OF NATURAL HISTORY. 175 



I can perceive only one plaufible objedion to this reafoning. If, 

 it may be faid, man were endowed with a natural language, this 

 language muft be univerfal ; from what fource, then, can the great 

 diverfity of languages in different nations, and tribes of the human 

 race, be derived ? The folution of this queftion depends not upon 

 metaphyfical arguments, but upon fadt and experience. I have had 

 confiderable opportunities of obferving the behaviour of children. 

 Infants, when very young, have nearly the fame modes of expref- 

 fing their pleafures and pains, their defires and averfions. Thefe 

 they communicate by voice, gefture, and feature ; and every infant, 

 whatever be the country, climate, or language, uniformly exprelTes 

 its feelings almoft in the fame manner. But, when they arrive at 

 nine or twelve months of age, a different fcene is exhibited. They 

 then, befide the general expreffions of feeling and defire, attempt to 

 give names to particular objedls. Here artifice begins. In thefe at- 

 tempts, previous to the capacity of imitating articulate founds, every 

 individual infant utters different founds, or rather gives different 

 names, to fignify the fame objeds of its defire or averfion. Befide 

 this natural attempt towards a nomenclature, infants, during the 

 period above mentioned, (for the time varies according to the health 

 and vivacity of the child), frequently make continued orations. 

 Thefe orations confift both of articulate and inarticulate founds, of 

 which no man can give an idea in writing. But moft men, and 

 every woman who has nurfed children, will perfedly underftand 

 what I cannot exprefs. From the fadt, that children adually utter 

 different founds, or give different names to denote the fame objeds, 

 I imagine, arifes all that diverfity of languages, which, by exhauft- 

 ing time and attention, retard the progrefs and improvement both of 

 Art and Science. If any number of children, or of folitary favages, 

 fliould chance to aflbciate, the names of objcds would foon be fet- 

 tled by imitation and confent. By obfervation and experience the 

 number of names would be augmented, as well as the qualities or 



attributes 



