In April 2016, news reports announced that the Canadian Federal Translation Bureau was slashing approximately 400 jobs in light of increased automation and a desire to—one would imagine—cut costs. In November 2016, Google announced the launch of Google Neural Machine Translation with a promise of “more accurate, fluent sentences in Google Translate.”1 The effects of these events were felt across translation schools in Canada: students enrolled in the translation programs started to seriously question their academic and career choices. They also began to worry about a bleak future, or, in some cases, a future that looked very different from what they had initially imagined. Some of them expressed not wanting to become ‘mere’ post-editors, spending their days editing machine output. Others wondered what opportunities would be available to them given the rapid acceleration of automation in fields related to language-service provision (i.e. Google Neural Machine Translation or GNMT). There is no denying that facing these questions is scary. For a number of years, as a profession, we could take comfort in the fact that automated technology was nowhere near as accurate or nuanced as human output. Today, as more examples of neural machine output will attest, it is nearly impossible to deny just how proficient some of these machine-generated translations truly are. Whereas people once used to say “This reads like Google Translate” as a pejorative statement; now, this assessment could read as a compliment. And whereas once credentials could be leveraged against lay translators, the profession has to take stock of the fact that some platforms relying on crowdsourced work simply do not care whether participants have expert degrees or not. If crowdsourced content pleases the end-users, does it matter who translates?
We have come to an important crossroads in the profession (and also in translation training and theorizing). We can no longer, as practitioners and trainers, deny the very real effects of automation. If we continue to maintain a naïve discourse on the quality of machine output (“it’s bad,” “human translation is always better”), we run the real risk of missing opportunities and becoming obsolete. In some translation courses and programs, the use of tools such as Google Translate is proscribed under the claim that to use such tools is to plagiarize content. While the discussion of plagiarism escapes the scope of this piece, one can wonder how effective training is when tools that are regularly used and encouraged in the marketplace are proscribed in the classroom. Such proscriptions have the unfortunate effect of ill-preparing students for the future challenges that await them. Banning these new applications also means that we are not training students how to use them critically and judiciously. It could also be ventured that we are undermining digital literacy as well.
It has been said—in Canada, more specifically—that a degree in translation should be marketed as a direct path to a career in language services exclusively, otherwise we would somehow be undercutting the profession. How so? Many students at the undergraduate level across Canadian campuses opt to enroll in programs in a wide array of disciplines, not necessarily because of where the degree will lead them professionally, but because the discipline interests them. Is this not what we call widening the ‘horizon of knowledge’ and encouraging intellectual curiosity? We could therefore explore how to expand translator training, with an eye to identifying where translation intersects with other disciplines, like computer programming, social media, and informatics, particularly with students who are interested in digital technology. This would not only entice students who would never have considered translation as an option, but also retain those who want to become translators and who are also aware of a shifting marketplace.
Expanding on this line of thought, the opportunity for disciplinary consilience between translation and social media studies is imminent (here, studies means theorizing, training, and praxis within both fields). Which municipal, provincial, or federal government entity does not have a Twitter account requiring translation of tweets, most of which currently escape the ‘expertise’ of machines? To date, no known automatic machine translation software can adequately render the complexities of hashtag translation without affecting proper indexing and user engagement. How many brands are trying to create cross-cultural ad campaigns using only emoji-based slogans? Presumably enough, given that companies are now looking to hire ‘emoji translators’ (cf. UK-based Today Translations recently hiring Keith Broni for exactly that type of position). Based on current job postings and internship descriptions, students will inevitably be confronted with translating digital content.
Finally, perhaps what might give translators the most pause is not so much machines as fellow humans. At a time when crowdsourcing is being used for even the simplest of tasks (e.g. choosing a restaurant or making a purchase), it is no surprise that translation is now increasingly outsourced to the crowd. What does this mean for the profession? What does it mean for remuneration? Julie McDonough Dolmaya, an Assistant Professor at York University’s Glendon College, and Claire Larsonneur, a lecturer at Université Paris 8, have both explored how remuneration models have to adapt to the digital landscape, precisely because of crowdsourcing initiatives, as well as how consumers now demand their content. We can think about the ethics of paying professionals in the form of ‘likes’ or ‘badges’ or other non-financial means. We can explore how subscription-based remuneration could be very advantageous. How can we stay competitive when virtually (and quite literally virtually) anyone can be a translator? Some might think that increasing automation and a shifting professional landscape signifies the death of the translator. The position here is that it does not. If we look at the evolution of the translation profession and industry, there have been several instances where technology was thought to be the kiss of death. Yet, translators are still here and gainfully employed. The digital landscape requires that we rethink traditional ways of doing and understanding. How can we work with automation in ways that create better translation services for end-users? How can we motivate students, many of whom have never known a time without social media, to leverage their inherent knowledge of those platforms in ways that will excite them? How can we redefine translation in an era where the scale of intercultural contact, whether online or offline, is absolutely unprecedented? These are the questions that we should be exploring as a profession.
Renée Desjardins is an Assistant Professor at the Université de Saint-Boniface in Winnipeg, Manitoba, Canada. She is the author of Translation and Social Media: In Theory, in Training and in Professional Practice (Palgrave Macmillan, 2017). She has worked as a freelance translator for over a decade, has worked as a reviser and also has professional experience in social media content creation and curation.