we come now to that great task which longfellow, after an early experiment, had dropped for years, and which he resumed after his wife’s death, largely for the sake of an absorbing occupation. eighteen years before, november 24, 1843, he had written to ferdinand freiligrath that he had translated sixteen cantos of dante, and there seems no reason to suppose that he had done aught farther in that direction until this new crisis. after resuming the work, he translated for a time a canto as each day’s task, and refers to this habit in his sonnet on the subject, where he says:—

“i enter here from day to day,

and leave my burden at this minster gate.”

the work was not fully completed until 1866, and was published in part during the following year.

the whole picture of the manner in which the work was done has long been familiar to the literary world, including the pleasing glimpse of 226 the little circle of cultivated friends, assembled evening after evening, to compare notes and suggest improvements. for many years this was regarded by students and critics as having been almost an ideal method for the production of a great work, and especially of a translation,—a task where there is always the original text at hand for reference. as time has gone on, however, the admiration for the completed work has gradually been mingled with a growing doubt whether this species of joint production was on the whole an ideal one, and whether, in fact, a less perfect work coming from a single mind might not surpass in freshness of quality, and therefore in successful effort, any joint product. longfellow had written long before to freiligrath that making a translation was “like running a ploughshare through the soil of one’s mind,”[92] and it would be plainly impossible to run ploughshares simultaneously through half a dozen different minds at precisely the same angle. the mind to decide on a phrase or an epithet, even in a translation, must, it would seem, be the mind from which the phrase or statement originally proceeded; a suggestion from a neighbor might sometimes be most felicitous, but quite as often more tame and guarded; and the influence of several neighbors collectively 227 might lie, as often happens in the outcome of an ordinary committee meeting, rather in the direction of caution than of vigor. longfellow’s own temperament was of the gracious and conciliatory type, by no means of the domineering quality; and it is certainly a noticeable outcome of all this joint effort at constructing a version of this great world-poem, that one of the two original delegates, professor norton, should ultimately have published a prose translation of his own. it is also to be observed that professor norton, in the original preface to his version, while praising several other translators, does not so much as mention the name of longfellow; and in his list of “aids to the study of the ‘divine comedy’” speaks only of longfellow’s notes and illustrations, which he praises as “admirable.” even lowell, the other original member of the conference, while in his “dante” essay he ranks longfellow’s as “the best” of the complete translations, applies the word “admirable” only to those fragmentary early versions, made for longfellow’s college classes twenty years before,—versions which the completed work was apparently intended to supersede.

far be it from me to imply that any disloyalty was shown on the part of these gentlemen either towards their eminent associate or towards the work on which they had shared his labors; it is 228 only that they surprise us a little by what they do not say. it may be that they do not praise the longfellow version because they confessedly had a share in it, yet this reason does not quite satisfy. nothing has been more noticeable in the popular reception of the completed work than the general preference of unsophisticated readers for those earlier translations thus heartily praised by lowell. there has been a general complaint that the later work does not possess for the english-speaking reader the charm exerted by the original over all who can read italian, while those earlier and fragmentary specimens had certainly possessed something of that charm.

those favorite versions, it must be remembered, were not the result of any co?perated labor, having been written by professor longfellow in an interleaved copy of dante which he used in the class room. they were three in number, all from the “purgatorio” and entitled by him respectively, “the celestial pilot,” “the terrestrial paradise,” and “beatrice.” they were first published in “voices of the night” (1839), and twenty-eight years had passed before the later versions appeared. those twenty-eight years had undoubtedly enhanced in width and depth mr. longfellow’s knowledge of the italian language; their labors and sorrows had matured 229 the strength of his mind; but it is not so clear that they had not in some degree diminished its freshness and vivacity, nor is it clear that the council of friendly critics would be an influence tending to replace just those gifts.

if a comparison is to be made between the earlier and later renderings, the best way would doubtless be to place them side by side in parallel columns; and while it would be inappropriate to present such a comparison here on any large scale, it may be worth while to take a passage at random to see the effect of the two methods. let us take, for instance, a passage from “purgatorio,” canto xxx. lines 22 and 23. they are thus in the original:—

“io vidi già nel cominciar del giorno

la parte oriental tutta rosata,

e l’altro ciel di bel sereno adorno.”

the following is longfellow’s translation of 1839, made by the man of thirty-two:—

“oft have i seen, at the approach of day,

the orient sky all stained with roseate hues,

and the other heaven with light serene adorned.”

the following is the later version, made by the man of sixty, after ample conference with friendly critics:—

“ere now have i beheld, as day began,

the eastern hemisphere all tinged with rose,

and the other heaven with fair serene adorned;”

230

i do not see how any english-speaking reader could hesitate for a moment in finding a charm far greater in the first version than in the second, or fail to recognize in it more of that quality which has made the name of dante immortal. if this be true, the only question that can be raised is whether this advantage has been won by a sacrifice of that degree of literalness which may fairly be demanded of a translation in poetic form. perfect and absolute literalness, it must be remembered, can only be expected of a prose version, and even after the most perfect metrical translation a prose version may be as needful as ever. let us consider for a moment the two examples as given above. it may be conceded at the outset that the adverb già is more strictly and carefully rendered by “ere” than by “oft,” but the difference is not important, as any one old enough to describe a daybreak has undoubtedly seen more than one. the difference between “the approach of day” and “as day began” is important, since the last moment of the approach coincides with the first moment of the beginning. in the second line, “la parte oriental” is both more literally and more tersely rendered by “the orient sky,” than by the more awkward expression “the eastern hemisphere,” unless it be claimed that “sky” does not sufficiently recognize the earth as seen 231 in the view; to which it may justly be replied that the word “hemisphere,” if applied only to the earth, equally omits the sky, and the two defects balance each other. “tinged with rose” is undoubtedly a briefer expression for the untranslatable “rosata” than “stained with roseate hues” would be. the last line of the three finds an identical rendering in the two versions, and while “bel sereno” is more literally rendered by “fair serene” than by “light serene,” yet the earlier phrase has the advantage of being better english, serene being there used as an adjective only, whereas in the later translation it is used as a noun, a practice generally regarded as obsolete in the dictionaries. even where the word is thus employed, they tell us, it does not describe the morning light, but indicates, like the french word “serein,” an evening dampness; as where daniel says, “the fogs and the serene offend us.” summing up the comparison, so far as this one example goes, it would seem that the revised version of longfellow has but very slight advantage over its predecessor, while the loss of vividness and charm is unquestionable.

to carry the test yet farther, let us compare the three lines, in their two successive versions, with the prose version of professor norton, which reads as follows: “i have seen ere now at the beginning of the day the eastern region all rosy, 232 while the rest of heaven was beautiful with fair, clear sky.” here the prose translator rightly discards the “oft” of the earlier longfellow version, but his “at the beginning” is surely nearer to the “at the approach” of the first version than to the less literal “as day began” of the second. the prose “the eastern region” conforms to the second version “the eastern hemisphere,” but surely the italian “la parte oriental” is more nearly met by “the orient sky” than by either of these heavier and more geographical substitutes, which have a flavor of the text-book. both the longfellow versions have “the other heaven,” which is a literal rendering of “l’altro ciel,” whereas “the rest of heaven” is a shade looser in expression, and “fair, clear sky” also forfeits the condensation of “light serene” or “fair serene,” of which two phrases the first seems the better, for reasons already given. on the whole, if we take professor norton’s prose translation as the standard, longfellow’s later version seems to me to gain scarcely anything upon the earlier in literalness, while it loses greatly in freshness and triumphant joyousness.

nor is this in any respect an unreasonable criticism. for what does a translation exist, after all, if not to draw us toward that quality in the original which the translator, even at his 233 best, can rarely reach? goethe says that “the translator is a person who introduces you to a veiled beauty; he makes you long for the loveliness behind the veil,” and we have in the notes to his “west-?stliche divan” the celebrated analysis of the three forms of translation. he there says, “translation is of three kinds: first, the prosaic prose translation, which is useful in enriching the language of the translator with new ideas, but gives up all poetic art, and reduces even the poetic enthusiasm to one level watery plain. secondly, the re-creation of the poem as a new poem, rejecting or altering all that seems foreign to the translator’s nationality, producing a paraphrase which might, in the primal sense of the word, be called a parody. and, thirdly, ... the highest and last, where one strives to make the translation identical with the original; so that one is not instead of the other, but in the place of the other. this sort of translation ... ‘approaches the interlinear version, and makes the understanding of the original a much easier task; thus we are led into the original,—yes, even driven in; and herein the great merit of this kind of translation lies.’”[93]

it may be doubted, however, whether longfellow, in his remarkable paper “on the translation of faust” 234 even if left to himself in making his version, could ever have reached the highest point attained by goethe, from the mere difference between the two languages with which he and his original had to deal. the charm of longfellow’s earlier versions is, after all, an english charm, and perhaps the quality of dante can no more be truthfully transmuted into this than we can transmute the charms of a spring morning into those of a summer afternoon, or violets into roses. goethe, it is well known, took for his model as to the language of “faust” the poetry of hans sachs, longfellow’s “cobbler bard;” and dante’s terse monosyllables were based upon the language of the people, which he first embodied in art. to mellow its refreshing brevities would perhaps be to destroy it, and that which mr. andrews finely says of the “faust” may be still more true of the “divina commedia,” that it “must remain, after all, the enchanted palace; and the bodies and the bones of those who in other days strove to pierce its encircling hedge lie scattered thickly about it.” so mr. w. c. lawton, himself an experienced translator from the greek, says of longfellow’s work, “his great version is but a partial success, for it essays the unattainable.”[94] but if it be possible to win this success, it is probably destined to be 235 done by one translator working singly and not in direct co?peration with others, however gifted or accomplished. every great literary work needs criticism from other eyes during its progress. nevertheless it will always remain doubtful whether any such work, even though it be a translation only, can be satisfactorily done by joint labor.

after all, when others have done their best, it is often necessary to fall back upon the french joubert for the final touch of criticism; and in his unequalled formula for translating homer, we find something not absolutely applicable to dantean translation, yet furnishing much food for thought. the following is the passage: “there will never be an endurable translation of homer, unless its words are chosen with skill and are full of variety, of freshness, and of charm. it is also essential that the diction should be as antique, as simple, as are the manners, the events, and the personages portrayed. with our modern style everything attitudinizes in homer, and his heroes seem fantastic figures which personate the grave and proud.”