2G0 Substitution of the Roman [No. 8, new series. 



ter for the other will leave matters precisely as it found them. If 

 the Indian k, t or I, or the Indian a, i, or e, happens to be pro- 

 nounced in one way only, the Roman k, t, and I, and the Roman a, 

 *, and e will in like manner have only one sound each. If the In- 

 dian characters represent, as they sometimes do, several different 

 sounds, it will simply be necessary to pronounce the corresponding 

 Roman characters in each of those different ways, as the usage of 

 the language may require. No existing advantage, therefore, will 

 in reality be forfeited, whilst, as was previously shown, many will 

 be gained by the use of the Roman character. 



• I may here add that the danger to the acquisition of a correct 

 pronunciation of the Native languages arising from the use of the 

 Roman character seems to be equally unreal. It is quite true that 

 the English characters will naturally in the first instance suggest 

 to the beginner their corresponding English sounds, but this is a 

 danger to which every person who commences to learn a new lan- 

 guage by book is exposed. The Englishman learning to read 

 French, the Frenchman learning to read German, the German 

 learning to read Hungarian, is constantly liable to pronounce old fa- 

 miliar letters in the old familiar manner. It will always be neces- 

 sary to learn pronunciation by the ear, not by the eye. But whilst 

 I admit that learners may be tempted to pronounce the Roman 

 characters not in the Indian but in the European manner, I ask are 

 not the Indian characters themselves pronounced at first in the Eu- 

 ropean manner by every learner ? As soon as the learner discovers 

 that such and such Indian characters are the equivalents of such 

 and such European ones, he forthwith supposes that the sounds 

 also must be identical. It is only after many errors that he learns 

 that, whatever the literary symbols of a language may be, the sounds 

 of the language are to be learnt exclusively from the speech of 

 the people. 



d. The last objection, and the most serious is, that there are 

 sounds in the Indian languages for which the Roman character 

 has no equivalent letters. This is undoubtedly a valid objection, 

 so far as it goes, and there are many persons who would give in 

 their adhesion to the scheme, if only they thought that this diffi- 

 culty could be satisfactorily disposed of. Supposing it to be prac- 



