62 COMPARATIVE LANGUAGE 



which in the course of time acquired the force and function of a 

 suffix. 



In syntax, the most important feature which has influenced the 

 methods of investigation has been the tendency to carry into practice 

 Humboldt's maxim that language is not so much an Ipyov as an 

 evepyeia, 1 in other words, to view language not solely as a collec- 

 tion of facts (spoken or written words or sentences) but rather as an 

 activity and a psychical process of which the spoken word itself is 

 only the outward and audible sign. This desire to turn from the fin- 

 ished product to an examination of the producing agency explains the 

 change from the logical to the psychological treatment of grammar. 



The arrangement of facts in grammars viewing language as 

 static is largely a classification of linguistic products according to 

 external similarities, similar to the Linnsean classification of plants. 

 Whether this be rougher or finer, whether the subdivisions be few 

 or many, the character of this classification remains essentially the 

 same, inasmuch as it is based upon the present external appearance 

 of things and often cannot take into consideration the genesis of 

 the very qualities according to which it classifies. Now, while such 

 a descriptive classification is necessary, useful, and sufficient for the 

 practical mastery of the details of a language, where the sole object 

 is acquaintance with the facts as they are or were, it is scientifically 

 insufficient because it fails to indicate how these objects came to be 

 what they are. More than that, it may be positively misleading 

 when it groups together facts which have an external similarity but 

 owe their existence to different causes. These genetic, differences a 

 descriptive classification at times veils and obscures. For while it is a 

 truism that a like combination of like forces must produce like effects, 

 it is no less true (though sometimes forgotten) that a different 

 combination of different forces may produce like effects also. It is 

 wholly wrong to work on the principle that like effects must neces- 

 sarily imply like causes. 2 Many illustrations might be given of the 

 grouping together of genetically different material under such general 

 descriptive heads as "assimilation," " anaptyxis," etc. And only 

 recently Meumann 3 has called attention to the abuse of the term 

 " metaphor " when applied to the variations in the semantic sphere 

 of the child's vocabulary, by showing how entirely different are 

 the psychological processes which underlie the creation of a poetical 

 or rhetorical metaphor from the so-called " generalizing tendency " of 

 the child. 



The desire to investigate processes of development rather than 

 classify finished products has affected semantic investigations in 



1 Delbriick, loc. tit. p. 45. 



2 Cf. Foy in IF, xn, 33. 



3 Die Sprache des Kindes, pp. 60-63. 



