255

CHAPTER V.

THE NATURE OF SELF-KNOWledge.

WHAT is that mode of knowledge which we call 'introspection,' self-knowledge, or self-consciousness? This is the question to which our discussion of 'Presentationism' led and it is indeed a question of primary and fundamental importance. The view which we take of the nature of this process, and of its relation to objective knowledge, determines issues of far-reaching significance. By 'objective knowledge' as I have fully explained elsewhere—is meant our knowledge of the world and universe around us; it is what Kant would have called 'cosmological' knowledge: our 'selfknowledge' is our knowledge of the functions and states which make up our conscious life,-in which of course objective knowledge is included.

1 As ordinarily used, 'introspection' has a specially psychological, and 'self-consciousness' a specially philosophical, significance. When selfconsciousness signifies the individual's knowledge of himself as an object, it is identical with what seems to be meant by introspection.' 'Selfknowledge,' as we shall see, is preferable to either of these terms. I have sometimes (after Dr Ward) used the phrase 'Subjective knowledge,' but this use of it is open to criticism; it has many misleading associations from which 'self-knowledge' is entirely free.

According to the view here set forth, self-knowledge is in no respect whatever analogous to a direct inspection of inner states'; it is a result of the same self-transcending tendency of thought (or intelligence) which makes possible the objective reference lying at the basis of our knowledge of the so-called 'external world. The whole of consciousness, the whole mind, is not exhausted in this objective knowledge; and therefore intelligence is not only aware (with more or less completeness and truth) of its own structure and laws, and its contents in the form of objective knowledge; but in virtue of its self-transcending tendency, it is aware (more or less completely) of the relations of mutual dependence which hold between the various modes of objective knowledge and the whole mind. "Self-consciousness," then, has degrees of truth, and has a real content which is more than the structure and content of objective knowledge. Some of those who develop Kantian principles in the light of Hegel tend to obliterate this distinction. In the second edition of his first Critique Kant had arrived at the conclusion that self-consciousness derives all its content from the structure of intelligence acting in objective knowledge. This is not all the truth even when we bring to light the most fundamental presuppositions and ultimate ideals of objective knowledge (e.g., the world as a systematic unity within which general laws obtain): self-consciousness is more than the consciousness of objective knowledge, actual or ideal. In self-knowledge we know ourselves not merely as systematic unities of related parts, but as centralised or focal unities,—each having the unity of a single life. Self-knowledge is therefore the primary and fundamental development; objective knowledge, though its

practical importance is indefinitely greater, is really subordinate; for only by reference to the former can the degree of truth of the categories of the latter be judged.1

§ 1. Whatever difficulty is found in accepting the view above expressed will arise probably from a misunderstanding of what I have here called the 'selftranscending tendency' of thought, and (elsewhere) its objective' or 'trans-subjective' reference. There is a curious philosophical prejudice or superstition-the effects of which can be traced far and wide through the whole history of philosophy-that in order to know a thing it is necessary to be that thing. It may, indeed, be argued that, for an intelligent being to know something of the world around him, the world and himself must both be rooted in one Being, and this Being must be rational, must be akin to his own. intelligence; I am far from rejecting this doctrine, although I reject any attempt to establish it which seems inadequate or superficial. But this doctrine is not at present in question; for, granting fully its truth, we cannot therefore say that the knowing subject simply is what it knows; we cannot identify the knowing with the known in order that it may be known. Psychologically, we cannot say that any constituent of consciousness, to be known, must itself be a cognitive fact, which is the theory involved in what has been called Presentationism.' Is it not of the essence of thought or knowledge to be representative or symbolic of something whose existence transcends it, something which is in a sense 'other' or 'more'

iii.

1 Cf. our discussion of some aspects of the Hegelian dialectic, in ch.

R

than the knowledge itself? What is symbolised or referred to or known is not to be identified with the process of reference, the knowledge itself. This view -the obvious truth of which must be apparent to every unprejudiced mind as soon as it is stated-we have elsewhere attempted to illustrate and defend at some length. Knowledge is essentially a process of reference; and all knowledge is direct in this sense,— that it refers directly to the reality known, the object referred to; and not indirectly through some substitute intervening between this and the knower. When a man reflects on his own states and when he reflects on something in the objective world-e.g., the motives of another man's conduct, or the nature of a chemical combination-his attitude to the 'object' is the same: the reference is equally direct, though the knowledge is necessarily incomplete, and may in some of its details be illusory.1

In saying that the intellectual reference in selfknowledge is direct, we are expressing a view which must be carefully distinguished from the misleading implications of certain traditional ways of describing knowledge. I refer to the idea that introspection is essentially a certain direction of the attention, and may be compared to a direct inspection of the objective contents of consciousness; to these contents, therefore, all that introspection reveals (including feeling and will) must belong. This is the inevitable result of certain current modes of describing knowledge. The traditional account recognises that cognition is

1 The bearing of these considerations on the problem with which our previous chapter closed--the problem (expressed in psychological terms) of the presentative cognition of super-presentational states in consciousness— is surely evident.

always essentially a relation or reference, and therefore involves a duality or distinction between the respective loci, so to speak, of the knowing and of the known: this is most true and important; but a favourite way of describing the relation is calculated to produce much confusion. Knowledge, it is usually said, involves a Subject knowing and an Object known; to this we may, with Dr Ward, add the implications of "a necessary dependence of the Subject on the Object as far as its knowing goes, and no such dependence of the Subject as far as its being goes."1 The sense in which all these propositions are true has been explained elsewhere; here we have to deal with some of the implications of that special form of expression which insists on describing knowledge as necessarily involving a duality of terms in the form of Subject knowing and Object known. The question at once arises: If this is true, how can the Subject know itself? Professor Ward has expressed the difficulty thus: "If we identify the two, we transcend our empirical conception of knowledge; a knowledge in which Subject and Object are one is at best a limiting case towards which we might perhaps conceive ourselves approximating in self-consciousness, and continuing to approximate indefinitely. . . . If, however, on the other hand, we regard the knowing Subject as distinct from the Object known, we require a second Subject, or at least a higher grade of consciousness" in order that the Subject may know itself. This line of thought rests. on the supposition that knowledge is analogous to a bond going across from the Subject to the Object, or, to use a less crude metaphor, that it is analogous to a light proceeding from the Subject and shed on the 1 Mind, N. S., No. 5: "Modern' Psychology."

1