3io
��INTERNATIONAL JOURNAL OF AMERICAN LINGUISTICS
��VOL. I
��(extrinsic energy) represent a period in the development of language when people used to speak 'H'itb more sentiment or to articulate less clearly. He might have added that both factors may easily be observed in modern child lan- guage, though I am far from asserting that this proves anything. In group I are also classed the reduplicated forms of numerous very old roots which could not possibly be classified accord- ing to their meaning, but which generally belong to the most indispensable part of the vocabulary. As usually, van Ginncken, after stimulating our interest to the utmost, leaves the subject ; attempting to construct a well thought out theory out of his sporadic remarks is quite as useless as trying to read by an occa- sional flash of lightning. In the first place his explanation of reduplications resulting from extrinsic energy is rather vague. That uncivi- lized peoples speak with more sentiment than we do, is not to be denied, but about the sta- tement that they articulate less clearly we should be glad to hear something more. Fur- ther, does van Ginneken mean that there has been a period in the development of language when all words occurred in a reduplicated form only? This must be our inference if it is true that the meaning of the words themselves had nothing whatever to do with the circumstance that they were reduplicated. But in this case it is obviously impossible to distinguish between the two groups, as any reduplicated form may have originated in the period when reduplica- tion was universal. It is evident that van Ginne- ken's theory cannot quite satisfy us, but this does not imply that his classification is wholly wrong. It will be admitted, I think, that at least one of his groups is really suggestive of a definite semantic category. This category comprises those cases of reduplication whose functions may be summarized by the general term increase: plural forms, distributives, inten- sity of action, continuity of action, repetition
��of action, customary action, superlatives etc. (for American examples see the Handbook of American Indian Languages I). With a view to the mental attitude of the speaker we may per- haps call them emphatic reduplications. If we now consider the numerous reduplications which are not immediately recognizable as belonging to the emphatic group, we meet with a striking variety : thus we find a. o. ono- matopoeia, " lallworter ", adjectives of color, shape and surface quality, nomina actons and acti ; further reduplication may express unrea- lity, imitation, playful activity. Among these various functions the last mentioned group seems to present itself as a semantic category indicating the idea of unreality. According to van Ginneken, this group has about the same function as the Indo-European vrddhi-deriva- tives, which he calls " allongements d'hesita- tion " in contradistinction to the " allonge- ments d'emphase ", which seem to be (psy- cologically) akin to our emphatic reduplica- tions. Further the onomatopoeia and " lall- worter ", whose common characteristic seems to consist in their emotional nature, cannot well be separated from the foregoing group, though here the emotional element is less obvious. No doubt van Ginneken's first class, with which the " lallworter" group brings us into touch again, is large enough to embrace both of them. However, there is no reason to regard these reduplications as " survivals " belonging to a period when reduplication was well-nigh universal because people used to speak with more sentiment : even to our " civilized " conceptions the character of the concerned words is quite sufficient to explain the senti- ment with which they were pronounced. The only kind of reduplications which undoubtedly originate of extrinsic energy are those found among the interjections; but these belong to all times and all peoples.
It is quite true that there remain a great
�� �
