Bear, Bamboo, Fog — the Homeland of Sino-Tibetan languages
In comparative linguistics, word roots from related languages in a family may reveal the common knowledge shared by the proto speakers of the language about their familiar natural environment. The lack of true cognates to signify a basic object from the nature across different languages within the same linguistic family leads to the suspicion that the earliest forebears of the proto-language may not have been familiarized with the said natural object before the split of the proto language into the extant languages. For example, Abram has been seeing automobiles and call them “car” by the time he had three kids, all his three kids would learn the word “car” and use it as a basic root to describe human-transporting machines even when they change the term with certain morphological adaptations. However, if Abram has never seen and called a “car” and all his kids learned about this machine on their own after they moved out, they may call the machine with their own convenient word root, e.g. auto-mobile, big wheel, transporter, vehi-cle, etc.
In earlier posts, I have talked about the competing hypothesis of the homeland of Sino-Tibetan language (ST): the northern China hypothesis vs. Sichuan-Himalaya hypothesis. The northern China hypothesis is really a catch-all umbrella term to put both central China Yangshao culture along the middle Yellow River (YR) and the upper YR Qinghai-Gansu area in the same designation. However, upper YR and middle YR are very distinct culturally and geographically, as well as in terms of their access to other contemporaneous cultures that bore the peopling of the Y haplogroups of C2, Q, and N. Linjia site of Majiayao culture at upper YR in Gansu is the earliest site for bronze metallurgy, suggesting an Inner Asia route of technological spread from the west. On the other hand, Yangshao culture as a millet farming population has acquired millet from the much earlier Transeurasian cultures in today’s Amur and Liao river areas (Robeets et al 2021), who had cultivated millet as hunter-gatherers and only needed to partly domesticate millet because its harvest cycle is a short period of 45 days. The spread of technological advances such as millet cultivation or bronze metallurgy does not mean a parallel spread of genetic and linguistic bearers. Archaeological findings that formed Majiayao culture may have appeared later than Yangshao and its people acquired millet farming from Yangshao, but they only imply the route of technological transmission, not a genetic and linguistic one. An archaeologist from the future may dig out sites in Guangdong with much advanced mastery and earlier adoption of semi-conductor technology, compared to Gansu, owing to Guangdong’s convenient maritime commerce route with the United States, then s/he may wrongly conclude that Guangdong precedes and leads the peopling of Gansu in a form of demic diffusion.
Back to my point, the spread of technology and artifacts is a separate phenomenon from the spread of language and cannot be used as a piece of foundational evidence for locating the homeland (urheimat) of a language family. Lauren Sagart, Zhang Menghan, He Guanglin, and many more have been forging a link between different operational levels without carefully differentiating their unique mechanisms.
Then back to my point in the first paragraph, we may locate the homeland of ST by tracing it to a natural environment, in which most languages in the ST family would have common etyma (root words) to describe the typical products of this natural environment. A large part of the etymological information in this post comes from Jacques Guillaume’s forthcoming article “Origins of Sino-Tibetan, bringing in evidence from history, biology and archaeology” and Matisoff (2009) “Stable Roots in Sino-Tibetan/Tibeto-Burman”, as well as from the STEDT database.
Many have suggested northern China as the Urheimat of ST and the demic diffusion of millet farming spread proto-ST from northern China to Tibet and Basuria. However, there is simply no common cognate for millet (both broomcorn and foxtail species) across ST. The Old Chinese 稷 tsik can only find a correspondence with Nungish cek. This oddity again suggests that Yangshao people, or whoever grew millet, did not speak a proto-ST language, and likely had learned millet cultivation from Trans-Eurasians in Manchuria, instead of inventing and later spreading it to other ST speakers. If the middle YR area in north China was the homeland of ST, then other ST clades would have known how to say “millet”.
What are some stable and reconstructible ST etymon then? They are:
#Bear.
In Old Chinese 熊 *gwem had reliable correspondences in Bodish dom, Burmo-Gyarongic təwam, Tani tum, etc. Another Old Chinese term 羆pje←*praj has correspondence with Gyarongic pri. The distribution of bears in East Asia (ursus tibetanus) precludes eastern China as the homeland of ST and points instead to Eastern Himalaya (including west Basuria, southern Gansu, Qinghai, and upper YR). The existence of another bear species, the Giant Panda, in western Basuria further corroborates the possibiltiy that Eastern Himalaya is the Urheimat of ST. Although earlier bear distribution before human devastation of the environment may extend further east in East Asia, bears’ preference for mountainous habitat and hibernation tendency makes Eastern Himalaya a more favored ecological niche. Because Giant Panda as a subspecies of bear has lost most of its habitat since entering the Holocene, other bears faced similar issue and were increasingly confined to the Tibetan Plateau, upper YR, and western Basuria.
#Monkey.
STEDT has a proto-form as m(y/r)uk, corresponding to Old Chinese mə-gˤ(r)o, which then survive as Mandarin 猕猴 (mihou) and Cantonese 马骝 (marou). This root word is widely found in other ST languages, such as Lolo-Burmese myuk, Jirel Tibetan moro, and yok in various Himalaya languages.
An interesting evolutionary biology study (Li Baoguo et al 2020) showed that Asian macaques entered East Asia from Yunnan and southeastern Himalaya (also pointing to western Basuria) before reaching into China proper. The early humanoids should have started to see monkeys first in Basuria and Eastern Himalaya. Having a common cognate shared across so many ST languages for monkey strongly suggests that the urheimat of ST is also within the same areal range. In addition, the lack of non-human primate fossils in northern China cautions against locating ST urheimat in northern China at all.
#Birch and Pine.
Old Chinese 桦 *ɢʷˁra corresponded to Tibetan gro.ga. For pine, Tibetan tʰaŋ, Gyarongic tutho, Tangkhulic mə.tɐŋ are related to Sinitic 蒸tɕiŋ←*təŋ.
Both birch and pine trees grow and thrive in cold climate. Its temperature tolerance is an upper limit of 30 Celsius. Today, its famous habitats in East Asia are Manchuria, Mongolia, Loess Plateau and southwestern China. Central Plain is simply not where it grows. In addition, the temperature in the Neolithic era is 3–5 Celsius higher than today, leaving the loconym 豫 to the Central Plain of China due to its elephant produce. The summer temperature of modern Central Plain often reaches 38 Celsius, making its neolithic upper temperature, which would have ranged from 40 to 45, a killer of any healthy birch growth. Alternatively, Eastern Himalaya and its slope ranges extending from Qinghai to Basuria would be the ideal habitat for birch forests. Heat also makes pine tree extremely susceptible to diseases, delisting lower YR and Central Plain from the ideal place for pine forests and making them unlikely a homeland to languages that share the etyma of birch and pine. The ecological needs of birch and pine do not, however, exclude Loess Plateau in northern China as a urheimat candidate.
#Bamboo.
Proto-ST (g-)pwɑ is constructed, from which Old Chinese 匪 pəjʔ derives. Bamboo is also a ST cognate according to Matisoff, only that he does not see it as pervasive enough.
While temperate flora like birch and pine do not prefer areas with high summer temperatures, subtropical flora such as bamboo also preclude further north in East Asia from the best Urheimat candidates. Thus, certain Yangshao sites such as Shimao, and Liao River Hongshan cultures would neither be the homeland of ST. This still does not exclude the southern Loess Plateau from the competition.
#Fog.
mway is both fog and cloud in Tibet-Burmese. For proto-ST, the cognate for fog is mruk, which is the Old Chinese kə.mok and mi̯ug (雾, or 霾). Corresponding pervasively with Baic mɯ²¹ko, Nuosu mu³³ hɔ⁵⁵, Gyalrong zdo mə, Tani doŋ muk, Tibetan rmuk.
Ding and Dong (2023) have found an interesting phenomenon that ST people living on higher altitudes tend to not distinguish cloud from fog in their language, due to the topologically identical position of both meteorological phenomena. The proto mway refers to both cloud and fog interchangeably. The existence of a co-lexified word cloud/fog in ST suggests locating its urheimat at somewhere with higher altitude. Again, the Eastern Himalaya appears to be a more likely place over northern China because the latter either lacked the high altitude (Central Plain) or high humidity (Loess Plateau) required to develop cloud/fog.
Although tracing the initial natural environment of stable cognate is not definitive in locating the homeland of the ST language, the evidence, when combined, points toward a more plausible area and helps us exclude those unlikely. When all ST languages share the common words of bear, bamboo, birch, pine, monkey, fog/cloud, and the early speakers have lived with or through these natural beings, isn’t it more likely than any other hypothesis that the homeland of ST is just Eastern Himalaya/Western Basuria, whose staple produce include bear, panda, monkey, bamboo, and birch and pines on mountain slopes, and is also cloudy and foggy all year around?
