When Taiwan Starts Speaking Like China
In June 2024, a Taiwanese university student compiled a list of the Mainland expressions he could no longer tolerate hearing. 外賣 (“takeout/delivery”) came fifth. 家人們 (“family,” used by influencers to address followers) came fourth. More irritating were children calling a bus 公交車 (“bus”) rather than the Taiwanese 公車, or referring to a video as 視頻 (“video”) rather than 影片. His greatest objection was reserved for 質量 (“quality”), since in Taiwanese usage the same characters traditionally refer to physical mass while 品質 means quality. What, he wondered, had happened to education if even primary-school children were learning their vocabulary from Douyin?
The responses produced more examples, including one commenter whose mother, a retired Chinese-language teacher, had unexpectedly begun using the Mainland word for “video.” Yet the original poster was willing to spare 內卷 (“involution”), a Mainland term for the exhausting escalation of competition in which everybody works harder without becoming much better off, because he thought Taiwan lacked an equally precise expression. His complaint was consequently selective. The Mainland word for “takeout” displaced a Taiwanese one without adding much; “involution” had become useful for describing Taiwanese schools and workplaces precisely because it named an experience that travelled more easily than the society in which the term had developed.
Mainland expressions are now familiar well beyond the people who deliberately follow Mainland politics or entertainment. A National Chengchi University study asked 479 Taiwanese Mandarin speakers about twenty popular Mainland terms and found very high comprehension, averaging 4.25 out of five, alongside an almost equally high ability to recognise their Mainland origin at 4.08. Attitudes towards actually using the terms were considerably cooler. Exposure had neither made the words foreign nor fully domesticated them: Taiwanese speakers could understand a term, recognise immediately that it came from Mainland China and still disagree over whether it belonged in Taiwanese speech. More unexpectedly, support for the Tsai government's cross-Strait policy had no significant relationship with respondents' linguistic attitudes towards the Mainland vocabulary. The political position respondents took towards China did not neatly predict what they were prepared to say.
For many young Taiwanese, these expressions are encountered while using Xiaohongshu for purposes with little obvious connection to China or cross-Strait politics. Taiwanese users do not have to seek out something recognisable as “Chinese culture” before encountering Mainland ways of describing clothes, relationships, studying or food; they encounter them while searching for the clothes, relationships, studying or food. A term picked up there can then pass through Taiwanese group chats and classrooms until the person using it may have learnt it from another Taiwanese speaker rather than directly from anybody in Mainland China. Its provenance—its recognisable Mainland origin—can survive even as the route through which people acquire it becomes increasingly Taiwanese.
That is also why disputes over 支語警察 (“Mainland-expression police”), the mocking label for people who correct Mainland usage, are rarely arguments about comprehension. Sociolinguist Hsi-Yao Su records ordinary online conversations in which the vocabulary used to discuss something as mundane as yoghurt becomes the subject of the discussion itself. Somebody identifies a Mainland expression, somebody else objects to the correction, and what began as an unremarkable use of Mandarin becomes an argument over cultural intrusion. The correction itself is only possible because the expression remains recognisable as Mainland even after it has become ordinary enough to appear in an otherwise unremarkable Taiwanese conversation.
The political backdrop has been moving in almost the opposite direction. National Chengchi University's long-running identity survey, which now extends from 1992 to June 2026, records the emergence of Taiwanese rather than Chinese identification as the dominant answer over the same decades in which digital contact with Mainland popular culture became dramatically easier. The comparison is imperfect—national identity cannot be inferred from vocabulary, and the survey covers the whole population rather than Xiaohongshu's young users—but that is partly what makes the linguistic evidence interesting. The boundary separating Taiwan politically has become clearer without requiring an equally clean cultural boundary in everyday life.
For Beijing, however, precisely this kind of cultural familiarity has sometimes been given political significance. In May 2025, Zhang Weiwei, director of Fudan University's China Institute, pointed to Taiwanese young people's enthusiasm for Xiaohongshu while arguing that, after unification, Taiwan would be easier to govern than Hong Kong because young Taiwanese were increasingly influenced by the Mainland. Taiwan's Mainland Affairs Council answered by warning schools about Mainland apps and the possibility of united-front influence. The same hours spent looking through restaurant recommendations, beauty advice or study posts could consequently appear in Zhang's account as evidence that Mainland influence was making Taiwan more amenable to eventual unification, and in the Council's response as precisely the kind of exposure Taiwanese schools should treat more cautiously.
American films or Korean music can acquire enormous audiences abroad without creating a dispute over whether their audiences therefore belong to an American or Korean nation. Mainland Chinese culture reaches Taiwan under different political conditions because Beijing's claim to Taiwan already gives cultural commonality a place within the question of national belonging. In official accounts of cross-Strait relations, the existence of separate governments since 1949 has not created separate Chinese and Taiwanese nations; shared ancestry, history, language and cultural traditions are instead invoked as evidence of a national community that persists despite political separation. When Taiwanese users adopt Mainland vocabulary or spend time on Mainland platforms, the activity can be interpreted from Beijing as a continuation—or renewed strengthening—of ties within an existing Chinese cultural community, even when the users themselves understand the same behaviour simply as participation in a useful or appealing part of Chinese-language digital culture. Its Mainland origin can consequently remain uncontested while the significance attached to that origin differs sharply between the person using it and the political actors interpreting its use.
Joseph Nye's account of soft power provides one way of separating these processes. Nye defines power as the ability to affect others in order to obtain preferred outcomes, which can be achieved not only through coercion or payment but by making those outcomes attractive to others. Culture can contribute to this attraction by shaping how foreign audiences understand a country and the possibilities associated with it, but its popularity becomes politically consequential only insofar as those perceptions affect what people subsequently prefer or do. A country's films, universities or digital culture can consequently become familiar and appealing abroad without bringing foreign preferences any closer to the outcomes that country wants. In Nye's formulation, cultural reach and political influence are connected through this process rather than treated as different names for the same phenomenon.
Zhang's interpretation requires the popularity of Xiaohongshu to produce something beyond the circulation Nye describes. Its Taiwanese audience establishes that Mainland cultural products can command attention across the Strait, while the spread of its vocabulary suggests that this contact can become embedded in everyday language rather than remaining passive consumption. Neither, however, establishes the change in preferences on which soft power ultimately depends. For Zhang's interpretation to hold in the political sense he gives it, familiarity with Mainland culture would have to do more than make particular expressions, creators or habits seem ordinary: it would have to alter how Taiwanese users understand the Mainland itself, and eventually make the political relationship sought by Beijing more acceptable to them. The available evidence instead separates cultural uptake from those political preferences. Respondents in the NCCU study could be highly familiar with Mainland vocabulary without their attitudes towards that vocabulary corresponding significantly to support for the government's cross-Strait policy, while the much broader expansion of Mainland digital contact has coincided with an increasingly predominant Taiwanese identity. Neither finding proves that cultural familiarity has no political effect, but both make it difficult to infer the direction or magnitude of that effect from cultural circulation alone.
The original poster had already made this distinction himself, although on the much smaller scale of a single word. His objection to words such as the Mainland term for “video” depended on recognising them as imports and preferring an established Taiwanese equivalent; he made the same judgement about “involution” and reached a different conclusion because the term expressed something for which he thought Taiwanese Mandarin lacked an equally precise alternative. Mainland provenance had therefore remained visible throughout the process of adoption. What had changed was the practical value he attached to it: once the expression described competition in Taiwanese schools and workplaces effectively enough, its origin was no longer sufficient reason to reject it. Linguistic adoption in this case tells us that a Mainland expression has become useful within a Taiwanese context, but it cannot by itself tell us whether the speaker has become more receptive to Mainland China beyond that context.
Taiwan makes the political meaning of cultural influence unusually difficult to infer because the boundaries involved are no longer moving in the same direction. Mainland vocabulary can become more familiar, more useful and more deeply embedded in Taiwanese speech while Taiwanese political identification becomes increasingly distinct; at the same time, the vocabulary's Mainland origin remains visible enough to provoke linguistic policing and to be interpreted by Beijing as evidence of continuing cultural connection. Describing this simply as either cultural convergence or political divergence misses what is happening between them. Chinese in origin, Chinese in everyday use and Chinese in political belonging are becoming increasingly separable categories, even though cross-Strait politics repeatedly brings them back into contact. Taiwanese teenagers speaking “Chinese internet” are therefore interesting for more than the vocabulary they borrow: their speech offers a particularly visible case of what happens when culture continues to cross a political boundary more easily than the political meanings historically attached to that culture.
Like it !