Such, then, are the two divergent modes of knowledge by which intelligence and instinct must be defined, from the standpoint of knowledge rather than that of action. But knowledge and action are here only two aspects of one and the same faculty. It is easy to see, indeed, that the second definition is only a new form of the first.
If instinct is, above all, the faculty of using an organized natural instrument, it must involve innate knowledge (potential or unconscious, it is true), both of this instrument and of the object to which it is applied. Instinct is therefore innate knowledge of a _thing_. But intelligence is the faculty of constructing unorganized--that is to say artificial--instruments. If, on its account, nature gives up endowing the living being with the instruments that may serve him, it is in order that the living being may be able to vary his construction according to circumstances. The essential function of intelligence is therefore to see the way out of a difficulty in any circumstances whatever, to find what is most suitable, what answers best the question asked. Hence it bears essentially on the relations between a given situation and the means of utilizing it. What is innate in intellect, therefore, is the tendency to establish relations, and this tendency implies the natural knowledge of certain very general relations, a kind of stuff that the activity of each particular intellect will cut up into more special relations. Where activity is directed toward manufacture, therefore, knowledge necessarily bears on relations. But this entirely _formal_ knowledge of intelligence has an immense advantage over the _material_ knowledge of instinct. A form, just because it is empty, may be filled at will with any number of things in turn, even with those that are of no use. So that a formal knowledge is not limited to what is practically useful, although it is in view of practical utility that it has made its appearance in the world. An intelligent being bears within himself the means to transcend his own nature.
He transcends himself, however, less than he wishes, less also than he imagines himself to do. The purely formal character of intelligence deprives it of the ballast necessary to enable it to settle itself on the objects that are of the most powerful interest to speculation. Instinct, on the contrary, has the desired materiality, but it is incapable of going so far in quest of its object; it does not speculate. Here we reach the point that most concerns our present inquiry. The difference that we shall now proceed to denote between instinct and intelligence is what the whole of this analysis was meant to bring out. We formulate it thus: _There are things that intelligence alone is able to seek, but which, by itself, it will never find. These things instinct alone could find; but it will never seek them._