with English usage, this sixth of the translating process becomes an exercise in the amendment of unidiomatic English. But the result is little more profitable than in the other case, for two reasons: One is, that the preliminary mustering of the main English words, and the puzzling over the constructions, absorb so much of the pupil's attention that the finished English rendering is little thought of. The other and principal reason is, that the English version, where attended to—as it must be in a good translation—is remembered only in connection with the Latin, and is not readily remembered when a natural object has to be described. A good translator has no facility in original composition, unless he has practised the art of composition by itself: the words used in translation do not occur as symbols for natural things, but only as equivalents for the Latin expressions. It was quite to be expected that Mr. Dasent would find good Latin scholars "utterly incapable of expressing themselves in their own language." The wonder would be if they found time to learn how to lay out felicitously their own thoughts and sentiments, while they acquired the art of felicitously translating the more or less skilful expression of the thoughts and sentiments of others.
Does the classical scholar acquire an abundance of words or skill in selecting; the right words? In translating, he must cast about over various words of cognate meaning for the word that will suit the passage. Does he thereby learn a wide command of synonymes, and a dexterity in seizing the aptest word to convey his meaning? He learns a command of synonymes, undoubtedly. But where does he get them? Not in Latin; but in his own remembered store, and in the pages of the English lexicographer, his starting-point being some English equivalent of a Latin word. As a learner of synonymes, he does no more, and can make no more progress, than the non-classical pupil that ransacks his memory and his dictionary with a similar object. He does not learn to seize the aptest words to convey his meaning. What he learns is, to seize the aptest words to represent particular Latin words in particular contexts—an entirely different thing. Mr. Dasent's evidence on this matter is very pointed. It is his express complaint of good Latin scholars, that "they have no choice of words" in English.
Does classical composition train in English composition? In translating English into Latin or Greek, the pupil must acquire a certain familiarity with a certain number of English words. If the English be good, so much the better for the pupil. If he is taught to twist and turn it about, so as to make idiomatic Latin out of idiomatic English, so much the better for him. But the advantage is no greater than he would have by keeping passages of good English some time in his memory for any purpose whatsoever.
Let us now consider what can be made of English as an instrument of education.
