Language learning and translation have always been complementary pillars of multilingualism in the European Union. Both have been affected by the increasing availability of machine translation in recent years: language learners now make use of free online machine translation to help them both understand and produce their foreign language or ‘L2’, but there are fears that uninformed use of the technology could actually undermine effective language learning; at the same time, machine translation is promoted as a technology that will change the face of professional translation, but the technical opacity of contemporary approaches, and the legal and ethical issues they raise, can make the participation of human translators in contemporary machine translation workflows particularly complicated. Meanwhile, those involved in both language and translation education are seeking ways to meet the challenges posed by the increasing use of machine translation. Against this background of flux, the strategic partnership “MultiTraiNMT - Machine Translation training for multilingual citizens” aims to develop, evaluate and disseminate open access materials that will lead to the enhancement of teaching and learning about machine translation among language learners, language teachers, trainee translators, translation teachers and professional translators across
Europe. The strategic partnership brings together experts at four European Universities -- the Universitat Autònoma de Barcelona, Université Grenoble-Alpes, Dublin City University and the Universitat d'Alacant, and two enterprises -- Prompsit Language Engineering and Xcelerator Machine Translations, and is supported by more than twenty associate partners in education and the translation industry, all of whom are interested in teaching and learning about the use of neural machine translation. The partnership aims specifically to develop an innovative syllabus in machine translation, and in particular machine translation based on currently popular deep learning techniques, also known as ‘neural machine translation’.
On completion we will have:
• developed explanations of deep learning and neural machine translation that are accessible to nontechnical audiences and are of particular relevance to language teachers and learners as multilingual citizens and trainee and professional translators;
• developed teaching materials that address both the technical foundations of machine learning—and especially deep learning—as used in machine translation, and the ethical, societal and professional implications of this approach;
• sourced data sets (and protocols for developing data sets) that teachers and learners can use in training their own machine translation systems;
• developed engaging activities that allow language learners and translators to co-construct knowledge about neural machine translation;
• developed a pedagogically-oriented neural machine translation platform that non-technical learners can use to gain insight into the internal workings of neural machine translation systems;
• tested and formally evaluated all the above materials at participating and partner universities;
• published the entire course (syllabus and links to associated materials) in an open-access ebook, which will also be available in print-on-demand format;
• disseminated knowledge of the project and the ebook across a wide variety of platforms.
In short, we will have developed an up-to-date syllabus in machine translation for use in European Higher Education and elsewhere, one that will allow students to develop the technical and ethical skills and competences required to become informed, critical users of contemporary machine translation in their own language learning and translation practice. In so doing, we will also have opened up the world of machine learning to language and translation students, their teachers and others, enhancing their ability to function as technologically competent, informed citizens in a multilingual Europe.