92 On Comparative Philology. 



that tlie want of agreement is no disproof o{ tlieir connexion ; and proba- 

 bilities seem to teach, that the existence of such agreement is of all argu- 

 ments the most powerful to prove a connexion. Certainly no well-informed 

 Philologists now doubt whether Gotliic, Sanskrit, Greek, and Latin, have 

 a common origin ; yet who would suspect English and French to be any 

 way connected at all, by a comparison of their conjugations ? Again, it is 

 notorious that the inflexions of nouns very easily rub off, dual numbers and 

 neuter genders vanish, prepositions are doubled and tripled, and new pre- 

 positions supersede the old ones ; all which pliBenomena are remarked in 

 the familiar instances of Greek and Latin, compared with their modern 

 dialects. Tt seems then unreasonable to reject the evidence of the kins- 

 manship of languages founded on positive agreement, by the argument that 

 their disparities must first be accounted for. Is it only by such reasoning 

 that some recent writers question the connexion of the Persian with the 

 Sanskrit ? 



On the other hand, it is certain that in the nations best known to us, a 

 great repugnance is displayed to borrowing ybm^n inJle,rions. Caprice or 

 convenience may introduce a few, (as the terminations ize and ism in 

 English,) but only among the more literary. Illiterate persons appear 

 unable to master the intricacy of foreign grammar. Similar in fundamental 

 principles as was the old German to the Latin, the languages could not 

 mix without a great destruction to the inflexions of both ; since neither 

 party will adopt the inflexions of the other. The lingua franca of the 

 Mediterranean shows the same thing. It is also remarked, that, while the 

 intolerant efforts of the Saracens to abolish the old Farsi language, has 

 introduced so large a body of Arabic words into modern Persian, no 

 Persian words submit to a single Arabic inflexion. Now this is a question 

 at once interesting and important to the whole inquiry, whether we are 

 justified in regarding the rule to be universal, that languages do not bor- 

 row from each other their affixes, suffixes, and their principles of inflection : 

 or whether it is to be believed possible that languages may have these in 

 common, and yet be unrelated by genealogy. 



If abstract reasonings will not settle the question, neither will any but a 

 very ample induction of fact settle it. For it will always remain possible, 

 that while some nations readily borrow words, but refuse to borrow inflexi- 

 ons ; others, very barbarous and very destitute of inflexions, might be 

 liappy to borrow these, while they did not want words. In fact, cliildren 

 who hear two languages readily adopt the inflexions of the one in speaking 

 the other, often selecting with care those words and modes of expression 

 which are at once easiest and shortest. The writer remembers sentences 

 such as the following, uttered by children who heard both English and 

 Arabic. — " No more doua, [medicine,] papa." — "I give you kisseyn, [two 

 kisses]." It may be asked, when a savage tribe, whose language, as their 

 minds, is as unformed as that of a child three years old, migrates in great 



