Among the many pronouncements that have shaped our understanding of literary translation, perhaps none is more often echoed than John Dryden’s preface to his version of the Aeneid. “I have endeavoured to make Virgil speak such English,” asserted Dryden, “as he would himself have spoken, if he had been born in England, and in this present Age.” No doubt Dryden’s achievement is to have made many of his contemporaries believe that he had impersonated the Latin poet. But this is merely a poetic sleight of hand. Dryden’s Virgil abandons the unrhymed verse of the Latin poem for English couplets while cribbing lines from a previous translator, the poet Sir John Denham. A skeptic might well wonder why Virgil should come back as Dryden instead of an epic poet who lived in the same period and wrote his epic without rhyme: John Milton. Should we not expect an English Virgil to be more attracted to the grand style of Paradise Lost?

The answer has less to do with a fancied reincarnation than with the fact that literary taste changes. And when it does change, a corresponding style of translation falls into disuse or is pre-empted, never to be adopted by leading translators (especially when, like Dryden, they happen to be Poet Laureate). By the end of the seventeenth century, the blank verse of Shakespeare and Milton had lost cultural capital to the couplet, so that a poet as talented and celebrated as Dryden could make the latter seem to be the most natural vehicle for a Latin poem written in a completely different verse form. The translator is no stand-in or ventriloquist for the foreign author, but a resourceful imitator who rewrites the original to appeal to another audience in a different language and culture, often in a different period. This audience ultimately takes priority, insuring that the verbal clothing the translator cuts for the foreign work never fits exactly.

The most questionable effect of Dryden’s assertion, to my mind, is that it winds up collapsing the translator’s labor into the foreign author’s, giving us no way to understand (let alone judge) how the translator has performed the crucial role of cultural go-between. To read a translation as a translation, as a work in its own right, we need a more practical sense of what a translator does. I would describe it as an attempt to compensate for an irreparable loss by controlling an exorbitant gain.

The foreign language is the first thing to go, the very sound and order of the words, and along with them all the resonance and allusiveness that they carry for the native reader. Simultaneously, merely by choosing words from another language, the translator adds an entirely new set of resonances and allusions designed to imitate the foreign text while making it comprehensible to a culturally different reader. These additional meanings may occasionally result from an actual insertion for clarity. But they in fact inhere in every choice that the translator makes, even when the translation sticks closely to the foreign words and conforms to current dictionary definitions. The translator must somehow control the unavoidable release of meanings that work only in the translating language. Apart from threatening to derail the project of imitation, these meanings always risk transforming what is foreign into something too familiar or simply irrelevant. The loss in translation remains invisible to any reader who doesn’t undertake a careful comparison to the foreign text-i.e., most of us. The gain is everywhere apparent, although only if the reader looks.

But usually we don’t look. Publishers, copy editors, reviewers have trained us, in effect, to value translations with the utmost fluency, an easy readability that makes them appear untranslated, giving the illusory impression that we are reading the original. We typically become aware of the translation only when we run across a bump on its surface, an unfamiliar word, an error in usage, a confused meaning that may seem unintentionally comical. Think of the bad English translations you’ve encountered abroad, the dry cleaner urging potential customers to “Drop your trousers here for best results,” the restaurant announcing that “Our wines leave you nothing to hope for,” the hotel advising its guests to “Please leave your values at the front desk.”

Our laughter at their ineffectiveness betrays a confidence, perhaps a complacence, in our native English proficiency. But something more instructive is revealed: we laugh only because we have sorted out the confusions, demonstrating quite clearly that readers of translations can perform several mental tasks at once. In reading to comprehend, we focus on both form and meaning, so that when the meaning turns obscure or ambiguous, we instantly clarify or untangle it by correcting the error in form, in word choice or grammar. Hence the first rule of reading translations: Don’t just read for meaning, but for language too; appreciate the formal features of the translation.

Savor the translator’s diction and phrasing, the distinctiveness of the style, the verbal subtleties that project a tone of voice and sketch the psychological contours of a character. Still-a reader may ask-don’t these qualities belong to the original? Not at all, certainly not in the sense that the translator just transfers them intact, without variation. They of course result from the translator’s imitation of the foreign text. But the fact remains that the translator has chosen every single word in the translation, whether or not a foreign word lies behind it. And the translator’s words, in our case, function only in English, releasing literary effects that may well exceed the language chosen by the foreign author.

Consider a passage from Margaret Jull Costa’s version of The Man of Feeling by the Spanish novelist Javier Marías. The narrator, an opera singer, is writing the story of his chance encounter with a woman:

I knew nothing at all about her history or past or life, apart from the scant information vouchsafed to me in Dato’s self-absorbed and fragmentary complaint during the first and only opportunity I had had to talk to him alone (too soon for my curiosity to have learned how to direct its questions) and from the enthusiastic remarks which, rarely and only in passing, she made about her brother, Roberto Monte, that recent émigré to South America.

The most striking feature of this sentence is its sheer length: it appears uncommonly long against the backdrop of current English-language fiction. It matches Marías’s Spanish, but we don’t need to know that to appreciate how effectively Costa constructs the English sentence, allowing it to unfold at a measured pace, embedding self-conscious qualifications at key points. She also chooses language that creates a slightly elevated, even precious tone, words like “scant,” “vouchsafed,” “émigré,” as well as phrases that display a punctilious care with grammar (“had had,” “to have learned”). With all these features, the translator has shrewdly avoided more colloquial English so as to fashion a somewhat affected character. The affectation is actually more pronounced in the English version because the Spanish is fairly common, using, for example, “me había permitido entender” (“had allowed me to understand”) instead of an archaic phrase like “vouchsafed to me.”

Costa’s sentence points to a second rule: Don’t expect translations to be written only in the current standard dialect; be open to linguistic variations. The translator’s hand becomes visible in deviations from the most commonly used forms of the translating language. Social and regional dialects, slang and obscenities, archaisms and neologisms, jargons and foreign borrowings tend to be language-specific, unlikely to travel well, their peculiar force difficult to render into other languages. Thus they show the translator at work, implementing a strategy to bring the foreign text into a different culture. Matthew Ward’s version of Albert Camus’s novel The Stranger opens with the surprising line, “Maman died today.” The context makes clear that the French “maman” means “mother.” Ward retains Camus’s use of the word, yet it means so much more in English: not only does it signal the childlike intimacy of the narrator’s relationship, but it tells us that we are reading a translation, a hybrid, not to be confused with the French work.

The translator’s language can also send down deep roots into the receiving culture, establishing suggestive connections to styles, genres, and texts that have already accumulated meaning there. This inevitable result of the translation process underlies a third rule: Don’t overlook connotations and cultural references; read them as another, pertinent layer of significance. An example is Dorothy Bussy’s inspired choice of Strait is the Gate for the title of André Gide’s La Porte étroite, the first of his novels to appear in English. Both titles allude to the Gospel of Luke, but the English one elegantly glances at the King James Bible (“Strive to enter in at the strait gate”). Bussy’s phrase invests Gide’s work with a cultural prestige that could not be achieved by referring to a less authoritative or less influential version of the Bible.

The connections inscribed in a translation are often stylistic, conjuring up literary genres or traditions that enhance and perhaps comment on the foreign text. Patrick Creagh’s style in Declares Pereira, his version of Antonio Tabucchi’s Italian novel, mines a rich vein of colloquialism that includes underworld argot, words and phrases like “bigwig,” “gagged,” “shady-looking characters,” “keep your eyes peeled,” and “skinny little shrimp.” On the one hand, this language fits the narrator, a veteran journalist who spent his career as a crime reporter; on the other hand, it fits the genre, a political thriller in which he finally resists the fascist dictatorship in 1930s Portugal by denouncing its crimes.

Creagh’s English doesn’t correspond perfectly to Tabucchi’s Italian. Where Tabucchi uses the standard dialect, Creagh occasionally shifts to colloquial usage while mixing in numerous Britishisms, some of which are a bit old-fashioned, specific to a pre-WWII lexicon (“doss-house,” “take a dekko,” “I’m in a pickle”). Yet in linking style to genre, in creating a period flavor, Creagh’s translating is far from arbitrary. The gangster-like lingo, in particular, is highly appropriate for a regime that relies on paramilitary thugs to intimidate and murder its citizens.

Creagh’s work really amounts to a remarkable tour de force. It serves as a reminder of what most readers implicitly know: a translation can never be identical to the foreign text or communicate it in some direct, untroubled manner, not even if the translator maintains a high degree of linguistic accuracy. But we may be less willing to accept a corollary: a translation is fundamentally incapable of providing its reader with an experience that equals or closely approximates the one that a native reader has with the foreign text. To provide this sort of experience, a translator would have to endow us with a lifelong immersion in the foreign language and literature. Only then can we read the translation with anything resembling the informed sensibility that a native reader brings to the original. Although translators are undoubtedly creative, they can’t make over their readers by giving them foreign identities.

What translators can do, however, is to write. We should view the translator as a special kind of writer, possessing not an originality that competes against the foreign author’s, but rather an art of mimicry, aided by a stylistic repertoire that taps into the literary resources of the translating language. A translation communicates not so much the foreign text as the translator’s interpretation, and the translator must be sufficiently expert and innovative to interpret the linguistic and cultural differences that constitute that text. When a foreign classic is retranslated, furthermore, we expect the translator to do something new to justify yet another version. And in raising the bar we might also expect the translator to be capable of describing this newness.

Richard Pevear and Larissa Volokhonsky’s version of The Brothers Karamazov utterly changed the understanding that many readers of English have long had of Dostoevsky’s novel. As Pevear explains in his preface, previous translators “revised, ‘corrected,’ or smoothed over the Russian writer’s idiosyncratic prose, removing much of the humor and distinctive voicing of the novel.” He and his collaborator sought “a truer rendering” that restored “phrases, mannerisms, verbal tics.” Their exemplary translation necessitates a fourth rule: Don’t skip an introductory essay written by a translator; read it first, as a statement of the interpretation that guides the translation and contributes to what is unique about it.

Nonetheless, the translator’s interpretation remains partial, both incomplete in omitting irrecoverable aspects of the foreign text and slanted towards what is intelligible and interesting in the receiving culture. It also reflects the cultural and financial interests of publishers, the gatekeepers who decisively exercise the power to admit or exclude foreign works. For an entire foreign literature is never translated, most of what has been translated rarely remains in print for very long, and everywhere translations of fiction far outnumber those of poetry, among other genres (hence the emphasis in my examples). Not only can’t we read a recently translated novel with a sense of how the foreign work draws on its native traditions, but uneven translation patterns can all too easily harden into misleading cultural stereotypes. Because the rate of translation into English is so low-roughly 2-4% of annual book output in the US and the UK, compared to 25% in countries like Italy and Spain-a reader may be unable to find a selection of translations from the same foreign author, even from the same language. This situation gives a special urgency to a fifth and final rule: Don’t take one translation as representative of an entire foreign literature; compare it to translations of other works from the same language.

Some languages and literatures are particularly undertranslated today. Take Arabic. Little Arabic writing is available in English, much less than Hebrew writing, for instance, undermining any effort to gauge the cultural impact of social and political developments in the Middle East. The Egyptian novelist Naguib Mahfouz deserves to be ranked among the most fascinating Arabic writers, but to regard him as the literary spokesman for the Arab world is undoubtedly a mistake. Mahfouz should be read alongside his countryman Abdel Hakim Qasim, whose Rites of Assent (translated by Peter Theroux) combines modernist techniques with Qur’anic allusions to interrogate Islamic fundamentalism, the forced conversion of an Egyptian Copt under the aegis of the Muslim Brotherhood. Qasim might then be juxtaposed to Sayed Kashua, whose Hebrew novel Dancing Arabs (in Miriam Shlesinger’s translation) incisively depicts the identity crisis of an Arab Israeli who, although raised in a family of militant anti-Zionists, tries to pass among Jews. Sometimes, to gain a broader view of the cultural situations that translation leaves behind, a reader must venture into neighboring languages and territories.

As these examples suggest, an aesthetic appreciation of translated literatures can powerfully illuminate the cultural differences that have led to political divisions and military conflicts. The fact is, however, that the current predicament of English translation doesn’t favor sharp distinctions between the literary and the political, the aesthetic and the sociological. English is the most translated language worldwide, but relatively little translated into, particularly given the size and profitability of the American and British publishing industries. Foreign publishers scramble to issue translations of English-language fiction, but publishers in the US and the UK tend not to reinvest their enormous profits from selling translation rights into translating foreign fiction. The figures are staggering, even if we set aside the immediate worldwide translation of bestsellers like Stephen King, Danielle Steel, and Tom Clancy and focus on authors with literary reputations. In France and Germany, for example, Joyce Carol Oates and Philip Roth each have translations of more than twenty works currently in print; in Italy and Germany, more than thirty titles by Charles Bukowski are available in translation (eighteen in France, fifteen in Spain). Rare is the contemporary foreign novelist whose body of work enjoys such representation and availability in English. In these circumstances, even to read a translation purely for its literary qualities can be seen as a political gesture, an act of resistance against long-standing publishing practices that have severely restricted our access to foreign literatures.

A translation ought to be read differently from an original composition precisely because it is not an original, because not only a foreign work, but a foreign culture is involved. My aim has been to describe ways of reading translations which increase rather than diminish the pleasures that only reading can offer. These pleasures involve primarily the linguistic, literary, and cultural dimensions of translations. But they might also include the devilish thrill that comes from resistance, from challenging the institutionalized power of cultural brokers like publishers, from staging a personal protest against the grossly unequal patterns of cultural exchange in which readers are unwittingly implicated. Read translations, although with an eye out for the translator’s work, with the awareness that the most a translation can give you is an insightful and eloquent interpretation of a foreign text, at once limited and enabled by the need to address the receiving culture. Publishers will catch on sooner or later. After all, it’s in their interest.