LANGUAGE. 87 



But it is evident that rational signs admit of almost 

 numberless degrees of complexity and elaboration ; so that 

 reason itself docs not present a greater variety of manifesta- 

 tions in this respect than does the symbolism whereby it is 

 expressed : an algebraical formula is included in the same 

 category of sign-making as the simplest gesture whereby we 

 intentionally communicate the simplest idea. Rational signs, 

 therefore, may be made by gesture, by tone, by articulation, 

 or by writing — using each of these words in its largest 

 sense.* 



The following schema may serve to show this classification 

 in a diagrammatic form — i.e. the classification which I have 

 myself arrived at, and which follows closely the one given by 

 Mr. Mivart. Indeed, there is no difference at all between the 

 two, save that I have endeavoured to express the distinction 

 between signs as intentional, unintentional, natural, conven- 

 tional, emotional, and intellectual. The subdivision of the 

 latter into denotative, connotative, denominative, and pre- 

 dicative, will be explained in Chapter VIII. 



• From this it will be seen that by u^ng such terms as " inference," "reason," 

 " rational," &c., in alludinj; to mental processes of the lower animals, I am in no 

 way prejudicing the question as to the distinction between man and brute. In 

 the higher region of recepts both the man and the brute attain in no small degree 

 to a perception of analogies or relations : this is inference or ratiocination in its 

 most direct form, and differs from the process as it takes place in the sphere of 

 conceptual thought only in that it is not itself an object of knowledge. But, 

 considered as a process of inference or ratiocination, I do not see that it should 

 make any difference in our terminology whether or not it happens to be itself an 

 object of knowledge. Therefore I do not follow those numerous writers who 

 restrict such terms to the higher exhibitions of the process, or to the ratiocination 

 which is concerned only with introspective thought. It may be a matter of straw- 

 splitting, but I think it is best to draw our distinctions where the distinctions 

 occur ; and I cannot see that it modifies the process of inference, as inference, 

 whether or not the mind, in virtue of a superadded faculty, is able to think about 

 the process as a process — not any more, for instance, than the process of associa- 

 tion is altered by its becoming itself an object of knowledge. Therefore, I hope I 

 have made it clear that in maintaining the rationality of brutes I am not arguing 

 for anything more than that they have the power, as Mr. Mivart himself allows, 

 of drawing "practical inferences." Hitherto, then, my difference with Mr. 

 Mivart — and, so far as I know, with all other modern writers who maintain the 

 ixiatiooality of brutes — is only one of terminology. 



