literatures, even to some degree those wrought by the hand of mere imitators like the Romans, depend upon conditions of social life, and, if not stationary or decaying, constantly throw out new forms of vitality, constantly enter new phases of art and criticism; and the fact that, in spite of this constant movement in each separate literature, in all literatures viewed together as productions of humanity, definition implies, and must at least provisionally assume, a degree of permanence which is too often secured off-hand by violently declaring selected ideas to be universal and independent, not only of social life in its myriad shapes, but even of space and time. Hereafter we shall have other opportunities for discussing these obstacles to the scientific study of literature—obstacles, it must be remembered, common to all the social sciences, political economy, jurisprudence, even logic, so far as the laws of thought are dependent on social evolution. At present, however, we shall be satisfied with two principles which may serve to guide our efforts to reach defined ideas of literature. (1) Our definition cannot cover an unlimited range of human life save at the expense of confusing perceptions of sense, emotions, thoughts, not only belonging to widely diverse social and physical conditions, but often directly conflicting in the form and spirit of their literary expression. (2) We must be ever prepared to forego our limited definitions of literature, or any species of literature, when we pass out of the conditions to which they are properly confined.
Bearing these principles in mind, we may be content to set out with a rough definition of literature, as consisting of works which, whether in verse or prose, are the handicraft of imagination rather than reflection, aim at the pleasure of the greatest possible number of the
