HARVARD STUDIES ΞΥΝ LIBRARY CLASSICAL PHILOLOGY EDITED BY A COMMITTEE OF THE CLASSICAL PREFATORY NOTE. THESE Studies are published by authority of Harvard University and are contributed chiefly by its instructors and graduates, although contributions from other sources are not excluded. The publication is supported by a fund of $6000, generously subscribed by the class of 1856. SOME QUESTIONS IN LATIN STEM FORMATION. THE By J. B. GREENOUGH. HE science of Linguistics is really less than a century old, a very short time indeed even to gather and coördinate any considerable body of facts, much less to schematize and explain them. So the complexity of the facts and the great significance of some of the minutest of them often lead men to subtlety of reasoning on subjects which would really find their explanation in the most obvious phenomena. The development of words from simpler elements, technically called stem formation, is a case of this kind. If you take up almost any book on this subject you are sure to find the most ingenious theories invented to account for things that ought to suggest their own explanation if looked at from the fundamental points of view already established in the science. No question has been more ingeniously discussed than that of the origin of the Latin gerund. Dozens of far-fetched and fine-spun theories have been worked out to account for its form and use. Yet it seems that the great fundamental principles as they are already settled beyond controversy are sufficient to account for all the facts about this seemingly difficult question. Let us examine a few of these principles. All agree that the Indo-European family of languages has developed its words by a process of stem formation in which significant elements, presumably verbal roots, have been enlarged and differentiated by the addition in sequence of other significant elements, mostly of pronominal origin, and that these combinations thus grown or made have been again and again subjected to the same process in a greater or less degree, but always following the same type according to the genius of the particular branch of the Indo-European family. The Northern European languages have generally short words, because this process of further formation was in some manner arrested early; the Southern European and the Asiatic languages of the stock have long ones, because the process was fostered and continued to a very great |