Imagine you are spying on your town or city, peering into malls, homes, computers, bookshelves, electronic devices carried on public transport. Where is literature? And if you can find it, where are literary translations? Where might they be read, talked about, or produced, if at all?
Mohsen Kafi has done some detective work of this kind in Wellington, New Zealand, where he finds no strong reason for pessimism (“Wellington Readers’ Perceptions”). There are still bookstores there, albeit with translations mostly stacked on rear shelves. There are embassies in town, so translations help showcase foreign cultures at public events. There is also a biennial Writers Week, where foreign authors are occasionally invited and translated. Further, some twenty percent of the population speaks a language in addition to the official English and Māori, so translations tend to appear in the reading histories of people who cross cultural boundaries. Some of those things might be found in your town or city—translations are all around us. Kafi did not look for active translators, though, since his focus was on reading. Moreover, he looked everywhere except where he himself was located: the university, which is surely where active translating is done in the teaching of foreign languages, where translations are actively used when engaging with foreign literatures, where students acquire ideas about what translations are and how they are to be approached, and where future literary translators might find some kind of haphazard training—translators are one of the products of a modern language department. Whatever is done there, at the university, seems likely to have some effect on the way translations function in the wider society.
Here I ask what combinations of knowledge and skills can be promoted in our teaching so that they enrich the social roles of translations. The combinations have become known as “translation literacy,” understood as the ability to make informed decisions about when to turn to translations, how to read them, how to compare them, when to trust them, when to intervene in them, and also, I suggest, how to produce them. The same term has been used to describe translating as an academic writing skill (Klitgård); as a combination of basic metalanguage, knowledge, and skills concerning how to translate (Takeda and Yamada); and as a process that starts by challenging students’ preconceptions of translation as a “linguistic matching game” (Baer 8). The term is also conceptually close to the ability to use machine translation intelligently (Bowker and Buitrago Ciro), where the fundamental starting point is similarly the need to break negative preconceptions about fixed mechanical relations. For a properly dynamic view of literacy, translation mostly involves choosing between alternatives. That is, when there are no alternatives, you are dealing with the rules of terminology or comparative grammar, not translation.
Beyond that initial awareness that translation is not simply a matching game, the development of translation literacy can be taken in several directions. Perhaps the most obvious and widespread is to compare different translations in the classroom, to engage in what Marilyn Gaddis Rose called “stereoscopic reading” (81, 90), inviting students to discuss variants, express preferences, and speculate on possible alternatives. Those would be aspects of a readerly translation literacy, a complex engagement that can help develop understandings of foreign languages and cultures. Much can be done to build on these insights and integrate a variety of active translation practices into the classroom, where translating is done in one way or another. Such integration can and should be promoted, I propose, because active translation is increasingly part of the way new generations interact with language. Our students are turning to online machine translation in the way that previous generations used dictionaries, most are accessing literary texts through audiovisual media, they are remarkably skilled at reading subtitles, some are fan-subbing, and so on.
For several years, I have been experimenting with activities that promote translation literacy in the classroom. Here I present a few of them. My classes have mostly been language-neutral, bringing together as many as seven languages other than English, allowing students to show one another what happens in their respective languages and cultures. Other classes have been language-specific, where the activities can explore more focused kinds of problem solving. Some of the courses have been in translator-training institutions, where active literacy is an obvious priority. What you can do depends on where you are and who you are teaching.
Here, then, are my suggestions for the classroom.
Use machine translation
The best way to show the limitations and advantages of online machine translation is to use it in class. Pick a rich translation problem, see how published translators have solved it, then see how machine translation deals with it. To break with the presupposition of symmetrical matching, compare different machine translation systems (see examples in Hadley et al.) and then the AI solutions such as those given by ChatGPT. In the case of well-known texts like Dante's Inferno, some machine translations will give lines by previous human translators; in other cases, the proposals are likely to be more literal than most human translations. Students can then generate their own translation by choosing from the various suggestions, clicking on the alternatives generated by a program like DeepL, then exporting and revising (“postediting”). And when they are stuck on a very difficult problem, they can make changes to the original text, exploring alternative means of expression (“pre-editing”) until the translation is acceptable. If the purpose is to test the limitations of machine translation, have the machine translate into the students’ first language; if there is more interest in exploring linguistic variation, ask students to make changes in the foreign language text; if it is a question of adapting to a special readership or medium, play with the prompts in automatic text generators, which are great as sources of ideas.
Guess the translation
One of the ways developers of machine translation claim parity with human translation is by showing people pairs of sentences and asking them to guess which is the machine translation: these days the guesses form no significant pattern (Hassan et al.). However, the claim does not work so well when whole documents are compared (Läubli et al.). This is something that can be tested in any class—it is another way of exploring the limitations of machine translation. An interesting variant on this, especially in a language-specific class, is to present an original text in the students’ first language and a translation of the same text into the students’ foreign language (possibly done by another group of students in the same class), but you reverse the names: present the translation as if it were the original. When the students point out the shortcomings of the ostensible translation with respect to what they think is the original, you then ask why they instinctively assume a translation is inferior. (I borrowed this trick from Andrew Chesterman.) If they do not assume inferiority, rejoice.
Explore the full range of translation solutions
Mirroring the initial fallacy of symmetrical matching is the terminal fallacy of permanent untranslatability: that the secrets of the original text remain forever locked inside it. The precept of untranslatability tends to ensue from a very narrow concept of translation, where translators are strangely not allowed to alter forms, explicate, add notes, or otherwise compensate for transitory gaps. How can this narrow translation concept be questioned? First, show the full range of possible translation solutions, from transliterating parts of the foreign text right through to expansion, contraction, adapting concepts, localizing references, and judicious editing. There are many typologies to choose from (see Pym). Then invite students to find examples of the various solution types—online translations of Alice in Wonderland are an easy place to start. And when something still seems untranslatable, ask students to borrow from the solutions found in published translations and to improve on them.
Identify and break norms
When students compare translations, a first task might be to spot infelicities and express personal preferences. A step beyond that is to identify translation norms, which could involve patterned decisions concerning the rendering of verse as prose, the copying of foreign literary forms, the legitimacy of omission, the use of in-text explications, or the adding of notes. Once students have identified different norms, they might try to guess why they are different, which could mean delving into the translators’ personal backgrounds and historical contexts. A more exciting activity might then be to take the norms and invite students to imitate them in a translation task. This sometimes means translating in a way that is different from what contemporary translators usually do: it is in line with Gideon Toury's pedagogical recommendation that we teach students to break norms. For example, since omissions are usually frowned upon in the context of contemporary translations into English, you might invite one group of students to adopt a norm from neoclassical French translators: leave out all the bits they do not really like. You can then ask the other groups to evaluate the result.
Translate for different readerships
Since all translations are done for a specific readership and for a given purpose in a particular place and time, they can all be altered to address alternative audiences and different purposes. This is a principle formulated in the 1930s by Lu Xun (see Lu and Qu 111–12). A first group activity here is to ask students to take one or two published translations, find out about the translators and the publishers, look at the paratexts, and try to ascertain something about the targeted readership and the purpose of the translation. Then ask them to envisage a different readership (perhaps a younger generation) and a different purpose (perhaps a change in gender perspective or political allegiance), and then to retranslate accordingly, with or without the aid of creative prompts fed into automatic text generators.
Localize
Different readerships may justify different degrees of adaptation, which can be experimented with. Adaptation involves everything from basic considerations about what to do with units of measurement and names of characters to large-scale conceptual rearrangements such as transforming Romeo and Juliet into West Side Story or Hamlet into The Lion King. Before discussing this huge topic, I get students to look at the way companies like Coca-Cola and Apple localize or standardize their websites in different languages, then to try to transfer those strategies to the translation of literature. A classicist might invite students to apply something like Quintilian's precepts for paraphrasing Aesop, but one could also invite them to translate like Coca-Cola.
Participate in online discussions
In addition to discussing and evaluating translations in class, students can follow online evaluations on sites like Goodreads and Amazon. A next step might be for them to contribute comments of their own, either to existing threads or to their own discussions, responding to one another. A further activity could be to check which translations are actually being sold by the online algorithms framing the discussions. In far too many cases, the electronic marketing fails to distinguish between translations.
Speak the written page
There is no reason why a classroom translation activity should be limited to the written replacing the written. Have students perform their translations orally, speaking them out loud—there are few better ways to test for infelicities (and it is a good technique for revision). And once they have done that, it takes just a few further steps to turn fragments of printed narrative into a full-fledged theatrical performance: page to stage.
Translate across media
Why stop at the spoken? I ask my students to work in groups to translate a poem into some other medium, any other medium, in ninety minutes. Most are great at putting together moving images, music, subtitles, and voice. One group produced a kite to render a Russian poem about words in the wind. Another translated a children's poem from French into Minecraft imagery and voice. Others produce animated cartoons, while in one case students sang their translation live to the whole class. I am constantly amazed not just at the technical prowess of younger generations but also at the creativity that remains untapped when we stick to the written page.
In all these activities, the pedagogical aim is not to produce a perfect translation and it is only incidentally to engage with a foreign language and culture. The main learning outcome is instead to have students explore and debate the inherent variability of translation itself. “A translation is something that has to be discussed,” said Peter Newmark long ago (20), and dynamic discussions should indeed be the gateway to active translation literacy. All the above activities are designed for student groups of three to five, who then report to the whole class for a general comparison and discussion. The important point is that those discussions can concern both the reception and the production of translations, such that the resulting literacy can have repercussions for the active use of translations in the wider community.
In highly multilingual societies, as most of the world's bigger cities have become, the role of translation is no longer to offer an isolated “auberge du lointain” (“inn of the distant”; my trans.), as Antoine Berman put it, just as a modern language department should no longer feel like a genteel suburb of cultural embassies. The mosaics and overlaps of languages are part of the places where we live, and communication between those languages must be part of what makes our shared life rich and pleasurable. As electronic technologies now allow virtually anyone to translate and to have their translations distributed, received, worked on, and retranslated, the role of translation literacy must be both active and passive, not merely a mode of reading.
Kafi's readership survey began as a doctoral thesis for which the main supervisor was a professor of German (Selection). It seems that some language-specific programs in that city, Wellington, have moved toward translation, perhaps as a way of gaining students, but also, laudably, as a mode of active engagement with the surrounding community.
