Another presupposition is often introduced by logicians: that of language; since it seems clear that, if man does not speak, he does not think. This presupposition also we accept, adding to it, however, a corollary, together with certain elucidations. The elucidations are: in the first place, that language must be taken in its genuine and complete reality; that is to say, it must not be arbitrarily restricted to certain of its manifestations, such as the vocal and articulate; nor be changed and falsified into a body of abstractions, such as the classes of Grammar or the words of the Vocabulary, conceived as these are in the fashion of a machine, which man sets in motion when he speaks. And, in the second place, by language is to be understood, not the whole body of discourses, taken all together and in confusion, into which (as will be seen in its place) logical elements enter; but only that determinate aspect of these discourses, in virtue of which they are properly called language. A deep-rooted error, which springs directly from the failure to make this distinction, is that of believing language to be constituted of logical elements; adducing as a proof of this that even in the smallest discourse are to be found the words _this, that, to be, to do,_ and the like, that is, logical concepts. But these concepts are by no means really to be found in every expression; and, even where they are to be found, the possibility of extracting them is no proof that they exhaust language. So true is this that those who cherish this conviction are afterwards obliged to leave over as a residue of their analysis, elements which they consider to be illogical and which they call _emphatic, complementary, colorative,_ or _musical_: a residue in which is concealed true language, which escapes that abstract analysis. Finally, the corollary is that if the concept of language is thus rectified, the presupposition made for Logic regarding language is not a _new_ presupposition, but is identical with that already made, when representations or intuitions were discussed. In truth, language in the strict sense, as we understand it, is equivalent to expression; and expression is identical with representation, since it is inconceivable that there should be a representation, which should not be expressed in some way, or an expression which should represent nothing, or be meaningless. The one would fail to be representation, and the other would not even be expression; that is to say, both must be and are, one and the same. [Sidenote: _Intuition and language as presuppositions._]