Multilingual Scholarly Communication and Translation Technologies
Multilingual Scholarly Communication and Translation Technologies
Lynne Bowker
As a country, Canada has two official languages – English and French – but it is also home to more than 70 Indigenous languages as well as numerous heritage languages. Multilingualism is an integral part of the fabric of Canada. In contrast, in recent decades, the world of scholarly communication has been largely focused on a single-language model, where English is the key language used for publishing and conferences. As a result, many graduate students and faculty in Canada (and beyond) participate in scholarly communication through an additional language, which can have costs. For instance, scholars may need more time to read and contribute to the academic literature when working in a less dominant language.
Groups such as Acfas (Canada’s largest French-language learned society) and Quebec’s French Language Commissioner have been advocating for a greater use of French as a language for research and publication. While a multilingual scholarly communication ecosystem would help to mitigate many of the inequities created by the single-language model that currently prevails, multilingualism nonetheless presents its own challenges. If researchers all publish in their own language, how will they be able to discover and access each other’s work? Can new AI-based translation technologies help?
—If researchers all publish in their own language, how will they be able to discover and access each other’s work? Can new AI-based translation technologies help?—
To investigate, we have developed a research program at the intersection of scholarly communication and translation technologies that aims to uncover the potential and limitations of these tools for research dissemination, as well as to support scholars so that they can get the most out of these tools while using them in an informed and responsible manner.
As part of this research program, we conducted a systematic literature review to find out whether and how translation tools are currently being used for scholarly communication. Some of the key findings that emerged include:
- Tools are currently being used mainly by non-Anglophone researchers to help them produce manuscripts for publication in English;
- Tools are mainly being applied to the production of written articles, while other types of communication (e.g. posters, conference presentations) are not well supported by current tools;
- Tool performance is uneven, with better quality translations for more widely used languages and popular subjects, and lower quality translation for more specialized topics and less widely used languages.
The gaps uncovered by this systematic literature review can help researchers to know where it could be fruitful to focus their efforts moving forward.
Another dimension of the research included developing a framework of machine translation literacy to help students, faculty, librarians and other actors in the scholarly community to become savvy users of translation technologies. Although free, online translation tools are very easy to access and employ – often, it’s just copy, paste, click! – their straightforward user interface can mask some of the underlying issues associated with their use.
Some key aspects of machine translation literacy include:
- Risk assessment: Asking questions such as: What are the consequences of a poor quality translation in this context? How likely is it that the subject matter and language pair that I need will be well represented in the tool’s training data? Is the content that I need to translate sensitive or confidential?
- Transparency: Understanding the importance of knowing where the training data comes from, as well as the importance of labelling machine-translated text and acknowledging tool use.
- Meaningful human-computer interaction: Learning how to craft effective prompts, write in a translation-friendly way, or verify and edit machine translated text.
Known by the title, The Machine Translation Literacy Project, this ongoing effort has so far resulted in the production of a number of open resources, such as:
- A series of four infographics highlighting key points about how translation technologies work, and how users can work with them to get better results (available in 14 languages);
- De-mystifying Translation – an open access textbook that helps people without a background in translation to understand some of the fundamentals of translation, including why translation is challenging for computers;
- Garbage in, garbage out! Get better translations by writing translation-friendly texts – an open educational resource (OER) that provides tips on how to prepare a text in a way that improves the quality of machine translation (also available in French);
- Translating for Canada, eh? – an OER that explains the basics of how different translation tools work, with a special focus on Canadian tools and language varieties (also available in French).
AI-based translation technologies have enormous potential, but their success in dealing with general texts in very widely used languages does not translate well to highly specialized content and less widely used languages. Likewise, the risks involved when these tools are employed by language professionals are not the same as the risks posed when they are in the hands of non-experts. The appetite for a multilingual scholarly communication ecosystem is growing, and so is the need for research on users and applications of translation technologies in this context. Much work remains to be done on improving discoverability of work in other languages, supporting multilingual conferences, enabling cross-language peer review, and more. Canada is a living lab for research and innovation at the crossroads of translation and scholarly communication – stay tuned for more exciting developments in this area!
Cite this article in APA as: Bowker, L. Multilingual scholarly communication and translation technologies. (2025, February 27). Information Matters, Vol. 5, Issue 2. https://informationmatters.org/2025/02/multilingual-scholarly-communication-and-translation-technologies/
Author
-
Professor and Canada Research Chair in Translation, Technologies and Society at Université Laval, Canada. Author of Machine Translation and Global Research (Emerald 2019) and De-mystifying Translation (Routledge 2023).
View all posts Professor