NATUEE OF ROOTS AND WORDS. 529 



2°. That in respect both to meaning and form, the determinative 

 cause of the preferential use of a particular meaning, or 

 form, was individual habit developed into general usage, 

 which caused similar objects, in course of time, to be 

 indicated, as a rule, by similar sounds, in the same com- 

 munity, and gave greater stability, and therefore greater 

 intelligibility, to language. 



3°. That the variation of meaning, the application of the same 

 names to different objects or ideas, could only take place 

 when the idea which was the primary cause of the use of 

 the particular name had been entirely forgotten, and had 

 become a mere arbitrary outward sign. 



4°. That primitive language, in order to be an intelligible means 

 of communication between man and man, must have dealt 

 only in conqrete or individual names and in particular 

 ideas, and that abstract names and general ideas were the 

 result of a subsequent process of comparison between the 

 different individuals, with regard to a number o^ particular 

 attributes common to many, which caused the general 

 resemblance to be perceived. 



Surely there is nothing miraculous in the direct invention of a 

 vehicle of communication, an engine of thought, so unstable, so 

 variable, so fluctuating as this, and yet so easily fixed by means so 

 natural and unconscious as habit and usage, and at the same time so 

 perfectly answering the purpose for which it was created or invented. 



We have, however, considered language only in one aspect — with 

 regard to the isolated word and its content. Now Professor Sayce, 

 in his very ingenious and interesting work on the "Principles of 

 Comparative Philology," has lately pointed out, with great force and 

 clearness, that language consists not only of words but of sentences. 

 The word bears the same relation to the sentence that letters do to 

 words. A letter is nothing by itself, nor can a word express thought, 

 except as a member of a sentence. The interjection can express 

 emotion, not thought ; and to this the imperative of the verb is akin 

 in usage, though not in origin. 



We have, therefore, as yet only completed half the task proposed ; 

 we have described the nature of primitive words as abstract and 

 isolated things, incapable of conveying thought. We have still to 



