1 72 THE EVOL UTION OF LANG UA GE. 



delicate appreciation of resemblance, the child eventu- 

 ally called all coins ' quack,' because on the back of a 

 French sou it had once seen the representation of an 

 eagle. Hence, to the child, the sign ' quack,' from 

 having originally hud a very specialized meaning, be- 

 came more and more extended in its significance, until 

 it now seems to designate such apparently different 

 objects as ' fly,' ' wine,' and ' coin.' " * 



The instructiveness of this, in showing the reason 

 why philology is often so helplessly at a loss in track- 

 ing far-strayed words to their original sense, is plain. 

 In the nature of the case, the onomatopoetic theory 

 can never be proved in more than a fraction of cases. 

 So cunning is the mind in associating ideas, so swift 

 in making new departures, that the clue to multitudes 

 of words must be obliterated by time, even if the first 

 forms and spellings of the words themselves remain 

 in their original integrity which rarely happens to 

 offer a feasible point to start the search from. 



But it is far from necessary to assume that all 

 words should have had a rational ancestry. On the 

 contrary many words are probably deliberate artifi- 

 cial inventions. When not only every human being, 

 but every savage and every child has the ability as 

 well as the right to call anything it likes by any name 

 it chooses, it is vain in every case to seek for any gen- 

 eral principle underlying the often arbitrary conjunc- 

 tions of letters and sounds which we call words. 

 Words cannot all at least be treated with the same 

 scientific regard as we would treat organic forms. 

 When dissected, in the nature of the case, they cannot 

 be expected to reveal specific structure such as one 

 1 Mental Evolution, p. 283. 



