In lieu of an abstract, here is a brief excerpt of the content:

  • The Liangdao skeleton and the dangers of overinterpretation
  • Robert Blust (bio)

The first thing to note with regard to the discovery of the so-called ‘Liangdao man’ is that any recipe used to ‘cook up’ an interpretation of the human history of East and Southeast Asia based on this fragile foundation will require a large pinch of speculation. My purpose in entering this discussion is not to enlarge the compass of the speculations already advanced, but rather to rein them in with some reminders of things about which we have much better observational control.

Following Norman and Mei (1976), Li maintains (2015, 224) that “There is linguistic evidence that the Austroasiatics inhabited the Yangtze delta and parts of the southeast coast [of China] during the first half of the first millennium B.C.” Regardless of how one regards the claim that modern forms of Min show traces of early borrowing from AA languages, these surely were not the only languages in southern China before political unification under Qin Shi Huang in the 3rd century BC initiated the southward expansion of the Han Chinese. Hmong-Mien languages are indigenous to the lower Yangtze region, and Kra-Dai languages have been in southern China since the earliest Chinese records. We have no idea how many other language families were among the ‘Hundred Yue’, which have since become absorbed and lost to history, but given the present linguistic picture, and the known process of sinicization in recent centuries, we have every reason to believe that southern China was linguistically more diverse in 6,000 BC than it is today. It is, therefore, one thing to argue that possible loanswords in modern Min are evidence for a significant population of AA speakers in southern China before the [End Page 242] Chinese, but it is a very different thing to imply, as Li (228) does, that these were the only pre-Chinese languages in southern China, and on this basis to conclude that “the Liangdao man (ca. 8,000 BP) was an Austric, rather than an “early Austronesian.””

Pursuing the argument from language a step further, if one accepts Austric, as Li (227) does, the location of the Austric homeland becomes critical. AA languages extend from the typologically distinct Munda group in northeast India, through the geographically isolated Khasian languages in Meghalaya, Assam and portions of Bangladesh, to the Palaungic and Mangic languages of northern Burma and southern China, to the main body of Mon-Khmer speakers in Thailand, Laos, Cambodia and Vietnam, to the Aslian languages of the interior Malay peninsula and the Nicobarese languages in the Bay of Bengal (for a thorough discussion of subgrouping issues affecting these languages, introduced by a map showing their distribution cf. Sidwell 2015; for what is still the most detailed linguistic map available for mainland Southeast Asia, cf. Lebar, Hickey and Musgrave 1964, pocket). As described by Sidwell (2015), the internal classification of AA is still unsettled. Pinnow (1959, 1963) recognized just two primary branches, 1. Western (Munda and Nahali), 2. Eastern (the rest); Diffloth (1974) recognized three, 1. Munda, 2. Nicobar Islands, 3. Mon-Khmer, Parkin (1991) recognized four, 1. Munda, 2. Nicobarese, 3. Aslian, 4. Mon-Khmer, and Sidwell (2015) favors no fewer than 11 primary groups. The one feature shared by nearly all classifications is the recognition of Munda as a primary branch. The Munda languages have evidently been in northeast India for a very long time, and any theory of an AA homeland must take their location into account. Placing the Austric homeland “in the east coast of South China” as suggested by Li (228) may help explain the settlement of Taiwan by Proto-Austronesian (PAN) speakers splitting off from a proposed Proto-Austric ancestor closer to their historical location, but it has the opposite effect in explaining the distribution of the primary branches of AA, most of which are located far from southeast China.

Since Sapir ([1916]1968) it has been recognized that the homeland for a collection of related languages is normally the area of greatest diversity. In some cases, as with Sinitic, this expectation has been masked by subsequent language leveling: there is good historical and [End Page 243...

pdf

Share