52 SOCIAL HEREDITY AND SOCIAL EVOLUTION 



The Records of Language-Building. — We can trace lan- 

 guage toward its beginnings by two general methods. 

 We may study the literature of the earlier ages, or 

 we may study the languages of the lower races of 

 living men. By either method the result is the same, 

 for by either method we find language becoming 

 simpler and simpler. The earliest written language 

 is vastly simpler than any written language of to- 

 day, and the languages of some savages are simpler 

 still. The simplification of language is seen first in 

 the fact that as we go backward many words disap- 

 pear. The words of the complex, highly developed 

 languages are largely coined out of combinations of 

 the earlier and simpler words, so that the farther we 

 go back the smaller become the number of words, 

 until even in the earliest written language their num- 

 ber seems surprisingly small. Sanskrit, represent- 

 ing, as it does, the oldest language with a literature, 

 becomes of special interest since to it we are able to 

 trace most of the later inflectional languages. The 

 languages derived from the Sanskrit are noted for 

 their wealth of words, and with this in mind it seems 

 hardly credible that one of the early philologists, 

 Miiller, at one time thought that he could trace the 

 whole language back to one hundred and twenty-one 

 original words, or *' roots, '* as he called them. It 

 must not be inferred that Miiller found any liter- 

 ature so primitive as to be made of such a small 

 number of words, since even the earliest Sanskrit 

 known contained many more. But he thought that 

 by analyzing the words of early Sanskrit he could 

 find evidence that they were themselves compounds 

 from older and simpler forms, and he finally reduced 

 these to one hundred and twenty-one. Other phi- 



