34 COMPARATIVE LANGUAGE 



The history of a given language belonging to a well-defined family can 

 be regarded only as a section of a long and continuous develop- 

 ment. 



Between historical grammar and comparative grammar there is 

 no essential difference. One may for convenience apply the term 

 historical grammar preferably to the study of the actually quotable 

 material belonging to different periods of the same language, and the 

 term comparative grammar where the evidence of cognate languages 

 is introduced. But both alike are historical and comparative. A 

 given form, meaning, or construction is traced back step by step to 

 the earliest stage of which there is historical evidence. And it is with 

 the object of taking a still further step in the same direction, of pene- 

 trating the prehistoric period, that one resorts to the comparison of 

 the cognate languages. There can be no more fundamental miscon- 

 ception of the purpose and value of comparative grammar than is 

 shown in the utterance of one of Germany's most eminent Hellenists 

 in the preface to a Greek grammar which is unrivaled for its collec- 

 tion of facts, though marred by too many antiquated explanations. 

 His words are substantially as follows : " The function of comparative 

 grammar is to compare, that is, to recognize the like and the unlike 

 in related languages, from which the explanation of the forms of the 

 individual language often results of itself, but not always, and when 

 it does not, the comparative grammarian has nevertheless fulfilled 

 his duty by the correct comparison. I do not regard it as the business 

 of comparative grammar to reconstruct the Indo-European, that is, 

 a language which is wholly hypothetical and of which no one knows 

 or ever will know when and where and by what sort of a people it 

 was spoken. How does such a language concern us? Still I will not 

 object if one wishes to write a grammar or even a lexicon of Indo- 

 European." Presumably it is the representative of a science, and 

 not one who is not even in sympathy with it, who is most competent 

 to define its scope, and it is safe to say that no professed represent- 

 ative of comparative grammar will accept any such limitation of its 

 function as is prescribed in the words quoted. 



Comparison is only a means to an end. The recognition that 

 a certain Greek form is the equivalent of a certain Latin form, or a 

 Sanskrit form of a Greek, may be interesting, but of what importance 

 in itself? Its value lies in the conclusions it enables us to draw as to 

 the parent form. The form of any one language will admit of various 

 possible origins, but the range of possibilities will not be the same 

 for each language, and by a process of exclusion we reduce these to 

 the one (or sometimes more than one) which satisfies the requirements 

 of all the related languages. Often the evidence is so complete and 

 conclusive that we feel as certain of the actual existence of the parent 

 form thus reconstructed as of the existence of the historical forms 



