2022 Volume 7 Pages 19-36
This paper revisits the concepts of intelligence, language, and language education by considering “cyborg,” “language games,” and “plurilingualism,” and discusses the use of machine translation in English education. The use of artificial intelligence (AI) does not negate human intelligence because human intelligence inherently involves the use of media and tools; humans may be described as “natural-born cyborgs.” An analysis of language into the specific activity of language games reveals that the usefulness of machine translation depends on particular language games. English-language education from a monolingual perspective discourages the use of machine translation because it involves learners’ native language (Japanese). However, educators must realize the diversity of the use and learning of English, ranging from a transient and spontaneous conversation to extended and reflective writing. Language education from a plurilingualism perspective encourages foreign-language learners to use all of their linguistic resources (including their native language skills) in various language games. A discussion on the introduction of machine translation into English education should be specific on a case-by-case basis, not categorical on an either-or basis.
This article’s author has been using machine translation in his university general-education English-writing courses since the 2020 academic year and has received favorable responses in student surveys.i Simultaneously, he has observed that some people regard machine translation as a panacea while others emotionally denounce it. The former claim that machine translation will render writing classes unnecessary, while the latter assert that the introduction of machine translation invalidates language education. In response to these views, the author pointed out the shortcomings of machine translation at the usage level (Yanase & Lees, 2022) and the conditions for the introduction of machine translation (Yanase, in print). Assuming that the views above derive from a simplistic understanding of relevant ideas, the current paper uses the keywords “cyborg,” “language games,” and “plurilingualism” to re-examine the concepts of intelligence, language, and language education. The current English version is issued as a “secondary publication,” a translation of the original Japanese edition, to demonstrate the machine-translation potential that it claims.ii
The first key word for the discussion is “cyborg.” A philosopher, Andy Clark (2004), coined the term “natural-born cyborg” to emphasize that humans exploit media or tools as part of themselves beside their own flesh and blood. Human intelligence has been irreversibly changed by spoken and written language, mathematics, typography, radio, television, and the Internet, among others, as sociologists such as Ong (1982) and Luhmann (2012) point out. Human intelligence without any of these media or tools is almost inconceivable now. The development of spoken language greatly changed how humans thought and acted, distinguishing them from other animals. Furthermore, the invention of the written language produced civilizations and influenced areas that extended far beyond those that voices could directly cover. In addition, the written-language culture promoted reflective thoughts and developed various academic disciplines. The medium of mathematics created an abstract world. Letterpress ensured the accurate dissemination of complex information in large quantities. Radio spread the voice of the “standard language” to every corner of the nation-state. Television produced virtual realities in which people shared a sense of unity with distant events. The Internet has brought an explosion of such information. People presently see the emergence of AI, a machine that creates information and knowledge, in this historical context. Clark expressed the idea that “human beings are born cyborgs” due to these essential roles that media and tools play in human intelligence.
Nevertheless, humans, from birth, are unexposed to other tools than spoken languages. When to provide children with which tools to develop intelligence are significant pedagogical questions. Educators must carefully consider when to introduce machine translation into foreign-language education. An appropriate analogy might be a calculator: students are encouraged to use calculators in math classes from the late stage of elementary school in Singapore (Nagasaki, ND); however, no countries allow children to use calculators from the beginning of math education from the belief that children first must acquire a sense of numbers with their own bodies.iii After acquiring that basic number sense, many countries encourage students to use calculators in complex mathematical processes to enhance their mathematical competence. The use of calculators is comparable to the use of machine translation. Machine translation can extend learners’ foreign-language skills if they use it after acquiring a sense of a foreign language. Regrettably, this paper cannot afford to discuss the optimal timing of its introduction into the curriculum.iv
2.2 Coexistence and collaboration with AIDespite disagreements on specific issues about AI, the principle remains unchanged that humans should learn to coexist and collaborate with AI. After decades of introducing AI to business enterprises, Khodabandeh (2022) states that the key for success is discovering concrete ways for AI and humans to coexist and collaborate, without antagonizing or ignoring each other. State-of-the-art machines are not essential for success. What is essential is machine-human collaboration to accomplish what humans or AI cannot do alone. The Economist, in its feature issue on AI (2022, June 9), views machine intelligence as distinct from human intelligence and argues that machine intelligence will complement human intelligence rather than threaten it, if appropriately used.
The author proposed three principles in the use of AI (Yanase, 2021). When applied to the issue of machine translation, the principles are as follows:
1 Machine translation only proposes a preliminary translation draft; it does not completely substitute human translation.
2 The draft that machine translation provides requires human reviewing and rewriting.
3 Humans must take the initiative and responsibility when using machine translation.
Essentially, foreign-language education aims to enhance learners' cognitive and practical abilities through the use of a medium or tool other than their native language. Foreign-language education then has no inherent reason to prohibit the use of AI as a tool. Machine translation enables learners to compare the target foreign language and their native language more deeply and to produce foreign-language texts much faster than before. After all, machine translation is already prevalent in many business operations (Sumida, 2022). Considering these, one may conclude that university English-writing classes should at least provide the students with some knowledge of machine translation.
To summarize this section, human intelligence entails the use of media and tools as an essential feature. The argument that the use of machine translation represents no human intelligence is an exaggeration. Another extreme view is that machine translation replaces human intelligence, because the former requires human intervention. Educators must avoid overly simplified views about intelligence and invent concrete ways for foreign-language learners/users and machine translation to coexist and cooperate with each other.
The second concept to reconsider is “language.” This section introduces the concept of “language games” for re-analysis. It argues that a single, uniform notion of “language” is inadequate when considering the introduction of machine translation.
3.1 Liberation from the framework of “language” and “writingWhen given a noun such as “language” or “writing,” people tend to instantly assume that there must be an entity corresponding to that noun. Moreover, guided by the set-theoretic idea, they conclude that they can define the entity by necessary and sufficient conditions.
In his early days, Wittgenstein employed the concept of “language” to analyze the world (Wittgenstein, 1922). Chomsky developed linguistics based on the concept of “language,” which does not refer to individual languages such as English and Japanese (Chomsky, 1965). In language education, the widespread framework comprises four skills: writing, reading, listening, and speaking. These notions of “language,” “four skills,” and “writing” have achieved certain intellectual results. The early Wittgenstein formulated the function of language in his logically conceived worldview. Chomsky's linguistics brilliantly extracted the characteristics of “language” as a formal system. The “four skills” framework has been institutionalized in curricula and testing, while educators discuss the use of machine translation in “writing.”
However, one should refrain from assuming that only one framework is valid. As practitioners, language educators should be aware of multiple perspectives and pragmatically choose an appropriate one. Whether machine translation should be used in “writing” is too crude a question. Debating on machine-translation use without distinguishing various aspects of writing is unproductive. Language educators need concepts that differ from “language” and “writing.” One such concept is a “language game.”
3.2 Language game as a form of lifeThe term “language game” was introduced by Wittgenstein in his later days (Wittgenstein, 1953) to emphasize that language use is part of an activity or form of life. Although both “reporting” and “praying” are examples of language use, they are very different “language games,” as is evident from the differences in the purpose, content, and manner of language use. The overlap of language use with a form of life means that a particular language game—a particular activity—becomes part of the language user as a person. Consider, for example, a Japanese person who starts reading English-language newspapers every morning. The person develops a body that is better tuned to the eye movement for English reading (the visual processing of ideographic Japanese and that of alphabetic English reading are very different.) The reading habit enhances the person’s visual-perception skills for English. Furthermore, the knowledge gained from that reading influences the person's thoughts and actions. The activity of reading English newspapers becomes more crucial for that person's life. The activity becomes a habit, and the habit becomes a form of life. Language is acquired when it is used in a person’s activities, and the use becomes part of that person's life. Language use results in “personal knowledge” (Polanyi, 1958).
Human use of language is not simply the input, manipulation, and output of codes. Human language use transforms the language user through the repetition of the activities involving particular words. It also incorporates the person as a member of the linguistic community that comprises people of similar tendencies to the activities. The people ultimately share the history of that linguistic community. In other words, human language use is not a mere formal processing of codes. It is a somatic, social, and historical activity—this is the understanding that derives from the concept of “language games.”
One of the requirements for using machine translation is the acquisition of the target language's comprehension vocabulary (i.e., vocabulary for understanding the language output from machine translation). For the comprehension vocabulary, one must know the particular emotional relationship between the word and the world in addition to the general referential relationship. Therefore, machine-translation use requires the user’s experience in language games. Post-editing, another indispensable part of machine-translation use, is to rewrite the machine-translation output from the overall perspective of how the word can affect the world. That requires a human who knows what it is like to live in language games. When considering machine-translation use, educators should employ the concept of “language games” rather than of “language.”
To summarize this section, educators should stop discussing machine translation with the uniform notion of writing. They should specify the preconditions for using machine translation (Yanase, in print). Machine-translation use necessitates language-game participation by the user as a person. Educators should analyze individual “language games” by describing them in concrete terms, not using a uniform and overly simplistic concept such as “language.”
The decision to introduce machine translation should be made by each institution and individual, in light of the balance between the public and individual nature of education. Machine translation will probably affect English learning profoundly. The common reference point in decision-making should be plurilingualism, which this section will explain. The section will then point out that plurilingual use of machine translation may develop transcultural communication.
4.1 The use of machine translation and a person’s form of lifeSome of the author’s students declined to use machine translation in class. Obvious examples include international students from non-English-speaking countries (whose first language the author could not read for comments) and students who spent several years in English-speaking countries before entering university (who were more experienced in writing essays in English). However, there were other types of students who did not use machine translation. Many of them probably felt as if the use denied their lifestyles as language learners.
One first-year undergraduate student never used machine translation. He did not even use Japanese words when writing the outline of an essay. The student was a very fluent English speaker by Japanese standards, who probably took pride in his oral proficiency. He could think quickly in English. The author acknowledged the value of his skills and yet suggested that the student might elaborate his ideas in Japanese, as his English vocabulary was more limited, nonetheless. However, he did not change his mind. Perhaps the student regarded the use of machine translation as a self-denial because he probably took painstaking effort to avoid the prevalent strategy in Japanese schools to produce English via Japanese.
A Japanese graduate student expressed his unwillingness to use machine translation when the author addressed the topic of AI in language education in a lecture. At the time of the lecture, he was writing his master's thesis in English and had already discovered that machine translation would outperform his English-writing ability. He felt embarrassed to use machine translation because it would undermine the value of his effort to learn English thus far.
Another graduate student who opposed machine-translation use was a native Chinese speaker who had studied in Japan for several years. She worked diligently to acquire the language and culture of Japan and was writing her master's thesis in Japanese when she expressed her view. She probably held the linguistic and cultural adaptation as a principle of her life in Japan. For her, the option of machine-translation use, i.e., writing a manuscript first in Chinese and then letting a machine automatically convert it to Japanese, likely seemed a denial of her years of dedication.
Some English teachers do not welcome machine-translation use, either. Many Japanese English teachers oppose machine translation—most of the author’s students are surprised when he promotes the use because they had met no English teachers who suggested such use. Perhaps those Japanese English teachers who ban the use of machine translation are, like the graduate student mentioned above, uncomfortable with the idea of a machine surpassing the English skills they earned with so much effort. Or perhaps they are afraid that their students’ secret use will threaten the authority of English teachers who virtually monopolize the knowledge of English in class.
Some native English-speaking teachers (NESTs) also fiercely denounce the introduction of machine translation into writing classes. NESTs, who have been privileged as “natives” in English education in Japan, will be disadvantaged if machine translation is introduced into writing classes, as many of them lack Japanese writing proficiency. The clichéd line, “No Japanese, please,” would lose its validity in the classroom. Some NESTs may feel uneasy about possible changes to their professional status.
Meanwhile, proponents of machine translation include many non-native English-speakers who are frustrated with the gap between their writing skills in English and those in their native language (the author probably falls in this category). Those people welcome machine translation because it transfers their native-language proficiency to English writing. They may regard machine translation as a “game changer” that empowers non-native English speakers’ restricted writing competence. Or they perhaps recognize it as a “great equalizer” that rectifies the overwhelming power inequality in the current world between native and non-native English speakers.
4.2 The choice of plurilingualism in public educationBehind much of the emotional denial of machine translation seems to lie monolingualism. Monolingualism holds that, for example, English learners must not use their native language when learning or using English. It also advocates that the ideal teacher is a native English speaker (native speakerism). However, this monolinguistic idea seems be too one-sided, given the diversity of language games and nature of public education.
Considering the diversity of language games, one should conclude that the monolingual use of English is appropriate in some but not other language games. For example, a conversation (speaking and listening sequences) requires prompt response. Participants must immediately respond to the flow of the conversation. Monolingualism is almost mandatory in this case because translation in the process significantly delays responses. To enhance conversational skills, learners should practice English without using their native language.
However, reading is not that simple. Even in speed reading, in which monolingual understanding is generally preferred, many learners strategically stop at challenging parts for precise understanding in their first language. Prohibiting such a strategy under the name of monolingualism would be an ideological admonition. In the current interview project with English users at his university, the author has confirmed various ways of using one’s native language in reading. One Japanese social scientist, who also serves as an editor of an international project, stated that he uses machine translation to speed-read English papers to obtain information (Kurosawa, 2021). He machine-translates English articles into Japanese, thereby increasing his reading speed by a factor of several to ten. Simultaneously, he reads essential English papers that require critical understanding without translation. It would be unpragmatic for monolingual ideals to deny diverse reading strategies.
Writing should also be considered in different language games. Sending a formulaic email quickly would not require the trouble of translating. However, writing an academic paper over a period of several months is an extended language game that involves idea-conception, reflection, draft-writing, and several revisioning. Prohibiting the use of one's native language during this long process makes little pragmatic sense. One of the interviewees in the aforementioned project, a graduate student who had published several international publications, stated that he uses Japanese to precisely organize thoughts for abstracts and introductions before writing in English (Kitagawa, 2021).
Thus, it is ineffective to adhere to monolingualism, at least in certain language games. Those who persist with a monolinguistic view might be under the influence of a colonialist prejudice without their awareness. Essentially, it makes little sense in foreign-language courses in public education to prohibit all learners from using the local language and require the exclusive use of the target foreign language, unless it is education for colonizationv. In this regard, plurilingualism proposed by the Council of Europe is worth noting. Below is an excerpt about the competence for which plurilingualism aims.
Plurilingual and pluricultural competence refers to the ability to use languages for the purposes of communication and to take part in intercultural interaction, where a person, viewed as a social agent has proficiency, of varying degrees, in several languages and experience of several cultures. This is not seen as the superposition or juxtaposition of distinct competences, but rather as the existence of a complex or even composite competence on which the user may draw. (Council of Europe, 2001, p. 168)
Worthy of attention here is the complex and composite nature of plurilingual and pluricultural competence. People have often assumed thus far that native-and foreign-language proficiencies are two separate entities that can be stacked like building blocks. However, various studies have shown that the influence of a native language remains in many people when using foreign languages. The word “composite” reminds us of Vygotsky's (2012) account of a water molecule, which is a compound of two hydrogen atoms and one oxygen atom. Although hydrogen and oxygen are inherently highly flammable, their flammability disappears when they constitute H2O. Similarly, the ability of a Japanese who has acquired English is probably a composite of his Japanese and English abilities; the person does not think like a monolingual English speaker or a monolingual Japanese speaker.
Many English teachers wish to consider a classroom as an “English-only space.” However, the real world is filled with multiple languages due to cultural, social, political, economic, and other factors. Daily conversations between English and Japanese speakers in Japan, for instance, may start in English only and then include occasional Japanese phrases, until they become mixtures of the two languages. Even in situations in which English is apparently the only working language, some non-native-English participants may deliberately add their local words to utterances in English. They may converse privately with their peers in their own language. The norm, “when using English, think and speak only in English,” does not represent the language use in the real world.
Plurilingualism describes the aim of language education as follows:
From this perspective [of plurilingualism], the aim of language education is profoundly modified. It is no longer seen as simply to achieve ‘mastery’ of one or two, or even three languages, each taken in isolation, with the ‘ideal native speaker’ as the ultimate model. Instead, the aim is to develop a linguistic repertory, in which all linguistic abilities have a place. This implies, of course, that the languages offered in educational institutions should be diversified and students given the opportunity to develop a plurilingual competence. (Council of Europe, 2001, p. 5)
There is no official agreement that English education in Japan is aimed at Japanese speakers of English who completely eradicate their Japanese identities and act like native English speakers when using English. English-language teaching in public education should, as plurilingualism advocates, encourage learners to use their linguistic resources (Japanese, English, and other languages) as social agents to communicate interculturallyvi.
The revisit of plurilingualism above suggests new changes. One is the reconsideration of the notion of autonomy. The use of tools or media does not immediately deny human autonomy, as discussed in the section on “natural-born cyborg.” Eyeglasses or a wheelchair do not negate the autonomy of the persons in need. Meanwhile, many academic associations in Japan categorically request non-native English-speaking contributors to have their English papers checked by native English speakers before submission. That requirement significantly diminishes the sense of autonomy of the users of English as a foreign language. AI, such as machine translation and grammar checkers, increases the likelihood of learners of English using English without depending on native English speakers. Such a change advances non-native English users’ autonomy.
The second change is the advancement of contrastive linguistic understanding. Translation encourages language learners to alternately think in their native and target languages and understand more about each language and culture. Without contrastive awareness, translation can distort communication. Increased knowledge of language and culture is critical when global exchange is on the rise.
Monolingualism and native-speakerism in foreign-language education are difficult to sustain given the diversity of language games. General consensus in public education should accord with plurilingualism, and further specific decisions should be left to the discretion of each educational institution, instructor, and learner, with their own autonomy and responsibility. As mentioned above, the author lets his learners decide on the use of machine translation. Recently, he further emphasizes that the use decision should be specific (when and where), and not categorical (whether-or-not). He also encourages them to simultaneously consider the issue from the long-term perspective of English proficiency enhancement and short-term perspective of essay completion. As the author has learned more about the pros and cons of machine translation from experience, he has strengthened his belief that a decision on the use of machine translation should be flexible, not ideological.
4.3 Development of transcultural language gamesThe proliferation of machine translation may change the idea of publishing in English. Thus far, non-native English speakers have been expected to follow the cultural conventions of English-speaking countries to publish in English. Non-native English researchers visiting English-speaking countries have been directed to “join the conversation” for publication. They have been instructed to place their writing in the dialogic context within the English-speaking academic community; they should not expect their papers to be published in English when they merely expressed ideas that they conceived in their local context.
The English-speaking sphere, although considerably international, has various tacit conventions, like any other cultural space. Katayama (2022) used the thought-provoking metaphor of a “hamlet” to describe cultural spheres, including the English-speaking one. The metaphor implies that each cultural space contains tacit assumptions that are not explicated in documents. A newcomer must learn its unwritten rules to survive and gain recognition within that hamlet. The English-speaking sphere, although much more liberal than others, expects foreign visitors to master its cultural manners. Foreign sojourners start with “legitimate peripheral participation” and gradually move into the core of the English-speaking sphere.
Therefore, publication in an English-speaking country inevitably requires foreigners to cite a large number of references in English to demonstrate the membership of the English-speaking sphere. It is difficult to imagine a paper whose references are mostly in foreign languages. Foreign-language speakers must incrementally enter the English-speaking culture to become residents of the hamlet to publish their papers. Although English is often regarded as a “global” language, its use naturally entails a commitment to the English-speaking culture.
Meanwhile, machine translation dramatically reduces the costs of time and money to produce English texts. Those who translate their local-language writing can produce English documents much faster than before.vii Foreign writers who outsource English translations to professionals can significantly reduce their expenses. Those who ask native English speakers to review their English translations can utilize AI technology to write English more independently. These changes may increase secondary publication in English.
Secondary publication in English refers to the English translation of an article written in a foreign language (the original or primary publication) for English audiences. It does not involve much commitment to the English-speaking community unlike the conventional English-language publication by foreigners described above. Writers do not need much cultural adaptation or a relearning of their issues in the English literature before publication. Secondary publication is conceived from local perspectives in non-English contexts and reaches English-speaking audiences. Such publishing may increase through the proliferation of machine translation.
The increase in secondary publication may eventually affect how original papers are written. Original papers may anticipate English readers in advance, although they are conceived in the context of the local language and culture.viii The current paper is one such example. While it originated in the context of English-language education in Japan, it was written in a particular Japanese style that the author believed fits machine translation. Furthermore, the storyline, including the paragraph construction, was created to be readily accessible to the distinct audiences of the two languages, to the best of the author’s ability. The author stays in his first language’s cultural sphere (Japanese) and yet modifies his style and story development for the sake of the cultural sphere of his second language (English).
Such communication style may be described as “transcultural communicationxi.” This concept contrasts with “intercultural communication.” Figure 1 below illustrates the difference.
Figure 1: Intercultural and transcultural communication
In intercultural communication, the two cultures (Cultures A and B) are presented with solid lines, indicating that the identities of the two cultures are firmly maintained. The two distinct cultures meet in a third place, such as a diplomatic negotiation table. The place is between (i.e., “inter-”) the two. Each side is aware that its own culture cannot dominate the place. Simultaneously, it does not surrender its own culture to the other. This paper defines such interaction as intercultural communication. Many representatives in diplomatic negotiations choose to maintain their own languages and employ translators despite their proficiency in the other party’s language. Such a stance may exemplify intercultural communication in which each side values its own culture.
In contrast, Cultures A and B in transcultural communication are drawn using dotted lines. The dots suggest that the cultural boundaries are somewhat looser; people in each culture may think and express beyond their cultural boundary. For example, as the members of Culture A communicate more with those of Culture B, the former may modify their patterns of cognition and action to be understood better by the latter, while never completely complying to the latter’s pattern. This paper uses the term transcultural communication to describe such communication.
In transcultural communication, Culture A’s members may use Culture B’s language (Language B) to more directly address Culture B’s members. Here, Cultures A and B are communicated in Language B alone, without mediation by interpreters. Yet, Culture A’s speakers do not abandon their culture when thinking and acting. Despite their use of Language B, Culture A’s speakers base themselves in Culture A. Nevertheless, they occasionally transcend it to adapt to the communication with Culture B’s speakers.
In the ideal situation of transcultural communication, Culture B’s speakers create no illusion from the apparent status of Language B as the common language. They are critically aware of the cultural differences. They understand that the Language B that Culture A’s speakers produce is based on Culture A. Furthermore, Culture B’s speakers modify their Language B to be better understood by Culture A’s speakers. The ideal transcultural communication promotes mutual recognition of cultural differences. It enhances each side to modify, but not deny, their culture for better communication. Consequently, each culture ceases to be rigid and becomes dynamic. Transcultural communication, which does not rivet cultural boundaries, will probably become more crucial as the global society promotes more exchanges.
However, in real-world situations, the common language (Language B) is often perceived to be more powerful and authoritative. Culture B’s speakers may impose their cultural framework and linguistic expressions without any consideration of Culture A’s speakers. The latter, who temporarily use Language B for communication, then must explicitly or implicitly request Culture A’s respect. Regrettably, such an attempt is often unsuccessful when Culture B’s speakers are monolingual English speakers with the mindset of colonialism or linguistic imperialism. Thus, a better understanding of the idea of transcultural communication is essential. In particular, specific knowledge of the patterns of language use in international communication (Kimura, 2021) is critical.
The secondary publication mentioned earlier is an instance of transcultural communication. The secondary publication of a Japanese article in English represents the view of a Japanese speaker who thinks and reads within the Japanese context. Inevitably, many English speakers prefer to read articles based on their familiar English-speaking culture. Secondary publications in English may not attract many readers. Nevertheless, some English-speaking readers who seek cultural diversity may want to read English papers conceptualized in non-English-speaking cultures.x An increase in secondary publication may promote particular Japanese writing styles that anticipate English translation. Such writing may eventually modify the mode of thought and expression in Japanese. Thus, machine translation, if it promotes secondary publication, may increase new language games of transcultural communication.
To sum up this section, it reported various attitudes to machine translation and confirmed that language choices reflect each user’s form of life. If a choice is a matter of lifestyle, particular decisions should be left to the discretion of each individual and institution. Monolingualism in foreign-language education is politically and practically difficult to maintain in public education unless the country in question seeks cultural colonization. A native language inevitably influences foreign-language use in the diversity of language games in the real world. As plurilingualism declares, the ultimate goal of foreign-language education should not be the monolingual native speaker. One of the by-products from the discussion on machine translation is a reappraisal of plurilingualism in public education, which aims to increase each learner’s linguistic and cultural repertoires. This section also suggested that plurilingual use of machine translation may advance cross-cultural communication.
In summary, this paper discussed that machine translation does not render English-language learning unnecessary or meaningless. Such an extreme conclusion derives from exceedingly simple views of intelligence, language, and language education. A complete affirmation of machine translation (i.e., no need for English education) or negation (i.e., a total ban of machine translation) does not reflect the language use in the real world. Those involved in English-language education should concretely discuss issues case-by-case, not categorically and abstractly.
Humans are cyborg-like animals in that they have achieved dramatic cultural development, compared to other animals, by mastering the use of medium and tools. The use of machine translation in education does not immediately render education inhuman or noneducational. Irrespective of how AI can expand their abilities, humans must experience language games themselves and convert that experience into personal knowledge to use a language.
This paper also re-examined the goal of English-language learning by referring to plurilingualism. The monolinguistic model of learners who always think, understand, and express themselves in the target foreign language alone is unattainable and not appropriate as an ideal of public education. It is desirable, both in terms of global equity (i.e., overcoming the hegemony of English) and particular circumstances of individual lives (i.e., the diversity of language games), that each individual broadens and adequately uses their own linguistic and cultural resources, as plurilingualism proposes.
A major challenge for education in the future will be the adaptation to coexist and cooperate with AI. This will be as significant as the 20th-century challenge of adapting to the factory mass-production system, or perhaps a more pressing one. Language research should concretely distinguish between various aspects of language use. Furthermore, it should consider the personal nature of language acquisition. The introduction of machine translation broadly affects English education and English use; it is not an issue limited to English writing. The discussion on machine translation may advance plurilingualism and transcultural communication.
In any case, the Pandora's box has been opened. Even in the process of writing this paper, news on AI continued, including on GPT-3, a language model that generates languages other than translation, and AI for image generation. It is unwise to hastily close the lid on this box out of emotional distress. The box should be kept open despite the pain from changing the status quo. At the bottom of the box, there must be a “hope:” the creation of a better society.
i In anonymous student surveys that the author’s department conducted, the author's courses received a reasonably high rating in one of the most indicative questions on the value of learning: “I felt this course was meaningful to me.” The results for the last two years (9 classes, 120 students in total) show that 87% of the students answered “A” (agree), 12% “B” (somewhat agree), 2% “C” (somewhat disagree), and 0% “D” (disagree) to the above question (the total is not 100% because the results were rounded off to the nearest whole number.)
ii The original Japanese edition was first machine-translated and then extensively rewritten by the author. In the final stage, the author’s revision received some advice from a professional proofreader.
iii Nevertheless, coin-like pieces are often used to teach children how to count numbers. Tool use seems to be virtually indispensable for human intelligence.
iv Calculators and machine translation differ significantly in two aspects. First, machine translation extends human intelligence in a far more complex manner than a calculator. Second, due to that complexity, machine translation is far less accurate in its results than a calculator—there is no correct answer to translation in the first place. Consider, for example, a calculator that provides correct answers only approximately 95% of the time; its user needs considerable computational intuition to effectively use the calculator that makes one mistake out of every 20 calculations. Since machine translation is not perfectly reliable now, the introduction of machine translation into the educational process requires much more careful consideration than the introduction of calculators.
v It is interesting to note that the Course of Study did not receive any significant political opposition when it instructed a monolingual approach in English classes in Japanese junior high and high schools. However, one hears many counterarguments that English-only classes are highly impractical given students’ (and some teachers’) English proficiency.
vi The term “intercultural” is taken from the first quote from the plurilingualism literature. This paper will introduce the term “transcultural communication” later to better describe the plurilingual and pluricultural idea.
vii However, if pre-editing is not sufficiently done, post-editing will be time-consuming, and it is sometimes better not to use machine translation. When describing the time advantages of machine translation here, the author assumes that the Japanese manuscript is written in a translation-friendly manner. Pre-editing will be a critical issue when considering the use of machine translation in English education. This also implies that English education may collaborate with Japanese-language education.
viii There has been a recent increase in the number of literary works that are written with the expectation from their conception that they will be translated into English.
This discussion on communication evidently leads to the issue of “translanguaging,” which the author cannot afford to discuss in detail now.
x Films may be suggestive in this regard. In the past, the Academy Awards in the U.S. only honored outstanding films made in languages other than English with the Academy Award for Best International Feature Film (formerly known as the “Academy Award for Best Foreign-Language Film” and, prior to 1956, as the “Special/Honorary Award”). In 2020, however, a foreign-language film produced in a non-English-speaking country (the Korean film, Parasite) won the Academy Award for Best Picture. This episode suggests the emergence of an increased interest in non-English cultures in the English-speaking sphere.
