On Translating Homer

by Matthew Arnold; Francis William Newman

Mr Clough’s hexameters are, as I have just said, too rough and irregular; and indeed a good model, on any considerable scale, of this metre, the English translator will nowhere find. He must not follow the model offered by Mr Longfellow in his pleasing and popular poem of Evangeline; for the merit of the manner and movement of Evangeline, when they are at their best, is to be tenderly elegant; and their fault, when they are at their worst, is to be lumbering; but Homer’s defect is not lumberingness, neither is tender elegance his excellence. The lumbering effect of most English hexameters is caused by their being much too dactylic[^[30]]; the translator must learn to use spondees freely. Mr Clough has done this, but he has not sufficiently observed another rule which the translator cannot follow too strictly; and that is, to have no lines which will not, as it is familiarly said, read themselves. This is of the last importance for rhythms with which the ear of the English public is not thoroughly acquainted. Lord Redesdale, in two papers on the subject of Greek and Roman metres, has some good remarks on the outrageous disregard of quantity in which English verse, trusting to its force of accent, is apt to indulge itself. The predominance of accent in our language is so great, that it would be pedantic not to avail oneself of it; and Lord Redesdale suggests rules which might easily be pushed too far. Still, it is undeniable that in English hexameters we generally force the quantity far too much; we rely on justification by accent with a security which is excessive. But not only do we abuse accent by shortening long syllables and lengthening short ones; we perpetually commit a far worse fault, by requiring the removal of the accent from its natural place to an unnatural one, in order to make our line scan. This is a fault, even when our metre is one which every English reader knows, and when we can see what we want and can correct the rhythm according to our wish; although it is a fault which a great master may sometimes commit knowingly to produce a desired effect, as Milton changes the natural accent on the word Tiresias in the line:

And Tíresias and Phineus, prophets old;

and then it ceases to be a fault, and becomes a beauty. But it is a real fault, when Chapman has:

By him the golden-throned Queen slept, the Queen of Deities;

for in this line, to make it scan, you have to take away the accent from the word Queen, on which it naturally falls, and to place it on throned, which would naturally be unaccented; and yet, after all, you get no peculiar effect or beauty of cadence to reward you. It is a real fault, when Mr Newman has:

Infatuate! O that thou wert lord to some other army—

for here again the reader is required, not for any special advantage to himself, but simply to save Mr Newman trouble, to place the accent on the insignificant word wert, where it has no business whatever. But it is still a greater fault, when Spenser has (to take a striking instance):

Wot ye why his mother with a veil hath covered his face?

for a hexameter; because here not only is the reader causelessly required to make havoc with the natural accentuation of the line in order to get it to run as a hexameter; but also he, in nine cases out of ten, will be utterly at a loss how to perform the process required, and the line will remain a mere monster for him. I repeat, it is advisable to construct all verses so that by reading them naturally—that is, according to the sense and legitimate accent,—the reader gets the right rhythm; but, for English hexameters, that they be so constructed is indispensable.

If the hexameter best helps the translator to the Homeric rapidity, what style may best help him to the Homeric plainness and directness? It is the merit of a metre appropriate to your subject, that it in some degree suggests and carries with itself a style appropriate to the subject; the elaborate and self-retarding style, which comes so naturally when your metre is the Miltonic blank verse, does not come naturally with the hexameter; is, indeed, alien to it. On the other hand, the hexameter has a natural dignity which repels both the jaunty style and the jog-trot style, to both of which the ballad-measure so easily lends itself. These are great advantages; and, perhaps, it is nearly enough to say to the translator who uses the hexameter that he cannot too religiously follow, in style, the inspiration of his metre. He will find that a loose and idiomatic grammar—a grammar which follows the essential rather than the formal logic of the thought—allies itself excellently with the hexameter; and that, while this sort of grammar ensures plainness and naturalness, it by no means comes short in nobleness. It is difficult to pronounce, certainly, what is idiomatic in the ancient literature of a language which, though still spoken, has long since entirely adopted, as modern Greek has adopted, modern idioms. Still one may, I think, clearly perceive that Homer’s grammatical style is idiomatic,—that it may even be called, not improperly, a loose grammatical style[^[31]]. Examples, however, of what I mean by a loose grammatical style, will be of more use to the translator if taken from English poetry than if taken from Homer. I call it, then, a loose and idiomatic grammar which Shakspeare uses in the last line of the following three: