which he has had to encounter. These have been far greater than he would have anticipated; nor is he at all sanguine that he has succeeded in overcoming them. Experience has made him feel that a translation, like a picture, is dependent for its effect on very minute touches; and that it is a work of infinite pains, to be returned to in many moods and viewed in different lights.
I. An English translation ought to be idiomatic and interesting, not only to the scholar, but to the unlearned reader. Its object should not simply be to render the words of one language into the words of another or to preserve the construction and order of the original; – this is the ambition of a schoolboy, who wishes to show that he has made a good use of his Dictionary and Grammar; but is quite unworthy of the translator, who seeks to produce on his reader an impression similar or nearly similar to that produced by the original. To him the feeling should be more important than the exact word. He should remember Dryden's quaint admonition not to 'lacquey by the side of his author, but to mount up behind him.' (Dedication to the Aeneis.) He must carry in his mind a comprehensive view of the whole work, of what has preceded and of what is to follow, – as well as of the meaning of particular passages. His version should be based, in the first instance, on an intimate knowledge of the text; but the precise order and arrangement of the words may be left to fade out of sight, when the translation begins to take shape. He must form a general idea of the two languages, and reduce the one to the terms of the other. His work should be rhythmical and varied, the right admixture of words and syllables, and even of letters, should be carefully attended to; above all, it should be equable in style. There must also be quantity, which is necessary in prose as well as in verse: clauses, sentences, paragraphs, must be in due proportion. Metre and even rhyme may be rarely admitted; though neither is a legitimate element of prose writing, they may help to lighten a cumbrous expression (Symp.). The translation should retain as far as possible the characteristic qualities of the ancient writer – his freedom, grace, simplicity, stateliness, weight, precision; or the best part of him will be lost to the English reader. It should read as an original work, and should also be the most faithful transcript which can be made of the language from which the translation is taken, consistently with the first requirement of all, that it be English. Further, the translation being English, it should also be perfectly intelligible in itself without reference to the Greek, the English being really the more lucid and exact of the two languages. In some respects it may be maintained that ordinary English writing, such as the newspaper article, is superior to Plato: at any rate it is couched in language which is very rarely obscure. On the other hand, the greatest writers of Greece, Thucydides, Plato, Aeschylus, Sophocles, Pindar, Demosthenes, are generally those which are found to be most difficult and to diverge most widely from the English idiom. The translator will often have to convert the more abstract Greek into the more concrete English, or vice versa, and he ought not to force upon one language the character of another. In some cases, where the order is confused, the expression feeble, the emphasis misplaced, or the sense somewhat faulty, he will not strive in his rendering to reproduce these characteristics, but will re-write the passage as his author would have written it at first, had he not been 'nodding'; and he will not hesitate to supply anything which, owing to the genius of the language or some accident of composition, is omitted in the Greek, but is necessary to make the English clear and consecutive.
It is difficult to harmonize all these conflicting elements. In a translation of Plato what may be termed the interests of the Greek and English are often at war with one another. In framing the English sentence we are insensibly diverted from the exact meaning of the Greek; when we return to the Greek we are apt to cramp and overlay the English. We substitute, we compromise, we give and take, we add a little here and leave out a little there. The translator may sometimes be allowed to sacrifice minute accuracy for the sake of clearness and sense. But he is not therefore at liberty to omit words and turns of expression which the English language is quite capable of supplying. He must be patient and self-controlled; he must not be easily run away with. Let him never allow the attraction of a favourite expression, or a sonorous cadence, to overpower his better judgment, or think much of an ornament which is out of keeping with the general character of his work. He must ever be casting his eyes upwards from the copy to the original, and down again from the original to the copy (Rep.). His calling is not held in much honour by the world of scholars; yet he himself may be excused for thinking it a kind of glory to have lived so many years in the companionship of one of the greatest of human intelligences, and in some degree, more perhaps than others, to have had the privilege of understanding him (Sir Joshua Reynolds' Lectures: Disc. xv.).
There are fundamental differences in Greek and English, of which some may be managed while others remain intractable. (1). The structure of the Greek language is partly adversative and alternative, and partly inferential; that is to say, the members of a sentence are either opposed to one another, or one of them expresses the cause or effect or condition or reason of another. The two tendencies may be called the horizontal and perpendicular lines of the language; and the opposition or inference is often much more one of words than of ideas. But modern languages have rubbed off this adversative and inferential form: they have fewer links of connection, there is less mortar in the interstices, and they are content to place sentences side by side, leaving their relation to one another to be gathered from their position or from the context. The difficulty of preserving the effect of the Greek is increased by the want of adversative and inferential particles in English, and by the nice sense of tautology which characterizes all modern languages. We cannot have two 'buts' or two 'fors' in the same sentence where the Greek repeats (Greek). There is a similar want of particles expressing the various gradations of objective and subjective thought – (Greek) and the like, which are so thickly scattered over the Greek page. Further, we can only realize to a very imperfect degree the common distinction between (Greek), and the combination of the two suggests a subtle shade of negation which cannot be expressed in English. And while English is more dependent than Greek upon the apposition of clauses and sentences, yet there is a difficulty in using this form of construction owing to the want of case endings. For the same reason there cannot be an equal variety in the order of words or an equal nicety of emphasis in English as in Greek.
(2) The formation of the sentence and of the paragraph greatly differs in Greek and English. The lines by which they are divided are generally much more marked in modern languages than in ancient. Both sentences and paragraphs are more precise and definite – they do not run into one another. They are also more regularly developed from within. The sentence marks another step in an argument or a narrative or a statement; in reading a paragraph we silently turn over the page and arrive at some new view or aspect of the subject. Whereas in Plato we are not always certain where a sentence begins and ends; and paragraphs are few and far between. The language is distributed in a different way, and less articulated than in English. For it was long before the true use of the period was attained by the classical writers both in poetry or prose; it was (Greek). The balance of sentences and the introduction of paragraphs at suitable intervals must not be neglected if the harmony of the English language is to be preserved. And still a caution has to be added on the other side, that we must avoid giving it a numerical or mechanical character.
(3) This, however, is not one of the greatest difficulties of the translator; much greater is that which arises from the restriction of the use of the genders. Men and women in English are masculine and feminine, and there is a similar distinction of sex in the words denoting animals; but all things else, whether outward objects or abstract ideas, are relegated to the class of neuters. Hardly in some flight of poetry do we ever endue any of them with the characteristics of a sentient being, and then only by speaking of them in the feminine gender. The virtues may be pictured in female forms, but they are not so described in language; a ship is humorously supposed to be the sailor's bride; more doubtful are the personifications of church and country as females. Now the genius of the Greek language is the opposite of this. The same tendency to personification which is seen in the Greek mythology is common also in the language; and genders are attributed to things as well as persons according to their various degrees of strength and weakness; or from fanciful resemblances to the male or female form, or some analogy too subtle to be discovered. When the gender of any object was once fixed, a similar gender was naturally assigned to similar objects, or to words of similar formation. This use of genders in the denotation of objects or ideas not only affects the words to which genders are attributed, but the words with which they are construed or connected,