The Italian aphorism, known widely in its negative formulation — traduttore, traditore, the translator as traitor — encapsulates a suspicion of literary translation that has been present in literary culture since at least the Renaissance. The proposition that the translator necessarily betrays the source text has been stated and restated in so many contexts that it has acquired, for many readers, the status of an obvious truth rather than a contestable claim. The scholarship on literary translation has, for the better part of a century, been engaged in the task of examining this claim with more precision — of asking, specifically, what is lost, by what mechanism, and whether the losses are as extensive and as inevitable as the commonplace suggests.
The evidence that emerges from this scholarly examination is considerably more complex than the aphorism allows. It is certainly the case that literary translation involves losses; this is, in a narrow sense, unavoidable, and no serious scholar of translation would deny it. What is equally clear, and considerably less often remarked upon, is that literary translation also involves discoveries — moments at which the translator’s effort to render a text in another language reveals possibilities in that text that were not visible from within the source language, or produces an aesthetic result in the target language that is in some respects superior to what an equivalently skilled writer working within the target-language tradition would have produced without the stimulus of the source.
The question of what is lost
The scholarly literature on translation loss identifies several distinct categories of what may be forfeited in the move from source to target language. The most frequently discussed is the loss of phonological texture — the sound pattern of the source language, which in a literary text is often integral to its effect. The rhyme schemes of formal verse cannot typically be reproduced in a target language without substantial distortion; the rhythmic patterns of metrically organised prose may or may not have equivalents in the target-language prosodic system; the sonic qualities of particular words — the onomatopoeic or euphonious effects that a skilled literary writer exploits — are rarely reproducible in translation.
The second major category of loss is what might be called cultural embeddedness — the network of allusions, associations, and assumed shared knowledge that a text in one language can draw on and that a reader in another language and cultural tradition cannot be presumed to possess. A literary text addressed to readers within the culture that produced it can assume that certain references will resonate, that certain tonal registers will be recognised, that certain ironic distances will be understood. The translator working for readers outside that cultural tradition cannot make equivalent assumptions, and the strategies available for closing the gap — footnotes, glosses, explanatory additions to the text — all carry costs in terms of the reading experience.
The problem of tone
Among the losses that practitioners of literary translation find most acute is the loss of tonal precision — the difficulty of reproducing, in the target language, the exact register in which the source text is pitched. Every literary language has developed, through its history, a set of tonal possibilities that are not straightforwardly reproducible in other languages. The formal and informal registers of, say, Japanese are structured differently from those of French; the distinction between archaic and contemporary diction is calibrated differently in German than in English; the ironic possibilities of a language with a long tradition of political indirection may not be reproducible in a language with a different political history.
The scholarship consistently identifies tonal accuracy as the quality that distinguishes the best literary translations from adequate ones, and tonal failure as one of the most common ways in which translations that are lexically accurate nonetheless fail as literary works. A translator who renders every word correctly but pitches the text in a register that does not correspond to the source — who makes a formal text colloquial, or a colloquial text stiff, or a comic text solemn — has produced a translation that will mislead readers about the nature of the source as profoundly as a translation that misrenders individual words.
What translation sometimes finds
The scholarly literature on what translation discovers — as opposed to what it loses — is less extensive, partly because the negative formulation is culturally entrenched and partly because the positive case requires a more nuanced argument. The evidence for translation as discovery takes several forms. The most documented is the phenomenon of the translation that supersedes its source in the target-language tradition — the translated work that becomes, in the language and literary culture into which it has been translated, a formative and influential text in its own right, shaping the subsequent development of that culture in ways that would not have been possible without the stimulus of the foreign original.
The history of Western literature provides extensive evidence of this phenomenon. The influence of Greek and Latin texts on the development of vernacular European literatures in the medieval and early modern periods was mediated almost entirely through translation; the originals were inaccessible to the majority of writers who were most influenced by them. The development of European Romanticism owed substantial debts to translations of Shakespeare and other British writers that, in several national traditions, arrived before the originals did. The history of the novel’s development as a form is inseparable from the history of the translation of exemplary novels across European literary cultures.
The translator as co-author
The scholarly tendency in translation studies over the past several decades has been to take increasingly seriously the idea of the translator as a creative agent rather than a mere conduit. The older model — in which the ideal translation was invisible, the translator self-effacing, the source text transparently reproduced in the target language — has given way, in most contemporary theoretical accounts, to a model in which the translator’s presence in the text is acknowledged as both inevitable and, at its best, a contribution rather than a contamination.
The evidence for this view is not merely theoretical. The practice of the greatest literary translators — those whose work has entered the literary tradition of the target language as a permanent part of it — is characteristically marked by an active and informed engagement with the target language’s own aesthetic resources. The translator who produces a rendering of Homer that is readable and affecting in English is not simply conveying information about the Greek; they are making a series of decisions about English literary form that draw on their knowledge and feeling for the target language at least as much as on their knowledge of the source. The result is, in a meaningful sense, a new work — not independent of the original but not reducible to it.
The politics of translation choice
Scholarship on the sociology of translation has drawn attention to the political and cultural dimensions of which texts are chosen for translation and which are not. The evidence is clear that the global translation market is highly asymmetrical: texts produced in English, French, German, and a small number of other widely-spoken languages circulate internationally in translation far more readily than texts produced in minority or less-prestigious languages. A writer working in Catalan, in Welsh, in Albanian, or in any of dozens of less globally distributed languages faces a structural disadvantage in accessing the international readership that translation can provide.
The consequences of this asymmetry extend beyond the career prospects of individual writers. They affect which literatures are visible to the international readership and which are not; which traditions influence the development of world literature and which remain largely invisible to it; which cultural perspectives are available to readers who depend on translation and which are effectively withheld. The scholarship indicates that the literary world’s understanding of itself — its sense of which works are canonical, which traditions are important, which innovations are significant — is substantially shaped by the translation choices that publishers, prize committees, and cultural institutions make, and that these choices reflect existing hierarchies of cultural prestige as much as they reflect the intrinsic qualities of the works in question.
What the practice requires
The evidence of comparative translation study suggests that the qualities required of a literary translator are, in their combination, exceptionally demanding. Proficiency in the source language must be sufficient to read the literary text with the kind of attention that its author brought to writing it — attending to implications, to tonal nuances, to the resonance of particular word choices within the author’s wider vocabulary, to the relationship between the individual work and the tradition it is engaging with. Proficiency in the target language must be sufficient to produce, within that language’s own aesthetic resources, a text that can sustain comparable literary attention. And the translator must possess, in addition to these linguistic competencies, the literary and cultural knowledge required to understand the full context of the source text and to make informed decisions about which aspects of that context can be transmitted to readers of the translation and which cannot.
The difficulty of finding individuals who possess all of these qualities simultaneously goes some way toward explaining why genuinely excellent literary translation is rare, why the best translators are as valuable to the literary culture as they are, and why the institutional and financial structures of publishing so often fail to reward the practice at a level commensurate with its difficulty and its cultural importance. The argument that translation is derivative work requiring less than original creation rests on a fundamental misunderstanding of what the best translation actually involves; the scholarship on the practice has been making this case for decades, and the evidence in support of it is, at this point, substantial and clear.
