Prehistoric and Savage Man 1 8 



o 



(2.) Sounds which are articulate but not rational, such 

 as the talk of parrots, or of certain idiots, who will repeat 

 without comprehending every phrase they hear. 



(3.) Sounds which are rational but not articulate, such 

 as the inarticulate ejaculations by which we sometimes 

 •express assent to or dissent from given propositions. 



(4.) Sounds which are both rational and articulate, and 

 such is ordinary human speech. 



(5.) Gestures which do not answer to rational conceptions, 

 but are merely the manifestations of emotions and feelings. 



(6.) Gestures which do answer to rational conceptions, 

 and are, therefore, 'external' but not 'oral' manifestations 

 of the ^mental word. 



A great deal too much stress has been laid upon the fact 

 of articulation, for this character of speech is merely due to 

 the breaking up of vowel sounds and consequent multiplica- 

 tion of distinguishable utterances ; various mere animals can 

 articulate, and it is quite conceivable that brutes might 

 (though as a fact they do not) so associate certain sensa- 

 tions and gratifications with definite articulate sounds, as 

 in a certain sense to 'speak' while remaining all the time 

 as utterly devoid as ever of rational language. Just as a dog 

 learns to 'beg' by associating certain gestures with 'sub- 

 sequent pleasurable sensations of joy,' so a parrot might 

 conceivably learn to ask for its dinner or to say, ' Open the 

 cage door, that Poll may take a walk,' while in no way 

 understanding either the words or the sentence thus uttered. 

 The late lamented George Henry Lewes, strong Evolutionist 

 as he was, clearly saw and admitted this distinction, and he 

 says,^ in his just published posthumous work: 'The dog, 

 although without language, may yet be taught so to connect 

 the sound "lion" with other feelings, that on hearing it 



1 Problems of Life and Mind. Third Series, Problems 2, 3, and 4, 1879. 

 p. 473. 



