Reimagine Foreign Language Teaching in the Age of AI

By Yong Zhao, Professor in Educational Leadership at the Melbourne Graduate School of Education

If machines can do better than most human foreign language learners in communicating in another language, perhaps it is time to rethink traditional foreign language courses, which have generally failed to make students competent in the target language, writes Professor Yong Zhao.

As ChatGPT and DALL-E grow viral, hopes and warnings about Artificial Intelligence or AI are growing. As AI-based computer programs, ChatGPT can write essays and DALL-E can generate artistic images. In other words, it seems that computers now can do human work, humanly creative work. While the technology will continue to grow more sophisticated for sure, conversations about the implications of these two creative AI applications have become wide ranging, affecting almost every aspects of human life, including education. For example, what should teaching of writing be like when machines can write? What is the nature of art education when machines can create art?

Another set of AI-based computer programs have not gained as much popularity as ChatGPT and DALL-E, but can have significant impact on foreign language education in our schools. These are machine translation software that automatically translates text or speech in one language into another. Machine translation has been in existence since the 1950s. With decades of advancement in theory and development, today’s machine translation has made significant progress. Popular machine translation tools include Google Translate, Microsoft Translator, DeepL, Alexa Translations, etc.

These tools vary a great deal in terms of the number of languages they can translate, ease of use, and accuracy. Some, such as Google Translate, can translate over 100 languages in the world, while others are covering much fewer languages. The quality of translation varies, so does the ease of use and access. Some of these translation tools are domain specific, meaning performing better in certain areas while others are more generic.

The purpose of this essay is not to evaluate the quality of machine translation applications. My primary interest is to start a conversation about the potential implications of machine translation for foreign language teaching. To be sure, some foreign language educators have used machine translation to assist their teaching and students’ learning. However, the conversation I wish to start is not so much about how to use machine translation in teaching, rather, it is about the purpose, curriculum, and pedagogy of foreign language education in schools.

A primary purpose of foreign language education is to teach students to become competent in the target language so that they can communicate with native speakers of the language. Thus, not surprisingly, all curriculum and pedagogy in foreign language courses have an almost exclusive focus on helping students to develop competence in listening, speaking, reading, and writing in the target language.  Whatever teaching methods are used, the expectation is to have students become proficient users of the target language.

The expectation is shared by teachers, school administrators, system leaders, parents, and the public. But the expectation is generally wishful for the majority of students who take a foreign language course either voluntarily or required. Despite teachers’ great and hard work and students’ genuine interests and efforts, most of the students are far from competent after one or four or even ten school years. Very few students can truly achieve native or near native speaker proficiency. In comparison to today’s machine translation, which is far from perfect, the majority of students are unable to do as well as translation machines in communicating with native speakers of the target language.

There are many reasons for the lack of development of true competence after so many years of studying the language, but three reasons are unchangeable. First, the amount of time of learning the target language is limited. In most schools, the max number of hours per week is three to five class hours. Assuming 40 weeks in a school year, the total number of hours is 120 to 200 class hours, which amounts to about 80 to 130 hours, far from the time needed to acquire proficiency in a second language. Second, there is typically no environment to authentically use the target language in most cases locally, which limits the real opportunities to practice the target language for communicative purposes. Third, the purpose of learning a language in schools has often been hijacked by testing. That is, for a significant majority of the learners, passing tests becomes more important than actually acquiring the language. Whatever the reason, the current context of foreign language teaching is unlikely to change and the hope that students in foreign language courses can become competent speakers is at best unrealistic.

Given that machine translation is already better at communicating today than most people who spent a few years studying foreign languages in schools and the technology will for certain continue to improve, it is important to rethink foreign language education courses. Of course, the most immediate, quick, but wrong reaction would be scrapping these courses. This reaction is wrong because it fails to understand the real meaning of studying foreign languages, which should have always been about culture instead of language for the majority of students.

Understanding a culture that is different from one’s own greatly expands one’s cognitive and psychological competence. Another culture, a different way of living, a different way of thinking, and a different way of feeling, challenges and comforts individuals who are used to thinking and feeling within only one culture. It helps people to go beyond what they are used to and experience something different, something uncertain, and something unexpected. This experience is challenging but also comforting because it presents more diverse ways of living, more options for human organization, and more possibilities in interacting with nature.

Culture should be the core content of foreign language teaching, but unfortunately linguistic competence has instead always been the primary focus. Assessment in foreign languages has always been on the mastering of vocabulary, grammar, and abilities in reading, writing, listening, and speaking. Given the difficult in achieving linguistic competence in foreign languages and the availability of increasingly sophisticated machine translation, it seems reasonable to rethink foreign language courses as primarily cultural courses, with limited expectations of linguistic proficiency in the target language.

Making culture the core of foreign language courses is more than learning about one culture. Rather, it is also about developing global competences. There are multiple definitions of global competence. For example, the Asia Society defines it as the ability to investigate the world, recognize perspectives, communicate ideas, and take action[1]. OCED’s PISA defines it as:

Global Competence is a multi-dimensional construct that requires a combination of knowledge, skills, attitudes and values successfully applied to global issues or intercultural situations. Global issues refer to those that affect all people, and have deep implications for current and future generations. Intercultural situations refer to face-to-face, virtual or mediated encounters with people who are perceived to be from a different cultural background.[2]

Whatever the definition, the main idea revolves around cultures and issues beyond one’s own culture in local communities. It is about human interdependence and interconnectedness.

Global competence has been accepted as a core competence for thriving in the 21st Century. In 2018, PISA conducted its first global competence assessment of 15-year-old students around the world.

It is difficult for schools to add new courses to teach global competence as schools are already crowded with too many required courses. The optimal place to teach global competence is foreign language courses, which should be about other cultures, other languages, and global issues beyond the local communities anyway. Of course, this would require changes in assessment, focus of content, and teaching methods in foreign language courses.

The purpose of this article, as stated earlier, is to start a conversation about the implications of the development of machine translation for foreign language teaching. The idea is very simple: if machines can do better than most human foreign language learners in communicating in another language and traditional foreign language courses have generally failed to make students competent in the target language, the primary focus of foreign language courses should be culture and global competence. This is for the majority of students. For the small percentage of students with strong interest in foreign languages, schools should offer different courses or encourage students to learn online from native or near-native speaking teachers.

If you'd like to hear more from the author and learn about the future of AI in education, Asialink will be running a webinar on March 22nd which you can find out more about here.


Professor Yong Zhao is a Professor of Educational Leadership at the Melbourne Graduate School of Education and a Distinguished Professor at the University of Kansas.

Author

Yong Zhao