REASON AND LANGUAGE. 185 



but it is not predicating any characters as belonging to 

 a dog, or performing any act oi judgmeni with regard to 

 a dog. Although the bird may never (or but rarely) 

 utter the name save when it sees a dog, this fact is 

 attributable to the laws of association acting only in the 

 receptual sphere. . . . Therefore, all my opponents must 

 allow that in one sense of the word there may be names 

 without concepts : whether as gestures or as words 

 (vocal gestures), there may be signs of things without 

 these signs presenting any vestige of predicative value. 

 Names of this kind I have called denotative : they are 

 marks affixed to objects, qualities, actions, etc., by re- 

 ceptual association alone." We freely concede that in 

 such a mere analogical sense vocal or motor phenomena 

 of the kind may be termed " names," and they are to a 

 certain sense signs, as smoke may be a sign of internal 

 heat in a volcano. 



He follows this by observing that such a name may 

 be "extended to denote also another thing, which is 

 seen [!] to belong to the same" class or kind," when they 

 become what he has called *^connotative" and in this con- 

 nection he refers back to his instance of the parrot and 

 the dog, which we have already* criticized, saying, f 

 " Even my parrot was able to extend its denotative 

 name for a particular dog to any other dog which it 

 happened to see — thus precisely resembling my child, 

 who habitually extended its first denotative name Star 

 to a candle." X But this we altogether deny, and must 



* See above, p. 157. t p. 180. 



X At p. 159 he had said, " On^ of my children learnt to say the 

 word Star. Soon after having acquired this word, she extended its 

 signification to other brightly shining objects, such as candles, gas- 



