244- THE ORIGIN OF HUMAN REASON 



of reasoning which can possibly coexist with the due ex- 

 pression of thoughts and inferences is the best. There- 

 fore, again, since the quickest and easiest signs are 

 articulate ones, in an ideal language, every sentence 

 should be capable of expression by a monosyllabic 

 word, and every inference by the utterance of three 

 monosyllables. 



It is not at all true, or a matter of course that " the 

 more that a single word thus assumed the functions now 

 discharged by several words when built into a proposi- 

 tion, the more generalized — that is to say, the less 

 defined — must have been its meaning." Such may or 

 may not have been the case, according to circumstances. 

 Mr. Romanes cites * various childish expressions to 

 support his view ; but, in the first place, primitive man 

 was not a child nor in the position of a child, and a very 

 young child does not adequately pourtray the mental 

 condition of an adult human ancestor, any more than 

 its body shows us what any adult human ancestor's 

 body was actually like. In the second place, supposing 

 a child does use the words, " Ta, ta," or " Ba-ba," or 

 " Bye-bye," in more senses than one, we may ask, why 

 should it not? It can do so quite as rationally as 

 when, being adult, it uses the one word ''box" in 

 several senses. 



Much that Mr. Romanes here urges might be ques- 

 tioned ; but for our purpose it is quite unnecessary so to 

 do. We have thus no objection, for argument's sake, to 

 concede that f " the earliest indications of grammar are 

 given by the simultaneous use of sentence- words and 

 * p. 296, t P- 297. 



