THE WITNESS OF PI/ILOLOGY. 299 



achieved by means of suffixes and terminations, though often 

 also by change of tone. We saw that, in an eadier stage, 

 the Aryan languages, too, could raise a root into a word, 

 without the aid of suffixes, and that, for instance, yudh, to 

 fight, could be used in the five senses of the act of fighting, 

 the agent of fighting, the instrument of fighting, the place of 

 fighting, and the result of fighting. For the sake of distinction, 

 however, as soon as the necessity began to be felt, the Aryan 

 language introduced derivative elements, mostly demonstrative 

 or pronominal," 



"The imperative may truly be called the most primitive 

 sentence, and it is important to observe how little in many 

 languages it deviates from what has been fixed upon as the 

 true form of a root. . . . va, weave, whether as a reminder 

 or as a command, would have as much right to be called a 

 sentence as when we say, 'Work,'/.^. 'Let us work.' . . . From 

 the use of a root in the imperative, or in the form of a general 

 assertion, there is a very easy transition to its employment in 

 other senses and for other purposes. ... A master requiring 

 his slaves to labour, and promising them their food in the 

 evening, would have no more to say than ' Dig — Feed,' and 

 this would be quite as intelligible as ' Dig, and you shall 

 have food,' or, as we now say, ' If you dig, you shall have 

 food.' " * 



Thus we may lay it down as a general doctrine or 

 well-substantiated principle of philological research, that 

 " Language begins with sentences ; not with single words ;" f 

 or that originally every word in and of itself required to 

 convey a meaning, after the manner of the early utterances 

 of children. " The sentence is the only unit which language 

 can know, and the ultimate starting-point of all our linguistic 

 researches. ... If the sentence is the unit of significant 

 speech, it i:j evident that all individual words must once have 

 been sentences; that is to say, when first used they must 

 each have implied or represented a sentence." % 



• Science 0/ Thought, 423-440. 

 t Saycc, Introduction^ ^c, i. III. \ Jbni., i. 113, 114. 



