. No other alternative exists. Conseqttently, 
 any definition that is a definition will involve an ontology. 
 Marvin's own definition is of the kind that in the third 
 chapter of this essay has been attributed to Hume and to 
 Kant. The entities that exist are those that are observed 
 and those that are connected with entities that are observed. 
 In other words, sense-data and entities coimected with 
 sense-data exist. If, however, the meaning of "connection" 
 is left undetermined, we cannot distinguish the entities that 
 are connected with sense-data from those that are not, and 
 so we have really a definition far from precise. If, on the 
 other hand, the meaning of "connection" is rendered 
 definite, as it is, for example, when Marvin suggests that 
 entities that enable us to predict observed facts are connected 
 
 "* Marvin: The Existential Proposition; Journal of Philosophy, etc., 
 
 vol. 8 (1911), p. 477- 
 
 lOI 
 
with them, then our definition of existence is selective. 
 For we are ruHng out of existence such entities as things- 
 in-themselves, entities that cannot be said to enable us to 
 predict observed facts. We are determining the content 
 of existence so as to exclude these entities that have often 
 been said to exist. We are laying down a definition of 
 existence that involves an ontology. 
 
 A definition of ''reality" must be arbitrary. But it does 
 
 not follow that a definition of reality is an absurdity. Let 
 
 us agree that none of the qualities in terms of which reality 
 
 may be defined ''has any intritisic connection with Be- 
 
 jj^g 'M25 Nolle the less it is possible to connect them with 
 
 being or subsistence so as to form a meaningful and stable 
 
 combination. The connection will be merely "extrinsic," 
 
 yet wt shall have a combination that is equivalent to 
 
 "reality" in the usage of some individual. And if we 
 
 mean by a definition a statement that gives the meaning 
 
 a term has at some time had, it will be a definition. Indeed, 
 
 in spite of the fact that a definition of existence means 
 
 the abandonment of an impartial attitude, "existence" must 
 
 be defined if it is to be a meaningful term. The definition 
 
 of "existence" I am about to give will explain only what 
 
 "existence" means for me and not what "existence" means 
 
 for all men. But it will cause "existence" to be a term 
 
 with a specific and definite meaning when it appears in my 
 
 writings. And to the extent to which my definition is 
 
 specific and unambiguous, it will enable one to tell just 
 
 which entities exist and just which ones do not exist, as 
 
 I am using that term. 
 
 For me "existence" and "reality" are synonymous. The 
 entities that exist are just those that are real. When I 
 am defining "existence," consequendy, I am at the same 
 time defining reality. Now, an entity to be real and exist- 
 ent, as I choose to use these terms, must l>e in time and 
 space. That is to say, I accept the view that only entities 
 in time and in space are real. We have seen, however, that 
 when reality is defined in terms of time and space, several 
 ambiguities remain to be cleared up. Let me therefore make 
 my definition more specific. Let me say that an entity to 
 
 '"W. H. Sheldon: The Demolition of Unreality; Journal of Phi- 
 losophy, etc., vol. 13 (1916), p. 320. 
 
 102 
 
 be real must have some rather definite position in time and 
 111 space; it cannot be vaguely located in all time and in 
 all space. That is, it must have definitive ubeity and not 
 circumscriptive ubeity. Time and space themselves I shall 
 not call real ; I shall restrict the term to the entities that are 
 in time and in space. Consequently, the time-space con- 
 tinuum, apart from the entities that are located in it, is 
 unreal. There was one other ambiguity that we found in 
 the view that entities in time and in space are real. We 
 have still to be told whether the reference is to entities 
 with a location in some space or to entities with a location 
 in the one objective space. It is the latter position that I 
 shall take. When I say that an entity to be real must be 
 in time and in space, I mean that it must be in the one 
 spatio-temporal system to which we usually refer. I shall 
 not call entities "real" that are in any space and in any time ; 
 they must be in the one space and time that is conceived 
 to be objective and not in some space and time that is con- 
 ceived to be subjective. 
 
 An entity to be real, then, must be in time and in space. 
 Furthermore, an entity to be real and existent must be con- 
 ceived to be one such that more of the subjects having 
 it as object of consciousness believe it than disbelieve it. 
 Let us suppose that ten people have an object called to their 
 attention and that five of them feel neither belief nor dis- 
 belief with respect to it. Then, if three of the others 
 believe the object under consideration while two of them 
 reject it, the object, provided it satisfies the other conditions 
 of existence, is a real object. If an object does not pretend 
 to have this quality, it is unreal. W^hen T think of phlogis- 
 ton, I am thinking of an entity that seems to be rejected by 
 the majority of individuals who think of it. So, too, with 
 the second chair that I see when I press the corner of my 
 eyeball. This chair is conceived as a chair that is not 
 generally an object of belief. Consequendy this chair is 
 immediately to be dubbed "unreal." 
 
 To be real, then, an entity must be in time and in space 
 and it must be an entity such that of the subjects having it 
 as object of consciousness more believe it than disbelieve 
 it. Undoubtedly much remains to be said before the mean- 
 ino- of these requirements becomes clear and unambiguous; 
 
 103 
 
and besides, there is a third requirement that I shall add 
 later in this chapter when I come to consider the notion of 
 independence. But for the present let me assume that these 
 are the sufficient conditions of ''reality" and "existence," 
 as I choose to use these terms. Let me assume, that is, 
 not only that entities must be conceived with these quahties 
 to be real, but also that all entities to which these qualities 
 belong are real. 
 
 We must now be careful to avoid the existing centaurs 
 and real golden mountains. It cannot be all entities that 
 are conceived to be entities such that more of the subjects 
 having- them as objects of consciousness believe them than 
 disbelieve them and that are conceived to be in time and 
 space that are real. For a centaur and a golden mountain 
 may be conceived with these qualities. Any entity may be 
 thought of along with these qualities, and so, if reality 
 were defined in this manner, the real world could be popu- 
 lated at will. Xo, a real entity is one to which position in 
 time and space belongs, and one that is really an entity such 
 that mr;rc of the subjects having it as object of conscious- 
 ness believe it than disbelieve it. But to what entities does 
 a position in time and space belong? The only way by 
 which we have been able to determine to what entities the 
 quality A "belongs" is by having recotirse to "belief." The 
 quality A "belongs" to an entity when this entity, being 
 conceived with the quality A, is an entity in which I believe. 
 Consequently, those entities belong in time and space in 
 which I belie\e when they are conceived to be in time and 
 space. And so a real entity, as I am using the term "real," 
 is an entity that ( i ) is conceived to have a position in time 
 and space, and (2) is conceived to be an entity such that 
 more of the sul)iects having it as object of consciousness 
 believe it than disbelieve it, and '(3) is an entity in which 
 T believe. Entities in which I do not believe when they are 
 conceived with these qualities are unreal. Those remain- 
 ing entities, that when so conceived are entities in which I 
 believe, are real ; and they alone are real. 
 
 It is in this sense that I choose to use the terms "reality" 
 and "existence." When I use them in this sense there will 
 be many entities that I shall be calling "real" that rithers 
 have called "unreal," and vice versa. I shall call the other 
 side of the moon real, though it is called unreal by one who 
 
 104 
 
 defines reality in terms of actual — and not possible — sense- 
 data. And I shall call Platonic Ideas when conceived as 
 supra-spatial unreal, notwithstanding the fact that they are 
 called real when "real" means "intelligible." It is also 
 probably true that some entities are real in my sense of the 
 word that are not usually called real in ordinary parlance. 
 Some violation of ordinary usage seems necessary, how- 
 ever we define "existence" and "realitv." For in common 
 speech these terms are applied very loosely and unsystem- 
 atically; consequently, a systematic and precise definition 
 will mark out a group of entities that cannot help but be 
 (lifi^erent in some particulars from that group of entities 
 ordinarily called ''real" or "existent." 
 
 My definition, in short, only gives the meaning "exist- 
 ence" and "reality" have for me. Relativity of this kind, 
 however, I do not think objectionable. For it is quite 
 different from the sort of relativity that is introduced in 
 so far as reality is defined in terms of belief. Let me show 
 this by comparing the two concepts "my native land" and 
 "the thinker's native land." "My native land" is relative 
 to me in that one must know something about me to know 
 which my native land is. But having once determined that 
 1 was born in the United States, my native land does not 
 depend on who it is that is doing the thinking. Whether 
 a Venezuelan or a Hottentot be doing the thinking, "my 
 native land" is still the United States. But "the thinker's 
 native land" does vary. There is this diff'erence, I think, 
 between defining the real as that which is believed and giv- 
 ing it a definition that is the definition I choose to give it. 
 For when it has once been decided what meaning I choose 
 to give "reality," the question whether an entity is to be 
 called "real" is one that can be decided on objective 
 
 grounds. 
 
 It is impossible, then, to find a definition of "reality" that 
 will make acceptable all of the senses in which the term 
 has been used. Any definition that we hit upon will, con- 
 sequently, be arbitrary. That is, it will conflict with some 
 meaning that has been given the term by metaphysicians or 
 with the usage of common sense, inchoate as that is. It 
 will not state the meaning that the term "existence" has for 
 all of us; it will state the meaning the term ha^ for the 
 individual who lays this definition down and accepts it. 
 
 105 
 
%f< 
 
 Wi^wS. 
 
 Yet there will be compensations for its lack of objectivity. 
 For if it is really a definition, a proposition that enables 
 us to pick out those objects that are to be named real and 
 those that are to be named unreal, it will be a touchstone 
 that will permit us to settle some of the problems that have 
 been troubling philosophy for centuries. 
 
 Let me first consider those entities that are not objects of 
 our perception, but which stand behind those objects and 
 may be said to correspond to them. As contrasted with 
 the objects of perception they are the thiiigs-in-themselves. 
 I do not see or feel or touch them, but the objects I do see 
 and feel and touch stand in a certain correspondence with 
 them At times these trans-experiential entities are con- 
 ceived to be in time and in space ; at other times they are 
 in time but not in space : at other times they are noumena 
 both non-temporal and non-spatial. But. however con- 
 ceived these things-in-themselves have had their existence 
 questioned. Now it has been asserted that there are such 
 things as things-in-themselves. now that such alleged enti- 
 ties do not really exist. How are these contradictory 
 acsertions to be judged? Why, obviously, the ontological 
 status of these entities depends on the meaning of exist- 
 ence "' If to exist means to l>e an object of sense-percep- 
 tion of course things-in-themselves do not exist. If "to 
 exist" means to have a place in the time-space continuum 
 then things-in-themselves when conceived as non-temporal 
 and non-spatial again do not exist. If, on the other hand, 
 "existence" is being used in the sense in which it points to 
 active entities, or to the causes of sense-perception, things- 
 in-themselves may very well be real. We must know in 
 what sense existence is being used before we can mtelli- 
 .r^ently judge whether things-in-themselves exist. Anfl 
 when we do once know what sort of existence is in ques- 
 tion the reality or unreality of these trans-experiential 
 entities should be therebv in large part determined. It may 
 not be entirely determined, for, firstly, the sense in which 
 "existence" is being u^cl may be a rather vague sense, one 
 that does not specify with any definiteness just what class 
 of entities it denotes, or, secondly, we may not know just 
 what is meant by things-in-themselves, and so may not be 
 sure that such entities have the <inalities that have been 
 asserted by definition to be the qualities ul existent things- 
 
 io6 
 
 An example of each will no doubt make these two remain- 
 ing sources of confusion clearer. If "existence" is defined 
 so as to include sense-data and entities that sense-data 
 imply, we are still left in doubt as to the status of things- 
 in-themselves, for we do not know whether the relation of 
 these entities to sense-data is of the kind that the definition 
 vaguely calls "implication." Or if "existence" means loca- 
 tion in time and space, we cannot tell whether trans-experi- 
 ential entities exist until we know whether the trans- experi- 
 ential entities whose existence is in question are conceived 
 as outside of time and space or not. But while the deter- 
 mination of the meaning of "existence" is not the sole 
 requirement for the solution of the problem as to the exist- 
 ence of things-in-themselves, it is an absolutely essential 
 requirement, though one that, so far as I know, is never 
 taken into consideration. 
 
 So much lor things-in-themselves. Another group ot 
 entities that have offered considerable trouble are the en- 
 tities in the past and present and future that are not objects 
 o* sense-perception and not the things-in-themselves behind 
 the objects of sense-perception. Julius Caesar. King Arthur 
 and the Knights of his Round Table, all the roses that blush 
 unseen the events that will occur and those that it is alleged 
 will occur in the year 2000,— all of these entities belong in 
 this class whether they be considered as objects of possible 
 experience, that is, as sense-data for some possible subject, 
 or as things-in-themselves behind such possible sense-data. 
 If we define "existence" in one way, some of these entities 
 mav well exist If we define it in another way, they are 
 all 'forthwith unreal. If to "exist" is to be an actual object 
 of sense-perception, these objects are of course unreal. But 
 suppose we say that an entity to exist must be one such 
 that it does not violate the law that there shall be at all 
 times the same quantity of matter. Then, so far as this 
 requirement is concerned, some of these objects that are not 
 perceived mav exist. Whatever definition of 'reality we 
 li-dit upon, tlie content of our world will be determined in 
 lan^e part bv that definition. After we have once decided 
 for^urselves what it shall mean to be "real," we shall m 
 large measure know whether things-in-themselves are real 
 and whether future objects and past ones are or are not real. 
 What we have just been considering have been entities 
 
 107 
 
that I am not perceiving. Sometimes, however, we talk 
 about objects that I am not thinking about. If I am not 
 thinking about these objects at all, if they are in no wise 
 objects of my consciousness, I do not see how I can hold 
 any opinion at all about their existence or their non-exist- 
 ence. When we take independence in this very radical 
 sense, I do not see how we can help but hold such proposi- 
 tions as: "There are entities of which I am in no wise 
 conscious" to be self-contradictory. To think anything at / 
 all about an entity, that entiiy must in some sense be an 
 object of my consciousness. I may not have its content m 
 all of its concreteness before me, but there must be some- 
 thing in my consciousness that gives me a hold on this entity 
 and^'enables me to call it "real" or ''unreal." Let me sup- 
 pose that yesterday I was thinking of Socrates drinkmg his 
 cup of hemlock. To-dav the definite content of yesterday's 
 thought has utterlv vanished. But I know that yesterday 
 I was thinking of something. All that I can think of 
 to-day in this matter of Socrates and his cup of hemlock 
 is an "indefinite something. I am not aware of the content 
 of this something, specifically and distinctly, but I am aware 
 of some of the properties of this something. I know per- 
 haps that it is not a picture by Rembrandt, and I know that 
 it was in all its concreteness the object of my thought yes- 
 terday What I am thinking of is not a "nothing" that 
 is without anv content; it is an indefinite ''something/' an 
 entity with some content but without any very full, definite, 
 and concrete content. The question now arising is: ''Do 
 these indefinite "somethings" exist o^ do only definite ob- 
 jects exists This, as 1 see it, is the question to which the 
 controversv between realism and subjectivism reduces. 
 These indefinite "somethings" are considered so very im- 
 portant that the theory that attributes reality to them is 
 called bv the general name: realism. So, tor the early 
 Schoolmen immaterial universals were the important enti- 
 ties, and the theory that attributes reality to them was called 
 by the general name : realism. 
 
 Perhaps I ought to give another example. I am not 
 a bacteriologist and it does not take me long to think over 
 the few notions I have on that subject. So I go on to 
 consider how many facts and theories are kno^nl to bac- 
 teriologists that I do not know. I am thinking about these 
 
 io8 
 
 facts and theories that I am not aware of in all their definite- 
 ness as I am aware of the Copernican hypothesis. But these 
 facts and theories that I am thinking about have some con- 
 tent. They are all about bacteria, their mode of life, their re- 
 lation to chemical reactions, their efifect on the taste of water, 
 on the health of the human body. And they have the quality 
 of being facts and theories that one could learn about if 
 one read through the books on the subject in a well equipped 
 library and conversed with the contemporary leaders in 
 bacteriological research. But then I think of the many 
 facts in this field that are not known and indeed never vvill 
 be known in any specific definite fulness of content. For 
 example, the number of bacteria there are in the world at 
 this present moment is without doubt a fact that will never 
 be known definitely. I think about this fact and others like 
 it and what I am thinking about again has some content. 
 These facts too are facts about bacteria, and facts that 
 have the quality of never being concretely and definitely 
 objects of consciousness. About the number of bacteria I 
 know something, I know the number to be very, very large, 
 and I know that the specific number,— whether the figure 
 in the unit column is a four or a seven,— will never be 
 known. I am thinking about an object with some content, 
 but one that will never be an object of consciousness with 
 any full and specific content. 
 
 Objects such as these, I say, are the objects that realism 
 asserts exist and that subjectivism asserts do not exist. 
 The quarrel between the two is not, as realists often think, 
 over entities that are hard and not subject to remodelling 
 at the hands of the thinking Ego; even a solipsist can con- 
 sider an object inevitable during the time that it is definitely 
 in consciousness. Nor is it, as subjectivists often think, 
 over entities that are not being thought about. For an 
 entity that is in no sense an object of my consciousness is 
 simply unthinkable, beyond the possibility of consideration ; 
 and realism— what, at least, I think realism ought to be— 
 is not a self-contradiction, it docs have some meaning. But 
 entities that are only vaguely and not definitely and specific- 
 ally objects of my consciousness, they are the mischief- 
 makers. Now shall we be realists or shall we be sub- 
 jectivists? Shall we say that these indefinite "somethings" 
 are real or shall we say that only definite entities are real ? 
 
 109 
 
The decision we reach will depend on the sense in which 
 we are using the term^ 'Veal-." If I lay down as part of my 
 definition of existence the requirement that an entity to be 
 real must be in considerable detail an object of my con- 
 sciousness, then I become a subjectivist forthwith. If I use 
 ''existence'' in a sense that does not exclude such indefinite 
 entities, I am a realist. Personally, I choose to use **exist- 
 ence" in a sense in which the term points not only to entities 
 that are in their full detail objects of my consciousness, 
 but also to entities that are, specifically, objects of some 
 consciousness, though / am not aware of them except 
 vaguely. That is, I shall define ''existence" so as to make 
 myself an objective idealist. 
 
 An entity is often said to be real if it is not dependent 
 on my consciousness of it.^"^ And an entity is independent 
 on mv consciousness of it if it persists when I am not think- 
 ing of it. What shall I say about the planet Mars? An 
 hour ago I was not thinking about the planet Mars. Conse- 
 quently if Mars existed an hour ago and has perdured 
 through the intervening hour, Mars, according to this de- 
 scription of "existence," is real. Now the question of liars' 
 existence in the past and persistence up to the present is a 
 question confronting me now. Shall I call the Mars of 
 an hour ago real or shall I call it unreal? In answering 
 this question the notion of existence I am discussing gives 
 us no help. It may be held that such a notion of existence 
 assumes that there arc some entities that are to be called 
 "existent," and so assumes the possibility of the Mars of an 
 hour ago being real. Now the Mars of an hour ago is an 
 object of my present consciousness; it was not an object of 
 consciousness for a subject contemporaneous with it. And 
 so I could not hold the Mars of an hour ago real if an entity 
 to be real had to be the object of a consciousness contem- 
 poraneous with it. Such a requirement would make it im- 
 possible for this past Mars to be real, and so would make 
 it impossible for there to be any "existent" entities accord- 
 ing to the notion of "existence'' we have undertaken to 
 elucidate. Consequently the most that we can get out of ^ 
 this notion of existence is the assertion that an entity to be 
 real need not be the object of a consciousness contem- 
 ^ Perry denies that the realist defines reality so.— New Realism, p. 117. 
 
 110 
 
 poraneous with k. Unless we draw from it, j addition ^ 
 the inference that an entity to be real must be perdurmg 
 
 '^ wtt iTave just given is one interpretation of inde- 
 i:>endence on consciousness. But independence on con- 
 Sousness is ambiguous, and the -terpreta.on of the 
 nhrase I have just given is not the most usual one. We 
 aXd u; to^another meaning of the phra. >f we come 
 back to our example of Socrates drmkmg 1ms cup of hem 
 lock Yesterday I was thinkmg oi him. To-day 1 haxe not 
 bet thinking Ji h,m, and indeed do not e-Uy reca w, t 
 it was that I was thinkmg ot yesterday. Now a real entity, 
 let us re ll s one, it i^ said, that is independent on con- 
 ic ousnes It is one that is there when I am "Ot thinkmg 
 about H And so Socrates and his cup of hemlock are real 
 tfXev are there now when I am not thinking about hem. 
 if I am no thinking about them at all, surely I can t say 
 Lt thev a e there I can't even talk about an entity that 
 s in twise an Object of my consciousness without -ser- 
 n<. a self-contradiction. It is not merely that an ego- 
 "entnc predicament prevents - from going^^^^^^^^^^^^^ 
 nhipcts of mv consciousness to entities that 1 can nevertne 
 les's asser t7be there. What lends plausibility to the notion 
 oTan elo-centric predicament is the fact that some entities 
 e^- vaguely Ld indefinitely objects o my conscio..- 
 ness, that they look beyond themselves, ^ to ^P^f ' ^^ 
 being manifestly incomplete, unfimshed and vag^J'^ ^ 
 ^en 1 am not thinking of Socrates and his cup of hemlock 
 1^ all lean say nothing about them. But suppose I am 
 otliLrg :7them definitely, but am -gt:^---- 
 
 1 a "something-; ^X::^'^^:^ ^^^^^^^ ^ 
 determmed. Let tnis oe wudL i u^nnlnrk Then 
 
 !Si« ^' lion 'o;".xis,.„«." if Socr.«s_ is .here^ow 
 
 .ague ;--f -£,i„„7hat some of the indefinite objects 
 :Tmrconsd:u:ne:: exist. It implies a meaning of the 
 
 III 
 
term 'Veal" that does not exclude all of these indefinite ob- 
 jects from the group of entities denoted by it. 
 
 For the epistemological dualist any thing-in-itself will 
 serve as an example of an indefinite ''something." For 
 when I have an "idea" of a thing-in-itself, that thing-in- 
 itself can not be wholly outside the scope of my conscious- 
 ness. If it were, 1 could not be aware of any of its 
 properties, not even of its correspondence with my idea. It 
 would be unthinkable. These trans-experiential entities, 
 consequently, are not wholly trans-experiential. Their full 
 and definite content is not experienced; but vaguely they 
 are known — some of their qualities are objects of conscious- 
 ness. And so these things-in-themselves belong to the class 
 of indefinite ''somethings." They can only be real if some 
 of these indefinite "somethings" can be real. 
 
 Perhaps I am overbold in daring to deal so cavalierly and 
 in such brief compass with such important subjects as the 
 realist-subjectivist controversy and the notion of self-trans- 
 cendent reference. But if I have not met all objections I 
 have at least stated my opinion that the entities under dispute 
 really are these indefinite "somethings" and not entities that 
 are in no sense objects of my consciousness. And feeling as 
 I do, 1 could not help listing this class of entities among the 
 classes of entities whose reality depends on the definition 
 of "real" that is implied. What common-sense thinks 
 about the reality of these indefinite "somethings" is rather 
 hard to determine. In general, I suppose, the realist is 
 justified in claiming the support of the man in the street. 
 Nevertheless, if an object is very inchoate and lacks all 
 content I think in ordinary parlance it will be called "un- 
 real." The God of a ''negative theology" is, I think, in 
 the ordinary usage of terms at the present time called "un- 
 real ;" and such a God is called "unreal" because such a God 
 can not be grasped, has too little content. A supra-spatial 
 God, on the other hand, that has the positive qualities of 
 all goodness and almightiness, is "real" according to com- 
 mon parlance, though just as far removed from perception 
 and just as surely not a member of the coherent system of 
 experience. "Existence" must be given a more definite 
 meaning than that which can be drawn from the customary 
 use of the word before we can determine the ontological 
 status of these vague and indefinite objects of conscious- 
 
 112 
 
 ness But when a sufficiently definite meaning is given this 
 term the status of these objects will stand out clear. ^ 
 
 All of the important philosophical questions I have just 
 mentioned are questions of existence. They arise rom 
 the assertion of the existence and counter-assertion of the 
 non-existence of one or another class of entities. And they 
 can be resolved, and indeed intelligently discussed, only 
 when we know what "existence" means. Now it has been 
 my thesis that "existence" does not mean the same thing 
 or anything near the same thing, to all of us. We can not 
 take the term as we find it used and extricate a definite 
 meaningful collection of terms that will be equivalent to it. 
 In tho'^e contexts in which the term does have a defimte 
 meaning, the meanings it has differ radically inter se from 
 context to context and from writer to writer. Consequent- 
 ly when we find a context in which "existence" is given a 
 definite meaning, we can not assert that the meamng found 
 there is the meaning of "existence;" we can only say tha 
 it is the meaning of "existence" for the writer laying it 
 down, and in the context in which it occurs. Each of these 
 many definitions of "existence," none of which is f/t. defi- 
 nition of "existence," determines the content of the^existen- 
 tial universe. Each writer, in using "existence m the 
 cense in which he chooses to use it, is peopling the real 
 world to suit himself. Consequently in one sense the real 
 world is not objectively there; its contents^ depend on the 
 sense in which we choose to use the word "real. 
 
 Since the earliest times of which we have record, philoso- 
 phers have been engaged in investigating the nature ot 
 realitv They have for the most part regarded their task 
 as a purely objective one. They have felt that reality was 
 there and that their task was simply to discover it and to 
 pick out its essential characteristics. They have made judg- 
 ments about existence and reality ; and they seem to have 
 regarded these judgments as genuine discoveries, discov- 
 erfes to be set forth in what Kant would call synthet 
 propositions. Yet if our thesis is correct, these judgi^ent 
 Lve implied an arbitrary and non-necessary definition o 
 reality. Thev have not been judgments he My of 
 which is objectively discovered. They have been judgmer^s 
 that have followed from a definition ot reality that has 
 IrtoraJly been assumed. Consequently an inquiry into 
 
 113 
 
the nature of reality is in an essential particular a mere 
 begging of the question. The metaphysician does not 
 simply discover the nature of reality; he finds it to be what 
 the particular definition he has tacitly assumed requires it 
 to be. To be sure, in following out the implications of a 
 given definition of reality, the metaphysician is dealing with 
 what are genuinely objective facts. But in starting from 
 one definition of reality rather than from another, he is 
 introducing into his results an element that is by no means 
 objective. There has in the past, however, been an almost 
 universal failure to distinguish between these two elements. 
 And so a considerable part of historical metaphysics has 
 involved a confusion between definitions and propositions 
 of fact. 
 
 Let us, however, recognize this necessary distinction, and 
 let us set ourselves to the task of defining reality. How 
 then shall we use the term '*real"? We ought not to use it 
 in a sense entirely different from any in which it has ever 
 been used before. To do so would be to cut ourselves off 
 from any possible readers and to prevent ourselves from 
 being taken seriously. But some specific definition of 
 '"existence" is essential. For since so many of our judg- 
 ments are existential judgments, unless "'existence" is given 
 some si>ecific meaning, these judgments are vague, indefinite 
 and worthless. And without some specific definition of 
 existence, moreover, the ontological status of those classes 
 of entities that have been troubling philosophy must remain 
 undetermined. 
 
 t*" 
 
 BIBLIOGRAPHY. 
 
 {To accompany Chapter i and Chapter 6), 
 
 Bradley, F. H. : 
 Bradley, F. H. : 
 Brentano : 
 
 Cohen, M. R. : 
 
 Dewey, J. : 
 Eckstein : 
 
 Friedrichs : 
 
 Holt, E. B. : 
 Hume : 
 Kant: 
 
 Lovejoy, A. O. : 
 Manan, W. T. : 
 Medicus, F. : 
 
 Moore, G. E. '• 
 Owen, R. B. : 
 
 Perry, R. B. : 
 
 Russell B.: 
 Sheldon,W.H.: 
 
 Appearance and Reality. 
 Essays on Truth and Reality. 
 Psychologic vom empirischen Standpunkte, 
 
 The^Use of the Words Real and Unreal, 
 Journ<il of Philosophy, etc, Vol. 13, 
 
 (1916) p. 635. . 
 
 Essays in Experimental Logic 
 Der Begriff des Dasems bei Julius Berg- 
 mann, Erlangen, 1902. ^.-^ 
 
 Beitrage zu einer Geschichte und Theone 
 
 des Existentialurteils, Prenzlau. 1906. 
 The Concept of Consciousness. 
 Treatise of Human Nature. . 
 
 Der einzig mogUche Beweisgrund zu einer 
 Demonstration des Dasems Gottes. 
 IVerke, ed. bv Hartenstem, Bd 2, p. 114. 
 Existence and Formal Logic; Jon^^^fof 
 Philosophy, etc.yol 8, (1911) P-^^o 
 The Existential Proposition; Journal of 
 Philosophy, etc, Vol. 8, (191 OP; 477- 
 Bemerkungen zum Problem der Existenz 
 Siematischer Gegenstande; Kantstu. 
 
 dien, Vol 19^ (i9i4)P-i- , ^ c; 
 The Nature of Judgment; Mmd, N. b. 
 
 Vol. 8, (1890) p. 17^- , , J 
 
 The Predicate^Real and Unreal ; Journal 
 
 of Philosophy, etc. Vol. 13, (iQi^) P- 
 
 A Realistic Theory of Independence; in 
 
 The New Realism, p. 9^- 
 Principles of Mathematics Vol. i. 
 The Demolition of Unreality; Journal g 
 
 Philosophy, etcVol 13, (1916) P- 3i»- 
 
 114 
 
 115 
 
VITA. 
 
 tS^'.'^'a?/^^'^*''.^''' ""'^^ ^<^™ '" Baltimore on July i, 
 p1^fdin'<^ r '"'^""■"^fh^ P"l>lic schools of Baltimore and 
 wh're in oifT "' J^'^/S""]"^--^- ^ went to Harvard 
 Znl / ^ ^' ""'"'^ '^^ ^"g^""^^ Of bachelor of arts 
 
 cum lamkm Economics. It was under the influence of a 
 fellow-student at Harvard, one who was always earnest 
 
 "vltrn"'"' ''"'-"'' '•'"•"''^^^^ °^ •^^ enthSa^r:: 
 now anarchism now socialism, now Nietzsche and now 
 
 n,rnT 7'r'i ^'^'^ ^ ^T^ ^" ^°"^'^" '"^■■e earnestly t°e 
 tZ '^Hn , '' '"!' '° ^^^' ^'"^ ini,K)rtance of ethics.^ At 
 John. Hopkms where I was a graduate student in phi- 
 osophy under Professor Lovejoy. from 1914 to ioi6 I 
 learned to appreciate the interest and significance of the 
 
 se'emsT: Z ^'^'^T'^^"^' ^"'' "-'-P^^y-s. Ethics Im 
 seems to me the most important branch of philosophy; but 
 metaphysics has come to engross my attention to a very 
 consideraole extent. During the summer season of 10,^ 
 aiiddunng the year 1916-1917. I studied at Columbia unde? 
 .nd h'p % '■'^' l^'oo^"^'-'^?'^- Montague. Bush, Costello, 
 Honf^; 1™"'"; /" September, 1917, I returned to John 
 rWr. R '-T ^ have completed my work for a doctor's 
 degree. Besides my principal field of study, philosophv. I 
 have done graduate work in political economy and political 
 science. ^ 
 
 I am particularly indebted to Professor Lovejoy under 
 whose direction my studies in philosophy have been carried 
 out He has tried hard to make me critical and circumspect 
 m dealing with philosophical problems and to make me 
 keen and acute in analysis. I am also grateful to the mem- 
 bers of the departmeut of philosophy at Columbia and to 
 niy tellow-students at Columbia and at Johns Hopkins for 
 having helped to make my four years as a graduate student 
 most instructive and most enjoyable. 
 
 
 lili 
 
 "1 
 
 k 
 
 116 
 
r.'^':^^''' UNIVERSITY LIBRARIES 
 
 provided by the libra J ru^s „r h"" ^" ^^'' °' borrowing, as 
 the Librarian in ch^Je '^ ''""'^' ^"^ge°>ent with 
 
 :;t:i:;;ir, 
 
 p..:i!i!\is 
 
 
 !:s.: 
 
 COLUMBIA UNIVERSITY 
 
 0032142633