Families of Formosan languages before Chinese colonization, per Blust (1999). Malayo-Polynesian (red) may lie within Eastern Formosan (purple). Note that the white section in the northwest of the country does not indicate a complete absence of aboriginal people from that part of Taiwan. On Chinese-language sources, this area is listed as the homeland of various Pingpu groups (e.g. the Kulon), and certain other groups (e.g. the Taokas) are arranged slightly differently than they are on the above map.
The Formosan languages are the languages of the indigenous peoples of Taiwan, all of which are Austronesian. The Taiwanese aborigines recognized by the government are about 2.3% of the island's population. However, far fewer can still speak their ancestral language because of centuries of language shift. Of the approximately 26 languages of the Taiwanese aborigines, at least ten are extinct, another four (perhaps five) are moribund, and all others are to some degree endangered.
The aboriginal languages of Taiwan have significance in historical linguistics since, in all likelihood, Taiwan was the place of origin of the entire Austronesian language family. According to linguist Robert Blust, the Formosan languages form nine of the ten principal branches of the Austronesian language family, while the one remaining principal branch contains nearly 1,200 Malayo-Polynesian languages found outside Taiwan. Although some other linguists disagree with some details of Blust's analysis, a broad consensus has coalesced around the conclusion that the Austronesian languages originated in Taiwan. The theory has been strengthened by recent studies in human population genetics, supporting also the matrilineal nature of the migration.
Based on recent archaeological evidence as well as linguistic evidence, Roger Blench (2014) considers the Austronesians in Taiwan to have been a melting pot of immigrants from various parts of the coast of eastern China that had been migrating to Taiwan by 4,000 B.P. These immigrants included people from the foxtail millet-cultivating Longshan culture of Shandong (with Longshan-type cultures found in southern Taiwan), the fishing-based Dapenkeng culture of coastal Fujian, and the Yuanshan culture of northernmost Taiwan which Blench suggests may have originated from the coast of Guangdong. Based on geography and cultural vocabulary, Blench believes that the Yuanshan people may have spoken Northeast Formosan languages. Thus, Blench believes that there is in fact no "apical" ancestor of Austronesian in the sense that there was no true single Proto-Austronesian language that gave rise to present-day Austronesian languages. Instead, multiple migrations of various pre-Austronesian peoples and languages from the Chinese mainland that were related but distinct came together to form what we now know as Austronesian in Taiwan. Hence, Blench considers the single-migration model to be inconsistent with both the archaeological and linguistic (lexical) evidence.
All Formosan languages are slowly being replaced by the culturally dominant Taiwanese Mandarin. In recent decades the Taiwanese government started an aboriginal reappreciation program that included the reintroduction of Formosan first language in Taiwanese schools. However, the results of this initiative have been disappointing.
In 2005, in order to preserve the language of the indigenous people of Taiwan, the council established a Romanized writing system for all Taiwan's aboriginal languages. The council has also helped with classes and language certification programs for members of the indigenous community and the Han Chinese to help the conservation movement.
It is often difficult to decide where to draw the boundary between a language and a dialect, causing some minor disagreement among scholars regarding the inventory of Formosan languages. There is even more uncertainty regarding possible extinct or assimilated Formosan peoples. Frequently cited examples of Formosan languages are given below, but the list should not be considered exhaustive.
|Amis||ami||5||'Amisay a Pangcah, Siwkolan, Pasawalian, Farangaw, Palidaw|
|Atayal||tay||6||Squliq, Skikun, Ts'ole', Ci'uli, Mayrinax, Plngawan||high dialect diversity, sometimes considered separate languages|
|Bunun||bnn||5||Takitudu, Takibakha, Takivatan, Takbanuaz, Isbukun||high dialect diversity|
|Kavalan||ckv||1||listed in some sources as moribund, though further analysis may show otherwise|
|Paiwan||pwn||4||Eastern, Northern, Central, Southern|
|Puyuma||pyu||4||Puyuma, Katratripul, Ulivelivek, Kasavakan|
|Rukai||dru||6||Ngudradrekay, Taromak Drekay, Teldreka, Thakongadavane, 'Oponoho|
|Seediq||trv||3||Tgdaya, Toda, (Truku)|
|Yami||tao||1||also called Tao. Not a member of the "Formosan languages", but a Malayo-Polynesian language.|
|Dialects||Native speakers, extinction date & notes|
|Babuza||bzg||3?||Babuza, Takoas, Favorlang (?).||0.|
|Taokas||(bzg)||3?||Babuza, Takoas, Favorlang (?).||0. Late 19th century. Ongoing revival efforts.|
|Favorlang||(bzg)||3?||Babuza, Takoas, Favorlang (?).||0.|
|Papora||ppu||2?||Papora, Hoanya (?).||0.|
|Hoanya||ppu||2?||Papora, Hoanya (?).||0.|
|Taivoan||tvx||1||0. Late 19th century. Ongoing revival efforts.|
|Siraya||fos||2?||Siraya, Makatao (?).||0. Late 19th century. Ongoing revival efforts.|
|Makatao||(fos)||2?||Siraya, Makatao (?).||0. Late 19th century. Ongoing revival efforts.|
Most Formosan languages display verb-initial word order (VSO (verb-subject-object) or VOS (verb-object-subject)) with the exception of some Northern Formosan languages, such as Thao, Saisiyat, and Pazih, possibly from influence from Chinese.
Tanan Rukai is the Formosan language with the largest number of phonemes with 23 consonants and 4 vowels containing length contrast, while Kanakanabu and Saaroa have the least number of phonemes with 13 consonants and 4 vowels (Blust 2009:165).
|*t||t, s||t, s, ʃ||t, θ||t, c (s)|
|*c||z [dz]||h||t||x, h|
|*g||k-, -z- [dz], -t||k-, -z- [ð], -z [ð]||k-, -ð-, -ð||k-|
|*ɣ||x||l [ḷ] (> Ø in Tonghœʔ)||ɬ||ɣ, r, Ø|
|*l||r||l [ḷ] (> Ø in Tonghœʔ)||r||l|
|*t||t, c||t, c||t, c||t||t||t||t, ʈ||tj [č], ts [c]|
|*c||s, Ø||c||θ, s, Ø||c ([s] in Central & South)||c||s||s||t|
|*q||Ø||ʔ||Ø||q (x in Ishbukun)||ɦ||Ø||ɦ||q|
|*b||v||v [β]||b||b||f||b||v [β]||v|
|*d||s||c||ḍ||d||r||z||d, z||dj [j], z|
|*j||s||c||d||d||r||z||d, z||dj [j], z|
|*g||k-, -ɬ-||k-, -l-, -l||g||k-, -Ø-, -Ø||k-, -n-, -n||k-, -n-, -n||h-, -d-, -d||g-, -d-, -d|
|*ɣ||r||r||r, Ø||l||l [ḷ]||ɣ||r||Ø|
|*l||Ø||Ø, l||ñ||h-, -Ø-, -Ø||l [ḷ]||r, ɣ||l [ḷ]||l|
|*b||b||p||b, -p||b, w|
|*d||d-, -l-, -d||h||d, -t||ḍ, r|
|*j||d-, -l-, -d||ch||j, -t||d|
|*g||k-, -l-, -d||Ø||d-, -r-, -r||g-, -r-, -r|
|*ɬ||n||ñ, n, l||l-/ñ-, -ñ-/-n-, -n||n|
The following table lists reflexes of Proto-Austronesian *j in various Formosan languages (Blust 2009:572).
|Saaroa||ɬ (-ɬ- only)|
|Atayal||r (in Squliq), g (sporadic), s (sporadic)|
|Sediq||y (-y- only), c (-c only)|
|Pazeh||z ([dz]) (-z- only), d (-d only)|
The following table lists reflexes of Proto-Austronesian *ʀ in various Formosan languages (Blust 2009:582).
|Kavalan||ʀ (contrastive uvular rhotic)|
|Atayal||g; r (before /i/)|
|Thao||lh (voiceless lateral)|
|Saisiyat||L (retroflex flap)|
Lenition patterns include (Blust 2009:604-605):
Li (2001) lists the geographical homelands for the following Formosan languages.