THE ORIGIN OF LANGUAGE 53 



lologists have not followed Miiller quite so far in 

 this simplification, and it is generally admitted to- 

 day that the oldest literature gives evidence of a 

 larger number of ''roots" than Miiller was at one 

 time inclined to accept. But even though the words 

 cannot all be reduced to quite such a small number, 

 it is universally admitted that the number of words 

 in early Sanskrit was surprisingly small, and that 

 they pointed toward a condition in which they were 

 fewer still in an earlier age. From such a condition 

 we can legitimately go a step farther back and 

 recognize the inevitable existence of a language so 

 simple that the few root words were all that existed. 

 Of such an early language, of course, literature could 

 only give us a hint, with here and there a suggestion 

 as to what the language was. Writing must have 

 come into existence ages after speaking. Many races 

 of men are not yet able to reduce their language to 

 writing ; and a language like Sanskrit, which had not 

 only reached the stage of writing but had even risen 

 to the grade of a literature, must have been an 

 immense distance from the primitive language. Con- 

 sidering the tremendous amount of time that must 

 have been required to convert the first spoken words 

 into a definite language, and then to have developed 

 writing and a literature, it is surprising to find it as 

 simple as it is. 



Even though Miiller 's roots can no longer be ac- 

 cepted as forming the foundation of all of Sanskrit, 

 it is certain that these one hundred and twenty-one 

 words were very prominent, and it will be instructive 

 to note briefly the ideas represented by these few 

 root words. They nearly all seem to be verbs, and 

 refer to types of action which the very crudest intelli- 



