REASON AND LANGUAGE. i6i 



does account for these performances. Whether he 

 admits this or not is, however, quite indifferent to us, 

 as we ground our whole argument, not on authority, 

 but on evidence. To say " half-past two " at the sight 

 of a coachman on whose appearance those words have 

 constantly been heard, is not " to apply words to desig- 

 nate an object," but to emit sounds with which the sight 

 of that object has become accidentally associated. 



Mr. Romanes next makes an altogether unwarrant- 

 able assertion which shows great confusion' of thought ; 

 he tells us that such inventions on the part of parrots 

 "often clearly have an onomatopoetic origin." Now, 

 onomatopoeia is a term used to denote the voluntary 

 employment of an imitation of sounds heard, to denote 

 the conception of the object which makes the sound 

 — as when a child calls a duck " quack-quack/' or when 

 the word " hiss," or something like it, has been employed 

 to express the idea of a hissing snake. Now, when 

 a parrot, which has often seen and heard corks drawn, 

 makes the sound of the drawing of a cork at the sight 

 of a bottle, such is no true case of onomatopoeia, as 

 there is no evidence of intention on the part of the 

 bird to use the sound as a name. 



Mr. Romanes ends the chapter by detailing evidence 

 to show the extent to which, under favourable circum- 

 stances, young children will invent arbitrary signs, 

 mostly of an articulate kind. Had we space we would 

 gladly cite these, as they are much to our purpose. We 

 maintain that man possesses, and always has possessed, 

 an instinct of language, whereby to express, and wherein 

 to incarnate, his spontaneously arising concepts. We 



M 



