REASON AND LANGUAGE, 157 



mentioned by any previous writer, although I should 

 think it is one that can scarcely have escaped the notice 

 of any attentive observer of these animals. I allude to 

 the aptitude which intelligent parrots display of extend- 

 ing their articulate signs from one object, quality, or 

 action, to another which happens to be strikingly 

 similar in kind. For example, one of the parrots which 

 I kept under observation in my own house learnt to 

 imitate the barking of a terrier, which also lived in the 

 house. After a time this barking was used by the parrot 

 as a denotative sound, or proper name, for the terrier — 

 i.e., whenever the bird saw the dog it used to bark, 

 whether or not the dog did so. Next, the parrot ceased 

 to apply this denotative name to that particular dog, 

 but invariably did so to any other, or unfamiliar, dog 

 which visited the house. Now, the fact that the parrot 

 ceased to bark when it saw my terrier after it had 

 begun to bark when it saw other dogs, clearly showed 

 that it distinguished between individual dogs, while 

 receptually perceiving their class resemblance. In other 

 words, the parrot's name for an individual dog became 

 extended into a generic name for all dogs." 



Now, as Mr. Romanes very often refers back to this 

 example, we must criticize the passage with some pains 

 and at some length. In the first place, as Mr. Romanes 

 has before remarked * — citing Dr. Wilks — it is common 

 enough for parrots to imitate on seeing a visitor some 

 words or noise he habitually makes, as it may imitate 

 the sound of cork-drawing on seeing a bottle. Barking 

 at the sight of the terrier is, then (as Mr. Romanes would 



* PP- 131, 132. 



