Corps de l’article

The temporality of translation

What defines a translated text as a translation? Since antiquity many answers have been given to this question, varying from one historical period to another, subject to changing ideas about the nature of language, textuality, and culture. Nevertheless, the definitions that have been advanced share a notion – whether explicitly stated or implied – of what I shall call the relative autonomy of translation. Translated texts are distinguished by their independence from two sorts of pre-existing compositions: the foreign-language texts that they translate and texts that were originally written in the translating language. Recognizing this autonomous status is essential for the study and practice of translation: it delimits translating as a form of textual production in its own right, requiring compositional methods and analytical concepts that differ to a significant extent from those applied to original texts. Yet the autonomy of translation must be described as no more than “relative,” never absolute, because translating is a derivative or second-order form of creation, intended to imitate or recreate a foreign-language text. And even though the precise relationship between a translation and a foreign text has been the object of historically variable accounts, that relationship remains a necessary category for any definition of translation as such.

Time is crucial in ensuring that a translation will be relatively autonomous from the foreign text that it translates. Because every text, whether original composition or translation, emerges in a cultural situation at a particular historical moment, it displays two temporal dimensions: one is synchronic, insofar as the text occupies a position in contemporaneous hierarchies of cultural materials and practices; the other is diachronic, insofar as the text enters into a relation to past materials and practices that may or may not have acquired the authority of cultural traditions in its own time. The temporality of a translation differs from that of the foreign text because languages and cultures undergo different forms and speeds of development. As a result, a translation reveals historical continuities and divergences between the two languages and cultures that it brings into contact. Furthermore, not only is every stage in the production of a translation profoundly marked by its historical moment, but its circulation and reception inevitably trace a history that is distinct from the destiny of the foreign text.

The historical nature of translation is first apparent in the succession of varying methods that define it within a single culture. In the Westöstlicher Diwan (The West-Eastern Divan, 1819), Goethe distinguished between three methods of translating poetry which were practiced by German translators in three different periods. The first, which he described as “a simple prosaic translation,” at once domesticated and homogenized the foreign text, “since prose totally cancels all peculiarities of any kind of poetic art” and “pulls poetic enthusiasm down to a kind of common water-level” (Lefevere 1992: 75-76). It was exemplified by Luther’s sixteenth-century version of the Bible. The second method, where “the translator really only tries to appropriate foreign content and reproduce it in his own sense,” was exemplified by Cristoph Martin Wieland’s prose translation of Shakespeare’s plays during the 1760s: Wieland’s “singular sense of taste and understanding,” wrote Goethe, “brought him close to antiquity and foreign countries only as far as he could still feel at ease,” only as far as their differences didn’t unsettle the values that dominated German-language culture in his time (ibid.: 76). The third method was described as an “approximation to the external form of the original,” a close adherence that imported foreign linguistic and cultural elements into German (ibid.: 77). Johann Heinrich Voss’s scrupulous versions of the Odyssey (1781) and the Iliad (1793) were thus responsible for introducing the hexameter into German poetry. Goethe acknowledged that the three translation methods “can be in effect applied simultaneously,” but he nonetheless regarded them as successive “epochs” which “are repeated and inverted in every literature” (ibid.).

Goethe clearly based his historical distinctions on the adequacy of the translation to the foreign text. And since none of the translators he cited would have considered their work less than accurate, his account suggests that changing translation methods reflect changing standards of accuracy. Indeed, what constitutes an accurate translation in one period may later come to be regarded not as a translation at all, but as an adaptation or even as a wholesale revision of the foreign text. In 1760, for instance, Abbé Prévost prefaced his French version of Samuel Richardson’s novel Pamela with the declaration that

I have not changed anything pertaining to the author’s intention, nor have I changed much in the manner in which he put that intention into words, and yet I have given his work a new face by ridding it of the flaccid excursions, the excessive descriptions, the useless conversations, and the misplaced musings.

Lefevere 1992: 39

Prévost assumed that an accurate translation conformed as much as possible to the foreign author’s “intention” and verbal “manner” or style. From the standpoint of current translation practices, however, his many revisions indicate that he wasn’t simply translating, but adapting and abridging as well. He in fact states that “the seven volumes of the English edition, which would amount to fourteen volumes in my own, have been reduced to four” (ibid.: 40). A French translator today might also ground a claim of accuracy on such categories as authorial intention and style, yet with the very different argument that a translation can conform to them only by rendering every word that the foreign author included in the foreign text.

What has changed since Prévost’s time, then, is not only a standard of accuracy as reflected in a particular translation method, but the interpretation of the conceptual categories on which that standard is based, as well as the interpretation and evaluation of the foreign text. For Prévost, the author’s intention was evidently realized in a formal and thematic essence (“the main points of his work”) that could be detached from the accidental and therefore dispensable “details”; hence he criticized Richardson’s novel for a “lack of proportion that undermines the reader’s interest,” a judgment in which he assessed the “proportion” by a comparison to his interpretation of the “main points” (ibid.: 39). Yet the concept of authorial originality that subsequently came to dominate western cultures treats every textual feature as the author’s self-expression. This concept, reinforced by the canonical status that Richardson’s work achieved in the British narrative tradition, prevents any contemporary translator from viewing an adaptation or abridgement as an accurate representation of the novelist’s intention.

Changing interpretations of foreign cultures lead to changes in the selection of foreign texts for translation. A social or political event on an international scale might prompt a reassessment of foreign literatures which attracts or renews the interest of publishers and translators. After the Second World War, British and American publishers issued numerous translations from the contemporary literatures of countries that were key participants in the conflict, notably France, Germany, Italy and Japan. The publishers’ aims were commercial to some extent, but also geopolitical: they were satisfying readers’ newly piqued curiosities about former allies and enemies and at the same time promoting cross-cultural understanding. An international award like the Nobel Prize for Literature usually sends publishers scrambling to make available works by the latest winner so as to capitalize on the enormous publicity that follows the announcement. In some cases, these works may have been translated but fell out of print. In other cases, especially where less translated languages like Arabic, Chinese and Polish are involved, a Nobel Laureate’s writing may have previously gone untranslated, despite the fact that it must have been well received in its original language to justify the writer’s consideration for the award.

More often than not, changing interpretations of foreign cultures are motivated by historical developments in the receiving situation. A translator or a publisher of translations may look abroad at a certain moment because of a conviction that the translating language and literature can benefit from foreign influences. A subsequent generation may in turn interpret those influences differently, in response to a changed cultural and political situation at home, and consequently a different foreign literature may be selected for translation. At the start of the twentieth century, the Catalan translator Josep Carner focused his work on canonical French and English writers, including Shakespeare and Molière, La Fontaine and Dickens, Lewis Carroll and Villiers de l’Isle-Adam (Ortín 1996: 105-107). His goal was to enrich the Catalan language and to create a modern Catalan literature in opposition to hegemonic Spanish culture, or what he called “the Castilian monopoly on the destinies of Spain” (Carner 1986: 77). Later in the century, after Catalunya had suffered decades of repression under Franco’s dictatorship, the Catalan publisher and translator Joan Sales judged Carner’s translations to be too narrowly directed to a learned audience. Sales, in his more embattled moments, wished to preserve the Catalan language and to expand the readership for Catalan writing, so he emphasized more accessible realistic novels, especially those associated with relatively minor cultures whose subordinate position resembled that of Catalunya: Provence, Sicily, South Africa (Bacardí 1998). Whereas Carner translated foreign literary works that introduced productive differences into Catalan, Sales sought works whose similarities represented possibilities for linguistic and cultural survival.

The time-lag that always intervenes between the production of a foreign text and its translation tends to be further complicated by the fact that different cultural traditions take shape within languages and cultures. Literary traditions differ because they possess distinctive styles and discourses, genres and conventions, but also because they establish unique affiliations with foreign literatures. The differences between a foreign literary work and the literary traditions in another language may be so great as to delay or altogether prevent translation. At least since the nineteenth century, British and American poetic traditions have been dominated by an aesthetic wherein the poet is assumed to express his or her personality in transparent language (see Easthope 1983). During the twentieth century, this dominance played a significant role in determining the selection of foreign poetries for translation into English. Foreign poets whose work seemed consistent with the aesthetic of authorial self-expression were in some cases repeatedly translated: notable examples are Baudelaire and Rilke, Lorca and Neruda. Foreign poets whose work pursued more impersonal linguistic experiments that pre-empted the illusion of transparency were neglected: the innovative poetry of Futurists such as Marinetti and Khlebnikov, for instance, did not appear in English until many decades after the poets’ deaths. An important factor in this delay was the reaction against modernism that occurred in British and American literary cultures during the 1950s, resulting in a return to traditional poetic forms and a deepened investment in the expressivist aesthetic (Perkins 1987; von Hallberg 1985).

Historical trends can have a special impact on the translation of pragmatic as well as literary texts. Changing patterns of tourism – in which the volume of tourists who visit or vacation in a specific area varies according to nationality – determine the languages into which museum brochures, restaurant menus, and even road signs are translated. Changing patterns of migration – in which the volume of immigrants and migrant workers, refugees and seekers of political asylum, varies according to nationality – likewise determine which languages are translated in automated bank tellers, advertisements, and government documents, but also which languages are orally interpreted in hospitals, law courts, and other social institutions. The continuing influx of Hispanic immigrants into the United States, to take one example, has changed the practices of a major commercial publisher: Random House recently created a Spanish-language division to publish original works and translations in a wide variety of genres, ranging from novels and biographies to reference works and self-help books. The emergence of a globalized capitalist economy in the second half of the twentieth century was accompanied by a marked increase in the translation of business reports and contracts, instruction manuals, and computer software, among other text types and media. The conditions under which translations are produced must be understood not only as social, encompassing the current formation of social relations within the receiving culture, but also as historical, since a social formation changes over time and gives rise to new translation practices.

Discursive strategies and linguistic change

The historicity of a translation is also apparent in its very linguistic texture, whether the discursive strategy is fluent or resistant and regardless of the fact that fluency is likely to involve the effacement or mere removal of historical markers. The translator’s lexical and syntactical choices are linked to specific periods in the history of the translating language, so that any translation mixes the present and past forms that constitute current usage. Obviously, the farther back in time the translation was produced, the more noticeable the historical dimension of its language will be. Yet this dimension can also be revealed in the most recent translations by examining them with the help of a historically oriented lexicon like the Oxford English Dictionary.

Consider the following extracts from three versions of an essay by Montaigne: the first by John Florio (1603), the second by Donald Frame (1958), and the third by M.A. Screech (1993). The mere juxtaposition of these translations heightens their historically specific features, yet the language of each is distinctive enough to repay close analysis on its own:

La plus commune façon d’amollir les coeurs de ceux qu’on a offencez, lors qu’ayant la vengeance en main, ils nous tiennent à leur mercy, c’est de les esmouvoir par submission, à commiseration et à pitié: Toutesfois la braverie, et la constance, moyens tous contraires, ont quelquefois servi à ce mesme effect.

Montaigne 1962: 1

The most usuall way to appease those minds we have offended (when revenge lies in their hands, and that we stand at their mercy) is, by submission to move them to commiseration and pitty: Nevertheless, courage, constancie, and resolution (meanes altogether opposite) have sometimes wrought the same effect.

Florio 1933: 3

The commonest way of softening the hearts of those we have offended, when, vengeance in hand, they hold us at their mercy, is by submission to move them to commiseration and pity. However, audacity and steadfastness – entirely contrary means – have sometimes served to produce the same effect.

Frame 1958: 3

The most common way of softening the hearts of those we have offended once they have us at their mercy with vengeance at hand is to move them to commiseration and pity by our submissiveness. Yet flat contrary means, bravery and steadfastness, have sometimes served to produce the same effect.

Screech 1993: 3

For a reader today, Florio’s Elizabethan English stands out conspicuously, not only because of his unstandardized spelling and punctuation, but because of his early modern lexicon and syntax. We readily recognize the period quality of words and phrases like “constancie” and “stand at their mercy,” as well as the practice of substituting “that” for the previous conjunction “when.” In addition, Florio’s use of “wrought” with the general meaning of “produced” is now obsolete, since this verb has come to be restricted to artistic or ornamental contexts. His reliance on current usage points to his cultivation of a fluent strategy that would make the translation immediately intelligible to his contemporaries. This aim can also be glimpsed in his use of parentheses to improve readability: they subordinate qualifying clauses and phrases that are not essential to what he apparently took to be the primary meaning of the sentences.

Frame’s version, in sharp contrast, is largely written in modern English. This is perhaps most noticeable in its lexical differences from Florio’s text, especially the use of the phrase “served to produce” in place of “wrought.” Yet “submission” in the sense of “submissiveness” had become an archaism by Frame’s time, a development suggested by the very appearance of this usage in Florio’s English. Similarly, Frame’s relatively complicated syntax, evident here in his tendency to embed phrases (“vengeance in hand,” “by submission”), endows his translation with a formality that is literary, if not quite archaic, and that is matched by his choice of the Latinate word “audacity.” Frame too sought a fluent strategy that would make his translation readable to his contemporaries: apart from his general adherence to current usage, he carefully inserted commas that clarify syntactical connections and thereby increase intelligibility. Nonetheless, the result was a historically specific version. Translating canonical authors with a formal style was actually a prevalent strategy among English-language translators during the 1950s, when archaisms were used occasionally for literary or poetical effect (Venuti 2000: 480).

Set against the previous two versions, Screech’s appears remarkably contemporary. His lexical choices avoid archaism: they consistently conform to current usage and rely on the most familiar forms, including some that have a conversational quality: “once they have us at their mercy,” “submissiveness,” “flat,” “bravery.” At points, he also departs from the French syntax to create more easily readable constructions. The shifts include the insertion of the preposition “with” and the possessive adjective “our,” both increasing cohesiveness; the movement of the phrase “by our submissiveness” to the end of the first sentence, making the syntax more continuous; and the reversal of the word order of the second sentence, whereby it begins with the explanatory phrase “flat contrary means.” The familiarity of Screech’s lexicon and the smooth linearity of his syntax guarantee that his version is now significantly more fluent than Frame’s as well as Florio’s.

This analysis, although rapid and selective in the linguistic features it has isolated for comment, allows us to draw several conclusions about the historicity of discursive strategies. It is first important to observe that the historical markers of the translations have no impact on their accuracy, that indeed the three translators seem to have applied the same standard in this respect: a close semantic correspondence to the French text that follows Montaigne’s lexicon and even his syntax where the differences between the two languages permit. The standard of accuracy can be called the same in each case even though Florio’s version is unique in including the word “resolution”: this is in fact another historical aspect of his translation, since it shows that he relied on an earlier edition of the French text (1580) which retained this subsequently deleted word. A second point is that although each translation develops a fluent strategy, their differences indicate that what constitutes fluency varies from one period to another, especially insofar as it depends on current usage and languages are constantly changing. Finally, despite the application of the same standard of accuracy, despite the development of the same discursive strategy, it is no exaggeration to say that Montaigne appears as a rather different writer in each translation, characterized by a different style and tone. Such differences make clear that a translation can be linguistically correct and yet offer nothing more than a representation of the foreign text that varies according to historical developments in the translating language and culture.

The language of pragmatic translations is similarly linked to their historical moments. Text types that are defined by topicality or currency, such as various kinds of journalism, are likely to be rendered, not only with current usage, but with the latest jargons and neologisms. The magazine b-guided, published in Spanish with accompanying English translations and sold in Spain, combines travel with cultural reporting aimed at a young adult readership: it serves as a guide to Spanish cities by focusing on such areas as art, fashion, and nightlife. The following extract describes a Barcelona shop that sells only denim clothing and bears the appropriately Hispanicized English name “Overales & Bluyines” (for “overalls and blue jeans”):

Ubicada en el Born, una de las zonas más dinámicas de la ciudad, centra su oferta en el denim. A su colección propia de series limitades se suma una selección de primeras marcas y un stock de segunda mano en constante renovación. Además, recuperan y personalizan piezas.

Located in the Born, one of the city’s most dynamic areas, and specialising in denim. Along with their own limited series, they stock a selection of top brands and a constantly updated second-hand collection. They also recycle garments, adding their own touch.

b-guided 2002: 154

The lexicon and syntax of the English version are readily comprehensible to a contemporary reader, even though some forms originated in earlier periods of the language. The word “brands” is derived from the nineteenth-century compound “brand-names”; the word “denim” first appeared in the late seventeenth century to signify a serge fabric, but was later applied to the coloured cotton twill that is familiar today. Most of the forms unmistakably date the translation in the present. The fashion term “collection,” meaning a line of clothes presented by a designer, is a twentieth-century usage, as is the term “recycle,” initially used for industrial materials but subsequently applied to a variety of objects, most recently clothing and refuse. The word “update” achieved wide circulation in English only after 1950. The translator’s intention to produce an effect of trendy informality is perhaps most evident in the shifts: the Spanish text opens with a grammatically complete sentence, whereas the English resorts to a more casual fragment; the Spanish uses “personalizan” (“personalize”) to describe the shop’s distinctive treatment of recycled clothes, whereas the English relies on an expansion, the phrase “adding their own touch,” which uses “touch” in the twentieth-century sense of a characteristic skill or ability. The Spanish text certainly aims for the same effect of trendiness with similarly recent usages, most obvious in the English loan-word “stock” and in the fashion jargon (“colección,” “marcas”). Still, it is clear that the two languages have developed in different ways, at different speeds.

So far we have construed the historical dimension of the translating language as an affiliation to current usage in a particular period. Some translators have taken a different route by deliberately inventing a historically specific language to produce certain effects, whether literary, cultural or social. Ezra Pound, for example, showed how an archaic foreign poem might be rendered by imitating an analogous poetry in the translating language which was equally archaic, even if the imitation didn’t establish a perfect stylistic or temporal fit. In “Guido’s Relations” (1929), he described his effort to translate the work of the thirteenth-century Italian poet Guido Cavalcanti by drawing on “pre-Elizabethan English,” the language used by such early sixteenth-century poets as Wyatt and Surrey (Anderson 1983: 250). Pound felt that the archaic style of his translations was useful in evoking qualities of Cavalcanti’s poetic language, “clarity and explicitness,” although the dense archaism of the English versions made clear that he had developed a modernist interpretation of the Italian texts which favored linguistic precision (Venuti 1995: 190-200). Perhaps the most important effect of Pound’s archaizing strategy was to historicize his translations, to suggest – indirectly, through his very choice of archaic English forms – that the Italian texts had been produced in a historically remote culture.

Translation traditions

When translation is considered from a historical perspective, it becomes possible to sketch traditions in which specific practices are repeatedly performed for decades, centuries, even millennia. The factor that historians most often use to codify a translation tradition is a discursive strategy. Thus Antoine Berman offered an “analytic” account of the “deforming tendencies” that manipulate the foreign text, such as clarification and expansion, arguing that they constitute “a two-millennium-old tradition” in the West (Berman 1985/2000: 286). For Berman, these discursive moves are “universals of deformation inherent in translating as such” (ibid.: 296), regardless of the norms that might obtain in a language and culture at a particular historical moment. He traced the universals back to classical antiquity:

From its very beginnings, western translation has been an embellishing restitution of meaning, based on the typically Platonic separation between spirit and letter, sense and word, content and form, the sensible and the non-sensible. When it is assumed today that translation (including non-literary translation) must produce a “clear” and “elegant” text (even if the original does not possess these qualities), the affirmation assumes the Platonic figure of translating, even if unconsciously. All the tendencies noted in the analytic lead to the same result: the production of a text that is more “clear,” more “elegant,” more “fluent,” more “pure” than the original. They are the destruction of the letter in favor of meaning.

ibid.: 296-297

Berman was obviously referring to what I have called the dominance of fluency, a strategy that aims to communicate a meaning for the foreign text (which is to say a particular interpretation of its meaning) at the expense of its formal features. Yet if this strategy is in fact universal, we must recognize that it also characterizes translation traditions in other parts of the world. The late Qing translator Yan Fu described several criteria for producing a good translation – “faithfulness (xin), comprehensibility (da) and elegance (ya)” – which also appeared in ancient Chinese translation theory, specifically in translations of Buddhist scripture during the third century A.D. (Chan 2004: 69). The dominance of fluency may well reflect not so much a western philosophical tradition as the millennia-long dominance of metaphysical thinking in human cultures, a privileging of the semantic “spirit” over the formal “letter” in language.

Berman’s universals would seem to undermine the very notion of a translation tradition, since they risk installing translation in a realm that transcends time and place. Yet his distinction between universals and norms can actually be useful in illuminating the historical dimension of translation practices: it allows for the likelihood that the deforming tendencies will vary from one period to another, taking different discursive shapes according to changing linguistic and cultural norms (see Toury 1995: 53-69). The relations between translation universals and norms, as Berman himself realized, are subject to historical variation. In some periods, the practices represented by each category may be consistent, overlapping and mutually reinforcing; in other periods, they may be contradictory and mutually interrogative; and in still others, they may display different degrees of consistency and contradiction. For the historian of translation, then, a careful contextualization of translation practices, distinguishing universals from both contemporary and past norms, becomes crucial in describing any tradition. For the translator, a historical perspective becomes crucial for developing a more critically aware approach to the current practices that any translator may happen to use, whether through sheer habit or because of a brief provided by a commissioner. “A translator without a historical consciousness,” wrote Berman, remains “a prisoner to his representation of translating and to those representations that convey the ‘social discourses’ of the moment,” where the term “discourses” can be defined as norms or dominant social values (Berman 1995: 61, my translation).

One such norm that is important for the history of translation is nationalism. The rise of modern translation traditions in western countries such as Great Britain, France, and Germany coincided with an increasing sense that languages and cultures are national in significance, expressive of the identities and destinies of national collectives. Hence, even though Abbé Prévost clearly deployed Berman’s universals of deformation, we might also recognize that his version of Richardson’s novel positioned him in a French neoclassical tradition that began in the seventeenth century with a translator like Nicolas Perrot d’Ablancourt, that continued in the eighteenth century with such translators as Antoine Houdar de la Motte and Pierre le Tourneur, and that culminated at the end of that century with the aesthetician Charles Batteux’s prescriptions for translation. Not only did this tradition routinely apply a fluent discursive strategy to classical and contemporary literatures so as to produce the illusion of transparency, but it also assumed that fluency is a timeless and universal value while inscribing foreign texts with the Enlightenment ideas that were then circulating in France.

Thus in 1640 Perrot d’Ablancourt felt that when rendering Tacitus’ extremely concise Latin into French “one is forced to add something to the thought to clarify it” and to avoid “offending the delicacy of our language and the correctness of reason” (Perrot d’Ablancourt 1640: preface, my translation). Nonetheless, the very need to revise the Latin text pointed to the fact that the translator’s concept of “reason” was historically situated, specific to current French culture, and not an essential trait of humanity that Tacitus might have been expected to share. In 1714 de la Motte was even more explicit in his assimilation of Homer to prevailing French values: “I have tried to ensure continuity of character,” he wrote, “since it is this point – which has become so well established in our time – to which the reader is most sensitive, and that also makes him the sternest judge” (Lefevere 1992: 30). In his 1769 version of Edward Young’s poems, Night Thoughts, le Tourneur similarly announced his “intention to distill from the English Young a French one to be read with pleasure and interest by French readers who would not have to ask themselves whether the book they were reading was a copy or an original” (ibid.: 39). In this context, Prévost can be seen as merely following a well-established practice of cultural assimilation when in his version of Clarissa he admitted that he “suppressed English customs where they appear shocking to other nations, or made them conform to customs prevalent in the rest of Europe” (ibid.: 40). Since he had written a French translation, however, the nation he had foremost in mind was undoubtedly France.

Batteux’s 1777 treatise, Principes de la littérature, summed up the French neoclassical tradition. He made no effort to resolve or conceal the various contradictions in its translation practices: at this late stage, the validity of these practices must have seemed self-evident, natural, far from questionable in their affiliation to a particular culture. Batteux’s first “principle” asserted that the translator must retain “all stylistic features” of the foreign text, starting with the “order” of “facts or arguments, since that order is the same in all languages and since it is tied to human nature” (ibid.: 118). Yet he finally rejected this principle to champion the fluency that long prevailed in French literature as well as translation:

we must totally abandon the style of the text we translate when meaning demands that we do so for the sake of clarity, when feeling demands it for the sake of vividness, or when harmony demands it for the sake of pleasure.

ibid.: 120

In the neoclassical tradition, as Berman noted, the universals of deforming translation were entirely consistent with contemporary literary norms in France. At the same time, however, this consistency implicitly subverted the Enlightenment notion of an essential “human nature” because the translation practices that recurred in the tradition were distinctively French, reflecting French literary and cultural values.

Translation traditions need not be characterized by continuity, by the repetition of similar practices over long stretches of time. They might also contain sharp discontinuities, agonistic episodes in which translators mount challenges to linguistic and cultural norms by deploying innovative strategies that nevertheless remain consistent with the universals of deformation. These challenges, moreover, might be inspired by theories and practices that were developed in a different language and culture. Because translation traditions traffic in the foreign, they are vulnerable to foreign influences which often prove to be the motors of linguistic and cultural change.

The Victorian period, for example, witnessed the publication of several translations that display a dense archaism drawn from different periods of English literature. These translations include Francis Newman’s version of the Iliad (1856), Dante Gabriel Rossetti’s versions of thirteenth-century Italian poetry, The Early Italian Poets (1861), and William Morris’s versions of classical epics, the Aeneid (1875) and the Odyssey (1887-8). In developing archaizing strategies, the translators were questioning the fluency that by that point had dominated British translation for more than two centuries. Newman in particular was following the example set by earlier German theorists and translators like Friedrich Schleiermacher who wished to communicate the linguistic and cultural differences of foreign texts. Newman’s goal, as he put it, was “to retain every peculiarity of the original, so far as I am able, with the greater care, the more foreign it may happen to be” (Newman 1856: xvi, his emphasis).

The Victorian translators generally believed that their strategies enabled them to achieve greater accuracy than a fluent discourse that assimilated the foreign text to the norms of the receiving culture. All the same, their translations exemplify some of Berman’s deforming tendencies insofar as their choices reveal significant shifts. Perhaps the most obvious shift involves the historical dimension of their lexicon and syntax: none of the foreign texts uses language that was archaic at the time of its composition. In Newman’s case, the shift was also prosodic: he cast his translation in the meter of the late medieval English folk ballads, departing from the Homeric hexameter because he argued that Homer’s “popular” style demands an English “poetry which aims to be antiquated and popular” (ibid.: xii). Thus, the archaism that justified the Victorian translators’ claim of accuracy simultaneously undermined that claim by deforming the foreign texts. Their challenge to fluency was short-lived, but it established a precedent that would later be followed by modernist translators like Pound.

Another factor that can be useful in sketching a translation tradition is the conceptual discourse or interpretation that translators inscribe in the foreign text. This interpretation can remain relatively stable within the receiving culture, housed in social institutions like the academy and leading to the adoption of particular discursive strategies. In 1860 Matthew Arnold’s view that only classical scholars were competent to judge English versions of the Iliad sparked a controversy with Newman which made clear Arnold’s investment in the prevailing academic reading of the Greek text. In Arnold’s formulation of this reading, “Homer is rapid in his movement, Homer is plain in his words and style, Homer is simple in his ideas, Homer is noble in his manner” (Arnold 1960: 141). On this interpretive basis Arnold criticized Newman’s version, arguing that the choice of the ballad as a metrical model was especially inappropriate: whereas “Homer’s manner and movement are always both noble and powerful,” Arnold wrote, “the ballad-manner and movement are often either jaunty and smart, so not noble; or jog-trot and humdrum, so not powerful” (ibid.: 128). He approved of the hexameter translations done by a Greek scholar at Eton, E.C. Hawtrey. In the twentieth century, imitations of the Greek hexameter in fact became the norm for Homeric translations. This form can be perceived in the widely circulated versions of American scholars such as Richmond Lattimore (1951) and Robert Fagles (1990). Fagles indicated the Victorian origins of the interpretation that guided his translation when he referred to Homer’s “speed, directness and simplicity that Matthew Arnold heard – and his nobility too, elusive yet undeniable, that Arnold chased but never really caught” (Fagles 1990: ix).

Translation traditions represent a history of receiving foreign texts which necessarily deviates from the reception of those texts in the language and culture where they were produced. The native significance and value of a foreign text are rarely reproduced when it is translated because the translation assumes a place in the different traditions of the receiving language and culture. These traditions not only assign another set of meanings to the translation, but allow it to exert a particular influence on subsequent writing in the receiving situation. Much depends on how the translation is circulated and read there, on its own reception, which is usually controlled by publishers, reviewers, scholars, and teachers in a variety of social institutions. The reception of a translation can even involve a concealment of its translated status. Readers, whether or not they know the foreign language, may respond to the translation as if it were in fact the foreign text. Other readers may take the translation for an original composition in the translating language.

The English reception of Petrarch’s poetry offers an especially revealing case. By the early sixteenth century, Petrarch had achieved canonical status in Italy because of commentators such as Pietro Bembo, who not only brought out an important edition of the poet’s works in 1501, but argued in his own essay Prose della volgar lingua (Writings in the Vernacular, 1525) that the language of Italian literature should be based on the Tuscan dialect used by fourteenth-century writers such as Boccaccio and Petrarch. Shortly thereafter, Sir Thomas Wyatt began to translate some of Petrarch’s sonnets into English, although his motives were far removed from literary ambition. Wyatt was more interested in serving as a courtier and diplomat under Henry VII, and so his translations were the fruit of his leisure, circulating only in manuscript and read by friends, lovers, fellow courtiers. The cultural uses to which Bembo and Wyatt had put Petrarch’s poetry were thus widely divergent.

In 1557 Wyatt’s translations were first published in a miscellany edited by Richard Tottel who, however, did not identify them as translations or anywhere mention Petrarch. As a publisher, Tottel was of course concerned about selling books, but his brief prefatory statement showed that his suppression of the Italian source (if in fact he were aware of that source) served another, ideological purpose:

That to haue wel written in verse, yea & in small parcelles, deserueth great praise, the workes of diuers Latines, Italians, and other doe proue sufficiently. That our tong is able in that kynde to do as praiseworthely as the rest, the honorable stile of the noble earle of Surrey, and the weightinesse of the depewitted sir Thomas Wyat the elders verse, with seuerall graces in sondry good Englishe writers, doe show abundantly.

Tottel 1557: A1r

Tottel’s appeal to his potential readers was based on a vernacular nationalism: he needed to present Wyatt’s poetry as original compositions to demonstrate that English writers were capable of competing favorably against foreign-language writers such as Petrarch. Wyatt’s translations were published as his own work, “to the honor of the Englishe tong, and for the profit of the studious of Englishe,” and since the miscellany went through many editions over the next three decades, it enabled Wyatt’s translations to influence Elizabethan poetry by initiating a craze for the Petrarchan love sonnet. Whereas Bembo had used Petrarch to recommend a particular Italian dialect for the creation of a national Italian literature, Tottel suppressed Wyatt’s Italian source to champion a national English literature. The histories of the foreign texts and the translations could not have been more different, even if nationalist agendas can be perceived in their receptions.

Translation traditions have also been deliberately constructed for various cultural and social purposes. In glancing back at the past, a translator might gather the work of predecessors into a sequence of which they themselves had no awareness during their lives. This invention of tradition might be done to validate the translator’s own work by aligning it to previous efforts that have gained cultural value while distinguishing it from them. This is evident in the commentaries of the most influential English translation theorist, John Dryden, who on more than one occasion presented his work by situating it in a tradition of English poetry translation. In the preface to his anthology, Ovid’s Epistles (1680), he described three translation methods, each of which he illustrated by citing a different translator:

First, that of Metaphrase, or turning an Authour word by word, and Line by Line, from one Language into another. Thus, or near this manner, was Horace his Art of Poetry translated by Ben. Johnson. The second way is that of Paraphrase, or Translation with Latitude, where the Authour is kept in view by the Translator, so as never to be lost, but his words are not so strictly follow’d as his sense, and that too is admitted to be amplyfied, but not alter’d. Such is Mr. Wallers Translation of Virgils Fourth Aeneid. The Third way is that of Imitation, where the Translator (if now he has not lost that Name) assumes the liberty not only to vary from the words and sence, but to forsake them both as he sees occasion: and taking only some general hints from the Original, to run division on the ground-work, as he pleases. Such is Mr. Cowleys practice in turning two Odes of Pindar, and one of Horace into English.

Dryden 1956: 182

Dryden’s examples constitute a chronological order: the three poets – Ben Jonson, Edmund Waller, and Abraham Cowley – represent three successive generations spanning the late sixteenth and early seventeenth centuries. By Dryden’s time, their poetry had assumed considerable authority, making it worthy of imitation. Dryden was thus tracing a canonical tradition of English literary translation by including only translators who were influential poets. For Dryden, moreover, “paraphrase” was the most effective approach because he aimed not simply to establish a semantic correspondence to the foreign text, but to develop a fluent discursive strategy. Hence, any resemblance between his and Goethe’s later description is misleading: whereas the German writer praised what Dryden termed “metaphrase” as the highest form of translating, the English poet-translator rejected it since “either perspicuity or gracefulness will frequently be wanting” (ibid.: 183).

Historical narratives

Although no history of translation can be written without extensive research into the past, the factual data that the historian collects do not themselves yield the significance of translation practices in a particular period or over time. Indeed, if the facts are simply arranged in chronological order, they lack any meaning that would explain or interpret them. This meaning comes, as Hayden White has argued, from the kind of “emplotment” that the facts are given by the historian (White 1978: 91-95; see also White 1973). In White’s words, “by a specific arrangement of the events reported in the documents, and without offense to the truth value of the facts selected, a given sequence of events can be emplotted in a number of different ways,” each of which carries explanatory force (White 1978: 61). White defines these “ways” as traditional narrative genres, namely romance, tragedy, comedy, and satire (for an account of the genres, see also Frye 1957). A history of translation, then, like any history, endows translation practices with significance through a specific narrative form or mixture of forms.

Goethe’s historical account of German translation methods is basically structured as a romance. He inserted them in an evolutionary or progressive narrative which, as is typical of the romantic genre, culminated in a sort of transcendence specific to translation:

we have lived through the third epoch, which could be called the highest and final one, namely the one in which the aim is to make the original identical with the translation, so that one should be valued not instead of the other, but in the other’s stead.

Lefevere 1992: 76

Since Goethe characterized the three methods according to their treatment of the linguistic and cultural differences of the foreign text, the third can be called the “highest” because it reproduced those differences so closely as to transcend them – or in fact to transcend the very distinction between foreign text and translation.

The narrative that informs a translation history, as this example suggests, turns on the particular factors that the historian selects to describe the chronological succession of translation practices. These factors are drawn from the basic constituents of any translation practice: discursive strategies and conceptual discourses, the translator’s agency, especially in relation to commissioning institutions and cultural norms, and the reception of the translated text. Goethe’s history also addressed reception, whereby he was able to invest his overall romantic narrative with what we might call a comedic subplot of reconciliation. For, as in the genre of comedy, Goethe sketched a plot in which obstacles that frustrate personal desire and strain social relations are finally removed and a new social formation emerges:

Originally this kind of translation [i.e., the third epoch] had to overcome the greatest resistance, since the translator who attaches himself closely to his original more or less abandons the originality of his own nation, with the result that a third essence comes into existence, and the taste of the multitude must first be shaped to accept it.

Ibid.: 76-77

Here the desire that motivates the translator to reproduce the differences of the foreign text encounters the “resistance” of readers in the receiving culture, but is subsequently satisfied in the creation of a new cultural constituency on a national scale.

Historical narratives that take the translator’s agency as their main focus can be especially complicated because of the many conditions, conscious and unconscious, individual and social, that shape the production of any translation. As a result, such narratives may reveal a contradictory combination of genres. Consider Susanne Starke’s account of the English women who translated German texts during the nineteenth century. Her focus on the translator’s agency is explicitly stated at the outset: her point of departure is Virginia Woolf’s view that she could achieve sufficient intellectual independence to become a writer only by rejecting or “killing” the “Angel in the House,” the submissive, self-effacing role that male-dominated Victorian society had established for women (Starke 1999: 31). Starke’s historical narrative thus begins as a romance wherein English women embark on careers as professional translators so as to emancipate themselves from the repressive gender hierarchy. Catherine and Susanna Winkworth, for example, were prevented by this hierarchical order from taking up the preaching or publishing careers for which their religious education might have prepared them. Yet they could become successful translators, as Starke explains:

Having made the decision to cut herself off from original discourse, Catherine gained lasting fame for her translations of German hymns, an endeavour to which she was well-suited. To preach or publish her religious beliefs was impossible; to translate the religious poetry of male authors, however, offered an ideal opportunity to communicate her own deepest convictions without articulating them herself.

my italics

Ibid.: 37

The italicized phrases run counter to the romantic narrative of emancipation and point to the presence of another genre, tragedy, in which the violation of a human or divine law leads to a downfall. The law faced by Victorian women with literary aspirations was patriarchy, which reserved original authorship for men. Since translation is a form of writing, even if derivative, these woman still risked a violation of the gender hierarchy.

Starke’s narrative oscillates between romance and tragedy as she quotes the women’s self-effacing statements, on the one hand, and documents their professional careers, on the other. Occasionally, the history dips into tragic irony, with the translators challenging gender roles “inadvertently”:

However much the female translators in question might have wished to distance themselves from what they perceived as a male role by securing themselves “behind the welcome defence of inverted commas,” and however much they might have wished to comply with what they would have considered to be an appropriate female role, they have nonetheless inadvertently slipped into the mode of literary professionalism.

Ibid.: 39

The construction of this sentence, the opposition between the repeated “however much” and the insistent “nonetheless,” suggests that the historian’s own “wish” is driving the romantic narrative while a scholarly commitment to documentation is pushing that narrative into a more tragic direction. In the end, Starke is unable to conclude with a romantic transcendence, asserting instead that “if we assume that the translators in question also killed that angel, we shall have to argue that they have accomplished the deed in disguise,” because they did it through translation “and without ever admitting it” (ibid.: 57-58). Irony ultimately governs the mixed genre of this translation history: any gender emancipation that might be glimpsed remains an unintended consequence.

It is satiric narratives, however, that typically display the interrogative power of irony. As a mode of emplotment in a translation history, satire is accompanied by skepticism towards translation practices and their cultural and social effects, often in opposition to the claims made by translators or the institutions that have commissioned their work. The satiric historian, as White observes, adopts “agnosticism or cynicism as a moral posture” (White 1978: 74).

An illuminating example is offered by Anthony Pym’s account of translation practices during the 1992 Olympics in Barcelona. Pym’s narrative is fundamentally satiric: he intends to criticize, often with a derisive tone, not only the nationalism fostered by the Olympic Games themselves, but any nationalistic investment in the particular languages of the participants. He reserves his most critical remarks for Catalunya, since the site of the Games led to the inclusion of Catalan among the official languages, along with English, French, and Castilian, requiring that the flood of information on the event be translated into all four languages.

The satiric form of Pym’s narrative is first noticeable in a succession of comments that question the nationalistic attitude towards Catalan taken by the government, the Generalitat of Catalunya:

The Generalitat itself has a very elaborate and expensive policy for the promotion and standardization of Catalan, surreptitiously combating the language’s regional varieties (although this is rarely admitted).

The students translating at the central press service worked into English and French and had their work revised by professionals. Yet no students were employed for written work into Catalan, given that only professionals could apparently write correct Catalan.

All the official languages were equal, but some had more translators and greater market demands than others.

Pym 2000: 214, 215

Pym’s rhetorical strategy is far from subtle: it relies on sniping (“surreptitiously combating,” “rarely admitted”), innuendo (“apparently”), and sarcasm (the juxtaposition of “equal” to the comparatives, “more” and “greater”). The cynical posture that underlies his narrative becomes most obvious in his conclusion: the relative “market demands” of the official languages, he argues, reduced the inclusion of Catalan to a purely “symbolic” gesture of “equality” (ibid.: 215, 216).

It is worth emphasizing that the data in Pym’s account can acquire a completely different significance if it is inserted into, say, a romantic narrative that is more sympathetic to Catalan culture. This genre would rather emphasize the historical conditions to which Pym gives virtually no attention: the Franco regime sought to repress Catalan through various prohibitions on its use in the press and in the schools, so that native proficiency in the language deteriorated and new generations needed to relearn the forms that had been standardized at the start of the twentieth century (Balcells 1996: 127, 143-144). No wonder, then – a romantic historian might argue – that the Generalitat takes an active role in preserving and developing a standardized form of the language, despite regional variations. No wonder that only professionals with a sure grasp on the language undertook the Catalan translations for the Olympics. And no wonder that Catalan, although spoken by roughly 10 million people, remains a minor language lacking the market demand of such other culturally and politically powerful languages as English, French, and Castilian. These points – the historian might continue – indicate that the symbolic value of Catalan’s official status at the Olympics constitutes an achievement that could not have been predicted even a decade earlier: it shows that Catalan has transcended historical obstacles to become a viable national language, a development that can also be seen today in a thriving publishing industry that issues many translations. Pym’s narrative works its satire, not only by treating the data with ironic skepticism, but by excluding other possible genres and explanations and by minimizing or simply suppressing other data.

Because literary genres shape historical narratives, scholars and translators must be both self-conscious and self-critical in their construction of histories to explain translation practices. Such narratives expose yet another facet of the relative autonomy of translation, insofar as they represent the intricate network of connections that exist between translated texts and translators, commissioners, and audiences in the receiving culture. The reception of a translation can continue long after its initial publication, furthermore, in scholarly commentaries and histories. And this reception will always be distinct, to an important degree, from the ways in which the foreign text is received in its own cultural situation.