THE WORLD AS IDEA. 53
tion rests on the world of perception as its ground of knowledge. Hence the class of abstract ideas is in this respect distinguished from other classes; in the latter the principle of sufficient reason always demands merely a relation to another idea of the same class, but in the case of abstract ideas, it at last demands a relation to an idea of another class.
Those concepts which, as has just been pointed out, are not immediately related to the world of perception, but only through the medium of one, or it may be several other concepts, have been called by preference abstracta, and those which have their ground immediately in the world of perception have been called concreta. But this last name is only loosely applicable to the concepts denoted by it, for they are always merely abstracta, and not ideas of perception. These names, which have originated in a very dim consciousness of the distinctions they imply, may yet, with this explanation, be retained. As examples of the first kind of concepts, i.e., abstracta in the fullest sense, we may take 'relation,' 'virtue,' 'investigation,' 'beginning,' and so on. As examples of the second kind, loosely called concreta, we may take such concepts as 'man,' 'stone,' 'horse,' &c. If it were not a somewhat too pictorial and therefore absurd simile, we might very appropriately call the latter the ground floor, and the former the upper stories of the building of reflection.[1]
It is not, as is commonly supposed, an essential characteristic of a concept that it should contain much under it, that is to say, that many ideas of perception, or it may be other abstract ideas, should stand to it in the relation of its ground of knowledge, i.e., be thought through it. This is merely a derived and secondary characteristic, and, as a matter of fact, does not always exist, though it must always exist potentially. This characteristic arises from the fact that a concept is an
- ↑ Cf. Ch. 5 and 6 of the Supplement.
