[Special Feature: Rethinking English Education] Yuka Akiyama: Rethinking English Education in the AI Era—What English Proficiency is Required for Engineering Students?

English education in Japan is currently at a major turning point. With the acceleration of globalization, the rapid development of AI technology, and the shift in policies of higher education institutions aimed at strengthening international competitiveness, various factors are complexly intertwined, making it an era to rethink the very nature of English education. In this context, Japanese universities are rapidly adopting EMI (English Medium Instruction), where subjects are taught in English as a medium of instruction (the language used to convey course content), with the goal of improving international competitiveness. This trend goes beyond a mere framework for language education and is closely related to the internationalization strategies of universities, exerting a significant influence on the direction of education itself. Furthermore, the remarkable progress of generative AI, such as ChatGPT, is shaking the foundations of traditional English education. For example, from the latter half of 2024, ChatGPT equipped with GPT-4 has realized real-time dialogue in multiple languages, including Japanese, and the possibility of AI itself becoming a foreign language learning partner has become a reality.

In such an era of rapid transformation, what should we English educators teach and what should we aim for? In this article, I will first explain the outline of English education implemented at the Graduate School of Engineering, The University of Tokyo. Next, I will consider the impact of the shift toward EMI in the Graduate School of Engineering and the rise of AI on the operation of the writing center and English speaking activities. Finally, I will examine the nature of English education in the AI era and explore what English proficiency is required for engineering students.

I belong to the International Education Promotion Section of the Institute for Innovation in International Engineering Education at the Graduate School of Engineering, The University of Tokyo. This institute is attached to the Graduate School of Engineering and develops classes and programs primarily for engineering students from third-year undergraduates to graduate students. Due to the need for classes and programs that emphasize engineering expertise, we support engineering students' English learning from multiple angles, including speaking classes to improve oral expression skills as engineers (for third-year undergraduates), academic writing and presentation classes to enhance research dissemination skills (for fourth-year undergraduates and graduate students), international collaboration classes to cultivate communication and problem-solving skills necessary for international joint research in cooperation with students from other universities (for all grades), as well as operating a writing center and holding international exchange events. However, since these are all elective courses or voluntary participation programs, fewer than half of the students are in contact with English on a daily basis through our classes.

Under these circumstances, the Graduate School of Engineering at The University of Tokyo decided on a policy to offer almost all graduate-level classes in English starting from 2026. This policy adopts a step-by-step approach, maintaining an environment where the basics of engineering are taught in Japanese during undergraduate years, while research activities are primarily conducted in English from graduate school onwards. Behind this decision are clear goals: improving competitiveness in international rankings, acquiring excellent international students from overseas, promoting international joint research, and encouraging Japanese students to be active on the international stage. In fact, internationality (the ratio of international students and the track record of international joint research) has become an important evaluation index in world university rankings, and major Asian universities are expanding EMI as if competing with each other.

However, research on EMI has also pointed out many negative aspects. These include, for example, a decline in the quality of education due to the English proficiency of faculty members, a decrease in student comprehension, and the superficiality of discussions. According to a study by Macaro et al. (2018), in EMI in non-English speaking regions, there are many cases where faculty members feel anxious about teaching in English, and the quality of classes deteriorates due to a decline in expressive power and responsiveness. Furthermore, on the student side, there is also the problem of being unable to allocate sufficient cognitive resources to critical thinking and creative problem-solving because they are too focused on understanding in English. Concerns such as "professional discussions do not deepen" and "questions decrease" are also heard from faculty members who are already implementing EMI at other universities. These suggest that EMI is not a mere substitution of language but has a significant impact on the quality and efficiency of education. Also, from a linguistic perspective, it cannot be denied that EMI reinforces aspects of linguistic imperialism. As Phillipson (1992) points out, promoting English as a "lingua franca" carries the risk of sacrificing linguistic and cultural diversity and uncritically accepting the values and ways of thinking of the English-speaking world.

Regarding my own experience, during my undergraduate years at International Christian University (ICU), I took more than half of my classes in English for the purpose of improving my English proficiency. I completed my master's and Doctoral Programs at graduate schools in the United States and have received professional education in English for a long time. However, currently, when I have the opportunity to learn similar content from Japanese professional books and textbooks in Japan, I am reminded once again of the ease of understanding and the high level of knowledge retention. I sometimes reflect on what kind of researcher I would have become if I had learned through a step-by-step approach, building a sufficient foundation of basic professional knowledge in my native language, Japanese, before moving on to advanced learning in English.

Increasing the number of subjects that international students can take is meaningful and may be a necessary measure to improve international rankings. Also, considering the current situation of the Graduate School of Engineering, where more than 40% are international students and the common language of many laboratories is English, the transition to EMI can be said to be an inevitable trend. However, should we promote EMI even if it means ignoring the advantage of Japan's environment where professional subjects can be studied in the first language (which is difficult in many non-English speaking countries such as Pakistan)? Will students who start research in a new field from graduate school or students who enter the University of Tokyo from other universities take classes in English without acquiring basic professional knowledge in Japanese? Will the academic advisor conduct seminars in the laboratory and provide opportunities to read papers in Japanese? Various questions arise.

Since the policy has already been decided, my role is to concentrate on supporting students and faculty and to assist in the smooth operation of EMI. Specifically, after conducting surveys and interviews to understand the needs of faculty and students, I would like to provide workshops and teaching materials to facilitate the progress of EMI. Through efforts to maximize the benefits of EMI and minimize its negative aspects, I intend to support the university's academic activities and dissemination capabilities.

Since arriving at the Graduate School of Engineering, The University of Tokyo in 2018, I have been engaged in the operation of the writing center. It is a program that provides feedback on English writing and workshops on writing, and although small in scale, it has supported the academic dissemination of University of Tokyo students. However, from the 2023 academic year, when AI began to be utilized by many students, the number of writing center users decreased significantly. The utilization rate, which had exceeded 90%, dropped to 21%, and questions were raised by the department's upper management (engineering professors) about the necessity of the writing center. One student said, "Since there is ChatGPT, I don't feel the need to have a person check it," and another student said, "AI allows me to ask questions many times without hesitation, and it responds 24 hours a day." From these voices, I strongly recognized the current situation where traditional writing centers are unable to adapt to the AI era.

To overcome this critical situation, we shifted our strategy in the 2024 academic year. Instead of viewing AI as an "enemy," we sought a path of "coexistence." Specifically, we held regular workshops and lectures to provide opportunities to learn paper-writing techniques using AI, while also clarifying the differentiation from AI and conveying the unique strengths of the writing center. For example, in a practical session on "how to improve English papers created by AI," we had students critically analyze AI-generated papers and point out unnatural logical developments, deviations from research field conventions (genres), and lack of expertise. As a result of publicizing that it is a service that can also provide advice on how to write papers using AI and continuing to emphasize how we can support what AI is not good at, the writing center utilization rate in the latter half of the 2024 academic year recovered to 89%.

From this experience, I realized the need to shift from traditional accuracy-oriented teaching methods (which, regardless of our intentions, are often perceived by students as correction services) to teaching "value that AI cannot provide," and the importance of making this known to students. In the future, I would like to continue supporting students by focusing on cultivating abilities that AI cannot easily replace, such as deep understanding and expressive power in specialized fields and awareness of genres.

In 2024, when coexistence with AI became commonplace, I introduced activities to converse with ChatGPT into my classes. Since ChatGPT did not have a built-in voice function in June of that year, it was a speaking assignment using Speech-to-text and Text-to-speech technologies from a Google Chrome extension (Mia AI), but AI was expected to support the Language aspect of Content and Language Integrated Learning (CLIL). The assignment was to dialogue with ChatGPT regarding the content of engineering lectures, and while students were amazed by the progress of AI, there were also critical opinions such as "it feels more like an audio version of a content confirmation test than a conversation." This is thought to be due to the unnaturalness of the dialogue.

However, in the latter half of the 2024 academic year, a voice mode was implemented in ChatGPT, and as of March 2025, dialogue with AI with extremely natural pronunciation is possible in Advanced Voice Mode. With the addition of a camera function, conversation while looking at an object has been realized. This is a phenomenon essential for language development called "joint attention." We have entered an era where AI answers when you ask "What is this? How do you use it?" while photographing an object with a camera. Furthermore, it has become possible to take pictures of news articles, graphs, etc., and have discussions about them. The constant presence of a dialogue partner capable of joint attention is a groundbreaking development in second language acquisition. This is because one of the biggest weaknesses in traditional Japanese English education was the "lack of speaking opportunities." In a situation where there are more than 30 students in a classroom, the speaking time per person is extremely limited. Also, many Japanese students tend not to try to speak English actively due to "concerns about making mistakes." The emergence of AI will radically improve this situation.

Nevertheless, there are fields that AI is not good at. Elements that make conversation natural, such as non-verbal communication, turn-taking, and back-channeling, still need to be learned through reciprocal interaction with humans. However, considering the speed of AI evolution in recent years, there is a high possibility that these weaknesses will be overcome in a few years. In such an era, how we English educators coexist with AI is an issue that requires continuous discussion.

Looking back at the trends in educational settings and the rapid development of AI technology over the past few years, I cannot help but think that English education for engineering students is being forced into a fundamental restructuring. In an era where AI is becoming an excellent interlocutor even in oral communication, how should we provide "value that AI cannot provide" to students who are not interested in learning the language itself? I would like to propose two important concepts for this challenge.

The first is "intercultural communicative competence." For engineering students, there are many situations where expertise and intercultural understanding intersect, such as collaboration in international research teams, communication in laboratories with researchers from different cultural backgrounds, and presentations at international conferences. And in such situations, English is often used as a lingua franca. In order to effectively advance intellectual exchange in such situations, understanding the other person's cultural background, empathy for diverse values, and the ability to recognize the differences between oneself and others are essential. True international communication cannot be achieved through language ability alone.

Byram (1997) proposes five components as constituents of intercultural communicative competence.

1. Attitudes (attitudes of openness and curiosity): Curiosity about different cultures, attitudes such as suspending judgment.

2. Knowledge (knowledge): Socio-cultural understanding of one's own and other cultures.

3. Skills of interpreting and relating: The ability to interpret phenomena of different cultures and relate them to one's own culture.

4. Skills of discovery and interaction: The ability to acquire new cultural knowledge in actual communication and use one's own knowledge, attitudes, etc.

5. Critical cultural awareness: The ability to critically evaluate cultural phenomena.

Among these, I believe that "1. Attitudes," "4. Skills of discovery and interaction," and "5. Critical cultural awareness" are human-specific abilities that AI cannot easily replace. This is because AI lacks physical experience in the real world, cultural identity, true emotions, and ethical judgments. "Attitudes" depend on the emotional foundation cultivated from actual cultural experiences, "Skills of discovery and interaction" require the power to react immediately based on one's own knowledge and attitudes in real-time communication, and the development of "Critical cultural awareness" requires real experience in independently evaluating the customs of one's own and other cultures within diverse value systems. Supporting the acquisition of these abilities is the value that human teachers can provide. For example, in the speaking class I teach called "Workshop towards Communicating Engineers," we implement activities where Japanese students and international student TAs collaborate to solve engineering problems. Through simulations of new product development by multinational engineering teams, we aim to have them experientially learn cultural differences in communication styles, decision-making processes, and time concepts, and acquire the power to independently evaluate the customs of their own and other cultures in real-time communication. I should add that when conducting such activities, it is important to break away from superficial comparisons and stereotypes that simply equate nation-states with cultural and linguistic systems (Akiyama, Y. & Ortega, L. 2024). In this class, we incorporate activities that also focus on diversity within each country and culture through dialogue with international student TAs from various regional and cultural backgrounds. Such practical training is a learning opportunity that is difficult for AI alone to provide and will become increasingly important in English education in the AI era.

The second is "acquiring a way of speaking that moves the other person's heart." Even if the content is inaccurate, speaking with confidence as if it were the truth in perfect English is something AI is good at, but such superficial confidence is not necessary. Rather, I believe that the attitude of actively trying to communicate even in imperfect and non-fluent English, the attitude of expressing one's own thoughts with dignity, and the heart that shows interest in the other person and wants to connect deeply with them are important for success in an international environment and are human qualities that AI cannot easily replace. In particular, Japanese students tend to think that "I should not speak unless it is perfect English" or "It is embarrassing to make a mistake," but for many engineering students, English is a common language, and there is no need to use native speakers as a norm. Rather, it is necessary to eliminate native speakerism and change their awareness so that their "content to convey and attitude to convey" are what make them attractive. In other words, shouldn't the focus of English education in engineering shift from grammatical accuracy and rich vocabulary to "heart-to-heart exchange activities" that connect with the heart in order to convey the content you want to convey to the other person in an easy-to-understand way?

In summary, the role of future English educators is not to compete with AI, but to foster human-specific values that AI cannot provide while effectively utilizing AI. The efforts toward EMI at the Graduate School of Engineering, The University of Tokyo, and the AI response of the writing center fundamentally have the same goal. That is to promote the "exchange of knowledge" through the common language of English.

Looking back at the tradition of Keio University as a school for English studies, what Yukichi Fukuzawa aimed for was not just the acquisition of English proficiency, but English education as "jitsugaku (science)" to incorporate Western knowledge and ideas and utilize them for the development of Japan. With the technological development of Japan, we have expanded our role from recipients of knowledge to disseminators. What modern engineers emphasize in particular is this dissemination power, and practicing it in a globalized world is today's challenge. I would like to rethink the nature of English education for engineering students by adding a modern perspective to these ideas. Specifically, I believe that two elements are important: "cultivating intercultural communicative competence" and "promoting heart-to-heart exchange activities." By incorporating these elements, we will be able to develop the exchange of knowledge into a human-specific activity suitable for the AI era. And English education will function as a foundation for that.

(References)

* Akiyama, Y., & Ortega, L. (2024). Coming out, heteronormativity, and possibilities of intercultural learning in a Google Hangouts telecollaboration. International Journal of Bilingual Education and Bilingualism. https://doi.org/10.1080/13670050.2024.2306388

* Macaro, E. (2018). English medium instruction. Oxford University Press.

* Phillipson, R. (1992). Linguistic imperialism. Oxford University Press.