What makes translators unfaithful?

“I believe in fidelity but I’m non-practising”

(Anonymous)

It is hard to meet a translator, even a good one, who is not convinced of the adage that “to translate is in part to betray.” But why?
To betray, there must obviously be two (spouses, fiancés, partners), and the partners are such within a bond that commits them beyond their actual conduct, a bond by virtue of which that conduct is judged: were there no duty and no possibility of being faithful, the bitterness suffered by the betrayed and the thrill that inflames the betrayer would vanish as well. I recall an acquaintance who, back in the seventies, would enthusiastically theorise the theoretical impossibility of betrayal. Love, he argued solemnly, is not made of contractual commitments. X and Y form a couple only in the moments when they are together; if and when Y couples with Z, the very word says so, he would add gleefully, one couple forms and the other dissolves; ergo, betrayal does not exist! The poor wretch later suffered in his own skin the marvellous precariousness of his logic, which nonetheless hints at an embarrassing question in translation itself.
On what condition can I know, while translating, that I am betraying the source language? That is the point. How, in short, can I be unfaithful if I do not have before my mind the entire meaning that I must faithfully carry across? I must, after all, understand that the meaning in question is distorted in the target language, and therefore grasp it immaculate in the source language. But if I grasp it in toto, why then can I not translate it? Perhaps because no corresponding term exists? If meaning is a transparent, self-evident entity that the mind seizes and onto which it sticks a sound and a conventional sign, why on earth should I not find in my own language a sign of that kind, an arbitrarily suitable garment? If, on the other hand, I do not grasp the meaning to be translated (or it is not even graspable), how can I realise that I am betraying it? Another translator, better than I, will certainly be able to notice any distortion. The consequence is, in every case, that a perfect mental fidelity remains the condition of translating. But no! Not even this is the consequence. A fidelity that admits of no violation is not even fidelity: with no possibility of erring, in love as in translation, the result is missing too!
A conclusion of this kind is born in philosophy classrooms, and in a certain idea of meaning that is formed there, certainly not in discussions among translators, in the obsessive search for the best version, in the jealousy with which, once one believes one has grasped the translator’s pearl, one awaits the printed version without telling anyone. What this idea of meaning is, where it comes from and what problems it generates, are all questions that here we leave to the philosophers[1].
Here we try to understand in what context, under what conditions, the perception of the “partial” untranslatability of words and sentences arises, and the consequent perception of betrayal; a betrayal sui generis, at times inevitable and without guilt, which produces inadequacy but not remorse. I think, moreover, that a few indications on this point also offer cues for discussing the abstract question dealt with in philosophy classrooms.

***

I should like to give three general and progressive indications, very briefly. The first (A) concerns terms that are falsely presumed to be untranslatable; these are in reality errors, sometimes generated by conceptual muddles in the source language, and which therefore call no conscious betrayal into play. The second (B) has to do with difficulties concerning style, and here genuine difficulties begin, difficulties that sometimes, though not always, involve the semantic aspect of words. The third (C) is the one in which a partial betrayal becomes inevitable. One aspect of (B) spills over into (C) as well, which should underline how the great difficulty of translation, always in general terms of course, concerns literature more than non-fiction. The examples offered are scanty to say the least, and it may well be that my observations appear debatable. What matters, beyond the correctness of the examples, is the kind of difficulty they reveal.
A) The examples, for (B) too, are drawn above all from the human sciences and from philosophy, especially in translations from German and English.
The first of these concerns the German verb erleben and its adjectival and nominal forms (for instance the noun Erlebnis). Besides the verb leben, which means vivere (to live), German also possesses the verb erleben, formed precisely from leben and the prefix er-, which often indicates the “success” or “fulfilment” of what is expressed by the base verb (blicken, for example, means to look, erblicken to glimpse). Now, in the translation of the great texts of the German phenomenological tradition, and in particular in the translations of the works of Edmund Husserl, it seemed that erleben had no counterpart in Italian, so much so that it was at times left untranslated in German. As though only the Italian words surrounding it could allow the reader to be prompted toward a concept that the Italian mind does not possess. If leben is vivere, then whatever could erleben mean, a successful life? In reality the busillis is rather simple to unravel. In Italian one can say both “ho vissuto bene per molti anni” and “ho vissuto momenti felici” (“I have lived well for many years” and “I have lived happy moments”). In the second case, the “happy moments” are what was once called the “internal object complement”: a subset of the set constituted by living. The moments lived are not, that is, something distinct from living, whereas, for example, the object of eating, an external object, is fortunately an entity quite distinct from eating. So while in Italian we always have one and the same verb, in German we have two, with erleben used precisely to express internal objects. Now, to the suspicion that the German erleben harboured who knows what semantic mantras, there was added, in translating Husserl, a difficulty owing to the contents expressed by the German philosopher. Husserl states in fact that what is erlebt (thus, for us, lived) are, for example, Ein Rot, ein Blau and so on, that is to say a red, a blue, and here, and in many analogous cases, the translator wavered, because to say in Italian that “one lives a red or a blue” seems to make little sense. It seems, rather, that one can have experience of a red or a blue, where the verb “esperire” precisely does not govern “internal objects”: what is experienced is not a subset of experience. And so the thought arose of putting into the translation of erleben both the word “vivere” and, at the same time, in order to make the sentence sensible when entities such as red or blue are erlebt, of adding “esperire”, which bears external objects. The hybrid that resulted was “esperire vivente” (a living experiencing, as if there were a dead one!) for the verb, and “esperienza vissuta” for the noun Erlebnis. Husserl’s phenomenological data, in Italian, are therefore at one and the same time moments of living (parts of experience) and objects of which one has experience. Had one merely translated erleben literally as vivere, the Italian reader would have understood everything Husserl means to say. Perhaps, as I believe, the difficulties of his proposal as well.

Two more quick examples. The first would reverse the situation to the advantage of our own language. Many have thought, and some still think, that the English language lacks a fundamental distinction in the cognitive domain. To the pair “conoscere/sapere”, present also in German, in French and so on, there would correspond in English the single verb “to know”! Here it would be the English philosophical tradition that betrays cognitive deficits. But the difference between “sapere” and “conoscere” is simply the difference between a verb that takes a completive clause (“sapere che …”) and one that takes a direct object (“conoscere x”), and English, obviously, distinguishes the two structures by having “to know” followed by a direct object (“I know it”) or by a completive clause (“I know that…”). If there is a difference between the two structures, it is fully expressed in English as well.
Finally, it is often suspected, the whole Italian tradition of moral philosophy seems to lack the “moral ought”. Where German distinguishes müssen from sollen, in Italian we have a single dovere (in English the package of modals is fatter still). And so the thought arose of translating müssen with dovere, and sollen with dover essere. But if the Germans have two words where we have one, why on earth should dover essere translate sein sollen and not sein müssen (where sein stands precisely for to be)? In dover essere one should already have understood that the dover is a sollen and not a müssen! In reality the supposed “moral ought”, the ought of obligation, is by no means absent from our language (perhaps it is more absent from behaviour). One of the basic meanings of sollen in German is that of rendering in “indirect speech” the imperatives of a “direct speech”. If I say to Ambrogio “per favore alzati!” (please get up!), reporting the injunction the next day I can say that I had told Ambrogio that he “doveva alzarsi”, and here in German I use precisely sollen and not müssen. Translations of Kant’s texts of moral philosophy are crammed with dover essere. It would have been far more fitting, without supposing the existence of a pure “moral ought”, simply to use dovere: when this answers to commands, laws, precepts (the Kantian imperatives, as it happens!) it is quite clear from the context.

***

B) The examples in (A) could go on, and perhaps to some readers those already given will even seem doubtful. But what they would exemplify is the idea that translative infidelity is not founded on the existence of meanings that a language possesses and that the translator mysteriously grasps with the mind, but for which it has no equivalent in the target language, in our case in Italian. A belief of this kind rests on a distorted (philosophical) conception of the nature of meaning.
Things change if problems of style come into play. As mentioned, in that case the problem does not arise (except for the elegance of the writing) when it occurs in the “scientific” domain, whereas things change, or may change, in the literary environment, particularly in poetry. Let us begin with an example drawn from sociology.
In the works of Niklas Luhmann, the greatest German sociologist of the last century, the verb ausdifferenzieren continually appears, along with its nominal and adjectival forms. Differenzieren is obviously in Italian differenziare, but the problem for its translators (Luhmann’s translations really are “a real b**ch”[2] ) is the prefix aus, so mysterious as to lead its translators to strike it out, leaving differenziare, differenziazione and so on (words, evidently, over which the Italian reader must not toil too much). Now the sense of ausdifferenzieren is in Italian the following monster: “autonomizzarsi attraverso differenziazione”, to make oneself autonomous through differentiation. The idea is that, according to Luhmann’s “systems theory”, religion, for example, is a system generated through the progressive differentiation of society, whose problems of a certain kind it resolves. Having reached a certain degree of complexity, however, religion turns into an autonomous system that conceives society itself as one of its own differentiations. It matters little here whether I have made myself clear. It is obvious that one cannot substitute for every occurrence of ausdifferenzieren and its nominal and adjectival forms the hideous autonomizzarsi attraverso …: Luhmann’s texts would have become even more indigestible and unreadable than they already are. But one could have explained the meaning of the term in a note and then added an asterisk to its occurrences, or else, again after clarifying it, coined a neologism. Style would certainly have suffered, but the German’s complex thought would have been carried over into Italian just as it is.
Imagine, then, a woman seated before an idle iron, and her man thinking: “non mi stira!”, she isn’t ironing for me! If the mi is interpreted as a dative of advantage, we are dealing with a crude chauvinist: the lady does not iron per lui, for his benefit! But if the dative is “ethical” (often used by Tuscan writers, by Palazzeschi among others), which Italian preserves from Latin and Latin from Greek, that mi indicates emotional participation and concern; the chauvinist male is thus transformed into a tender and solicitous companion. In English or in German, which do not have the ethical dative, one would need to explain what it is about, but in translation the shared concern of the tender little man is of course lost. Or think of the English verb to rubberneck, used among other things for the hapless souls who stop in the opposite lane to contemplate an accident. How is one to render in Italian the grotesque, cartoon-like image of a rubber neck stretching with cynical greed toward the other side? In these and in myriad analogous cases the failure to render the image and the feelings, however fully it may be understood, strips from the translation elements that merge with meaning. The shortcomings of a language, moreover, can also bear fruit. La coscienza di Zeno by Svevo is a title untranslatable into English and German, languages that possess distinct terms for “cognitive” conscience and “moral” conscience. Zeno’s is both in one, and the synthesis is lost.
Clearly the interplay of style, sounds and sensations is the very soul of poetry. In “il divino del pian silenzio verde” from Il bove by Carducci, at least four or five figures of speech play off one another (anastrophe, hyperbaton, hypallage and so on), with the last, “synaesthesia”, giving the underlying tone to the whole. In synaesthesia, a figure of speech but also an alteration of the perceptual system, a certain property is attributed to incongruent sensory qualities, so that a colour can become crisp and a tactile sensation sugary. In Carducci’s line verde, which belongs to pian, follows silenzio, creating an atmosphere of intimate peace, with a colour that accompanies the rapt absence of sounds. I know of no translations of the line, but I can well imagine that much of the original, its sounds, metaphors and syntax, the very substance of the line, is lost. I leave to the reader, finally, the abracadabra that might render into Italian the leap between dentals and fricatives of the famous incipit of Nabokov’s Lolita: “Lolita […]. My sin, my soul […]: the tip of the tongue taking a trip of three steps down the palate to tap, at three, on the teeth. Lo. Lee. Ta.”

***

C) The problems we have called “of style” therefore cause little trouble when it is a matter of rendering abstract concepts or theories, where one can paraphrase the missing word. They begin to cause trouble when what is at stake is the rendering of the expressiveness of the source language. They then explode into outright infidelities in literature, and above all in poetry, where expression merges with meaning. Here betrayal, in the great translator, is inevitably joined by creativity. It is no accident that translations of poets made by poets, even when the meaning is distorted (see Quasimodo’s famous translation of the Greek Lyric poets), keep the lyric rendering high.
But is it possible to understand one or more meanings and betray them in translating, apart from questions of style, of sound, of metaphorical reach, as happens in poetry? Is it possible, I mean, without falling back into the idea that, if I understand the meaning in the source language, I must then be able to render it in the target one? And without having to fall back into the whimsical idea that I understand a little mental sprite that has, who knows why, no corresponding sound and sign to clothe it in the translation?
The fact is that words acquire their semantic volume, the aspects of their meaning, in the various contexts of use in which they are applied. Like those who use them, words live a complex and changing life that is then handed down from generation to generation and that, in many cases, vanishes into the past. If the objective entities at which science aims, every science, natural or human, are, at least in intention, points of reference of truths that do not change and do not vary according to latitude, the feelings, the habits, the value judgements, the social recognition, the expectations, above all the illocutionary force (the effects we want them to produce) that words express vary from culture to culture and, within each, vary over time. In this sense words live within an environment and, upon their use in a specific circumstance, there weigh the further uses in which they are expressed and exchanged in order to communicate. In the burden of uses that a word possesses in each single employment, in the different situations, there resounds, so to speak, its history, made up precisely of the many and varied occasions on which it is spoken or written. This multiform life, I think, accompanies it in every single occurrence. To disturb certainly corresponds in Italian to our disturbare, but in English it is also used to express our turbare; a person deeply disturbed is a person profondamente turbata (not mentally disturbata!), and the English verb can also be used in the meaning of to move or to shift, for example a body. I do not at all mean to claim that, in disturbing a meeting, the idea of inner turmoil or of shifting an object comes to the mind of the English speaker: I don’t want to disturb him means non voglio disturbarlo. What I mean is that a word, at least in some cases, is affected by the whole environment in which it is used, as our past weighs on our actions. And it is this environment, at least a large part of its echoes, that is lost in translation.
One might object, while accepting the semantic richness of a term in the sense described, that whatever example I use to clarify the matter, reporting among other things the indications of a good dictionary, I am still explaining in italiano the meaning of the term, and therefore render it perfectly, indicating the contexts of use and using paraphrases. The fact is that the words which serve to define and evoke the use of the foreign term also live by an environment of their own, different from the foreign one they mean to illustrate. If this is so, something is lost beyond all efforts, and what is lost is an integral part of a literary work. The best way to understand this situation is to call to mind our ability, more or less marked depending on sensibility, to identify, from a sentence uttered on a certain occasion, the character of an individual who speaks our language, to derive from a sentence the way in which he stands toward others, even his values and preferences, and the emotions that accompany them. Behind this ability lies a myriad of conventions that we intuitively grasp because they belong to us in the first person, because we are accustomed to living the word in analogous situations. Against such an ability, consider the difficulty of grasping with equal sureness the linguistic register of someone who speaks a language different from our own, even one we command reasonably well. If we now suppose a translator who commands both languages perfectly in their respective contexts of use, perhaps because he has long lived in both countries, we can imagine the perfect comprehension of meanings accompanied to a certain degree by the impossibility of translating them, of pouring a historical environment and its expressive potential into a different environment. The difficulty becomes fundamental, to repeat, when it is a matter of a literary text that is, time and again, the most intense expression of its own culture. An ideal translator who perfectly commands the life of two cultures will have in part to reproduce the semantic environment he translates, as far as the target language allows, and in part to recreate what is lost, keeping as far as possible analogies and shared aspects. There is therefore a degree to which the translator enters as an actor into literary production.
As for me, very far from a competence of that kind (indeed I sometimes translate non-fiction), the pleasure of reading literary texts in a language different from my own is also the pleasure of entering a world to be explored and imagined, trying to grasp at least some resonances of the kinds of life that I have not lived.

Luigi Cimmino

Notes

[1]The metaphor of the sound or graphic sign that, like a garment, is conventionally attached to meaning generates a problem famously highlighted by W.V.O. Quine, that of the structural “indeterminacy of translation”. Reduced to the bone, it comes down to the following. Imagine an anthropologist who comes into contact with the members of a tribe whose language he does not know. The anthropologist has no direct access to mental meanings, which are accessible only in the first person; he can only observe the conventional sounds and signs that conventionally express them. Now, Quine tries to show how any outward behaviour of the subjects he observes, beginning with their verbal behaviour, is translatable by mutually incompatible sentences. The evidence available for the translation leaves it inevitably indeterminate. After all, how is one to reach the meaning, if it is walled up behind its conventional physical garments? Quine’s next step is to make clear that, this being so, such indeterminacy extends to the anthropologist’s very language, calling into question the starting point of the experiment: I myself can cast doubt, and for the same reasons, on the meaning of the speakers of my own language, and thus of the language I believed I commanded. It seems that only the possibility remains of accessing my own meanings by internal inspection, in the first person; but if mental meanings were known solely and only through a kind of internal self-consciousness, within a mental theatre of which I would be the sole spectator, a) the consequences would be those of solipsism (I alone know that I have a mind and meanings) and of its paradoxes; b) the internal observation of meaningful mental entities is a myth, no one has experience of it. Quine’s conclusion, as a good naturalist, is that “mental meanings” do not exist and that language too, like nature, is composed exclusively of sequences of causally connected physical entities. To refute Quine’s experiment, which I hold leads to consequences as unpalatable as those of solipsism, one would have to show that the linguistic garment, in the various natural languages, directly reveals the characteristics of the bodies it wraps, as much as, and far more than, the clothes that wrap human bodies do. As said, we here abandon the convolutions of the philosophers, while noting that Quine’s argument has had important consequences in contemporary philosophy and that its assumptions are various and complex. Let us trust the translator’s starting problem, which is after all the datum observed by anthropologists too: the problem is that of rendering manifest meanings as well as possible, not of trans-lating, of carrying across from one language to another mental will-o’-the-wisps mysteriously connected to sounds and signs.

[2]I thank my wife for the stylistic suggestion.

Huck and Montalbano: Translating Dialectical Variety
There is a missing page in the Italian edition of the Adventures of Huckleberry Finn. The second page of The Library of America edition of Mark Twain’s masterpiece, presents an eleven-line note signed by The Author under the title EXPLANATORY:

In this book a number of dialects are used, to wit: the
Missouri Negro dialect; the extremest form of the back-
woods South-Western dialect; the ordinary “Pike County”
dialect, and four modified varieties of this last.
The shadings have not been done in a hap-hazard fashion,
or by guess-work; but pains-takingly,  and with the trustworthy
guidance and support of personal familiarity with these several
forms of speech.
I make this explanation for the reason that without it many
readers would suppose that all these characters were trying
to talk alike and not succeeding.

Published in 1963, translated and prefaced by Enzo Giachino, the prestigious Giulio Einaudi Millenni edition of Twain’s masterpiece is complete, except for this explanatory note from the author. Its absence is easily explained. Giachino’s translation does not attempt to render the dialectical variety that Twain worked so hard to create in the original. Here is one famous dialogue between Huck and Jim, from their discussion of the fate of the heir to the French crown who is rumoured to have escaped to America:

“I told about . . . his little boy the dolphin, that would a been a king
but they took him and shut him up in jail. . . .
But some says he got out and got away, and come to America.”
“Dat’s good! But he’ll be pooty lonesome – dey ain’t no kings
here, is dey, Huck?”
“No.”
“Den he cain’t git no situation. What he gwyne to do?”
“Well, I don’t know, Some of them gets on the police and some
of them learns people how to talk French.”
“Why, Huck, doan de French people talk de same way we does?”
“No Jim; you wouldn’t understand a word they said – not a single
word.”
. . .
“Is a Frenchman a man?
“Yes.”
“Well, den! Dad blame it, why doan’ he talk like a man? You answer me dat!”

It is evident even from this brief dialogue that Huck and Jim understand each other and that their ways of talking have a lot in common. But there are also some obvious differences: Huck says “that”, “the” “get” and “don’t” while Jim says “dat”, “de,” “git” and “doan’,” not to mention the wonderful “gwyne” for “gonna”.
Now look and listen to the same dialogue in the Italian translation:

“Gli parlo . . . del suo bambino, il Delfino, che doveva diventar
anche lui re, ma l’hanno preso, e chiuso in prigione . . .
Ma altri dicono che è uscito, è potuto scappare e è venuto in America.”
“Così va meglio. Ma certo che si sentirà piuttosto solo. Qui non
ci sono mica dei re, vero, Huck?
“No.”
“Allora non può farsi una posizione. Cosa può fare?
Be’, non so. Alcuni fanno i poliziotti, altri imparano alla gente a
parlare francese.”
“Come, Huck, ma i francesi non parlano come noi?
“No, Jim, non capiresti una parola di quello che dicono, non una sola parola.” . . .
“E un francese è forse un uomo?”
“Sì.”
“E allora, che il signore lo benedica, perché non parla anche lui come un uomo?
Rispondetemi un po’, se ce la fate.”

Retranslating this Italian dialogue back into English gives the measure of the standard quality of the translated dialogue compared to the original.

“I talk to him . . . about his little boy, the dauphin, who was supposed
to be a king too, but they caught him and shut him in prison. . . .
But others say he got out and was able to escape and came to America.”
“That’s better. But he must certainly feel rather lonely. There sure aren’t
any kings here, right Huck?”
“No.”
“Then he can’t make a position for himself. What can he do?”
Well, I don’t know. Some join the police, others learn people to
speak French.”
“What, Huck, but don’t the French talk like us?”
“No, Jim, you wouldn’t understand one word of what they say, not
one word.”
. . .
“And isn’t a Frenchman perhaps a man?”
“Yes.”
“So then, God bless him, why doesn’t he talk like a man too?
Answer me that, if you can.”

In Italian, Huck uses imparare instead of insegnare as he uses learn instead of teach in English, but other than that his grammar is fine and so is Jim’s. More importantly, there are no differences in their ways of talking. Huck and Jim both speak nearly letter-  perfect standard Italian. Or to put it as Twain might have, whether or not they are trying to talk alike they are certainly succeeding.
Does this failure to represent Twin’s meticulous recreation of dialect amount to a glaring weakness in Giachino’s translation? To the extent the use of standard Italian fails to render the richness of the language of Twain’s narrative one might answer in the affirmative. Indeed, critical assessments of Twain’s novel have consistently highlighted its brilliant use of language. T.S. Eliot called the style of the novel a “new discovery in the English language,” remarking that there is “no sentence or phrase to destroy the illusion that these are Huck’s own words.”  And Richard Chase, in The American Novel and Its Tradition, observes that “The Adventures of Huckleberry Finn delights the reader first and last by its language. The book makes a music of words which is beautifully sustained and modulated to the very end.”
But supposing Giachino, or any translator, had decided to try to represent Twain’s use of multiple dialects in his translation, how could he have done it? Perhaps he could have had Huck and Jim and the other dialect speakers use an ungrammatical Italian. But dialect is not only a question of non-standard grammar. A dialect also has its own lexicon and phonetics, and even its grammar, while different, is no less complex or cohesive than the grammar of the standard. Then how about using Southern Italian dialects, Neapolitan or Sicilian say, to represent the various Southern American dialects in the novel? In a certain sense, it might well be said that Neapolitan is to Italian what “backwoods South-Western” is to American English.
If you put yourself in the position of an Italian reader of the novel, however, it immediately becomes clear that the analogous relationship of dialect to standard cannot sustain the choice to substitute one dialect for another. As Twain emphasizes in his EXPLANATORY, the dialects he has chosen to re-create are tied to and evocative of place, race, and history in ways that standard American English is not. It would take an extraordinary willing suspension of disbelief for an Italian reader to accept a black American slave or a Missouri-bred white boy holding forth in Neapolitan.
Strangely enough, however, like standard English or French, standard Italian does not require the same leap of faith. Although it is obviously the product of a different place and culture than Twain’s English, Italian is not, precisely because it has become a national standard, evocative of any specific geo-cultural location. Giachino’s American characters can speak standard Italian without distracting the Italian reader from the Americanness of their story because their language does not evoke a specific region or cultural history. The reader knows that the only reason they are speaking Italian is because otherwise s/he couldn’t understand what they are saying. The standard language gives the reader access to the characters’ different world without getting in the way, as their speaking in a recognizable regional dialect certainly would.
Of course, the same holds for translations in the opposite direction, as in the case of the American versions of Andrea Camilleri’s Montalbano novels. One of the reasons for Camilleri’s enormous success among Italian readers is his brilliant interweaving of Sicilian dialect into his otherwise standard Italian narratives. Even the made-for-TV film versions of the stories use some of the recurring words and phrases: “Montalbano, sono,” “macari” for “anche” (also) or “taliare” per “guardare” (look) or “spiare” per “domandare” (ask).

Here is a brief example of the Siculo-Italian interweaving from La voce del violino with the dialect words in bold:

Montalbano trasi nella Chiesa gremita, la funziona era appena prin-
cipiata. Si taliò torno torno, non racconiscì nisciuno. Tamburrano
doveva essere in prima fila, vicino al tabbuto davanti all’altare maggiore.
Il commissario decise di restarsene dov’era, allato al portone d’ingresso:
avrebbe stretto la mano a Tamburrano quando il feretro nisciva dalla chiesa.
Alle prime parole del parrino dopo già tanto che la messa procedeva.
ebbe uno sbalzo.

Now the same passage from Stephen Sartorelli’s translation:

Montalbano entered the crowded church. The service had just begun.
He looked around and recognized no one. Tamburrano must have been
in the first row, near the coffin in front of the main altar. The inspector
decided to remain where he was, near the entrance. He would shake
Tamburrano’s hand when the coffin was being carried out of the church.
When the priest finally opened his mouth after the Mass had been
going on for some time, Montalbano gave a start.

While Camilleri’s narrator peppers his account with about 10 percent Sicilian words and phrases, Sartorelli’s narrator uses standard English. The Italian reader might encounter some difficulty reading smoothly through the narrator’s description but his impression of the narrative voice will be flavored by all the associations and allusions expressed by the peculiar regional musicality of the language. The American reader of the translation will have access – through the proper names, the ceremony, the architecture of the church, other elements of the story – to the foreign place and culture but the access will be limited by the absence of the special musicality of the dialect.
Another regular character in the Montalbano mysteries, the entry level police officer and station receptionist Catarella, offers us another point of view on the question of dialect translation. Here is a dialogue between Catarella and Inspector Montalbano, again from La voce del violino, which resonates curiously with Huck and Jim’s dialogue about Frenchmen:

“Domando pirdonanza e compressione, dottori. . . . Tre giorni passati
cercarono proprio lei di lei, dottori, lei non c’era, però io me lo scordai
di farle riferenza.”
“Da dove telefonavano?”
“Dalla Flòrida, dottori. . . .
“Levami una curiosità, come vi siete parlati?”
“E come dovevamo parlarci? In taliano, dottori.”
“Ti hanno detto che volevano?”
“Certo, tutto di ogni cosa mi dissero. Dissero così che morse la
mogliera del vice-questore Tamburrano.”
[Montalbano] tirò un sospiro di sollievo. . . . Non dalla Flòrida aveevano
Telefonato, ma dal commissariato di Floridia, vicino a Siracusa.

At first glance, it seems that Catarella is speaking in dialect while Montalbano, a college graduate, uses standard Italian. A closer reading, however, reveals that, rather than dialect, Catarella is speaking bad Italian. He is a sort of Sicilian Mr. Malaprop, and Sartorelli renders Catarella’s mangled Italian with an analogous brand of mangled English.

“’Beckin’ pardon, Chief, for the ‘sturbance. . . . Tree days ago somebody
calls for you, Chief, wanted a talk t’ you in poisson, but you wasn’t ‘ere ‘n
I forgotta reference it to you.”
“Where were they calling from?”
“From Flòrida, Chief. . . .”
“Tell me something. What language did you speak with them?”
“What langwitch was I aposta speak? We spoke ‘Talian, Chief.”
“Did they tell you what they wanted?”
“Sure, they tol’ me everyting about one ting. They said as how Vice Commissioner
Tamburrano’s wife was killed.”
“Montalbano breathed a sigh of relief. They’d called not from Flòrida but from
police headquarters in the town of Floridia, near Siracusa.

So Sartorelli was able to render the difference in the way of talking between Montalbano and Catarella, but it is important to observe that what is at stake here is not dialectal variety but the difference between correct and incorrect standard language. Catarella’s malapropisms evoke not regional specificity but comically poor language skills. This kind of language difference is easier for the translator to represent, although one might remark that in some instances, at least to my ear, Catarella’s erroneous speech – “in poisson”, “everyting about one ting” – sounds a little too much like vaudeville Brooklynese, and to the extent that we hear that local resonance, Catarella may sound out of place.
What can we conclude from these examples taken from the Italian and English translations of these two masters of dialect, Twain and Camilleri? Rather than a definitive conclusion, I prefer to hazard the hypothesis that dialectical variety may indeed be one of those things that is inevitably lost in translation. Beyond that, it strikes me that this hypothesis, if true, may also be somewhat paradoxical. Our inability to translate dialectical variety may deny readers access to the very aspects of foreign cultures that are expressed by the peculiar and fascinating musicality of natural human speech.

Gregory Conti

Luigi Cimmino teaches Gnoseology and Ethical Paradigms at the University of Perugia.

Gregory Conti teaches English at the University of Perugia. His translations include works by Rosetta Loy, Giuseppe Berto, Sebastiano Vassalli, and Paolo Rumiz. For a full list of publications, visit www.gregoryconti.com.