314 RECENT PROGRESS OF THE THEORY OP VISIOIT. 



There is a most striking analogy between the entire 

 range of processes which we have been discussing, and 

 another System of Signs, which is not given by nature 

 but arbitrarily chosen, and which must undoubtedly be 

 learned before it is understood. I mean the words of our 

 mother tongue. 



Learning liow to speak is obviously a much more 

 difficult task than acquiring a foreign language in after 

 life. First, the child has to guess that the sounds it 

 hears are intended to be signs at all ; next, the meaning 

 of each separate sound must be found out, by the same 

 kind of induction as the meaning of the sensations of 

 sight or touch ; and yet we see children by the end 

 of their first year already understanding certain words 

 and phrases, even if they are not yet able to repeat 

 them. We may sometimes observe the same in dogs. 



Now this connection between Names and Objects, which 

 demonstrably must be learnt^ becomes just as firm and 

 indestructible as that between Sensations and the Objects 

 which produce them. We cannot help thinking of the 

 usual signification of a word, even when it is used 

 exceptionably in some other sense ; we cannot help feeling 

 the mental emotions which a fictitious narrative calls 

 forth, even when we know that it is not true ; just in the 

 same way as we cannot get rid of the normal signification 

 of the sensations produced by any illusion of the senses, 

 even when we know that they are not real. 



There is one other point of comparison which is worth 

 notice. The elementary signs of language are only twenty- 

 six letters, and yet what wonderfully varied meanings 

 can we express and communicate by their combination ! 

 Consider, in comparison with this, the enormous number 

 of elementary signs with which the machinery of sight 

 is provided. We may take the number of fibres in the 

 optic nerves as two hundred and fifty thousand. Each 



