Skip to content
Language Lineage

Proto-Austronesian

The reconstructed ancestor of more than 1,200 languages. It had a consonant system rich in contrasts (*C, *t, *N, *n, *S, *s) that only the languages of Taiwan still keep apart.

c. 4000–3000 BCE (reconstructed)Taiwan
On this page
  1. Origins
  2. Why Taiwan?
  3. Daughter Branches
  4. Where its languages are spoken
  5. Turning points
  6. Contested classifications

01

Origins

Proto-Austronesian is reconstructed, like Proto-Germanic, by comparing its descendants. The vocabulary rebuilt this way describes a farming and fishing society: *pajay 'rice (plant)' (Malay padi), *Rumaq 'house' (Malay rumah), *Sapuy 'fire' (Malay api, Hawaiian ahi). It also includes basic words that have survived almost unchanged across 20,000 kilometres, such as *mata 'eye' and *lima 'five', still mata and lima in Malay and maka and lima in Hawaiian.

Most linguists and archaeologists place the language in Taiwan, about five to six thousand years ago. Its speakers' ancestors had probably crossed the Taiwan Strait from the coast of what is now southern China, bringing rice and millet farming. No relative on the mainland has been proven, so this tree's roots are drawn dashed.

02

Why Taiwan?

The argument is linguistic before it is archaeological. The Austronesian languages of Taiwan are few, but between them they make up most of the family's primary branches. All the rest, from Madagascar to Hawaii, belong to a single branch, Malayo-Polynesian. A family's place of origin is usually where its deepest splits are found, just as English's greatest dialect diversity is in England and not in Australia.

Taiwan's languages also keep sound contrasts that every language outside Taiwan has merged. Most Formosan languages still distinguish reflexes of *C from *t and *N from *n, while Malayo-Polynesian merged each pair. Archaeology agrees broadly: red-slipped pottery and other traces move from Taiwan into the northern Philippines around 2000 BCE.

03

Daughter Branches

The first division is between the Formosan languages, which stayed in Taiwan, and Malayo-Polynesian, which left. 'Formosan' is a geographical label and not a branch: Taiwan's languages form perhaps nine or ten separate primary branches. This tree draws them as one limb for legibility and records the dispute under Contested Classifications.

On the map

Where its languages are spoken

Timor-LesteTaiwanBruneiSingaporeSamoaTongaAmerican SamoaCook Is.Fr. PolynesiaWallis and Futuna Is.
Spoken todayShading marks whole countries, not the regions within them.
  1. PhilippinesPhilippine c. 105 million
  2. IndonesiaPhilippine c. 2–3 million · Chamic c. 3.5 million · Malayic c. 80 million · Barito c. 1 million · Central Malayo-Polynesian c. 4–5 million
  3. MalaysiaPhilippine · Malayic c. 17 million
  4. Timor-LesteCentral Malayo-Polynesian c. 1 million
  5. SingaporeMalayic c. 350,000
  6. BruneiMalayic c. 300,000
  7. CambodiaChamic c. 300,000
  8. SamoaPolynesian c. 170,000
14 more countries
  1. TongaPolynesian c. 100,000
  2. American SamoaPolynesian c. 50,000
  3. TuvaluPolynesian c. 10,000
  4. ChinaChamic c. 4,000
  5. MadagascarBarito c. 25–30 million
  6. ThailandMalayic c. 1–3 million
  7. VietnamChamic c. 1–1.3 million
  8. ChilePolynesian
  9. Cook IslandsPolynesian
  10. French PolynesiaPolynesian
  11. New ZealandPolynesian
  12. TaiwanFormosan languages · Philippine
  13. United StatesPolynesian
  14. Wallis and FutunaPolynesian

Figures are rough first-language estimates, rounded, from censuses and Ethnologue. Counts for small or scattered communities are especially uncertain.

Rings along the limb

Turning points

Oldest first.

  1. Splitc. 4000–3500 BCEContested

    Across the Taiwan Strait

    Farmers from the coast of southern China settle Taiwan, and their language develops into Proto-Austronesian. The link to the mainland is archaeological, not linguistic: no Austronesian relative survives there, and which mainland culture it was is debated.

    Sources (2)
    • Bellwood, Peter. 2005. First Farmers: The Origins of Agricultural Societies. Oxford: Blackwell.
    • Blust, Robert. 2013. The Austronesian Languages, revised ed. Canberra: Asia-Pacific Linguistics.
  2. Splitc. 3000–2500 BCEContested

    Breakup within Taiwan

    Proto-Austronesian splits into dialect groups across the island. One of them later leaves Taiwan as the ancestor of Malayo-Polynesian. Both the order of the splits and the number of primary branches are disputed.

    Sources (1)
    • Blust, Robert. 1999. Subgrouping, circularity and extinction: some issues in Austronesian comparative linguistics. In Selected Papers from the Eighth International Conference on Austronesian Linguistics, ed. Elizabeth Zeitoun & Paul Jen-kuei Li. Taipei: Academia Sinica.

Outside the tree

Contested classifications

These disagreements may never be settled. Each position is given with its grounds, and no side is declared correct.

  • Minority position

    Austric

    Austronesian and Austroasiatic (Mon-Khmer, Munda, Vietnamese) descend from one ancestor

    Evidence for

    Wilhelm Schmidt proposed it in 1906 on the basis of shared word-building morphology, including a causative prefix pa- and infixes such as -m- and -n- found in both families, together with some basic vocabulary. Lawrence Reid has argued that the morphological parallels are too specific to be chance.

    Evidence against

    Few regular sound correspondences have been demonstrated, and the shared words are few enough to be explained by chance or ancient contact. Infixes of this kind also occur in unrelated languages. Most specialists regard Austric as plausible at best but unproven.

    Sources (2)
    • Schmidt, Wilhelm. 1906. Die Mon-Khmer-Völker: ein Bindeglied zwischen Völkern Zentralasiens und Austronesiens. Braunschweig: Vieweg.
    • Reid, Lawrence A. 2005. The current status of Austric: A review and evaluation of the lexical and morphosyntactic evidence. In The Peopling of East Asia, ed. Laurent Sagart, Roger Blench & Alicia Sanchez-Mazas. London: RoutledgeCurzon.
  • Minority position

    Austro-Tai

    Austronesian and Kra-Dai (Thai, Lao, Zhuang and relatives) share an ancestor

    Evidence for

    Kra-Dai shares basic vocabulary with Austronesian, often matching the end of the Austronesian word: Proto-Tai ta 'eye' beside Austronesian *mata, and taay 'die' beside *m-atay. Ostapirat has shown regular sound correspondences for a core set of such words. Sagart goes further and places Kra-Dai inside Austronesian, as an offshoot from Taiwan.

    Evidence against

    Kra-Dai is tonal and isolating, and centuries of contact with Chinese have overlaid its vocabulary, so the evidence rests on a limited set of words that could reflect early borrowing. Supporters disagree among themselves whether Kra-Dai is a sister of Austronesian or a branch within it. The proposal is taken more seriously than Austric but is not established.

    Sources (3)
    • Benedict, Paul K. 1942. Thai, Kadai, and Indonesian: A new alignment in Southeastern Asia. American Anthropologist 44: 576–601.
    • Ostapirat, Weera. 2005. Kra-Dai and Austronesian: Notes on phonological correspondences and vocabulary distribution. In The Peopling of East Asia, ed. Laurent Sagart, Roger Blench & Alicia Sanchez-Mazas. London: RoutledgeCurzon.
    • Sagart, Laurent. 2004. The higher phylogeny of Austronesian and the position of Tai-Kadai. Oceanic Linguistics 43(2).