Hoabinhian

The Hoabinhian is a lithic techno-complex: stone-tool assemblages found across Southeast Asia from the late Pleistocene to the Holocene, dated to c. 10,000–2000 BCE and attributed to hunter-gatherer societies whose changes in technology over time are poorly understood. French archaeologists working in northern Vietnam first used the term Hòa Bình culture for Holocene assemblages from rock shelters; the English “Hoabinhian” came to mean assemblages of mainland Southeast Asia with flaked cobble tools. It once meant a particular ethnic group of a limited period; recent work, such as Shoocongdej’s, uses it for artefacts of certain forms.

Its age has been pushed back. Colani and others assumed it was Holocene because the animals in its layers were living species, and Chester Gorman’s radiocarbon dates confirmed a Holocene age. At a conference in Hanoi in 1994 Vietnamese archaeologists presented Hoabinhian artefacts 17,000 years old, and in 2016 a rock shelter at Xiaodong in Yunnan, China, produced Hoabinhian artefacts dating from 41,500 BCE. Below Hoabinhian layers at several sites Hà Văn Tấn found an older flake technology he named the Nguomian, after the Ngườm rock shelter in Thái Nguyên province; the Hoabinhian is also claimed to continue the Sonvian technology.

Ancient DNA from Pha Faen in Laos (7888 ± 40 BP) and Gua Cha in Malaysia (4319 ± 64 BP) shows a lineage related to the main East Asian lineage but deeply diverged from it, of unknown Upper Paleolithic origin.

Hoabinhian people lived as hunter-gatherers, oriented towards rock shelters near freshwater streams in upland limestone country, though shell middens show at least one other way of life. Their food remains are much alike from site to site: living species of shellfish, fish, and small to medium-sized mammals. They drew food from three settings: limestone mountains gave land snails and small mammals; streams, small rivers, swamps and lakes gave snails and fish; and the valley floors gave nuts, fruits, fungi, vegetables, wild cereals and wild mammals.

At Da But in Vietnam, occupied from about the fifth or sixth millennium BCE to the end of the third, Viet (2004) found that people ate mostly mountain shellfish, nuts and fruit, including an as-yet-unnamed freshwater clam of the genus Corbicula, which lives in swamps and lakes. Differences between the tools of different regions largely follow local resources and changing environments.

Stone tools are also read for how people moved. Studies by Marwick and Shoocongdej at the Tham Lod and Ban Rai rock shelters in Thailand applied behavioural ecology models: at Tham Lod much pre-processing of stone went with residential mobility and a low intensity of occupation, a contradiction that a “multiple optima” model, allowing more than one best strategy, has been proposed to explain.

The Hoabinhian is known almost entirely from stone tools and food remains, and nothing in that record states what its makers believed. No writing survives, and the two people whose remains have yielded ancient DNA, at Pha Faen in Laos and Gua Cha in Malaysia, are known from their genomes rather than from any described rite.

Some finds hint at uses beyond food. Grinding stones made on rounded cobbles often occur with iron oxide, though what was ground on them is not recorded. At Spirit Cave Chester Gorman suggested that plants such as areca (betel), pepper, canarium and candlenut were used as foods, condiments, stimulants and for lighting, and stressed that the remains show an early, sophisticated use of species still culturally important in Southeast Asia.

The Hoabinhian is defined by its stone tools. In 1932 the First Congress of Prehistorians of the Far East described it as a culture of tools flaked with varied workmanship, often worked on one face only, with hammerstones, implements of sub-triangular section, discs, short axes, almond-shaped pieces and an appreciable number of bone tools. Chester Gorman’s 1970 definition stressed a mainly one-sided flaking tradition on water-rounded pebbles and the large flakes struck from them; sumatraliths, pebbles flaked all over one side; grinding stones; many flakes used without further shaping; and edge-ground tools and cord-marked pottery in the upper layers.

Madeleine Colani’s typology was elaborate: she sorted 82 artefacts from Sao Dong into 28 types. In 1964 J. M. Matthews measured unifacial cobble tools from Sai Yok rock shelter in west-central Thailand and found that Colani’s types did not exist: the tools form a continuous range of shapes and sizes. The Bacsonian, often treated as a variant, has more edge-ground cobble tools (c. 8000–4000 BCE). Marwick’s experimental assemblage of 2008 showed the value of recording where cortex lies on a flake’s back to measure how intensively stone was worked. Spirit Cave’s Hoabinhian levels also produced cord-marked pottery, though not until after 6800 BCE.

Hoabinhian sites are found across mainland and island Southeast Asia. More than 120 are known in Vietnam, a concentration that reflects intensive research there rather than a centre of Hoabinhian activity. Sites in Terengganu, Sumatra, Thailand, Laos, Myanmar and Cambodia have also been called Hoabinhian, though the quality of their descriptions varies, and some archaeologists see isolated Hoabinhian elements in Nepal, South China, Taiwan and Australia.

Whether Hoabinhian people farmed has been argued over. Gorman reported almond, Terminalia, betel, bean or pea, water caltrop, pepper, candlenut, a cucumber and other plants from Spirit Cave layers of c. 9800–8500 BP. None differed from wild forms, and he wrote that whether they were early cultigens remained to be established. In 1972 W. G. Solheim wrote in Scientific American of “an advanced knowledge of horticulture” there and put incipient agriculture in Southeast Asia at about 20,000 BCE; though he called his reconstruction “largely hypothetical”, his overstatement led to inflated claims of Hoabinhian agriculture.

Farming came with movements of people between southern China and Southeast Asia, with which the Neolithic there is associated. Neolithic samples from the mainland carry mostly East Asian-related ancestry like that of ancient southern China, with a smaller share of Hoabinhian-related ancestry.

Hoabinhian ancestry, described as a “deeply branching East Asian ancestry”, is taken to represent the native pre-Neolithic hunter-gatherers of much of Southeast Asia. It shows high genetic affinity with the Andamanese Onge and Jarawa, the Semang (also called “Malaysian Negritos”) and the Maniq, as well as with the Upper Paleolithic Tianyuan man of northern China and the Jōmon people of Japan. Among living peoples the Maniq have one of the highest shares of Hoabinhian-related ancestry, alongside about 35% East Asian-related ancestry. In modern populations a mix of East Asian and deeply diverged Hoabinhian-related ancestry is most closely tied to Austroasiatic-speaking groups, and a 2026 study found that modern mainland Southeast Asians carry less of it than ancient samples, perhaps through later contact with South or East Asian-related populations.

The name itself has been debated. In 1994 archaeologists meeting in Hanoi voted to keep it, to call the Hoabinhian an industry rather than a culture or techno-complex, to keep the term “sumatralith”, to speak of cobble rather than pebble tools, and not to call it “Mesolithic”. Spirit Cave, despite the inflated claims made for it, remains a site with well-preserved evidence of how people lived and of the environment of their time.