382 THE POPULAR SCIENCE MONTHLY 



Contre nous de la tyrannie l'etendard sanglaut est lev6 — 



though it is so easy to rethink it as " Against us tyranny's standard 

 of blood is raised." 



No, the phrase is the unit. Of this I once had an almost convincing 

 demonstration. A class was beginning Cicero's essay on old age and, as 

 there were no books at hand, I began to read the first chapter aloud, 

 slowly, and by phrases, with the result that I secured in this oral way a 

 much better translation than the class brought up next day for their 

 prepared lesson. If any one wants to convince himself of the superiority 

 of Virgil in the narrative style to any other Latin poet, I know of no 

 better test than reading him aloud by phrases: his brief phrases — not 

 his words — move with a simplicity and naturalness not unlike the prose 

 style of King James's version. 



XI. Analytic versus Synthetic Languages. 



We can not resist the impression that modern Greek and Italian, as they 

 are but the ruins and vestiges of the languages in which Demosthenes and 

 Cicero spoke, afford by comparison but miserable accommodation for thought. 

 From our extremely small experience of the speech of the world we judge that, 

 in the case of the few languages which we know, evolution has proceeded back- 

 wards: the better organized, and therefore, from the evolutionary standpoint, 

 the higher language has given place to the lower. . . . Greek and Latin were 

 not made by cultured Greeks and Eomans. The language took form in the 

 converse of their illiterate ancestors. Literature, upon which the beginnings of 

 culture rest, closes language building in the larger sense. Zulu is a more highly 

 flexional language than Greek. . . . The language of the Zulus is not great 

 because it is complex in form. Every language becomes great when greatly 

 used— Greek from Demosthenes 's mouth; English from Milton's pen. . . . The 

 only evolutionary tendency in language which we can recognize is this tendency 

 towards analysis, toward dismemberment. So great an authority as Sir Charles 

 Eliot, vice-chancellor of Sheffield University, who perhaps knows a greater 

 variety of languages than any other man, from Portuguese to Eussian, from 

 Turkish to Japanese, languages of central Africa and of the Polynesian Islands, 

 tells me that he considers that this progress favors thought. Gender, number, 

 case hamper language, restrict its flexibility, impede thought. A monosyllabic 

 root-language, such as Chinese or Burmese, is a swifter and more precise solvent 

 of thought than are the highly inflected Bantu tongues. If this be true — and 

 it does not seem to me open to doubt — it is easier to think in English than 

 in Latin. 



I have not the least doubt that it is easier for Sir Charles Eliot, 

 Professor Hill or me to think in English than in Latin. The great 

 educative value I assign to the study of Latin lies precisely therein. 

 The rethinking of Latin into English can not fail to be tremendously 

 more difficult than the rethinking of any modern cultural tongue into 

 English. But Professor Hill is dealing with postulates, not demon- 

 strations. Who shall show us that Cicero's Verrines cost more effort 

 in the thinking and phrasing or appealed less simply and directly to his 

 audience than Burke's " Impeachment of Warren Hastings " ? 



