Glass CN Book Digitized by the Internet Archive in 2011 with funding from The Library of Congress http://www.archive.org/details/personalidealismOOstur Lf % U PERSONAL IDEALISM •9- PERSONAL IDEALISM PHILOSOPHICAL ESSAYS BY EIGHT MEMBERS OF THE UNIVERSITY OF OXFORD EDITED BY HENRY STURT MACMILLAN AND CO., Limited NEW YORK : THE MACMILLAN COMPANY 1902 A U rights reserved % ^S7> PREFACE This volume originated in the conversations and discus- sions of a group of friends drawn together primarily by their membership in the Oxford Philosophical Society. The Society was started in the spring of 1898, and among some of the most regular attendants at its meetings a certain sympathy of view soon declared itself. In the course of two years the trend of opinion had grown so definite as to suggest to me the project of a volume of essays. Among those who- seemed likely to contribute I circulated a programme which made it the object of our volume " to represent a tendency in contemporary thinking, to signalise one phase or aspect in the development of Oxford idealism." That tendency was summed up in a phrase which I thought I was originating at the time I wrote the programme, though it seems to have occurred independently to others. 1 It is the phrase we have chosen for our title, " Personal Idealism." For me our volume fulfils the purpose with which it was projected so far as it develops and defends the principle of personality. Personality, one would have supposed, ought never to have needed special advocacy in this self-assertive country of ours. And yet by some of the leading thinkers of our day it has been neglected ; while by others it has been bitterly attacked. What makes its vindication the more urgent is 1 Prof. Howison uses it to characterise the metaphysical theory of his Limits of Evolution, published last year. V vi PERSONAL IDEALISM that attacks have come from two different sides. One adversary tells each of us : " You are a transitory resultant of physical processes" ; and the other: " You are an unreal appearance of the Absolute." Naturalism and Absolutism, antagonistic as they seem to be, combine in assuring us that personality is an illusion. Naturalism and Absolutism, then, are the adversaries against whom the personal idealist has to strive ; but the manner of the strife must be different in each case. Per- sonal Idealism is a development of the mode of thought which has dominated Oxford for the last thirty years ; it is not a renunciation of it. And thus it continues in the main the Oxford polemic against Naturalism. To it and to Naturalism there is no ground common, except that both appeal to experience to justify their interpretations of the world. Thus against this adversary the argument must take the form of showing that from naturalistic premises no tolerable interpretation of the cardinal facts of our experience can be made. If it be asked what are those cardinal facts, I should answer : Those which are essential to the conduct of our individual life and the maintenance of the social fabric. They are summarily recognised in the credo that we are free moral agents in a sense which cannot apply to what is merely natural. Round this formula of conviction are grouped the ques- tions debated with Naturalism in our volume. They are the reality of human freedom, the limitations of the evolu- tionary hypothesis, the validity of the moral valuation, and the justification of that working enthusiasm for ideals which Naturalism, fatalistic if it is to be logical, must deride as a generous illusion. If these crucial questions be decided in our favour, the system of Naturalism is con- demned. Accordingly, where Naturalism confronted us, we were not unfrequently obliged to take the aggressive and carry PREFACE vii the war far into the enemy's country. But in the other essays a different line of action has been taken. The Absolutist is a more insidious, perhaps more dangerous adversary, just because we seem to have more in common with him. He professes to agree with us in the funda- mental conviction that the universe is ultimately spiritual ; against the naturalist it was just this conviction which had to be vindicated. We decided, then, to meet the Absolu- tist with what may be called a rivalry of construction. Absolutism has been before the world for a century, more or less. It has put forth its account of knowledge, of morals, and of art ; and that account, suggestive though it is, has not satisfied the generality of thinking men. If the grounds of dissatisfaction be demanded, I can only give the apparently simple and hackneyed, but still fundamental answer, that Absolutism does not accord with the facts. Thus, instead of entering upon the intricate task of refuting Absolutism, we have felt free to adopt the more congenial plan of offering specimens of constructive work on a principle which does more justice to experience. Our essays are but specimens. They indicate lines of thought which could not be worked out fully in the space allowed. But they are extensive enough, let us hope, to enable the reader to judge whether their general line of interpretation is not more promising than that of Absolutism. It may be objected that we are wrong in assuming that Absolutism cannot be reconciled with the principle of per- sonality. In reply two points of incompatibility may be specified shortly; further particularity is impossible without a much fuller statement, more especially since Absolutism is not so much a definite system as an aggregate of tendencies without a universally acknowledged expositor. The two points in respect of which Absolutism tends 1 to 1 I use a guarded phrase, because what follows is not entirely true of exponents of Absolutism so distinguished as Prof. Henry Jones and Prof. Royce. viii PERSONAL IDEALISM be most unsatisfactory are, first, its way of criticising human experience, not from the standpoint of human experience, but from the visionary and impracticable standpoint of an absolute experience ; and, secondly, its refusal to recognise adequately the volitional side of human nature. Both matters are dealt with in the essay on Error which stands first in the volume. There it is shown that error and truth are not dependent upon the Absolute ; in other words that we can know with certainty without knowing the absolute whole of Reality ; and that, if we err, it is by human criteria, not by a theory of the Absolute, that we measure the degree of our error. Further, in regard to volition, the same essay shows that error is relative, not to the content of knowledge only, but also to its intent, i.e., the intention of the agent in setting out upon his search for knowledge. The reader may be left to trace for himself the wider operation of these principles. In conclusion there is one feature in our essays to which I would venture to call attention as constituting what to my mind is the most valuable feature of their method ; that is, the frequency of their appeal to ex- perience. The current antithesis between a spiritual philosophy and empiricism is thoroughly mischievous. If personal life be what is best known and closest to us, surely the study of common experience will prove it so. ' Empirical idealism ' is still something of a paradox ; I should like to see it regarded as a truism. H. S. CONTENTS ESSAY PAGE I. Error. By G. F. Stout, M.A., Wilde Reader in Mental Philosophy i II. Axioms as Postulates. By F. C. S. Schiller, M.A., Fellow and Assistant Tutor of Corpus Christi College ...... 47 III. The Problem of Freedom in its Relation to Psychology. By W. R. Boyce Gibson, M.A., Queen's College, Lecturer in London University . 134 IV. The Limits of Evolution. By G. E. Underhill, M.A., Fellow and Senior Tutor of Magdalen College 193 V. Origin and Validity in Ethics. By R. R. Marett, M.A., Fellow and Tutor of Exeter College . 221 VI. Art and Personality. By Henry Sturt, M.A., Queen's College . . . . .288 VII. The Future of Ethics : Effort or Abstention ? By F. W. Bussell, D.D., Fellow, Tutor, and Vice- Principal of Brasenose College . . . 336 VIII. Personality, Human and Divine. By Hastings Rashdall, D.Litt, Fellow and Tutor of New College 369 Note. — Each writer is responsible solely for his own essay. IX ERROR 1 By G. F. Stout SYNOPSIS I. In Error, what is unreal seems to be thought of in the same way as the real is thought of when we truly know it. How is this possible ? As an essential preparation for answering this question we must first deal with another. Do other modes- of thinking exist besides those which can be properly said to be either true or false ? There are two such modes, (i) Indeterminate or problematic thinking. (2) Thinking of mere appearance without affirming it to be real. 2 and 3. To think indeterminately is to think of something as one of a group of alternatives, without deciding which. The indeterminateness lies in not deciding which ; and so far as the indeterminateness extends there is neither truth nor error. Whatever is thus indeterminately thought of belongs to the Intent of consciousness. The term Content should be reserved for what is determinately presented. In cognitive process, indeterminate thinking takes the form of questioning as a mental attitude essentially analogous to questioning. Interrogative thinking is the way we think of something when we are interested in knowing it, but do not yet know it either truly or falsely. Its distinc- tive characteristic is that the decision between alternatives is sought for in the independent reality of the total object in which we are interested. This object is regarded as having a determinate constitution of its own, independently of what we may think about it. We are active in cognitive process only in compelling the object to reveal its nature. The activity is experimental ; its result is determined for us and not by us. In the play of fancy, on the contrary, we do not seek to conform our thought to the predetermined constitution of our object. We select alternatives as we please, and to this extent make the object instead of adapting ourselves to its independent nature. 1 Throughout this essay I am deeply indebted to the criticisms and sugges- tions of Professor Cook Wilson. In particular, I have substantially adopted his account of the distinction between abstract terms and adjectives, in place of a less satisfactory view of my own. B 2 G. F. STOUT i 4. Besides indeterminate thinking there is yet another mode of thinking which is neither true nor false. It consists in thinking of mere appear- ance without taking it for real. This happens, for example, in the play of fancy. Mere appearance consists in those features of an object of con- sciousness which are due merely to the special conditions, psychological and psychophysical, of its presentation, and do not therefore belong to its independent reality. 5 and 6. Error occurs when what is merely apparent, appears to belong to an independent reality in the same way as its other real features. The conditions under which this occurs may be divided under two heads. (1) Confusion. (2) Ignorance and inadvertence. Ignorance or inadvertence are present in every error, Confusion only in some. 7. It follows from the very nature of error that it cannot exist unless the mind is dealing with something independently real. Hence, some truth is presupposed in every error as its necessary condition. 8. There are limits to the possibility of error. There can be no error unless in relation to a corresponding reality, which is an object of thought for him who is deceived. Further, this reality must be capable of being thought of without the qualification which is said to be illusory. Hence, among other results, we may affirm that abstract objects cannot be illusory unless they contain an internal discrepancy. For, they are considered merely for themselves, and not as the adjectives of any other reality in relation to which they can be illusory. So far as the abstract object is merely a selected feature of actual existence, it is not merely not illusory ; it is real. It is something concerning which we can think truly or falsely. 9. But the constructive activity of the mind variously transforms and modifies the abstract object, in ways which may have no counterpart in the actual. To this extent, the abstract object may be relatively unreal. None the less, such mental constructions, so far as they belong to scientific method, are experimental in their character and purpose. They serve to elicit the real nature of the object as an actual feature of actual existence. Thus abstract thinking, even when it is construc- tive, gives rise to judgments concerning -what is real. These judgments may at least be free from the error of ignorance. For the mind may require no other data to operate on in answering its questions except those that are already contained in the formulation of them. Errors of confusion and inadvertence may still occur. But even these are avoidable by simplifying the problems raised. Thus, abstract thinking yields a body of certain knowledge, to. Certainty, then, is attainable. It exists when a question is made to answer itself, so as to render doubt meaningless. When this is so the real is present to consciousness, as the illusory can never be. I. The General Nature of Error § 1. The question raised in the present essay is funda- mentally the same as that discussed in Plato's Thecetetus. The Thecztetus may be described as a dialogue on Theory of Knowledge. But the central problem did not take the same shape for Plato as it does for most modern epistemo- logists since the time of Descartes. What the moderns i ERROR 3 trouble themselves about is the nature and possibility of knowledge in general. How, they ask, can a particular individual be in such relation to a reality which transcends and includes his own existence as to know it. Can he know it otherwise than through the affections of his own consciousness which it produces ? If it can only be known in this way, can it be said to be known at all ? Are not his own mental states the only existences which are really cognised ? Questions of this sort occupy modern philo- sophers, and they have given rise to the Critique of Pure Reason, among other results. But I cannot see any evidence that in this form they gave much trouble to Plato. The nature and possibility 6f knowledge would probably not have constituted a problem for him at all, had it not been for the existence of error. That we can know was for him a matter of course, and it was also a matter of course that we may be ignorant. But he was puzzled by the conception of something intermediate be- tween knowing and not knowing. If an object is present to consciousness, it is pro tanto known ; if it is not present to consciousness, it is not known. But in so far as it is known there can be no error, because the knowledge merely consists in its presence to consciousness. And again, in so far as it is not known there can be no error, for what is not known is not present to conscious- ness : it is to consciousness as if it were non-existent, and therefore the conscious subject as such cannot even make a mistake concerning it. Hence we cannot be in error either in respect to what we know or to what we don't know, and there seems to be no third alternative. This is Plato's problem, and ours is fundamentally akin to it. For with him we must assert that, in knowing, the object known must be somehow thought of, and in this sense present to consciousness. The grand lesson of the history of Philosophy is just that all attempts to explain knowledge on any other assumption tumble to pieces in ruinous incoherence, and that from the nature of the case they must do so. The only form such attempts can take 4 G. F. STOUT i is to treat knowledge simply as a case of resemblance, conformity, or causality, between something we are conscious of and something we are not conscious of. What we are conscious of we may be said to know immediately. But the something we are not conscious of, how can that be known. The only possible pretence of an answer is that the knowing of it is wholly constituted by its somehow resembling, or corresponding to or causing what is actually present to consciousness. But this pretended answer in all its forms is utterly indefensible. The supposed conformity, resemblance, or causality is nothing to us unless we are in some manner aware of it. If I am to think of A as resembling B or as corresponding to it or as causing it, I must think of B as well as of A. Both A and B must be in some way present to my consciousness. The very distinction of truth and error involves this. Truth is frequently defined as the agreement, and error as the disagreement, of thought with reality. But this definition, taken barely as it stands, is defective and mis- leading. It omits to state that the reality with which thought is to agree or disagree must itself be thought of, and that the thinker must intend to think of it as it is. Otherwise there can be neither truth nor error. I may imagine a dragon, and it may be a fact that dragons do not actually exist. But if I do not intend to think of something which actually exists, I am not deceived. And, on the same supposition, the actual existence of dragons exactly resembling what I imagine would not make my thought true. It would be a curious coincidence and nothing more. So in general, if we assume a sort of inner circle of presented objects, and an outside circle of unpresented realities, we may suppose that the presented objects are similar or dissimilar to the real existences, or that in some other way they correspond or fail to correspond to them. But the resemblance or correspondence would not be truth and the dissimilarity or non-correspondence would not be error. Even to have a chance of making a mistake we must think of something real and we must intend to think i ERROR 5 of it as it really is. The mistake always consists in investing it, contrary to our intention, with features which do not really belong to it. And just here lies the essential problem. For these illusory features seem to be present to cognitive consciousness in the same manner as the real features are. 1 How then is it possible that they should be unreal. This is our problem, and evidently it is closely akin to that raised by Plato. But there is a difference and the difference is important. Our difficulty arises from the fact that when we are in error what is unreal appears to be present to consciousness in the same manner as what is real is presented when we truly know. While the erroneous belief is actually being held, the illusory object seems in no way to differ for the conscious subject from a real object. The distinction only arises when the conscious subject has discovered his mistake, and then the error as such has ceased to exist. The essential point is not merely that both the illusory and the real features are presented, but also that they are both presented as real and both believed to be real. It is not enough to say that they are both really appearances. We must add that they are both apparent realities. Now the question did not take this shape for Plato. The difficulty which he emphasises is not that what is unreal may be present to consciousness in the same way as what is real. The stumbling-block for him is rather that it is present to consciousness at all. For what is present to consciousness must, according to him, be known ; and if it is known, how can it be unreal ? On the other hand if it is not present to consciousness, it is simply unknown. Thus there would seem to be no room for that something intermediate between knowing and being ignorant which is called error. Before proceeding to deal with our own special difficulty it will be well to examine the Platonic assump- 1 It will be found in the sequel that I admit cases where the conditions which make error possible are absent, and in these cases the real is present to conscious- ness in a different manner from that in which the unreal is capable of being presented. M 6 G. F. STOUT i tion that whatever is in any way present to consciousness, whatever is in any way thought of, is known — unless indeed error be an exception. Besides knowing and being mistaken it is also possible merely to be aware of a mere appearance which not being taken for reality is therefore not mistaken for reality. This is a point to which we shall recur at a later stage. For the present I wish to draw attention to another mode of thinking which is neither knowing, nor mere appearance, nor error. II. Intent and Content § 2. Cognitive process involves a transition or at- tempted transition from ignorance to knowledge, and where we are trying to make this transition there may be an intermediate state which is neither knowledge, nor ignorance, nor error. We may be interested in knowing what we do not as yet know. But we cannot be interested in knowing what we do not think of at all. In what way then do we think of anything before we know it or appear to know it ? I reply that it is an object of interrogative or quasi-interrogative consciousness. It is thought of as being one and only a certain one of a series or group of alternatives, though which it is we leave undecided. Sometimes the question is quite definite. The alternatives are all separately formulated. Thus we may ask — Is this triangle right-angled, acute, or obtuse? In putting the question we seek for only a certain one of the three alternatives, but until the answer is found we do not know which of them we are in search of; we do not know it although we think of it. Sometimes the question is only partially definite ; only some alternatives or perhaps only one of them is separately formulated. Thus we may ask — Has he gone to London, or where else ? Sometimes again, the question is indefinite. What is sought is merely thought of as belonging to a group or series of alternatives of a certain kind, which are not i ERROR 7 separately formulated. Suppose that I am watching the , movements of a bird. My mental attitude is essentially ^ of the interrogative type even though I shape no definite question. I am virtually asking, — what will the bird do next ? The bird may do this, that, or the other, and I may not formulate the alternatives. But whatever changes in its position or posture may actually occur, are for that very reason what I am interested in knowing before I know them. I am looking for the determinate while it is as yet undetermined for me. Or, to take an illustration of a different kind. I have to find the number which results from multiplying 1947 by 413. At the outset I do not know what the number is, and yet there is a sense in which I may be said to think of it. I think of it determinately as the number which is to be obtained by a certain process. So far I may be said to know about it. But the knowledge about it is not knowledge of what it is. Yet this is what I aim at knowing, and therefore I must in some sense think of it. I think of it indeterminately. I think of it as being 'a certain one of a series of alternative numbers, which I do not separately formulate. So far I have considered only cases in which know- ledge is sought before it is found, so that the transition from the indeterminate to the determinate comes as the answer to a question definite or indefinite. But there are instances in which this is not so. There are instances in which the answer seems to forestall the question. A picture falls while I am writing. I was not previously thinking of the picture at all, but of something quite different. My attention is only drawn to the picture by its fall. But the picture then becomes distinguished as subject from its fall as predicate. This means that the picture is thought of as it might have existed for consciousness before the fall took place. It is regarded as relatively undetermined and the predicate as a deter- mination of it. The fall of the picture comes before consciousness as if it were the answer to a question. 1 1 Of course if we suppose that the noise of the fall first awakens 8 G. F. STOUT i )h. The relation of subject and predicate is essentially- analogous to what it would have been if we had pre- viously been watching to see what would happen to the picture. In this and similar instances, there is an actual dis- tinction of subject and predicate essentially analogous to that of question and answer. But in a very large part of our cognitive experience no such distinction is actually made. I look, let us say, at my book-shelves, and I am aware of the books as being on the shelves and of the shelves as containing the books. But I do not formulate verbally or otherwise the propositions : — " The books are on the shelves," or " The shelves contain the books." Neither the books nor the shelves are regarded as re- latively indeterminate and as receiving fresh determination in the fact that one of them stands in a certain relation to the other. Again, I may meet a friend and begin to talk to him on some political topic, proceeding on the assump- tion that he agrees with me. I find that he does not, and only then do I wake up to the fact that I have been making an assumption. And it is only at this point that the distinction of subject and predicate emerges. Such latent or unformulated presuppositions are constantly present in our mental life. They are constantly involved in the putting of questions. They are constantly involved in the conception of the subjects to which we attach pre- dicates, and also in the conception of the predicates. The nature, function, and varieties of this kind of cog- nitive consciousness we cannot here discuss. It is suffi- cient for our purpose to note that all such cognitions are capable of being translated into the subject -predicate form, without loss or distortion of meaning. Further, this translation is necessary if we are to submit them to logical examination. In particular, we cannot otherwise deal with any question relating to their truth or falsity. The disjunction, true or false, does not present itself to the question — What is falling ? before we think of the picture, the fall is subject and the picture predicate. But I do not think that this account of the matter always holds good in such cases. i ERROR 9 consciousness until we distinguish subject and predicate. In the absence of this distinction there is only uncon- scious presupposing or assuming. But when the dis- tinction is made it is essentially analogous to that of question and answer. So far as our thought is indeterminate there can be neither truth nor error. But it must be remembered that our thought is never purely indeterminate. A question always limits the range of alternatives within which its answer is sought ; and the question itself may be infected with error. A man for instance may set out to find the square root of two. In the formulation of the question he leaves it undetermined what special numerical value the root of two has. But he assumes that it has some determinate numerical value. To this extent his question is infected with error, and it can have no real answer unless it is reshaped. If he seems to himself to find an answer, he does but commit a further error. What he thinks he wants to know, is not what he really wants to know. Hence in finding what he really wants to know he must alter the form of his question. This leads me to make a suggestion in terminology. The term 'content of thought' is perpetually being used with perplexing vagueness. I propose to restrict its application. We cannot, without doing violence to language, say that the indeterminate, as such, is part of the content of thought. For it is precisely what the thought does not contain, but only intends to contain. On the contrary, we can say with perfect propriety that it belongs to the intent of the thought. It is what the conscious subject intends when its selective interest singles out this or that object. From this point of view we can deal advantageously with a number of logical and epistemological problems. For instance it throws light on the proposed division of propositions into analytic and synthetic. Whatever can be regarded as a judgment, whether expressed in words or not, is and must be both analytic and synthetic. It is synthetic as regards content and analytic as regards intent. io G. F. STOUT i While I am watching a bird, whatever movement it may- make next belongs to the intent of my thought, even before it occurs. It is what I intend to observe. But the special change of posture or position does not enter into the con- tent of my thought until it actually takes place under my eyes. Hence each step in the process is synthetic as re- gards content though analytic as regards intent. This holds generally for all predication which is not mere tautology. If the predicate did not belong to the intent of its subject, there would be nothing to connect it with this special subject rather than with any other. If it already formed part of the content there would be no advance and there- fore no predication at all. From the same point of view, we may regard error as being directly or indirectly a discrepancy between the intent and content of cognitive consciousness. Sometimes the discrepancy lies in a latent assumption. The initial question which determines the intent of thought may itself be infected with error, as in the example of a man setting out to find the square root of two. In such cases it would seem that a man cannot reach truth unless he finds something which he does not seek. But the reason is that there is already a discrepancy between intent and content in the very formulation of his initial question. The man is interested in formulating an answerable question, and he fails to do so. Similarly wherever error occurs there is always an express or implied discrepancy between intent and content. It follows that truth and error are essentially relative to the interest of the subject. To put a question seriously is to want to know the answer. A person cannot be right or wrong without reference to some interest or purpose. A man wanders about a town which is quite unfamiliar without any definite aim except to pass the time. Just in so far as he has no definite aim he cannot go astray. He is equally right whether he takes a turn which leads to the market-place or one which leads to the park. If he wants to amuse himself by sight-seeing it may be a i ERROR 1 1 mistake for him to go in this direction rather than in that. But if he does not care for sight-seeing, he cannot commit this error. On the other hand if his business demands that he should reach the market-place by a certain time, it may be a definite blunder for him to take the turn which leads to the park. In this example the interest is primarily practical and the blunder is a practical blunder. But the same principle holds good for all Tightness and wrongness even in matters which appear purely theoretical. Our thought can be true or false only in relation to the object which we mean or intend. And we mean or intend that object because we are, from whatever motive, interested in it rather than in other things. If a man says that the sun rises and sets, he may refer only to the behaviour of the visible appearance of the sun, as seen from the earth's surface. In that case you do not convict him of error when you remind him that it is the earth which moves and not the sun. For you are referring to something in which he was not interested when he made the statement. Error is defeat. We mean to do one thing and we actually do another. So far as the error is merely theoretical what we mean to do is to think of a. certain thing as it is, and what we actually do is to think of it as it is not. This implies that the thing we think of has a constitu- tion of its own independent of our thinking — a constitution to which our thinking may or may not conform. A question is only possible on the assumption that it has an answer predetermined by the nature of the object of inquiry. It is this feature which marks off the interroga- tive consciousness peculiar to cognitive process from the form of indeterminate thinking which is found in the play of fancy. While the play of fancy is proceeding, its object is at any moment only partially determined in consciousness, and each step in advance consists in fixing on one alternative to the exclusion of others. But the intent of imaginative thinking is different from that of cognitive, and consequently the decision between com- peting alternatives is otherwise made. An examination 12 G. F. STOUT i of this difference will carry us a step farther in our inquiry. III. Imaginative and Cognitive Process S 3. Imaginary objects as such are creatures of our own making. When we make up a fairy-tale for a child the resulting object of consciousness is merely the work of the mind, and it is not taken by us for anything else. In the development of intent into content, of indeterminate into determinate thinking, the decision among alternatives is made merely as we please, whatever be our motive. It depends purely on subjective selection so far as the process is imaginative. It is necessary to add this saving clause. For no imaginative process is merely imaginative. Even in the wildest play of fancy, the range of subjective selection is restricted by limiting conditions. Gnomes must not be made to fly, or giants to live in flower-cups. Thackeray's freedom of selection in composing Vanity Fair was circumscribed by his purpose of giving a faithful repre- sentation of certain phases of human life. In so far as such limiting conditions operate, the mental attitude is not merely imaginative. It is imaginative only in so far as the limiting conditions still leave open a free field for the loose play of subjective selection. This freedom of subjective selection is absent in cognitive process. Instead of deciding between alterna- tives according to his own good pleasure, the conscious subject seeks to have a decision imposed upon him independently of his wish or will. It is true that cognitive process may include a varied play of sub- jective selection. But there is one thing which must not be determined by subjective selection. It is the deciding which among a group of alternative qualifica- tions is to be ascribed to the object we are interested in knowing. In cognitive process as such we are active merely in order that we may be passive. Our activity is successful i ERROR 13 only in so far as its result is determined for us and not by us. In this sense we may say that the work of the mind when its interest is cognitive has an experimental character. What is ordinarily called an experiment is a typical case of this mental attitude. A chemist applies a test to a substance. The application of the test is his own doing. But the result does not depend on him : he must simply await it. Yet he was active only in order to obtain this result. He was active that he might enable himself to be passive. He was active in order to give the object an opportunity of manifesting its own independent nature. His activity essentially consists in the shaping of a question so as to wrest an answer from the object of inquiry. In all cognitive process the mental attitude is essentially analogous. Suppose that I am interested in knowing whether any number of terms in the series 1 +2 + 4 +^, etc., have for their sum the number 2. I may proceed by actually adding. This is a mental experiment, but it turns out to be unsuccessful. It does not transform my initial question into a shape in which it wrests its own answer from its object. By adding any given number of terms I find that the sum is less than two. But the doubt always remains whether by taking more terms I may not reach a different result. Under this mode of treatment my object refuses to mani- fest its nature so as to answer my question. I fail to obtain an answer by waiting for data which I have not got — by waiting till some number of terms shall present itself having 2 for their sum. Accordingly I re- sort to another form of experiment. I appeal to ex- perience a priori, instead of experience a posteriori. Instead of looking for data which I have not got, I try to obtain an answer by manipulating the data which I already possess in the very conception of the series as such, and of the number 2. I fix attention on the form of serial transition, and I inquire whether this is capable of yielding a term such as will make 2 when it is added to the sum of preceding terms. I find that i 4 G. F. STOUT i such a term must be equal to the term that precedes it, and that according to the law of the series each term is the half of that which precedes it. Hence no number of terms can have 2 as their sum. My experiment is successful. It translates my question into a shape in which it compels an answer from its object. Suppose again that I am verifying the statement that two straight lines cannot enclose a space. I conceive two lines as straight, ignoring all else but their being lines and their being straight. I then consider the varying changes of relative position of which they are capable, and I find by trial that only certain general kinds of variation are possible. If I think of them as not meeting at all, they refuse to enclose a space. The same is true when they are thought of as meeting at one point only. But if they meet at more than one point they insist on coinciding at all points. This result of my experiment does not depend on my activity ; it is deter- mined for me by the nature of the object on which I operate, by the constitution of space and of straight lines. It will be seen that I have included under the term experiment two very different groups of cases. To the first group belong such instances as the application of a chemical test. Their distinctive character is that an answer to the question raised cannot be obtained merely by operating on the data which are already presupposed in putting the question itself. When I am watching to see what a bird will do next, the decision does not come merely from a consideration of what I already know about the bird. The decision is given by a posteriori experience. On the other hand, if I want to know whether two straight lines can enclose a space I need no data except lines, straightness, and space as such. I can shape my question by mentally operating on these data so that it answers itself. The decision is given by a priori experience. But both results obtained a priori and those obtained a postei'iori are equally due to an experi- i ERROR 1 5 mental process, to an activity that exists in order that it may be determined by its object. IV. Mere Appearance and Reality § 4. All error consists in taking for real what is mere appearance. In order to solve the problem of error we must therefore discover the meaning of this distinction between mere appearance and reality. We are now in a position to take this step. We have a clue in the foregoing discussion of the nature of the imaginary object as such. The imaginary object as such is unreal and we see quite clearly wherein its unreality consists. It is unreal inasmuch as its imaginary features as such have no being independently of the psychical process by which they come to be presented to the individual consciousness. They are merely the work of the mind, merely the product of subjective selection and. they are therefore mere appearances. But though they are mere appearances, they are not therefore illusory or deceptive. They are not deceptive, because they are not taken for real. While the purely imaginative attitude is maintained, they are not taken either for real or unreal. The question does not arise, because in imagination as such we are not interested in the constitution of an object as independent of the process by which we come to apprehend it. 1 On the other hand when the question is raised whether what we merely imagine has this independent being, we commit no error if we refuse to affirm that it has. Mere appear- ance is not error so long as we abstain from confusing it with reality. The imaginary object is only one case of mere appear- ance. It is the case in which the nature of what is presented to consciousness is determined merely by the psychical process of subjective selection. But there is always mere appearance when and so far as the nature of 1 The fact that the object is merely imaginary is not attended to. We do not contrast it as unreal with something else as real. If we are externally reminded of its unreality, the flow of fancy is disturbed. The flow of fancy is also disturbed if we are called on to believe that our fancies are facts. The whole question of reality or unreality is foreign to the imaginative attitude. 1 6 G. F. STOUT i a presented object is determined merely by the psycho- logical conditions of its presentation, whatever these may be. There is always mere appearance when and so far as a presented object has features due merely to the special conditions of the flow of individual consciousness as one particular existence among others, connected with a particular organism and affected by varying circum- stances of time and place. In ordinary sense-perception the thing perceived is constantly presented under modifications due to the varying conditions of the perceptual process. But what we are interested in knowing is the thing so far as it has a constitution of its own independent of these conditions. Hence whatever qualifications of the object are recognised as having their source merely in the conditions of its presentation are pro tanto contrasted with its reality as being merely its appearances. An object looked at through a microscope is presented as much larger and as containing far more detail than when seen by the naked eye. But the thing itself remains the same size and contains just the same amount and kind of detail. The difference is due merely to conditions affect- ing the process of perception, and it is therefore merely apparent. On the other hand, the details which become visible when we use the microscope, and which were previously invisible, are ascribed to the real object. The parts of the object being viewed under uniform perceptual conditions, whatever differences are presented must be due to it, and not to the conditions of its presentation. The visible extension of a surface increases or diminishes according as I approach or recede from it, and the visible configuration of things varies according to the point of view from which I look at them. But these changes being merely due to the varying position of my body and its parts are regarded as mere appearances so far as they are noted at all. 1 If I close my eyes or look 1 To a large extent they pass unnoted. We have acquired the habit of ignoring them. So far as this is the case, they are not apprehended as appear- ances of the thing perceived. i ERROR 17 away, objects, previously seen, disappear from view. But this being due merely to the closing of my eyes or my turning them in another direction is no real change in the things. They are really just as they would have been if I had continued to look at them. It is important to notice that in cases of this kind the mere appearance is not to be identified with any actual sense-presentation. The appearance is due to a certain interpretation of the sensible content of perception, suggested by previous experience. When we see a stick partially immersed in a pool, the visual presentation is such as to suggest a bend in the stick itself. Even while we are denying that the stick itself is bent, we are thinking of a bend in it. Otherwise the act of denial would be impossible. This being understood, it is easy to see that all cases of mere appearance are in principle analogous to the examples drawn from sense-perception. Mere appearance exists wherever anything is thought of as having a character which does not belong to it inde- pendently of the psychical process by which it is appre- hended. Unless this character is affirmed of its inde- pendent reality, there is no error. If a man denies that two lines are commensurable, or if he questions whether they are so or not, their commensurableness must have been suggested to his mind. If the lines are really incommensurable, this suggestion is mere appearance. Should he affirm them to be commensurable he is in error. We now pass to two important points of principle. In the first place it should be clearly understood that mere appearance is a qualification of the object appre- hended and not of the mind which apprehends it. There is here a complication due to an ambiguity in the term, appearance. It may mean either the presenting of a certain appearance or the appearance presented. The last sense is that in which I have hitherto used the word in speaking of mere appearance. A stick, partly immersed in a pool, appears bent in the sense that it presents the appearance of being bent. The bend is the appearance C 1 8 G. F. STOUT i presented. Now the presenting of this appearance is an adjective of the stick as an independent reality. The stick which is really straight really presents the appearance of being bent. It does not merely appear to appear bent : it really appears so. Given the psychological and psychophysical conditions of its presentation, it is part of its independently real nature that it should wear this appearance. But the apparent bend is not a qualifica- tion of the independently real stick. It is a qualification of a total object constituted by the real stick so far as it is present to consciousness and also by certain other presented features which are due merely to the special conditions under which the real stick is apprehended. Mere appearance is in no sense an adjective of the cogni- tive subject. The person to whom a straight staff appears as bent when it is partially dipped in a pool is not himself apparently bent on that account, either bodily or mentally. He who imagines a golden mountain is not himself the appearance of a golden mountain : his psychical processes are not apparently golden or mountainous. The existence of mere appearance is not that of a psychical fact or event except in the special case where the real object thought of happens to be itself of a psychical nature. In the second place, the distinction between mere appearance and reality is relative to the special object we are interested in. In ordinary sense-perception we are interested in the objects perceived so far as they have a constitution independent of the variable conditions bodily and mental of the perceptual process. Contrast this with the special case of a beginner learning to draw from models. For him what in ordinary sense-perception is mere appearance becomes the reality. He has to repro- duce merely what the object looks like from the point of view at which he sees it. And he finds this a hard task. The visual presentation is apt to be apprehended by him as having qualifications which do not belong to its own independent constitution, but are merely due to the conditions of his own psychical processes in relation to it. His established habit of attending only to physical i ERROR 19 magnitude and configuration leads him to think of physical fact even in attempting to think only of the sensory presentation. Thus a child in drawing the profile of a face will put in two eyes. But the physical fact so far as it is unseen does not belong to the reality of the visual presentation. It is therefore mere appearance relatively to this reality, and in so far as it is confused with this reality, it is not only mere appearance but error. V. Special Conditions of Error § 5. Having defined what we mean by mere appear- ance we have now only one more step to take in order to account for error. We have to show how the mere appear- ance of anything comes to be confused with its reality. It is clear from the previous discussion that there can be neither truth nor falsehood except in so far as the mind is dealing with an object which has a constitution predetermined independently of the psychical process by which it is cognised. Such logical puzzles as the Litigiosus and Crocodilus involve an attempt to affirm or deny something which is not really predetermined independently of the affirma- tion or denial of it. In the Litigiosus the judgment to be formed is supposed to be part of the reality to which thought must adjust itself in forming it. Euathlus was a pupil of Protagoras in Rhetoric. He paid half the fee demanded by his teacher before receiving lessons and agreed to pay the remainder after his first lawsuit if he won it. His first lawsuit was one in which Protagoras sued him for the money. The jury found themselves in what appeared a hopeless perplexity. It seemed as if they could not affirm either side to be in the right without putting that side in the wrong. The difficulty arose from the attempt to conform their decision to a determination of the real which had no existence independently of the decision itself. Apart from the judgment which they were endeavouring to form, the reality was indeterminate and it could not therefore determine their thought in the 20 G. F. STOUT i process of judging. The Crocodilus illustrates the same principle in a different way. A crocodile had seized a child, but promised the mother that if she told him truly whether or not he was going to give it back, he would restore it. There would be no difficulty here if the mother's guess were supposed to refer to an intention which the crocodile had already formed. But he is assumed to hold himself free to regulate his conduct according to what she may happen to say, and so to falsify her statement at will. There is therefore no predetermined reality to which her thought can conform or fail to conform ; which alternative is real, is not pre- determined independently of her own affirmation of one of them. Hence an essential condition of either true or false judgment is wanting. One consequence of the general principle is that a proposition cannot contain any statement concerning its own truth or falsity. Before the proposition is made in one sense or another its own truth or falsity is not a predetermined fact to which thought can adjust itself. Thus if a man says, " The statement I am now making is false," he is not making a statement at all. On the other hand, he would be speaking significantly and truly if he said " The statement I am now making contains nine words." For he can count each word after determining to use it. His pre- cedent determination to use the word is an independent fact which he does not make in the act of affirming it. For error to exist the mind must work in such a way as to defeat its own purpose. Its interest must lie in conforming its thought to the predetermined constitution of some real object. It must be endeavouring to think of this as it is independently of the psychological conditions of the thinking process itself. And yet, in the very attempt to do so, it must qualify its object by features which are merely due to such psychological conditions. I cannot pretend to give anything approaching a full analysis of the various special circumstances which give rise to this confusion of appearance and reality. But i ERROR 21 the following indication will serve to illustrate the general principle involved. Errors may be roughly classified under two heads which we may designate (i) as errors of confusion, and (2) as errors of ignorance, inadvertence, and forget- fulness. All errors involve a confusion of appearance and reality. But this confusion is the error itself, not a condition determining its occurrence. When we speak of an error of confusion, we mean an error which not only is a confusion, but has its source in a confusion. Again, all errors involve some ignorance, inadvertence, or forgetfulness. Whenever any one makes a mistake, there is something unknown or unheeded which would have saved him from error if he had known and taken account of it. But we can distinguish between cases in which ignorance or inadvertence or forgetfulness are the sole or the main source of the erroneousness of a belief, and those in which another and a positive condition plays a prominent part. This other positive condition is what I call confusion. I shall begin by explaining wherein it consists, and illustrate it by typical examples. (1) Errors of Confusion § 6. There is a confusion wherever our cognitive judg- ment is determined by something else than the precise object which we are interested in knowing. We mean to wrest a decision from just this object concerning which the question is raised ; but owing to psychological conditions, other factors intervene without our noticing their opera- tion and determine, or contribute to determine, our thought. Optical illusions supply many examples. I must content myself with one very simple illustration of this kind. / In the above figure there are two straight lines, a b and ef\ the part c d is marked off on a b, and the 22 G. F. STOUT i part g h on e f. c d is really equal to g h. But most persons on a cursory glance would judge it to be longer. The reason is that though we mean to compare only the absolute length of c d with the absolute length of g h, yet without our knowing it, other factors help to determine the result. These are the relative length of c d as compared with a b, and the relative length of g h as compared with e f. This example is typical. In all such instances we mean our judgment to depend on comparison of two magnitudes as presented to the eye. But these magnitudes are presented in more or less intimate union with other items so as to form with these a group which the attention naturally apprehends as a whole. Hence there is a difficulty in mentally isolating the magnitudes themselves from the contexts in which they occur so as to compare these magnitudes only. We seek to be determined by the nature of the object which we are interested in knowing, but we escape our own notice in being determined by something else. This is confusion. Another most prolific source of confusion is found in pre- formed association. All associations are in them- selves facts of the individual mind and not attributes of anything else. If the idea of smoke always calls up in my mind the idea of fire as its source, this is something which is true of me, and not of the fire or the smoke as independent realities. It might seem from this that whenever our judgment of truth and falsehood is deter- mined by association, we commit a confusion. But this is not so ; for it is the function of association to record the results of past experience ; and when the results recorded are strictly relevant to the object we are interested in knowing, and to the special question at issue, there is no confusion. The association between 12x12 and equality to 144 registers the result of previous multiplication of 12 by 1 2. There is therefore no confusion in allowing it to determine our cognitive judgment. But the associative mechanism may become deranged so that 12x12 calls i ERROR 23 up 154 instead of 144. In that case to rely on it as a record involves an error of confusion. It often happens that certain connections of ideas are insistently and persistently obtruded on consciousness owing to associations which have not been formed through experiences relevant to the question at issue. So long and so far as their irrelevance is unknown or unheeded, the irrelevant association determines the course of our thought in the same way as the relevant. Take by way of illustration an argument recently used by an earth-flattener. The earth must be flat ; otherwise the water in the Suez Canal would flow out at both ends. The associations operative in this case, are those due to experience of spherical bodies situated on the earth's surface. Whenever the earth-flattener thinks of the earth as a globe, inveterate custom drives him to think of it as he has been used to think of all the other globes, of which he has had experience. But the question at issue relates to the earth as distinguished from bodies on its surface. Hence a fallacy of confusion. One effect of repeated advertisements such as those of Beecham's pills, covering several columns of a newspaper, is to produce this kind of illusion. Self-praise is no recommendation. But self-praise skilfully and obtrusively reiterated may suffice to produce an association of ideas which influences belief. 1 Errors due to ambiguity of words come under this head. A word is associated with diverse though allied meanings, and, as we go on using it in what aims at being continuous thought, one meaning insensibly substitutes itself for another. Being unaware of the shifting of our object from A to A f we go on assuming that what we have found to be true of A is true of A'. We begin for instance by talking of opponents of government, meaning 1 Many persons have a prejudice against advertisements. I share this prejudice myself. And yet the obtrusive vividness and persistent reiteration of some of them does now and then produce in me a momentary tendency to believe which might easily become an actual belief if I were not on my guard. Allitera- tive and rhetorical contrast often help to stamp in the association. " Pink pills for pale people " is a good instance. Of course the whole effect of advertisement cannot be explained in this way. 24 G. F. STOUT i advocates of anarchy, and we proceed to apply what we have said of these to opponents of some existing government. " Bias " is a source of confusion distinct from irrelevant association, though the two frequently co-operate to produce error. Bias exists so far as there is a tendency to accept one answer to a question rather than another because this answer obtrudes itself on consciousness through its connection with the emotions, sentiments, desires, etc. of the subject or in one word, because it is specially interesting. The interest is most frequently agreeable. But it may also be disagreeable. In return- ing home after the discovery of the famous footprint, Robinson Crusoe's terror caused him to mistake every bush and tree, and to fancy every stump at a distance to be a man. To say that a man's mind is intensely occupied in escaping or guarding against danger, is equivalent to saying that he is intensely interested in finding out what the danger is and where it lies. Hence he will be on the alert for signs and indications of peril. He will therefore attend to features of his environment which would otherwise have passed unnoted, and he will neglect others which he would otherwise have attended to. Thus fear may influence belief by determin- ing what data are, or are not, taken into account. By excluding relevant data it may give rise to error of inadvertence. But besides this the data which fear selects are also emphasised by it. They obtrude them- selves with an insistent vivacity proportioned to the intensity of the emotion. This insistent vivacity directly contributes to determine belief and becomes a source of error of confusion. In view of current statements this last point needs to be argued. The prevailing view appears to be that errors due to bias are merely errors of inadvertence. Dr. Ward, for example, strongly takes up this position. " Emotion and desire," he remarks, " are frequent indirect causes of subjective certainty, in so far as they determine the constituents of consciousness at the moment — pack the i ERROR 25 jury or suborn the witnesses as it were. But the ground of certainty is in all cases some quality or some relation of these presentations inter se. In a sense, therefore, the ground of all certainty is objective — in the sense, that is, of being something at least directly and immediately determined for the subject and not by him." l What Ward's argument really proves is that subjective bias cannot be recognised by the subject himself as a ground or reason for believing. It does not follow that it may not directly influence belief through confusion. In cases of confusion we seek control proceeding from the nature of our object, and we find our thought determined by something else which we fail to distinguish from the objective control we are in search of. Now there seems to be no reason why subjective interest should not, in this way, mask itself as objective control. Connection with emotion and desire may give to certain ideas a persistent obtrusiveness which is not always adequately traced to its source. But this persistent obtrusiveness, when and so far as it is not traced to its source in emotion and desire, must appear as if it arose from the nature of the object. It will thus appear to the subject as something which determines him and is not determined by him. This confusion may assume three forms. In the first place there are instances in which it is very difficult to dis- cover any other cause of belief except subjective bias. The person who holds the belief cannot assign any reason for it except that he feels it to be true. Sometimes, no doubt, there may be in such cases an objective ground which the believer finds it impossible to express or indicate to others. But there are instances in which the sole or the main factor seems to be subjective bias. What is believed obtrudes itself upon consciousness vividly and persistently because of its peculiar kind and degree of interest so that it is difficult to frame the idea of alternative possibilities save in a comparatively faint, imperfect, and intermittent way. The second class of cases is less problematical. I refer 1 Article on "Psychology," in Ency. Brit. p. 83. 26 G. F. STOUT i to instances in which there are relevant reasons for belief but reasons which are inadequate to account for the actual degree of assurance, apart from the co-operation of bias. A regards B with hatred and jealousy so that the mere imagination of B's disgrace or ruin has a fascination for him. Something occurs which would produce in an impartial person a suspicion that B had been behaving in a disgraceful way. A at once believes the worst with unwavering decision and tenacity. It may be that the impartial person, who only entertains a suspicion, has just as restricted a view of the evidence as A. The restriction may be due to ignorance or indifference in his case, and mental preoccupation in A's. But for both the relevant evidence may be virtually the same. The difference is that in A's mind it is reinforced and sustained by subjective bias which he does not sufficiently allow for. In a third class of instances irrelevant association co- operates with subjective bias. This is perhaps the most fertile source of superstitions and of those savage beliefs of which superstitions are survivals. Take for example the tendency which some uneducated persons and even some who are educated find irresistible, to think of their bodies as still sentient after death. Sit tibi terra levis is more than a metaphor. It points back to the belief that the weight of the superincumbent earth actually distresses the corpse. It is a Mahometan superstition that the believing dead suffer when the unhallowed foot of a Christian treads on their graves. In the old Norse legends to lay hands on the treasure hidden in the tomb of a chief is to run a serious risk of rousing its owner from his long sleep to defend his possessions. Perhaps there are few people who look forward to their own funeral without figuring themselves to be present at it not only in body but in mind. This whole point of view is in part due to a firmly established association arising from the intimate connection of mind and body during life. But besides this we must also take into account the gruesome fascination of such ideas. Their vivid and absorbing interest makes it difficult to sret rid of them, and this i ERROR 27 persistent obtrusiveness in so far as it is not traced to its source in psychological conditions contributes to determine belief in their reality. (2) Errors of Ignorance and Inadvertence We turn now from the error of confusion to the error of mere ignorance, which must be taken to include all forgetfulness or inadvertence. As I have before pointed out, all error involves some ignorance or inadvertence ; but in the case of confusion there is also some other positive ground of the erroneousness of the belief. An irrelevant condition operates as if it were relevant. It would not do so, if we were fully and persistently aware of its presence and influence, and to this extent the error of confusion is one of ignorance or inadvertence ; but the ignorance and inadvertence is not the sole cause of error. There is also the undetected influence of the irrelevant factor determining the course of thought. In the error of mere ignorance or inadvertence, on the other hand, the sole ground of the erroneousness of the belief lies in the insufficiency of the data, at the time when it is formed. But here we must guard against a misapprehension. The error is not identical with the ignorance or inadvertence. It is a belief having a positive content of its own. Nor is it correct to say even that the determining cause of this belief lies in the ignorance or inadvertence. Mere negation or privation cannot be the sole ground of any positive result. What directly determines belief is the data which are presented, not anything which is un- presented, and we must add to these as another positive condition the urgency of the interest which demands a decision and will not permit of a suspense of judgment. It is these factors which are operative in producing the belief. Ignorance and inadvertence account only for its erroneousness. In all cognitive process we seek to be determined by the nature of our object. But if the object is only partially known, what is unknown may be relevant so that if it had been known 28 G. F. STOUT i and heeded another decision would have been imposed on us. As an example of error due to mere ignorance, I may refer to a personal experience of my own. Some time ago I set out to visit a friend who, as I assumed, was living in Furnival's Inn. I found on arrival that the whole building had been pulled down. My error in this case was not due to any confusion. The evidence on which I was relying was all relevant and such as I still continue to trust on similar occasions. I went wrong simply because certain events had been occurring since my previous visit to Furnival's Inn without my knowing of them. Inadvertence is not sharply divided from mere ignorance. It includes all failure to bring to conscious- ness knowledge, already acquired and capable of recall, at the time when it is required for determining our decision. It may also be taken to include other failures to take into account knowledge which would have been immediately and easily accessible if we had turned our attention in the right direction. Mill gives many examples under the head " Fallacies of Non-observation." From him I quote the following : — " John Wesley, while he commemorates the triumph of sulphur and supplication over his bodily infirmities, forgets to appreciate the resuscitating influence of four months' repose from his apostolic labours." Wesley knew that he had taken rest and also that rest has commonly a recuperative effect in such cases. His failure lay in omitting to take these facts into account owing to subjective bias, as an amateur physician with crochets and as a religious enthusiast. So far as error is traceable to ignorance or inadvertence, it is perhaps abstractedly possible to conceive that it might have been avoided by an absolute suspense of judgment. I might have refused to count on the con- tinued existence of Furnival's Inn, or even on the chance of it, on the ground that I did not know all that had happened in relation to it, since I saw it last. But such suspense of judgment cannot be uncompromisingly main- i ERROR 29 tained as a general attitude throughout our whole mental life. It would be equivalent to a refusal to live at all. Any one who carried out the principle consistently would not say " this is a chair " when he saw one. He would rather say, " This is what, if my memory serves me right, I am accustomed to regard as the visual appearance of a chair." In thus cutting off the chance of error we should at the same time cut off the chance of truth. In order to advance either in theory or practice, we must presume — bet on our partial knowledge. We must take the risk due to an unexplored remainder of conditions which may be relevant to the issue we have to decide on. But there is another alternative. A mental attitude is possible intermediate between absolute suspense of judgment and undoubting acceptance of a proposition as true. We may judge that the balance of evidence is in favour of the proposition. Instead of unreservedly expecting to find Furnival's Inn, I might have said to myself that it was a hundred to one I should find it. So far as this proposition has a practical significance as a guide to action it can only mean that I should be right in relying on similar evidence in 99 cases out of 100. But such an attitude does not really evade the possibility of error arising from ignorance and inadvertence. For (1) we are liable to go wrong even in the estimate of probabilities. There are, for example, vulgar errors of this kind which mathematical theory corrects. (2) In determining the probability of this or that proposition, we proceed on the basis of a preformed body of beliefs which are themselves liable to be erroneous. In particular, we are apt to assume undoubtingly that our view of competing alterna- tive is virtually exhaustive, when it is really not so. But we cannot be always sifting these latent presuppositions to the bottom. If we constantly endeavoured to do so in a thorough-going way, it would be impossible to meet the emergencies of practical life or even to make effective progress in knowledge. It is a psychological impossibility to assume and maintain a dubitative attitude at every point where ignorance or inadvertence are capable of 3