The first principles of knowledge

Preface

A FEW words will be enough to put exactly before the reader the object at which the present volume aims. A well-known criticism on the Aristotelian Logic is the complaint, that it provides for the consistency of thought with thought, but not for the consistency of thought with things ; that it secures right processes upon given or assumed materials, but does not guarantee the materials upon which the processes are conducted. To supply the want thus indicated, several modern logicians have curtailed or omitted portions of the old Logic, and added new chapters, of which the following headings may serve as specimens, taken from Mr. Bain's work : " Uniformity and Laws of Nature," " Elimination of Cause and Eifect," " Experimental Methods," "Frustration of the Methods," " Chance and its Eliminations," " Secondary Laws, Empirical and Derivative," " Explanation of Nature," " Hypotheses," " Classification," " Logic

•/i PREFACE.

of Mathematics," ** Logic of Physics," " Logic of Chemistry," " Logic of Biology," " Logic of Rhetoric," " Logic of Pohtics," ** Logic of Medicine.'* These titles show the kind of addition that now-adays is asked, beyond the simple bill of fare found in the Aldrich who satisfied the students of a past generation, and to many even afforded more than they wanted.

It is unfortunate that those who in this country were, perhaps, the loudest in their clamours that logic should take account of the reality which hitherto it had seemed to neglect, should have embraced a system of philosophy which is fatal to firm belief in any reality beyond thought itself. Messrs. Mill and Bain assuredly have not directly tended to take men out of idealism, and make them realists. Yet the former was explicit enough in his demands :' " I conceive it to be true, that Logic is not the theory of thought as thought, but of valid thought: not of thinking, but of correct thinking. . . . In no case can the thinking be valid unless the concepts, judgments, and conclusions resulting from it are conformable to fact. And in no case can we satisfy ourselves that they are so by looking merely at the relations of one part of the train of thought with another. We must ascend to the original

* Examination, c. xx. pp. 397, seq. (2nd Edit.)

PREFACE. vll

sources, the presentations of experience, and examine the train of thought in relation to these."

Little as the modern representatives of the Schoolmen are satisfied, either with the spirit of Mr. Mill's demand, or with the mode of his own response to it, they have deemed it well worth while, not indeed to change the old Logic, but tO' add to it a new book. Pure Logic remains substantially what it was, and is justified in its position^ It assumes, as all other sciences do and must, that human thought has, in general, objective reality '^ and on this most legitimate assumption it proceeds to lay down the laws of orderly, consistent thinking. The newly added part of Logic, often called Material, Applied or Critical, takes for its special purpose to defend the objective reality of thought. It is thus an assertion of a form of realism, as against idealism, and is called in this book the Philosophy of Certitude. For the whole question comes to this : what reasonable account can be given of man's claim to have real certainty about things ? What are the ultimate grounds for holding, that man may regard his knowledge about objects as undoubtedly correct ? Scientifically to draw out the account here demanded is a work appositely described by the title, The First Principles of Knowledge, viii PREFACE.

An endeavour has been made throughout these pages, while stating the sound, traditional principles of certitude, to bring them into constant contact with the antagonist principles, more particularly with the principles of Hume and the pure empirics. It is not true that the only possible philosophy is a history of the opinions which, at various times, have prevailed ; but it is true, that the modern spirit will not be satisfied without a statement of how controversies stand on questions which are notoriously disputed. The truth as made manifest in conflict, is what has to be exhibited : and this necessity, whether exactly desirable or not, must stand as explanation or apology to those, whose own special tastes might prompt them to desire a simple exposition of scholastic doctrine apart from the encumbrance of adverse systems. Scholasticism must now be militant, and that, not only with a view to outsiders, but with a riew to retaining its own clients, who cannot fail to come across much in modern literature, for the understanding and the consequent rejection of which some direct preparation is needful.

Readers not already familiar with the questions here discussed, would do well at first to leave alone the notes which are printed in smaller type, and concentrate attention on the positive doctrine, the

PREFACE. \m,

importance of which must be judged, not by the length of its statement, but by the weight of the words. The matter is eminently one which is best conveyed in a few precise sentences, the full import of which must be mastered by leisurely consideration.

Preface To The Second Edition

In spite of paragraphs which expressly assert the contrary, some readers have declared their suspicion, or more than suspicion, that the writer, giving up the cause of a purely natural certitude, has fallen back upon revelation to fill up gaps in his argument which otherwise he could not bridge over. A careful inspection will show, that the power of unaided reason to attain the truth, even about God and man in relation to God, is openly maintained.

But besides natural knowledge, it is as certain as the certainty of Christianity that a supernatural knowledge is within human reach ; and of this, though it does not belong to logic as such, some mention is necessary in a logical course, in order to give completeness to the treatment. The Pelagian controversy fully brought out the fact

PREFACE.

to which we refer. Pelagius himself was not very clear as to his own doctrines, but his statements in general committed him to the assertion of such an efficacy in unassisted nature towards the accomplishing of its own salvation, as was' in direct conflict with Holy Scripture. While in the East the heresiarchs had busied their subtle minds over the highly mysterious natures of the Trinity and of the Incarnation, he, with a truly British instinct, asked himself the homely question : How far in the saving of our souls are we able to manage the business by ourselves alone ? The answer given by the Church most unmistakeably was, that possible as is the act of natural faith, yet the saving faith requisite for the present order of Providence must be supernatural, resting on grace which has for its double office to enlighten the understanding and to urge on the will. Here, therefore, is the reason why in the course of our logical treatise reference is occasionally made to revelation, without any derogation from the natural powers of the human intellect when working within its own sphere.

The second matter on which the attitude taken up by this volume needs a word of explanation, is in regard to idealism. No doubt it is against idealism that the polemical part of this treatise is largely directed. But the error takes so many

PREFACE. "

shapes, or rather in the hands of most authors it is left so very shapeless, that in a small volume to single out individual systems is impossible. Therefore the refutation has been attempted in very general terms, leaving open to each idealist the subterfuge that his particular views have not been combated. At any rate against all adversaries alike we may urge the most complete neglect on their part — a neglect as complete as that of some destructive critics of the Bible — to follow the good example of the schoolmen in stating and answering the main objections to their own theories. One consequence of such omission is that we find idealists, among themselves, complaining on this very head ; as, for instance, when Hegel lays it to the charge of Schelling that the " intellectual intuition of the absolute," from which the latter starts his philosophy, is quite unwarranted in reason ; after which salutary admonition to another, Hegel himself is not deterred from grounding the whole of his own system on a process equally unwarranted. Kant, who declared himself to have been "awakened out of his dogmatic slumber " by Hume, had been roused only to fall into another slumber not less dogmatic; so that Hegel and Schelling were but copying their master in their neglect to consider objections. Exactly this contempt for difficulties and xii PREFACE.

this groundlessness of assumption are what stagger us in the works of ideahsts. They seem to rely on certain catch-phrases for carrying them over the most mountainous obstacles, telling us that the object as out of knowledge cannot be known, and that therefore there is no knowledge of things as existing independently of the mind ; that the difference between subject and object is fully satisfied by supposing a double aspect of one reality, which in itself is sometimes said to be above and beyond all such difference, being the identity of the two; that the object can never be placed, as dualists pretend to place it, on a line of equality with the subject which posits it by distinguishing itself from its antithesis; that new knowledge can enter man's mind only because the universal mind communicates itself to the finite and realizes itself in an organism ; that the mind can know only what itself has caused or constructed, and that hence the constructive imagination is the highest human faculty ; that the development of thought is but the reverse side of the development of things, so that thought and thing are identical, and each system of thought is true for its own stage of evolution, error being nothing but relatively imperfect knowledge. One and all of these principles are open to the gravest objections, yet idealists who hold a greater or a

PREFACE. in

less number of them can rarely be brought seriously to meet objectors on these grounds ; they are satisfied with a mere repetition of their stock phrase. Our refutation of their position is given simply from an analysis of what knowledge reports about its own nature, which certainly is that realism is the right interpretation ; and a realism, be it observed, which is not confined simply to the assertion of an external world of matter, but includes also realities of a higher order.

Chapter I. — Definition of Truth . . . . i

M II. — In what act of the Mind a Truth may be

FOUND completely POSSESSED . . I4

„ HI. — Definition of Certitude and of the states

OF Mind falling short of Certitude . 42 „ IV. — Kinds and Degrees of Certitude . , 50

„ V. — Metaphysical and Physical Certitudb . 68 „ VI — The order of precedence between Natural

and Philosophic Certitude . . 108

„ VII. — The charge of Discord (or at least of

WANT OF co-operation) BETWEEN NATURAL

AND Philosophic Certitude . .119

„ VIII. — Universal Scepticism .... 134 „ IX. — Cartesian Doubt .... 148

„ X. — The Primary Facts and Principles of the

Logician ..... 164

„ XI. — Retrospect and Prospect . . .183

„ XII. — The rejection of Various Theories about

the Ultimate Criterion of Certitude . 188 „ XIII. — Evidence as the ultimate objective Criterion OF Truth .... 216 M XIV. — ^The Origin of Error in the Understanding 232

CONTENTS

Part Ii.— Special Treatment Of Certitude. Page

Chaptbr I. — Short Introduction 249

„ II. — The Trustworthiness of the Senses . 258

„ III. — Objectivity of Ideas, whether singular or

UNIVERSAL . 301

„ IV. — Exaggerated Realism, Nominalism, and

CONCEPTUALISM ..... 332

„ V. — Consciousness . . . • . 340

., VII. — Belief on Human Testimony . . , 377

„VIII. — Belief on Divine Thstimoht . • . 389

THE FIRST PRINCIPLES OF KNOWLEDGE

Part I .

The Nature of Certitude in General.

CHAPTER I. definition of truth.

Synopsis.

1. Three kinds of truth.

2. Definition of truth as found in knowledge, (a) The com-

mon-sense view, agreeing with the scholastic definition. {b) Assigned reasons for modifying and even radically altering that definition, (c) Assertion of the old definition, a course which the rest of this work must defend, but for which a little may be said at the outset.

3. Definition of error, as a corollary to the definition of truth. Addenda.

I. Truth is commonly divided into truth of things, truth of thought about things, and truth in the outward expression of our thought about things. The first kind of truth is called ontological, the third moral, and each of these is discussed in separate volumes of the present series. It is with the second member of this division, about what is often styled logical truth, that the treatise which we are here beginning is concerned. What true knowledge is, and how its possession by the human B

NATURE OF CERTITUDE.

intellect can be vindicated, these are the questions specially calling for our investigation.

2. (a) An ordinary man, if asked to explain what true knowledge is, would reply simply, that knowledge was true when the thing was really such as we thought it to be. He would thus agree with the definition of the schoolmen, *' Truth is an equation or a conformity of thought to thing."'

(6) But a matter which, at first sight, seems thus readily settled, presents, on reflexion, a number of difficulties, which some have regarded as so serious as to upset the plain man's view, backed though it be, by other philosophies, and by the massive volumes of scholastic philosophy which the centuries have piled one on the top of another. For when the case is more narrowly sifted, are we not driven to make some awkward inquiries : How can mental images be like outer objects, especially material objects ? How are sensations like the external bodies which stimulate them ? When several senses give their reports of one object, as there is no likeness between the several reports, say between the taste of an orange and its colour, how can they be all like the object ? What is the conclusion from the notorious fact that in different persons, and even in the same person at different times, any one of the five senses may bear divergent testimonies ? Besides, even if our knowledge were like what we call its object, to observe this correspondence must be an act of comparison, which not we, but some one else, must make : for we at least

1 Tongiorgi, Institutiones Philosophica, Vol. I. nn. 370. seq.

DEFINITION OF TRUTH.

cannot compare the thing as known to us, with the thing as out of our knowledge. Such an attempt on our part would be preposterous, a fraudulent endeavour to assert, for our essentially relative faculties, an absolute validity.

Moved by these considerations, a number of modern philosophers dare to claim for human knowledge only some correspondence with its object which is less than that of likeness, and is describable as symbolic: just as a mathematical formula, though not like, may yet symbolize, the path of a cannon-ball. Thus, in the words of Mr. Frederick Harrison, " our scientific conceptions have a very good working correspondence with the assumed reality without ; but we have no means of knowing whether the absolute correspondence between them be great or small, or whether there be any absolute correspondence at all." Mr. H. Spencer,* in expounding his theory of " transfigured realism," sets forth still more clearly the position of the " relativist," that is, of the philosopher who denies that we can know things absolutely as they are. Maintaining that " resistance, as disclosed by opposition to our energies, is the only species of external activity which we are obliged to think of as subjectively and objectively the same," still even here he will not positively affirm that knowledge is like the object. And for ordinary objects his teaching is this : " If x and y are the two uniformly connected properties in some

2 Mr. Spencer's doctrine may be seen in his Psychology, Part I. c. xix. ; First Principles, Part I. c. ii.

NATURE OF CERTITUDE.

outer object, while a and b are the effects which they produce in our consciousness ; the sole need is that a and b, and the relation between them, shall always answer to x and y, and the relation between them. It matters not if a and b are both like x and y, or not ; could they be exactly identical with them we should not be one whit the better, and their total dissimilarity is no disadvantage." ^ In other words, if for every definite change in objects, there is one constant change in the mind which is affected by that object, then this is enough, without any resemblance of thought to thing; concomitant variation suffices.

(c) Against this theory the time-honoured definition of knowledge must be re-affirmed. The objections raised against it are only the old arguments in favour of complete distrust in the power of man to attain truth ; and to refute them will be the main purpose of all that follows in this volume. At the outset, this book defines truth of intellect to be "the conformity of thought to thing " : subsequently its one grand aim will be gradually to make good the definition. Whilst patiently awaiting the development of a long line of argument, the reader may find some consolation in a few declarations that can be offered him at once. First of all, when knowledge is said to be a sort of " equation " of mind to thing, it is not meant that knowledge, in order to be true,, must exhaust the whole object : a partial knowledge is true, as far as it goes, especially when it is recognized as only partial.

3 First Principles. Part I. c. iv. | 25.

DEFINITION OF TRUTH.

Next, the likeness which is asserted is quite sui generis. An idea, — to use the word at present in its broad sense of any act of knowledge, — is not a dead picture, but something effected by and in the living, cognitive mind ; it is a thing with a conscious meaning of its own:"* it is, as Spinoza says, self-assertive or self-referent ; it is what the schoolmen sometimes call a sigmim quo, a sign which taken, not in its isolation as a mere phenomenon, but as it exists in the mind, is the knowledge of the thing signified. Thus it differs from a sigmim ex quo, a sign which has first to be known, that from it the mind may travel to the thing signified. To quote Father Liberatore : " The signum a quo is that which, by being first known, leads on the mind to the knowledge of the thing it signifies. Another way of signifying is presented to us in those inward signs which do not come before the mind as objects of its perceptions, but which, by informing the cognitive faculty, effect actual knowledge. These latter may be called signa in quibus [or, signa qicibus] , or also, * formal signs,' in that they do not represent objects as previously known, but are forms determining the mind to perceive the object. To this category belong mental concepts." 5 Hence the representative, significative, or meaning power of an idea is not photographic, nor anything analogous to photography ; and to fancy it so is the cardinal

4 Hence Hume greatly errs : " The reference of an idea to an object is an extraneous denomination, of which in itself it bears no mark or character." (Treatise, Part I. sec. vii. pp. 327, 330.)

5 " Signum ex quo quod prius in se cognoscatur et ex sui cognitione potentiam ducat in cognitionem rei significatae. Alter modus significandi locum habet in signis internis, quae non se offerunt ut objecta.

NATURE OF CERTITUDE.

error of scepticism. The idea depicts its object in no way open to our artists : neither by similarity of substance, nor of colour, nor of outline, nor by any mode of material portraiture. The process is so peculiarly mental or spiritual, that illustrations borrowed from matter are more calculated to mislead than to direct. The uniqueness of the phenomenon is essentially its strangeness, for we cannot explain it by reduction to any familiar class. Yet the strangeness is welcome as serving, in another treatise, to show the inadequacy of the materialistic hypothesis.

A difficulty, raised by Cousin, really amounts to no more than a matter of words. He says^ that an idea cannot be " an image " of an object, because only material representations can be " images ; " that we cannot strictly speak of " likeness " between spiritual objects, but only between material : and that therefore, if we do call ideas " images " bearing the "likeness" of their objects, we are talking not properly but metaphorically. As our knowledge begins in sense, so far all our spiritual ideas may be said to be conveyed in metaphorical terms ; it seems however a fair procedure to regard a term as no longer metaphorical, when we no longer advert to the figurative meaning, but pass straight to the main object. Thus, in speaking of moral rectitude

sed informando potentiam earn efficiunt cognoscentem in actu. Hsec dici possunt signa in quihus [vel signa quibus] , vel etiam signa formalia : quia non repraesentant tanqaam objecta prius cognita sed tanquam formae determinantes potentiam ad perceptionem objecti. Hujusmodi est conceptus mentis." {Logica, Pars I. c. i. n. 5.) 6 Histoire dt la Philosophie, leqon. 21 me, et alibi passim.

DEFINITION OF TRUTH.

we hardly refer to the image of a man keeping the straight path as he walks ; but we go at once to the notion of right conduct. Similarly, whatever may have been the origin of the terms, we can now claim to apply the word "image" or "likeness" straight to spiritual resemblances. However, if any one should insist on seeing a trace of metaphor left here, the point is not worth controverting.

We assume, therefore, that ideas are, in the language of Aristotle, ofioKo/jbaTa tmv TrpayfjuaTcov, — "hkenesses of things," and so stand contrasted with words which are conventional signs : whereby a special meaning is given to the saying, that man is a microcosm.7 Not only does man sum up the several constituents of our Cosmos by uniting together mineral, vegetable, and animal nature; but by knowing all things, he, in a manner, reproduces, or becomes all things. Homo est quod est, says the materialist scoffingly; translating the words, " Man is what he eats, what his food makes him." False as a complete statement, this is true as a partial statement ; and equally true is it, homo est quod scit — " Man is what he knows." In this sense St. Thomas writes anima est quodammodo omnia — " The soul is in some sense everything ; " which is the repetition of Aristotle,^ 97 '^^XV '^^ ovra ttco? eVrt iravTa — " The soul is in a manner all things."

3. From the definition of true knowledge before

given, a corollary as to the nature of error may be

gathered. Not any absence of likeness between

thought and thing is straightway falsehood : rather

7 Silvester Maurus, Quastiones PhilosophictB, qusestio vi. 8 De Anima, Lib. III. c. viii. i.

NATURE OF CERTITUDE.

such mere absence is ignorance. Before downright error is reached, there must be not only want of conformity but positive deformity. For knowledge, however limited, is true knowledge so long as it does not transgress or deny its own limits — a fact highly important to finite intelligences like ourselves who can but " know in part." Be it our consolation, then, to remember that the opposition between knowledge and ignorance is only what the logicians call a " contradictory ; " while it is not till we have gone as far as the " contrary " opposition that we commit error.

Addenda. (i) The mysteriousness of the act of knowledge, and its apparent impossibility on any material analogy, were points that arrested the attention of the early speculators. Whereas St. Thomas' argued that because the mind was capable of becoming cognizant of bodies it could not itself be corporeal, some of the old Greeks had pushed to its extremes the principle, Like is known by like — ofioui 6fioiot,<i yijvaxTKeTai,. Empedocles, for example, had said, " We perceive earth by means of earth, water by means of water, air by means of air, fire by means of fire, love by means of love, and strife by means of strife," where love and strife stand for what we call attractive and repulsive forces. Others spoke either of the eye sending out its influences to the object, or of the object emitting its ecScciXa, or minute images to the eye, which,

Like little films from outer surface torn, In mid air, to and fro, are lightly borne.* « la, q. 75. art. 2.

{Lucretius, iv. 31, 32.

DEFINITION OF TRUTH.

All such conceptions are the follies of a crude materialism ; and a long way the better course is, while admitting that how knowledge is possible is inscrutable to us, yet to insist that the fact is manifest to experience. The reaction of our faculties to their appropriate stimuli must simply be accepted for what it declares itself to be, namely, not any kind of a reaction, but the special reaction which must be called cognitive.

(2) Preferring theory to fact, and a theory the very arguments for which rest on an assumption which is just the contrary to what they are fancied to prove, a number of modern writers quite set aside the doctrine that we have knowledge like to the realities outside of •ourselves. From America comes the voice of Mr. Borden Browne, telling us, as an introduction to his volume on Metaphysics, that because we cannot compare thought with being, therefore, " truth cannot be viewed as the correspondence of thought and thing, but as the universally valid in our thought of the thing. That is the true conception of reality, which grasps the common to all, and not the special to one ; " so that the test of truth is " the necessity of the conception and the inner harmony between several conceptions. It is not the lack of harmony between our conceptions and reality which disturbs, but the discord of our conceptions among themselves." A like utterance we have from a German author, according to whom " truth does not consist in any sort of correspondence between our thought and the things outside us, but in a character that belongs to our mode of putting together our internal experiences. Our thoughts are true, when their nature, as internal events, is understood ; when they are placed in equal relation to the rest of experience. The criterion of truth is the feeling of universality and necessity in the ultimate axioms." In our own country we have

NATURE OF CERTITUDE.

some authors completely rejecting the doctrine that truth is conformity of mind to thing; while others, using the same words, are less thorough in their divergence from us, though sufficiently divergent to be in decided opposition. They distinguish between perception and thought, so as to make thought more especially a matter of subjective forms without ascertainable objective validity. " Truth relatively to the human mind," writes Mansel,3 " cannot be defined as a conformity with its object ; for to us the object exists only as it is known by one or other faculty. Hence material truth consists rather in the conformity of the object as represented in thought with the object as presented in intuition ; while logical truth consists in the conformity of thought to its own laws." With these words may be compared the following from another countryman of ours, Mr. S. Hodgson: "Without thought no truth, without perception no reality. By reality I understand the actual existence of the object, its actual presence to consciousness.^ Reality is not greater after thought than before ; thought has transformed it into a new shape, has given it new relations, but has added nothing to its real existence. Truth, on the other hand, is a product of thought, the form which an object assumes after investigation, and is thus greater after thought than before it. Reality depends on the relation between objects and consciousness: truth on the relation between objects in consciousness." The difference here is partly a matter of words, but it is also a matter of fundamental doctrine ; and with regard to each of the authors cited, it is sufficient for present purposes if the reader

3 Prolegomena Logica, c. vi. in fine; also Mansel's Aldrich, Appendix M. p. 277. (3rd Ed.)

4 See Green on Thought as constitutive of the Reality of the World Introduction to Hume, § 173. The passage from Mr. Hodgson is found in his book on Time and Space, p. 352.

DEFINITION OF TRUTH.

understands them as representatives of a now widespread revolt against the scholastic definition, " All truth in cognition consists of the assimilation or conformity of the mind with the object."'

(3) The consequences of the doctrine logically carried out, that ideas are mere symbols, are very fatal to all religious belief, as well as to everything worth calling true knowledge: and though it is often protested, that a doctrine is not to be judged by its inconvenient logical consequences, but by its intrinsic truth, yet there are consequences so manifestly bad, as to afford evidence that the premisses, whence they are drawn, cannot be sound. If our knowledge of things is what adversaries say it is, then it is not genuine know ledge at all : and this some of themselves admit. Take, for instance, Lange's confession, in words gathered from a long declaration : " All our knowledge of nature is, in fact, no knowledge at all, and affords us merely the substitute for an explanation. The intelligible world is a world of poesy, and precisely upon this fact rests its worth and nobility. No thought is so calculated to reconcile poesy and science as the thought, that all our reality is only appearance."^ So, too, Mr. Spencer 7 is perpetually harping on the string, " ultimate religious ideas, and ultimate scientific ideas are mere symbols of the actual, not cognitions ; " even " the personality of which each is conscious, and of which the existence is to each a fact beyond all others the most certain, cannot be truly known at all : knowledge of it is forbidden by the very nature of thought." It is well that the reader should thus be brought