Prehistory & Neolithic

Through the Ages

1. The Stone Record

What Was Dug, and Where

What survives of the valley’s prehistory is stone and fired clay, recovered from a handful of sites. Where it survives is itself a fact about the Brahmaputra. The river rebuilds its bed every season and reworks whatever is left on the plain, so almost every early find comes from the hills at the valley’s edge rather than from its floor.

Before the polished stone

The oldest tools come from the hills. In the Rongram valley of the Garo hills, in present Meghalaya, surveys from the 1970s onward recovered handaxes and cleavers of an Abbevillio-Acheulean cast, the earliest stone tools known from the northeast, and later a microlithic phase with small worked tools and a crude handmade pottery. The Garo hills have yielded more Palaeolithic-looking tools than anywhere else in the region. Almost none of them come out of a deposit with a secure sequence, so an attribution as bold as the Middle Pleistocene rests on the shape of the tools rather than on a dated layer.

The Neolithic signature

Two things define the age here. The first is the shouldered celt, an axe or adze with a cut-back haft. The second is cord-marked and basket-impressed pottery. Cord-marking is rare in the Neolithic of the Indian mainland and dominant across China and Southeast Asia, which places the valley’s Neolithic in an eastern world rather than a South Asian one. The valley is walled but not sealed. The Patkai, the Brahmaputra gorges and the southern hills are narrow gates onto Burma, Tibet and the Bay of Bengal, and the material came through them.

Daojali Hading, in the North Cachar hills, was dug in the early 1960s by M.C. Goswami and T.C. Sharma of Gauhati University, the first stratified Neolithic deposit excavated in the northeast. It yielded shouldered celts, four kinds of pottery, the grinders and querns of an everyday kitchen, tools of hard local fossil wood, and a piece of jadeite with no source in these hills, generally read as carried in from the direction of China. Sarutaru, a hillock in the Kamrup country about twenty-five kilometres southeast of Guwahati , was dug between 1967 and 1973 and gave nine ground celts of slate and sandstone with cord and basket-impressed pottery close to the Daojali ware. Selbalgiri in the Garo hills is a third. Hills and valley floor yield the same material, and the pots and grinders point to grain and to settled work.

2. Dates and Gaps

An Uncertain Calendar

Typology put Daojali Hading’s deposit in the fourth or fifth millennium BCE. A luminescence date taken from the deposit itself returned about 2,700 years, near the start of the first millennium BCE, and some workers date the site later still, into the early common era. Thousands of years separate the readings, so any firm span for the Assam Neolithic is provisional.

Between the Neolithic and the first kingdoms the record thins to almost nothing. Copper, bronze and iron horizons, well attested in Bengal to the west and in Southeast Asia to the east, are barely traced in the northeast at all, and whether those sites are absent or merely undug is unsettled. The one excavated thing in the gap is Ambari, in the heart of Guwahati , found by accident in 1969 during digging for a bank’s foundations, whose earliest levels of the Sunga-Kushan horizon reach back to around the second century BCE and were the first material of that age ever recovered in the northeast.

Excavated brick foundations at the Ambari site: low walls of laid brick, a stepped tank structure to the right, cleared earth between them, part of the dig under a blue tarpaulin — Plate 1.The Ambari dig, Guwahati. The Ambari mound under excavation in the heart of Guwahati, the brick foundations of the buried city exposed. The mound's earliest levels reach back to around the second century BCE.Photograph: ComparingQuantities · CC BY-SA · Wikimedia Commons

3. The Peoples Who Came Through

Migration, Speech and What Survives

The stone record is anonymous. No tool or pot carries a name, and nothing in the archaeology ties an assemblage to any community now living. What can be traced is speech.

The streams of speech

The oldest identifiable layer is Austroasiatic, a family generally taken to be the region’s earliest, running from eastern India through mainland Southeast Asia. Its only speakers in the northeast are the Khasi and Pnar of the southern hills, of its Khasic branch, living as a linguistic island amid Tibeto-Burman neighbours and cut off by hundreds of miles from the nearest related tongue. Tibeto-Burman is the largest family, and its Bodo-Garo branch is the valley’s oldest broad layer. The Bodo, Dimasa, Rabha, Tiwa, Deori, Sonowal Kachari, Mech and Koch-Rajbongshi hold the plains, and the Garo the western hills. The Karbi of the central hills belong to another branch, and the Mising, of the Tani group, came down from the northern hills much later.

What the living world keeps

None of this can be dated to the Neolithic, and the link is one of descent and practice rather than of stratigraphy. What can be said is narrower. The shifting hill cultivation the Assamese call jhum still works the ground as the Neolithic did, and the Wangala of the Garo hills is its harvest festival. The Baisagu of the Bodo and the Ali-aye-ligang of the Mising turn on sowing, and at Jonbeel Mela the Tiwa and their hill neighbours still barter goods without money, the surviving form of the hill-and-plain exchange that carried jadeite to a hillside in North Cachar. These were the peoples holding the valley when writing arrived, and the Bodo-Kachari stock was still its largest layer when the first kings cut their names into copperplate. The age of writing begins with Kamarupa , in the middle of the first millennium CE.