244 ROMANCE LANGUAGES 



ago, and also to appreciate the immense progress which has been 

 accomplished in the last thirty or forty years in working the domain 

 which the master had so brilliantly explored. 



The Romance languages are nothing else than Latin modified 

 differently according to times and places. But to what Latin do we 

 refer? To the Vulgar Latin, assuredly, to Raynouard's langue romane, 

 which was at first almost homogeneous throughout the Roman 

 Empire. Diez was convinced of this fact, and all that he says on the 

 subject in the first part of his grammar is very sensible; it is evident, 

 however, that he had the lexicographical elements more in view than 

 matters of grammatical structure. But he deliberately refrained 

 from any attempt to tell us how and under what circumstances 

 the local changes occurred which have transformed Vulgar Latin into 

 the infinite variety of the Romance idioms. Here was a question 

 which had been much debated, and one to which various solutions 

 had already been proposed. Some believed that Latin had under- 

 gone profound changes through contact with Germanic or Slavic 

 languages at the period of the invasions in the fifth century, and 

 comparisons with chemical compounds were made which conveniently 

 veiled the weakness of the historical and linguistic arguments in- 

 voked. This was the opinion put forward by Muratori and upheld 

 by Littre", forty or fifty years ago. Others held, with greater prob- 

 ability, that the local variations of Latin must have existed in an 

 even more remote period, and that we must attribute the first 

 changes to the linguistic habits of the Celts, Iberians, Ligurians, etc., 

 of Gaul, Spain, and Italy, habits of which these populations had 

 not been able to rid themselves in learning to speak Latin. This is 

 the theory once defended, with more energy than weight of proof, 

 by Fauriel. It has since been revived and supported with more 

 definite arguments by eminent linguists, among whom it is sufficient 

 to name Prof. Ascoli. But Diez was concerned with facts that could 

 be proved; he had no great liking for questions whose answers 

 involved too large a proportion of the hypothetical. Rather than 

 to continue debating these doubtful questions, what was needed, 

 if the historical method was to be employed, was to reduce as 

 much as possible the space still vacant between Latin as it was 

 known in the classic authors that is, written Latin, which had not 

 greatly changed since the first century and the Romance languages, 

 which did not make their appearance before the ninth or tenth 

 centuries. In this vacant space there was the Vulgar Latin, about 

 which very little was known. 



On still other subjects Diez had left work for his successors. 

 History and geography touch philology on several sides; these 

 sciences mutually aid and support one another. At first this was 

 not well understood. To what boundaries did the Roman conquest 



