Corps de l’article

When practitioners and theorists of translation talk about translatability or untranslatability, what is uppermost in their minds is often the semantic level. They generally agree that, while the denotations of the original readily lend themselves to the process of “carrying over or across,” which is the meaning of transferre (“to carry over or across”), the Latin word from which the English translate is derived (Simpson 1968: 611), connotations, associations, and nuances relating to the semantic level prove more elusive. Yet, compared with musicality, all semantic features are easy fare.

By “musicality,” I mean phonological features that contribute to the sum total of the original’s meaning. It can be “music” in the conventional sense of the word, with all those qualities associated with what is the mellifluous, melodious, or pleasing to the ear; at the same time, it can encompass phonological features that are deliberately cacophonic or discordant, employed by the addresser to get his message across more effectively, more memorably, or with greater emphasis. When a Frenchman says, “Quand on est cantonais, on est né à Canton” (literally “When one is Contonese, one was born in Canton,” meaning “Cantonese were born in Canton”), the communication of the message depends as much on the echoing of the same sounds in “Quand on est cantonais, on est nécanton” as on the signifiés of the individual lexical items on the semantic level. When the target language has no corresponding phonological items to reproduce the same effect, the translation cannot be considered to have adequately carried over the total meaning of the message. Similarly, when a Spaniard says, “Del dicho al hecho hay mucho trecho” (literally “From the word to the deed, there is much distance,” meaning “Its is easier said than done”), a large part of the meaning is conveyed by the interplay of phonological features, such as the repetition of the vowels “e” and “o” and of the voiceless post-aveolar affricate “–ch-” (/tʃ/), which requires more physical effort on the part of the addresser when making the utterance, suggesting strenuousness and conditioning the addresser in such a way that he is compelled, through kinaesthesia, to feel the distance (“trecho”) between the “word” (“dicho”) and the “deed” (“hecho”). In the English translation, there are no such phonological features. Saying it aloud without much physical effort, one is aware of a suggestiveness opposite to that of the original: instead of feeling “much distance” (“mucho trecho”), one feels “little distance” (“poco trecho”).

With respect to the significance of phonological features in the communication of a message, Italian can perhaps provide us with the largest number of everyday examples. Thumb through an Italian dictionary, and we will find saying after saying in which the communication of the message depends heavily – or even hinges – on phonological features: “Poca brigata, vita beata” (literally “Fewer people, blessed life,” meaning “Small company makes for happy life”); “Patti chiari, amici cari” (literally “Clear agreement, dear friends,” meaning “Terms clearly defined in advance prevents quarrels later”); “Cielo a pecorelle, acqua a catinelle” (literally “A sky with fleeces, water in washhand basins,” meaning “A sky with fleece-clouds forebodes plenty of rain, that is, raining cats and dogs”); “Moglie e buoi dei paesi tuoi” (literally “Wife and oxen of your hometown,” meaning “If you must marry, marry a woman of your hometown”); “Aprile, dolce dormire”(literally “April, sweet sleep,” meaning “In April, one sleeps best”); “Il soverchio rompe il coperchio” (literally “Excess breaks the cover,” meaning “Excess is harmful”); “Chi di spada ferisce, di spada perisce” (He who hurts with the sword dies by the sword); “Casa mia, casa mia, per piccina che tu sia, tu mi sembri una badia” (literally “My home, my home, small as you are, to me you seem an abbey,” meaning “Home, sweet home”)…. With a master translator, it may be possible to reproduce in the target language certain phonological features approximating those of the original; yet, no matter how ingenious the translator is, the success is bound to be partial because no two languages are phonologically identical.

For this reason, therefore, of all the problems posed by a source-language text, musicality is certainly one of the most recalcitrant – perhaps the most recalcitrant – a translator has to face. When the writer of the source-language text confines himself to linguistic devices on the semantic level, translation can be relatively easy; when he draws heavily on phonological features of the language with which he is working, especially in longer stretches of discourse, the abilities of the translator can be taxed to the full. With masters of language like Dante and Shakespeare, the problem can border on the completely insurmountable. Take the following passage from Dante’s Divine Comedy, for example:

A l’alta fantasia qui mancb possa;

 ma già volgeva il mio disio e il velle,

 sì come rota ch’igualmente è mossa,

l’amor che move il sole e l’altre stelle.

Dante 1960: 798

As the climax and concluding lines of the Italian masterpiece, the quotation works powerfully not only on the semantic level, but also on the phonological level, making full use of the interplay of vowels (“a,” “i,” “o,” “e”), especially in the last line, where “a,” “o,” and “e” set off a pleasing resonation, echoing one another in a rich, mellifluous pattern of sounds that drives home through its suggestiveness the harmony which the Pilgrim has reached with the Holy Trinity in the Empyrean.

For a translator, this kind of musicality, which is musicality of the first order, is the most formidable,[1] especially in respect of interfamily translation, that is, translation between two different language families, such as the Indo-European and the Sino-Tibetan. In translating the four lines quoted above into Chinese, I have retained Dante’s terza rima, and re-created a metre of my own to complement what is conveyed by the semantic signifiés in the original:

forme: 012995aro001n.pngHuang 2003 Tiantangpian: 511

However, because of the vast gap between Italian and Chinese, which belongs respectively to the Indo-European and the Sino-Tibetan family, a large part of the original’s musicality is lost beyond redemption, particularly with respect to the last line of the translation, in which one can no longer hear the satisfying echoing of the vowels so consummately deployed by Dante.

Moving from interfamily to intrafamily translation, one will see the linguistic gap narrowing:

 Here force failed my high fantasy; but my

desire and will were moved already – like

a wheel revolving uniformly – by

 the Love that moves the sun and the other stars.

Mandelbaum Paradiso 1984: 303

In the English translation, one can hear an echo, albeit faint and remote, of the original’s music in “fantasy” (for “fantasia”), “will” (for “velle”), “were moved” (for “move”), “revolving” (for “mossa,” but echoing “volgeva”), and “stars” (for “stelle”), which become more audible when played against the frequency of Chinese. However, compared with those members of the same family that have a closer affinity with Italian, English is still not too competent a “performer.” This is because English is still one or two degrees removed from Italian: while Italian belongs to the Italic branch, English belongs to the German.

For the same reason, German, another member of the German branch, fares no better:

Der hohen fantaseîn brach hie verträuen:

 Doch schon bewog mein >sehnen< und mein >gerne<

 Rade geleich, das rollt auf bahnen neuen:

Minne, di wiegt die sonne und andern sterne.

Borchardt 1967: 467

To be sure, one can hear a faint echo of the original in “fantaseîn” (for “fantasia”) “sonne” (for “sole”), and “sterne” (for “stelle”), but the grating “brach,” “doch,” “geleich,” and, particularly, the guttural “rade” (for “rota”) have all but destroyed the original’s mellifluousness and harmony.

By looking closely at the German version, one can see that the “infelicitousness” is due not to some personal failing on Borchardt’s part; working within the “limitations” of German,[2] Borchardt has proved to be a competent translator. However, because of the great difference between the German and Italian phonological systems, the translator has no alternative but to make the best of an impossible job. This point will be better illustrated if we look at another German version of the same lines:

Die hohe Bildkraft mußte hier versagen,

 Doch schon bewegte meinen Wunch und Willen,

 So wie ein Rad in gleichender Bewegung

Die Liebe, die beweget Sonn und Sterne.

Gmelin 1949 Das Paradies: 401

The harsh-sounding “Bilfkraft,” “versagen,” and “Rad,” the voiced alveolar fricative /z/ in “versagen,” “So,” and “Sonn,” the voiceless velar fricative /x/ in “Doch,” the voiceless post-aveolar fricative /ʃ/ in “schon,” “Wunsch,” “gleichender,” and “Sterne,” and the voiced labio-dental fricative /v/ in “Wunsch” and “Willen” all “conspire” effectively to “sabotage” the original melody.[3] It is clear, then, that any inability on Borchardt’s part to replay the original music is due to linguistic rather than personal factors.

When one enters the Italic branch, one’s ears will immediately be greeted by less “alien” sounds, sounds in which one can recognize a closer kinship with Italian. Take Jacqueline Risset’s French translation of the same lines:

 Ici la haute fantaisie perdit sa puissance;

mais déjà il tournait mon désir et vouloir

tout comme roue également poussée,

 l’amour qui meut le soleil et les autres étoiles.

Risset 1990 Le Paradis: 315

In addition to “fantasie” (for “fantasia”), which plays more or less the same role as the English “fantasy” and the German “fantaseîn,” “désir”(for “disio”), vouloir (for “velle”), “également” (for “igualmente”), “l’amour” (for “l’amor”), and “soleil” (for “sole”) have moved much closer to the Italian music than do their counterparts in the English and German versions. With exactly the same pronunciation as its Italian counterpart “e” (/e/), the French conjunction “et” (/e/) reminds the reader that he is listening to a gamut with a frequency not too dissimilar from that of the original. If one reads the English, German, and French versions aloud, one will also be aware that, as a whole, the French version is softer, more “feminine,” and more in tune with the mood and condition to which Dante the Pilgrim has risen, due, no doubt, to the fact that both Italian and French are Romance languages descended from Latin.

Nevertheless, despite the more competent “performance” of French in replaying Dante’s music, it is in Spanish and Latin that one finds the closest frequency to that of Dante’s celestial music. Take a Spanish version of the four quoted lines:

Ya mi alta fantasía fué impotente;

 mas cual rueda que gira por sus huellas,

 el mío y su querer movió igualmente,

el amor que al sol mueve y las estrellas.

Mitre 1938: 307

The similarity between Spanish and Italian can instantly be recognized even by someone who speaks neither language: “alta,” “igualmente,” and “amor” have the same spelling as their Italian counterparts; “fantasía” (for “fantasia”) and “mío” (for “mio”) differ from the Italian words only in the presence of an accent; “movió,” “sol,” and “mueve,” though spelt differently, can readily be linked to the Italian “move” and “sole.” Because of this close kinship, the Spanish music is just as mellifluous and harmonious as the Italian original, suggesting with equal effectiveness the beatitude granted to Dante the Pilgrim, a state which is best described by the famous line spoken by Piccarda Donati in the sphere of the Moon: “E ’n la sua volondade è nostra pace” (Dante 1960: 690).[4]

To see whether the musicality reproduced by Mitre is accidental or due to some other factors, we can look at another version in the same language. Published in 1968, the Spanish version, in prose, is by Arce:

Pero a mi fantasía faltó fuerza; y ya deseo y voluntad giraban como rueda con uniformidad, impulsados por el Amor que mueve al sol y al las demás estrellas.

Arce 1968: 451

Words like “pero,” “a,” “mi,” “fantasía,” “deseo,” “voluntad,” “giraban,” “como,” “rueda,” “uniformidad,” “impulsados,” “por,” “el,” “Amor,” “que,” “mueve”…readily remind the reader that he is listening to almost the same piece of music as the Divine Comedy. To those who read both German and Spanish, the word “rueda” (meaning “wheel,” for the Italian “rota”) is especially instructive: the gliding from “u” to “e,” which is highly mellifluous in the context, contrasts sharply with the German “Rad,” which, being guttural and harsh, fails to fit in with the celestial music intended by Dante. Working with German, which has a phonetic pattern suited to a different kind of music, a translator rendering the last canto of the Paradiso has really very little room for manoeuvre.[5] Working with Spanish, a translator has a phonological system made up of very similar musical notes as Dante’s Italian; when he is called upon to sing the heavenly music of the last canto of the Paradiso, he will be able to rise to the challenge with little effort.

With Latin, the story is more or less the same, except that, while Spanish is related to Italian like one sibling to another, Latin is related to Italian like a father to his child. For this reason, one can expect to hear music of a frequency very similar to that of Italian when the great lines of the Paradiso are translated into Latin:

Alte phantasie hic defecit posse;

 Sed iam involvebat meum velle, [divinum] velle,

 Sicut rota que equaliter movetur,

Amor qui movet solem et alias stellas.

Serravalle 1891: 1213

Except for a few words like “hic” (here), “iam” (already), and “sicut” (just as), all the other words are either similar to or identical with their counterparts in the Italian original. As Latin is Italian’s father, it is just natural to see in it the image of the son.

In the preceding paragraphs, the discussion has centred on Italian as the source language, which has made English and German appear to be incompetent candidates for the task of reproducing musicality. When the source language is English, we will see the situation reversed: in translating musicality in the English-Italian direction, it will be the turn of the Italian language to have its “inadequacies” exposed.

For the sake of a more objective comparison, let us look at the work of Shakespeare, a poet comparable to Dante in stature.[6] In one of his great tragedies, Macbeth, the Bard of Avon has three witches prepare a charm with all sorts of horrid ingredients that can give the audience the creeps. The setting is a cavern. When the three witches enter, there is thunder, as is the case with Act 1, Scene 1, Act 1, Scene 3, and Act 3, Scene 5. The world evoked is the antipode of the Paradiso:

Second Witch.

Fillet of a fenny snake,

In the cauldron boil and bake;

Eye of newt, and toe of frog,

Wool of bat, and tongue of dog,

Adder’s fork, and blind-worm’s sting,

Lizard’s leg, and howlet’s wing,

For a charm of powerful trouble,

Like a hell-broth boil and bubble….

Third Witch.

Scale of dragon, tooth of wolf,

Witches’ mummy, maw and gulf

Of the ravin’d salt-sea shark,

Root of hemlock digg’d i’ the dark....

Shakespeare 1974: 860

The moment they are spoken, these lines will start working simultaneously on two levels: the semantic and the phonological. On the semantic level, the signifiants will, through the signifiés, evoke a complex of images, associations, and connotations that complement and reinforce one another, thereby contributing to the eerie and bizarre atmosphere intended by Shakespeare. Phonologically, the sounds, from whole lines down to the single phoneme, co-operate to suggest what the semantic units are powerfully evoking. Take the sharp monosyllabic words “snake” (/sneɡk/), “bake” (/beɡk/), “frog” (/frɒl/), “bat” (/bæt/), “tongue” (/tʌŋ/), “dog” (/dɒɡ/), “fork” (/fɔːk/), “sting” (/stɡŋ/), “leg” (/leɡ/), “shark” (/ʃɑːk/), and “dark” (/dɑːk/). Apart from the vowels, the velar plosives /k/ and /ɡ/, the alveolar plosive /t/, and the velar nasal /ŋ/, together with the strenuous “-zard” (/zəd/) in “lizard” (/ˈlɪzəd/), the explosive “bʌb-” (/bžb/) in “bubble” (/ˈbʌb(ə)l/), the open and jarring “-lock” (/lɒk/) in “hemlock” (/ˈhemlɒk/), and the sharp, short “dig-” (/dɪɡ/) in “digg’d (/dɪɡd/), all help the audience “hear” the eerie, the bizarre, the mysterious, and the horrid;[7] what was originally abstract has become something concrete, something clearly audible. In his The Four Quartets: Little Gidding,” Eliot (1963: 221)) describes “every phrase / And sentence that is right” as one in which “every word is at home, /Taking its place to support the others, /…The complete consort dancing together.” After listening to the quotation from Shakespeare, one is tempted to conclude that it must have been the kind of “phrase” and “sentence” Eliot had in mind when he was writing the above lines.

As the greatest English poet, Shakespeare is able to exploit the phonological system of the English language more skilfully than any other English poet. In writing the witches’ lines, he must have had complete mastery of his tools, namely, all the sounds of the English language available to him. When called upon to translate the musicality created by these sounds into a language made up of a different phonological system, a translator may employ phonological features that appear to correspond to those of the original, using poetic devices common to both the source and target languages, such as rhyme and alliteration.[8] However, since the target language is made up of a different phonological system, its “limitations” will immediately become obvious in the translation:


Bolli e cuoci nella pentola

tu, filetto di serpente;

ramarro occhiuto, pollici di rana,

vampiro peloso e lingua di cane.

Forca di serpe, aculeo d’orbetto

zampe ed ali, ramarri e civette,

magie potenti, gorgogli del male,

bollite tutti nel brodo infernale....


Dente di lupo, dragone a scaglie,

mummia di strega, gonfie frattaglie

del pescecane d’acqua salata,

cicuta al buio disradicata....

Gassman, 1983: 92-93

In terms of sound effect, the Italian version makes a widely different impact on the listener: the rich pattern of vowels (“o,” “i,” “u,” “e,” and “a”) resonates across the lines and echo one another in a music that is full and mellow, deviating drastically from Shakespeare’s music: there is no longer the discordance that works in unison with the semantic level;[9] the Stravinskian cacophony in the original has given way to a Mozartian harmony. While Mozartian harmony has its proper place in Canto 33 of Dante’s Paradiso, with respect to the Macbeth scene, it is just too “beautiful.”

From the above discussion, it can be seen that, in respect of musicality, intrafamily translation is much easier than interfamily translation. In interfamily translation, musicality of a different kind may be created by the translator’s ingenious handling of a different phonological system, but it is no longer musicality of the same gamut, much less musicality of the same notes. When one moves from interfamily to intrafamily translation, a higher degree of phonic fidelity can be achieved: the closer the affinity, the higher the phonic fidelity. Within the same family of languages, it can be further seen that intrabranch translation, as is the case with Italian-Spanish translation, yields musicality of an even higher degree of phonic fidelity. Lastly, if one wishes to make even finer distinctions, one can say that, within the same branch, musicality lends itself most readily to translation between languages with the closest kinship, such as Italian and Spanish.

In Florence, there is a church called Santa Maria Novella. Even to one who is not a native speaker of Italian, the sound of the name gets across like a melody, working as it does on the interplay of a pleasing pattern of vowels. Say it in English: “New St. Mary” (/njuː/ /sənt/ /ˈmeəri/), or, worse still, in Chinese: forme: 012995aro002n.png (/ɕin/ /ʂəŋ/ /ma/ /li/ /ia/), and the music vanishes into thin air.[10]Così la neve al sol si disigilla; / così al vento ne le foglie levi / si perdea la sentenza di Sibilla” (Dante 1960: 796) – “Thus the snow loses its imprint in the sun; thus in the wind on the light leaves the Sibyl’s oracle was lost” (Sinclair 1971: 481). Say it in Spanish: “Santa Maria Novela,” and the music remains intact. In the light of this difference, therefore, translators working between languages that belong to two different families or branches can only hope to produce a musical ersatz.