Bilingual 2030 and the elephant

In a recent exchange over Bilingual 2030, Gerhard Erasmus (“Bilingual 2030,” Sept. 5, page 8) declared that “time’s up for opinions.”

However, the time is hardly up for ideas about how to confront the elephant in the room — the core problem barely mentioned in the discussion.

The elephant does not speak.

Contributors Paul Garrelfs and Daniel Eyrich pointed upstream to junior-high school and earlier, with Eyrich noting that what we assess influences what teachers and students prioritize.

Exactly: Taiwan wants university students to learn academic subjects through English as the medium of instruction (EMI), yet its school system gives them remarkably little incentive to develop spontaneous spoken English. How does it make sense to expect students to learn academic subjects in English when they struggle to formulate spoken questions?

The General Scholastic Ability Test English section assesses reading and writing, but not speaking. Yet the Ministry of Education’s statements and curriculum guidelines have repeatedly stressed using English for interpersonal communication — that developing students’ spoken English matters.

Taiwan’s principal university entrance exam, through its design, tells students, teachers and parents otherwise.

It is as if the system assumes that because high-school students can develop substantial reading competence despite having little speaking practice, this particular receptive competence will somehow carry them through university EMI lectures. Reading a difficult English text at one’s own pace — pausing, rereading and reconsidering — and processing a chain of academic ideas in real time are fundamentally different tasks. Listening proficiency requires extensive comprehensible listening input, while speaking and interaction add productive practice and opportunities for deeper processing of spoken language. Students need both; Taiwan’s school English system cannot reasonably expect strong EMI readiness while giving either one short shrift.

In his influential, research-based Four Strands framework, Paul Nation, one of the world’s leading authorities on second-language vocabulary, recommends roughly equal attention to meaning-focused input, meaning-focused output, deliberate language study and fluency development. Deliberate study is necessary, but focusing on it almost exclusively is like building a diet around just one food group.

Education First’s 2025 English Proficiency Index reports substantially higher speaking scores in Malaysia and the Philippines — where historical circumstances have left English with strong social and educational functions. Taiwan lacks that environment. For Bilingual 2030 to achieve its goals, the ministry must create in junior-high schools the productive speaking practice society does not naturally provide — and make speaking count.

An obvious objection: How can schools possibly test speaking? Individual schools might lack the resources, but national testing at intervals need not be left to individual schools; the technology already exists to administer speaking tests nationally at scale.

An extensive international research literature documents the linguistic difficulties students can face when they undertake EMI without sufficient English proficiency. As Jye Smallwood (“Basic misunderstanding of EMI,” Aug. 9, page 8) wrote, EMI is not a substitute for the English proficiency needed to participate in it successfully.

That does not mean the proper choice is no EMI. The proper choice is bridges.

A few Taiwanese educators are already experimenting with such bridges. At National Chiayi University, Lin Keng-ta (林庚達) teaches the core concepts of difficult engineering material in Mandarin, after which Chen Chien-chung (陳建中) of Marian University in the US explains the same material in English. Students can approach already-familiar professional content through a second language.

Meanwhile, National Taiwan University offers one-credit “Mini-EMI” courses — condensed or introductory versions of regular academic courses designed to ease students into learning and discussing subject matter in English.

Why not extend these types of approaches to junior-high school education? Suppose students complete a history course in Mandarin and, during the summer, those choosing a more ambitious English pathway take a shorter English extension built on material they already understand. Students could examine another source, compare interpretations, discuss causes and complete a project in English. Such an approach could also be tried in other subjects.

Also, why not handle Taiwan’s approximately 6,000-word junior-high-school reference vocabulary more dynamically? The ministry could designate a core portion as productive vocabulary, supplement it with common chunks and collocations and regularly have students build “language islands” — short, prepared passages about their own lives and interests. That would serve as a flexible foundation for real interactions.

Erasmus is right that we need evidence. So run pilot programs and gather it: measure speaking development, subject retention and subsequent readiness for EMI. Student choice should be part of such experiments, too: productive language learning requires sustained participation, so the effects of motivation and student buy-in should themselves be measured.

Taiwan’s education system can produce outstanding results, as demonstrated by its consistently top-tier math, science and reading performances in international assessments like the Programme for International Student Assessment. These achievements hardly suggest, though, that the same system can deliver an outcome it neither cultivates nor directly assesses.

The broad evidence is already clear about what Bilingual 2030 most needs to achieve its goals: much more speaking practice and consequential testing of speaking ability at the junior-high school level, along with appropriate scaffolding for EMI at university whenever students are not yet linguistically ready.

What is needed now are more ideas about how best to accomplish these things, not an end to such opinions. Then put the most promising ideas to the test — and get the elephant to speak.

Kevin Fitzpatrick has taught English in Taiwan for 35 years.