Written by Amy Pei-jung Lee.
Image credit: Elizabeth Zeitoun.
Taiwan’s linguistic landscape is characterised by remarkable diversity, with a rich heritage of Formosan languages spoken by Indigenous peoples across the island. Historically, these languages were transmitted orally from generation to generation. Approximately 24 Formosan languages are known to have been spoken in the past centuries in Taiwan (including Keta(n)galan, Basay, Taokas, Papora, Babuza, Hoanya, Siraya, Makattao, Taivoan, Kavalan, Pazeh-Kaxabu, Thao, Atayal, Saisiyat, Bunun, Tsou, Saaroa, Kanakanavu, Rukai, Paiwan, Puyuma, Amis, Seediq, Yami).
The arrival of external powers approximately 400 years ago, beginning with the Europeans, followed by the Japanese and later the Chinese, introduced successive forms of colonisation that drastically altered this situation. Today, nearly half of the languages listed above are already extinct, and the remaining languages are rapidly declining, with the exception of Siraya, whose status has shifted from extinct to dormant over the past two decades. This decline can be traced to a series of language policies that were implemented during Taiwan’s colonial periods and under Nationalist rule, which severely suppressed Indigenous languages in favour of other dominant languages, particularly Japanese and Mandarin Chinese.
In response to these challenges, Taiwan has seen a shift in language policy since the mid-1990s, with increasing efforts to revitalise and promote Indigenous languages. A combination of governmental policies, institutional programs, and grassroots community efforts has supported these revitalisation initiatives. Key milestones in this ongoing process have included the development of textbooks for Indigenous language instruction, the establishment of language proficiency tests, the compilation of dictionaries, and the transcriptions of texts. The implementation of these initiatives highlighted the urgent need for standardised orthographic conventions, despite the fact that attempts to transcribe the Indigenous languages date back to the arrival of the Europeans.
The Formosan languages, which are toneless, typically exhibit relatively simple phonemic inventories, consisting of around twenty consonants and generally four vowels. These systems usually include a series of voiceless and voiced stops, fricatives, nasals, liquids, and glides, together with the vowel system, although variations have arisen through phonological change. Despite the relative simplicity of these phonemic systems, the development of writing systems for Formosan languages has been a complex and contentious issue, as thoroughly summarised in I-sin Ng’s contribution (2024). The introduction of Roman-based scripts by Dutch missionaries in the 17th century, since no Spanish records have survived, was later followed by subsequent attempts to represent these languages by using Chinese characters, Japanese syllabaries, and even Mandarin phonetic symbols (Bopomofo).
Many scholars and language activists have argued that Romanised writing systems provide the most suitable and practical means of representing the phonological structure of oral languages. In the early 1990s, the Ministry of Education commissioned linguist Paul Jen-kuei Li to design appropriate orthographic systems for Taiwan’s Indigenous languages. His pioneering work provided detailed descriptions of the phoneme inventories of the nine languages officially recognised at the time, outlining dialectal variation where present and proposing corresponding Romanised representations.
However, the orthographic correspondences he devised were not widely adopted for at least two reasons. First, the Mandarin-only policy established in 1947, which severely restricted the use of Indigenous languages, was partially counterbalanced by the efforts of religious institutions, which played a critical role in language documentation during the period of martial law. Missionary activities often involved the production of written materials such as Bible translations, folktales, myths, dictionaries, and primers for local populations. Second, in the early 1990s, shortly after the end of decades of martial law, the forces of language revitalisation, brought about by the democratisation and liberalisation of Taiwan, had not yet fully taken root in society, and the need for standardised orthographic conventions had not yet been widely recognised.
As mentioned above, by the early 2000s, it had become clear that the development of textbooks and language proficiency exams, along with the emergence of the internet and online corpora, highlighted the need for a consistent and widely accepted writing system. This was essential in providing both learners and educators with reliable and effective tools for transcriptions and transmission. The case of Siraya also demonstrates that written materials can play a crucial role in the preservation and revitalisation of endangered languages, including those that are moribund or even extinct.
In response to this new situation, a series of public meetings was convened in 2003 onwards to address the issue of orthographic standardisation. These meetings brought together linguists and native speakers. Eventually, they led to the promulgation, on December 15, 2005, by the Ministry of Education and the Council of Indigenous Peoples, of a revised set of orthographic systems for the 12 recognised Formosan languages at the time, divided into 42 dialects.
While this represented a significant step forward, a major difficulty lay in the failure to establish a single unified system across these dialects. Instead, separate orthographic conventions were adopted for each of the 42 dialects associated with the 12 languages and their dialectal variants. This fragmentation resulted partly from previously writing traditions introduced by churches or earlier linguistic studies, such as Tung’s (1964) grammar of Tsou. In other cases, it stemmed from misunderstandings regarding certain phonemic distinctions, whether non-existent or misinterpreted, which created additional challenges for standardisation and consistency. These issues, in turn, further complicated orthographic learning, language teaching, and broader revitalisation efforts.
By the early 2010s, it had become clear that a truly unified orthographic system applicable to all Formosan languages was needed to ensure consistency and facilitate language revitalisation efforts. Linguists and language advocates have since continued to work toward developing such a system, with the ultimate goals of establishing a definitive and standardised writing system that would be widely accepted by Indigenous communities. However, these ongoing efforts have long encountered challenges from native speakers and language teachers without a background in historical linguistics. Their parochial perspectives on dialectal variation, often accompanied by a preference for particular pronunciation norms, have led to continuing disputes over orthographic conventions.
Both progress and setbacks mark the history of Formosan language revitalisation in Taiwan. Although significant advances have been made in recent decades, considerable work remains to be done, particularly in the standardisation of writing systems. Addressing this challenge is essential to ensuring the long-term survival of Taiwan’s Indigenous languages.
Having participated in this revitalisation process for nearly twenty years, I survey in this paper the challenges involved in the choice of particular orthographic letters and symbols from a linguistic perspective and put forward several recommendations aimed at promoting greater orthographic standardisation as a guide for future policy decisions.
Please read the full article in the International Journal of Taiwan Studies 9.1 (March 2026): https://brill.com/view/journals/ijts/9/1/ijts.9.issue-1.xml.
Amy Pei-jung Lee is Professor in the Department of Indigenous Languages and Communication at National Dong Hwa University (NDHU) in Hualien, Taiwan. Her research focuses on Formosan languages, with interests in descriptive, cognitive, and historical linguistics. Based on long-term fieldwork and collaboration with community speakers, her work examines topics such as phonology, reduplication, metaphorical euphemism, ideophones, the language of olfaction, and loanwords. She is also engaged in Indigenous language revitalisation. She explores ways of connecting linguistic research with digital media and emerging technologies to support the documentation and promotion of Indigenous languages in Taiwan.
This article was published as part of the International Journal of Taiwan Studies-Taiwan Insight special issue on ‘Indigenous Language Policies in Taiwan and Beyond’.