But alongside of this passion for simplification there exists a sister passion, which in some minds--though they perhaps form the minority--is its rival. This is the passion for distinguishing; it is the impulse to be _acquainted_ with the parts rather than to comprehend the whole. Loyalty to clearness and integrity of perception, dislike of blurred outlines, of vague identifications, are its characteristics. It loves to recognize particulars in their full completeness, and the more of these it can carry the happier it is. It prefers any amount of incoherence, abruptness, and fragmentariness (so long as the literal details of the separate facts are saved) to an abstract way of conceiving things that, while it simplifies them, dissolves away at the same time their concrete fulness. Clearness and simplicity thus set up rival claims, and make a real dilemma for the thinker. A man's philosophic attitude is determined by the balance in him of these two cravings. No system of philosophy can hope to be universally accepted among men which grossly violates either need, or {67} entirely subordinates the one to the other. The fate of Spinosa, with his barren union of all things in one substance, on the one hand; that of Hume, with his equally barren 'looseness and separateness' of everything, on the other,--neither philosopher owning any strict and systematic disciples to-day, each being to posterity a warning as well as a stimulus,--show us that the only possible philosophy must be a compromise between an abstract monotony and a concrete heterogeneity. But the only way to mediate between diversity and unity is to class the diverse items as cases of a common essence which you discover in them. Classification of things into extensive 'kinds' is thus the first step; and classification of their relations and conduct into extensive 'laws' is the last step, in their philosophic unification. A completed theoretic philosophy can thus never be anything more than a completed classification of the world's ingredients; and its results must always be abstract, since the basis of every classification is the abstract essence embedded in the living fact,--the rest of the living fact being for the time ignored by the classifier. This means that none of our explanations are complete. They subsume things under heads wider or more familiar; but the last heads, whether of things or of their connections, are mere abstract genera, data which we just find in things and write down.