But the philosopher will perhaps refuse to found a theory of knowledge on such considerations. They will be repugnant to him, because the mathematical order, being order, will appear to him to contain something positive. It is in vain that we assert that this order produces itself automatically by the interruption of the inverse order, that it is this very interruption. The idea persists, none the less, that _there might be no order at all_, and that the mathematical order of things, being a conquest over disorder, possesses a positive reality. In examining this point, we shall see what a prominent part the idea of _disorder_ plays in problems relative to the theory of knowledge. It does not appear explicitly, and that is why it escapes our attention. It is, however, with the criticism of this idea that a theory of knowledge ought to begin, for if the great problem is to know why and how reality submits itself to an order, it is because the absence of every kind of order appears possible or conceivable. It is this absence of order that realists and idealists alike believe they are thinking of--the realist when he speaks of the regularity that "objective" laws actually impose on a virtual disorder of nature, the idealist when he supposes a "sensuous manifold" which is coördinated (and consequently itself without order) under the organizing influence of our understanding. The idea of disorder, in the sense of _absence of order_, is then what must be analyzed first. Philosophy borrows it from daily life. And it is unquestionable that, when ordinarily we speak of disorder, we are thinking of something. But of what?