The act of translation and the discipline of art history are not obvious bedfellows, yet have a long and close relationship. Just as it might be argued that all reading is translation, so all art history might be seen to be grounded in ekphrasis, the construction of texts that are able to describe, evoke, and analyse the work of visual art. As the critical theorist W.J.T. Mitchell has noted: ‘Insofar as art history is a verbal representation of visual representation, it is an elevation of ekphrasis to a disciplinary principle’ (Reference MitchellMitchell 1994: 157). And from this base it is only a short step to comparisons between the ekphrastic writing of the art historian and the activity of the translator. Indeed, as defined in a recently published article, ekphrasis ‘translate[s] the visual and sensual nature of a work of art into a linguistic formulation capable of being voiced in a discursive argument. The act of translation is central’ (Reference ElsnerElsner 2010: 12; see also Reference VenutiVenuti 2010). Even the most doggedly attentive attempt to render an image in text is bound, however, to fall short, in that the relationship between the two semiotic systems, visual and verbal, simultaneously embraces both resemblance and difference. The text mimes the picture but can never be it or become it. As Reference EcoUmberto Eco (2001: 97) assures us: ‘The practice of ekphrasis makes it possible to convey an image in words, but no ekphrasis of Raphael's The Wedding of the Virgin could convey the sense of perspective perceived by the viewer, the flowing lines that manifest the position of the bodies, or the tenuous harmony of the colours.’ Given the impossibility of even partial success in this enterprise, pessimistic art historians might be tempted to abandon the exercise entirely. In adopting this fatalistic position, they would be endorsing an equally bleak and extreme conclusion drawn in the realm of translation studies, namely the Sapir-Whorf hypothesis, which insists that ‘communication between two people who do not share the same native language is impossible’ (Reference Schogt, Schulte and BiguenetSchogt 1992: 195).

Yet art history, like scholarship, is based on the desire to say something about a specific object or set of objects, and the need to do this across semiotic codes, from painting to text, for example, is an inescapable problem. To counter the dark and strangely silent world of Sapir-Whorf, it could be argued, following the insights of Peirce, that the meaning of any semiotic item is dependent on ‘the translation of a sign into another system of signs’. In simple terms, Raphael's painting takes on meaning when translated into such verbal signs as ‘virgin’ and ‘wedding’. In linguistic terms, this represents an intersemiotic translation, moving from one sign system to another, from image to text.

Figure 1. Raphael, Lo sposalizio della Vergine, 1504, Pinacoteca di Brera, Milan, Italy.

More directly, interlingual translation – translation from one language to another – is a mainstay of art history. As currently practised, art history is a Western preoccupation, and the great predominance of university departments are to be found in the USA and Europe.^{Footnote 1} Within this Western orbit, English is currently by far the dominant language, a condition confirmed by a recent census of journals devoted to art history and cognate fields. The US leads the way with 123 journals, and is followed by the UK (45), Australia (17), Canada (12), New Zealand (5) and South Africa (4). Unsurprisingly, given this geographical spread, English is the principal vehicle, with ca. 200 journals, 35 in Dutch, and 19 in German. No other European language can point to more than 10 journals (Reference Elkins and BrzyskiElkins 2007: 62–3). With such a marked dependence on one language, translation has played an essential role in defining the canon of Western art, as defined by such writers as Pliny and Vitruvius, Vasari and Serlio, Winckelmann and Hegel, Wölfflin and Riegl. Very clearly, art history in the English-speaking world has always been dependent on translation: from Greek and Latin in antiquity, from Latin and Italian in the Renaissance, from French in the Enlightenment, and from German in the nineteenth and early twentieth centuries.

An interesting overview of the centrality of translation in establishing the canon of Western art history can be gained from an analysis of the three magisterial volumes of source texts on European art, published as Art in Theory: An Anthology of Changing Ideas (Reference Harrison, Wood and GaigerHarrison, Wood & Gaiger 1998–2000). Each volume includes over 250 texts, more than half of which are translations of documents originally published in languages other than English.^{Footnote 2} Two things emerge from an analysis of this selection of key texts. First, without translation there would be no art history; and second, English has become an increasingly dominant voice in this field over the last century, a theme to which this essay will return.^{Footnote 3}

When art history gained academic currency in the later nineteenth century, it was doubly inflected by language. The subject matter was articulated along linguistic divisions, often reflecting national boundaries, and works of art were ordered, exhibited, and studied in the context of national schools: Italian, Spanish, French, etc. At the same time, schools of art writing developed in the major European languages. According to the standard historiography, German was the principal language of the discipline in the early years, superseded in the second half of the twentieth century by English. Only after the enforced exile of the German-speaking intelligentsia by the National Socialist Party in the 1930s did the linguistic balance in art history, as in many other academic disciplines, shift from German to English, greatly aided by the arrival in Britain and the USA of such eminent historians as Ernst Gombrich, Erwin Panofsky, Nikolaus Pevsner, and Aby Warburg.

While the notion of a national school offers a convenient shorthand, it is a crude and insensitive device, which has problems in accommodating the great mobility enjoyed by eminent artists over the centuries, and the power of patronage to attract talent across local and national borders. More worryingly, it is open to exploitation by the crudest forces of nationalism and racism. An extreme example can be found in a book by Dagobert Frey, an Austrian scholar who held the chair of Breslau from 1931 until 1945. In a book entitled Englisches Wesen in der Kunst (The English Character in Art), published rather remarkably in Reference Frey1942, when Germany was at war with Great Britain, Frey labours to establish that formal artistic characteristics are dependent on race and blood lines: ‘In England, too, the tension in the development [of the arts] and the intrinsic dynamics of this development must not be overlooked. They are first and foremost conditioned by the racial composition of the inhabitants. Already in the early Middle Ages, the essential difference between the Mediterranean-Celtic and the Nordic-Anglo Saxon peoples became sharply apparent’ (Reference Frey1942: 442). Although this, as noted above, is an extreme example, it points in its excesses to the problems of classifying art according to national or racial groupings. With the globalization of artistic practice and discourse, the position becomes even more untenable.

A significant route out of the Western-centric and nationalist mindset was opened in the 1970s and 1980s. Led by Edward Said's Orientalism (1978), postcolonial studies instigated close investigations of the role of language and translation within the structures of colonization, and in the production of power and knowledge. These studies made clear that translated knowledge masks inequalities: more texts are translated from the dominant culture (centre) into the colonized culture (periphery) than vice versa; translators working into the dominated culture, ‘integrate foreign objects without question’ (Reference RobinsonRobinson 1997: 36), while those translating into the hegemonic culture incline towards translation as domestication, whereby the foreignness of the source text is, as far as possible, masked and polished out of existence in favour of known and habitual linguistic norms. The insights derived from interlingual translation also offered a valuable metaphor through which to investigate and to describe broader forms of cultural encounter and transformation. First adopted by anthropologists, the term ‘cultural translation’ describes what happens when one culture tries to make sense of another. A famous example is the anthropologist Laura Bohanna's account of ‘how she told the story of Hamlet to group of Tivs in Africa, and heard the story corrected by the elders until it finally matched the patterns of their Tiv culture’ (Reference BurkeBurke 2007: 8). Translation is always a process of decontextualization and recontextualization: it enriches the receiving culture, but may signify violence, loss, or abuse of the source text.

In all cases the translator functions not merely as a vehicle for saying the same thing in different sign systems, but as an interpreter. And as an interpreter, the translator has multiple options, which can be seen to form two interlinked axes of choice. The linguistic axis runs from the aim to cling as closely as possible to the original text at one extreme to an almost entirely free reading at the other, in which the source text survives merely as a catalyst and armature. The second, cultural, axis offers assimilation at one end, with the translation domesticated as far as possible to conform with the linguistic, social, and cultural conventions of the target culture. At its other end, this axis shuns the known and habitual, and rejoices instead in the cultural differences made manifest in the source text. The resulting, foreignizing translation aims at challenging and revitalizing the cultural and linguistic boundaries of the target culture by insisting on difference rather than homogenization. This hermeneutic model of translation is radically transformative, a violent, ethnocentric process, which can be abusive in its rejection of the straight interpretation of the signifier.

Postcolonial theory has taught us that the hegemonic culture usually favours a domesticating approach to translation, but it has also shown us the emptiness of this assimilative position. As Gayatri Chakravorty Spivak noted in the context of translating Bengali poetry: ‘I must resist both the solemnity of chaste Victorian prose and the forced simplicity of “plain English,” that have imposed themselves as the norm … Translation is the most intimate act of reading. I surrender to the text when I translate’ (Reference Spivak and Venuti2004: 370). The alternative to tame domestication is a conscious and spontaneous foreignization that engages vigorously with the particularities and idiosyncrasies of the source text, which it seeks to transmit into the target language. Through the act of translation, the centre should be provincialized and the edges brought towards the centre. In the process, our engagement with text and image will allow us to explore and formulate emotions and concepts that we would otherwise not have.

Clearly, the options open to the translator run from the ultra-faithful to the ultra-inventive, from word-for-word to sense-for-sense, and from the literal to the literary. The perfect solution, an interlinear translation, would embrace all these options, with the original text in the source language accompanied by a transliteration, a vernacular translation into the target language, and any number of comments or layers of analysis. In most instances, however, both the translator and the art historian must commit to a particular position with regard both to the source text and to the target language. Translation theory offers little assistance in this choice, as both extremes find vigorous advocacy. In his essay Die Aufgabe des Übersetzers (The Task of the Translator), originally published as the preface to his 1923 translation of Baudelaire's Tableaux parisiens, Walter Benjamin argues for translation as the necessary continuation (Fortleben) of the original rather than a secondary, subservient activity. This turns the translator into a creative power carrying on and extending the thought processes of the author, rather than jogging along behind as some sort of faithful servant. Umberto Eco, in contrast, insists that all enrichment of the text through translations should be avoided, arguing instead for translation as an act of negotiation, in which a middle path and a compromise solution is sought between the source text and the target text. Eco, therefore, is trying to narrow the gap as far as possible between source and target text, while Benjamin urges that it be left vibratingly open.

The general condition addressed in postcolonial studies of translation as power is exemplified within the world of art history. Not only is it predominantly made in the English-speaking world, but also written from an entirely Western-centric viewpoint, which privileges Western art over the art of the non-Western world. Ernst Gombrich's The Story of Art offers a striking example. Translated into more than 30 languages, the volume takes a Eurocentric view of art, leading the reader from ancient Egypt, Greece, Rome through to the Renaissance and modernism in the Western world. With his focus firmly on the West and its classical origins, Gombrich delivers a compressed and fleeting account of Islamic art and Asian art in just two chapters, diminishing thereby the importance of artistic achievements rooted in non-Western cultures. A more recent survey of modern art, Art Since 1900. Modernism, Antimodernism, Postmodernism (2004) written by the members of the US-based October group, focuses principally on Western art, but also incorporates sections on twentieth-century art produced in Russia, Brazil, Mexico, and Japan. Although international in scope, its mindset is firmly rooted in North America. As Reference Piotrowski, Bru, Baetens and HjartarsoPiotr Piotrowski (2009: 50) has argued, the authors of this survey do not ‘deconstruct the relations between the centre and the margins in the world history of modern art … The art produced outside the centres of Western Europe and the United States are described within, as it were, the Western paradigm.’ Their approach results in a ‘vertical’ narrative, which establishes Western cities as the centres from which all power and influence radiate out. The recognition of artists working at the periphery depends on their relationship with the centre, established through exhibitions organized in the West, and in texts published in the West. The ‘periphery’, therefore, is understood not on its own terms, but always as the ‘other’ from the viewpoint of the centre.

The imbalances of cultural power in today's world of scholarship in art history are obvious: with English established as the lingua franca, British and North American art history is largely mono-lingual. It is little receptive to foreign scholarship, tends to mistrust second-language texts as in any way authentic, or even dependable, while its own scholarly output – theories, values, canons – is readily translated into the non-English speaking world. For those outside the ‘centre’, it is almost a commonplace that the only way to be taken seriously in the international community of art history is to write in, or to be translated into, English (or secondarily German or French). To be accepted for translation and publication in English, non-English speaking authors need to be willing to assimilate their writing to the expectations, standards, and values of the English-language publishers located at the ‘centre’.

What actually happens when a text crosses a linguistic border and what can art history learn from this process? The act of translating is never a passive one-way procedure that reproduces an original in another language, but is invariably an active and transformative process that challenges the purity of the original. Inevitably, it also opens the door to simple mistranslations and to cultural misunderstandings. One of the most notorious instances of mistranslation in art history can be found in Sigmund Freud's seminal essay Leonardo da Vinci and a Memory of his Childhood (1910). Freud's starting point was da Vinci's childhood memory, as recorded in his notebook, of being struck on the lips by the tail of a vulture. This, Freud argued, revealed da Vinci's unconscious sexual fantasies, which, when sublimated, fed into the artist's work. The word ‘vulture’ is used by Freud to relate da Vinci's memory to myths about a bisexual Egyptian goddess; it also prompts Freud to see the shape of a vulture in the folds of the clothes in da Vinci's painting Madonna and Child with St Anne. This entire interpretative apparatus, however, is ungrounded, as it is built upon a mistranslation into German of the Italian word ‘nibbio’ — ‘kite’ — as ‘vulture’ (Reference ShapiroShapiro 1968).

The synchronic nature of translation can also lead to misunderstanding, in that the decisions made in transferring an idea or intention from one language to another may well be valid only at a given moment as they are linked to the habits of a particular speech community at a certain point in time. As the literary critic Reference Steiner and VenutiGeorge Steiner (2004: 195) has argued, the receiving native culture is ‘not a vacuum but already extant and crowded’ so that there are ‘innumerable shadings of assimilation and placement of the newly acquired, ranging from a complete domestication, an at-homeness at the core of the kind which cultural history ascribes to, say, Luther's Bible … all the way to the permanent strangeness and marginality of an artifact such as Nabokov's English-language Onegin’. Art History can offer a variety of interesting cases. For example, in 1919, William Morris's views on the advantages of romantic love over marriage contracts fell foul of contemporary Roman Catholic sensibilities in Catalonia. The text that caused the offence was Morris's utopian tract News from Nowhere (1890) in a Catalan translation by the urbanist Cebrià de Montoliu published in 1918 in a book collection issued under the imprint of the County Council in Barcelona under the direction of Eugeni d’Ors, then in charge of the Education Department. The translation was damned as unworthy of publication by the more conservative forces in the city, who saw it as an apologia for free love. As a result, the unfortunate d’Ors was dismissed from his official position (Reference CalveraCalvera 2002: 88).

In many instances, the translator has to add extra material to the original content in order to make it comprehensible in the target culture. The judgement of how much or how little needs to be added is a key issue in translation, as it pulls the source material ever further away from the site of its creation and domesticates it in a foreign environment. Within the Western canon of art history this is relatively straightforward, for although London and Londres are quite different concepts in their respective linguistic contexts, the two cultures in which they sit are cognate and well-acquainted. The challenge of difference becomes infinitely more intractable in a global context, where an easy acquaintance between source and target cultures is the exception rather than the rule.

The most obvious and direct way of rendering a text as a translator or a painting as an art historian is a word-for-word, image-for-image account. In ancient Rome, for example, translation meant an absolutely strict and literal adherence to the original, and this convention survived well into the Middle Ages. Boethius, for example, defends literal translation in his commentary on Porphyry's Eisagoge, penned around AD 510, by insisting: ‘I fear that I shall commit the fault of the faithful interpreter when I render each word by a word corresponding to it’ (Reference Robinson and FranceRobinson 2000: 16). In the context of art history, this might be compared with a text by Libanius from the fourth century AD, describing a painting in the Council House at Antioch:

For the reasons intimated above, however, it is clear that neither the literal translation nor the purely descriptive account of the painting, regardless of its ekphrastic intent, can create a second text that stimulates the same intellectual and emotional responses as the original. They are at best transmutations or adaptions, analogous efforts produced with different implements. Working backwards, however, it would not be possible to reconstruct the original from the transmutation.

The alternative to tame domestication is a conscious and spontaneous foreignization that engages vigorously with the particularities and idiosyncrasies of the source text, which it seeks to transmit into the target language. Such insights offer art history a heightened awareness and the critical mechanisms with which to negotiate the competing demands of centre and periphery. As argued above, the centre should be provincialized and the edges brought towards the centre. More prosaically, an engagement with work first made in other semiotic languages or codes demands a fidelity of reading that precludes indifference. It will stretch the visual and linguistic boundaries of the target culture and, paradoxically, offer the very best practical training in writing in one's native language, as it sharpens the historian's sense of responsibility as mediator, with literary, ethical, and political responsibilities, not only to the source text but also to the readership.

As pointed out by Peter Burke, the linguistic metaphor has two advantages: it insists that the artist has work to do to make the foreign legible and interesting to the receiving audience. It also avoids judgmental terms, such as misappropriation, misunderstanding, thereby offering a degree of neutrality. These advantages also apply to art history. For instance, the distinctions between domesticating and foreignizing approaches in translation can be helpful to examine hybrid forms of art – from Russian cubo-futurism, Moscow conceptual art, or to Peruvian baroque architecture which combines indigenous elements with Catholic forms and symbolism. In addition, translation as a concept is also useful to expose the motivations, strategies and tactics used whenever one culture attempts to represent another. The pioneering article by Linda Nochlin, ‘The Imaginary Orient’ (1983) springs to mind. The first art historian to engage with Edward Said's Orientalism (1978), she looked at nineteenth century paintings of the so-called Orient by Eugène Delacroix and Jean-Léon Gérôme, arguing that these tell more about the arrogant and formulaic expectations of the colonizers – that the locals are corrupt, deviant, dangerous, and generally depraved – than about real life in the Middle East. Facing the same dilemmas as a translator of texts, the artists mediate between the foreign material and their native audience, and attempt to reconcile fidelity to the original with intelligibility to the viewer at home in France. For instance, Gérôme's The Snake Charmer emphasizes the social, political and moral inferiority of the East, and by implication French superiority, a notion perfectly established within French culture of the time. The adopted style, marked by close attention to detail and polished surfaces, shows no traces of the artist's handiwork, lending the painting a degree of photographic realism, which persuades the viewer of the truthfulness of the scene. However, ‘literalness’, as August Wilhelm Schlegel once noted in the context of translation, ‘is a long way from fidelity’.

How then can a heightened awareness of the challenges, delights, and dangers of translation help art history as a discipline to become more pluralist, multi-directional, and horizontal in terms of its authors, viewpoints, and subject matter? While the data above paints a pessimistic picture, there are a number of initiatives to counter the hegemony of Anglo-American art history and to invite in other voices. February 2009 saw the publication of the first issue of Art in Translation, a new e-journal that publishes English translations of critical essays and articles from foreign-language journals and catalogues ranging across all fields of art history. While not attempting encyclopedic coverage, the journal acts as a window onto the practice of art history and visual culture in the non-English speaking world. The texts, which have already been published in their source languages, are drawn from all areas of the visual arts (painting and drawing, sculpture, architecture, design, installation works, electronic media, art theory) and introduce the English-language readership to new worlds of scholarship and writing in the visual arts. The journal also includes a selection of key texts from earlier decades that have never before been available in English. The mission of Art in Translation is to open, via the act of translation, new vistas into historical and current scholarship and production in the visual arts worldwide. The journal itself, and the collaborative effort involved in producing it, acts as a significant catalyst in connecting scholars worldwide, in bringing important scholarship to broader attention, and in making art history a more broad-based discipline. The journal brings attention and repute to texts that are undeservedly unknown both to the audience of native English speakers and to the even larger audience that has English as a second or third language. At least two major art historical associations now support tri- or quattro-lingual publications, thereby reinstating the polyglot, transnational character of art history. Also in 2009, EAM (‘European Avant-Garde and Modernism Studies’) published its first trilingual volume of a collection of essays written in English, French, and German, with the aim to compare ‘French, German and British, and also Northern and Southern as well as Central and Eastern European findings in avant-garde and modernism studies’ (Reference BruBru et al 2009, Preface). The collection further seeks to ‘bring out the complexity of the European avant-garde and modernism by relating it to Europe's intricate history, multiculturalism and multilingualism’ (ibid, back cover text). Another recent initiative is the journal of RIHA (International Association of Research Institutes in the History of Art), which invites submissions of articles, via its member institutions, in one of its four ‘official’ languages: French, English, Spanish, or German. A conference entitled ‘Art in Translation’, held in Reykjavik in May 2010, invited scholars and artists to reflect on translation, understood in the broadest sense, within art and art history, especially in the contemporary context. Further evidence of a growing awareness of the isolation of non-Western art history can be seen in the programme of the forthcoming CAA conference to be held in New York in 2011. One session specifically invites non-Western art historians, working outside the West, to present papers on any kind of art historical subject, but preferably providing insight into approaches or methodological issues conditioned by the scholar's position outside the West.

Figure 2. Jean-Léon Gérôme, The Snake Charmer, 1886, Clark Institute, Williamstown, MA, USA.

Clearly, the very process of selecting papers for conferences and texts for translation and publication is an exercise of power, but one informed, it is to be hoped, by the heightened awareness of power relations stimulated by the act of translation itself, and by an engagement with the cultural implications of translation.