Prehistoric Settlement and Subsistence

Prehistory in the Mithila–Vajji–Anga study region cannot be written as the early chapter of a community whose later name is already known. The archaeological record precedes the historical identities, states, literary languages and social classifications that dominate later periods. Accordingly, this chapter asks a narrower but more defensible set of questions: where are prehistoric occupations actually documented; what kinds of settlement and subsistence can be reconstructed from stratified evidence; what remains uncertain because alluvial processes have buried, eroded or displaced sites; and which wider Middle Ganga or Himalayan comparisons illuminate the problem without becoming substitutes for local evidence. The principal direct anchors used here are the Patu/Rato Khola lithic complex in eastern Nepal, Chechar–Kutubpur in present Vaishali district, and Chirand in north Bihar. Wider sites such as Lahuradewa and Senuwar are used comparatively and are labelled as such. 2.1 Prehistory without retrospective identities Terms such as Mithila, Videha, Vajji and Anga belong to historical geographies whose documentary visibility is much later than the Stone Age occupations considered here. A lithic scatter in the Siwalik foothills is not evidence of an early ‘Maithil’ population, and a Neolithic settlement in modern Vaishali district is not automatically a settlement of the later Vajjian republic. Modern political borders are equally anachronistic as cultural boundaries for the early Holocene. The historian may use present district and national names to locate evidence, but those labels describe where the site is now, not who prehistoric inhabitants believed themselves to be. This distinction is more than terminological caution. It changes the kinds of conclusions that can responsibly be drawn. Archaeology can reconstruct technologies, diets, settlement forms, seasonality, environmental choices and sometimes exchange or mobility. It is usually much less secure for language, ethnicity, caste, political identity or named religious affiliation in periods without writing. Those later categories will enter this volume only when the evidence capable of supporting them appears. 2.2 The evidence problem of an alluvial country The north-Bihar plain is a difficult archive. The rivers described in Chapter 1 have repeatedly migrated, aggraded, incised and reworked their floodplains. A prehistoric occupation surface may be sealed beneath metres of younger sediment, truncated by a later channel, exposed only briefly on an eroding bank, or HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II destroyed by agriculture and construction. Consequently, low site counts in an actively alluviating zone cannot be translated directly into low prehistoric population. Archaeological visibility is partly a geomorphological outcome. The contrast with foothill terraces is instructive. In the Siwaliks, artefacts may survive on older terrace surfaces or within soils that can be related to Quaternary landforms; on a young floodplain, equivalent occupations may be deeply buried or displaced. This asymmetry helps explain why the prehistoric record of Nepal’s Siwalik margins and the excavated river-bank mounds of Bihar looks discontinuous even though human use of the wider landscape need not have been discontinuous. 2.3 An evidence hierarchy for Chapter 2 The strongest claims in this chapter rest on excavated or closely contextualized material: stratigraphic position, structures, radiometric samples, lithic assemblages, pottery, bone and antler objects, botanical remains and faunal remains. A second level consists of inferences repeated across more than one secure context—for example, the association of wattle-and-daub architecture with settled or recurrent occupation. A third level uses wider regional comparisons, such as the archaeobotany of the Middle Ganga Plain, to frame questions that local excavation has not yet answered. The chapter does not use later literary traditions as evidence for prehistoric settlement. This hierarchy is particularly important for chronology. The category ‘Neolithic’ is not a single date shared across South Asia. Local sequences differ; published radiocarbon determinations may be internally inconsistent; old excavations did not always use present sampling protocols; and technological traits can persist across chronological boundaries. Dates are therefore presented as ranges or as the published assessment of a particular context, not as a universal timetable. GAJENDRA THAKUR Figure 4 — Evidence control for prehistoric reconstruction: direct context, supported inference, comparison, and rejected retrojection (colour schematic). 2.4 Eastern Nepal: the Patu / Rato Khola prehistoric record The most important prehistoric evidence close to the Nepal side of the later Mithila region comes from the Siwalik foothills of Mahottari district. Gudrun Corvinus’s surveys at Patu and along the Rato Khola established a large cobble-tool industry associated with older river terraces. The Patu 2 locality was described as a particularly dense concentration or activity area. The assemblage includes abundant flakes, choppers, core-scrapers and adze- or axe-like tools made largely from locally available cobbles. Corvinus’s work is foundational because it connects artefacts to terrace geomorphology rather than treating isolated stone tools as self-dating objects (Corvinus 1987, 1989, 2007). Later synthesis has emphasized that Patu is best understood in a Late Pleistocene–Early Holocene sub- Himalayan context rather than forced into an old, uniform ‘Lower Palaeolithic chopper’ category. Gaillard, Singh and Dambricourt-Malassé describe the Patu assemblage as a macrolithic industry with strong flake representation, unifacial and bifacial adze-like pieces, choppers and core-scrapers; a charcoal sample from the artefact-bearing horizon yielded an age around 7 ka BP and therefore a minimum age for the industry. The chronology is early Holocene, but the industry is not evidence for farming. It is evidence for human occupation and technological adaptation in a forested, stream-rich foothill environment (Gaillard, Singh and Dambricourt-Malassé 2011). This matters to the history of the Nepal–India borderland in a negative as well as a positive sense. It demonstrates prehistoric human use of landscapes north of the present international boundary, while offering no basis for assigning those people to later Videha, Maithili, Madhesi or other identities. The archaeological continuity to be investigated is therefore occupation of an ecological corridor, not continuity of a named community. 2.5 Palaeolithic and Mesolithic labels: why terminology must remain provisional Older South Asian prehistory often classified stone-tool industries primarily by typology. In the Siwaliks this produced long debates around ‘Soanian’, Acheulean, chopper-chopping and later macrolithic assemblages. Modern work gives greater weight to geological surface, stratigraphy, raw-material economy, reduction sequence and absolute dating. Patu is a good example of why the change matters: cobble tools alone do not make an assemblage ancient, and similar tool forms can recur when local river cobbles encourage comparable technical solutions. For this volume, therefore, the labels Palaeolithic, Mesolithic and Neolithic are used as archaeological conveniences tied to the cited literature, not as rigid social stages. ‘Mesolithic’ at Patu describes a scholarly placement of a lithic tradition; it does not imply that the inhabitants followed the same subsistence regime as every other Mesolithic community in South Asia. Likewise, ‘Neolithic’ at Chirand and Chechar–Kutubpur refers to archaeological packages in which settled life, food production and characteristic material technologies become visible; it does not mean that foraging suddenly ceased. 2.6 North Bihar’s strongest Neolithic anchors Within Bihar, the excavated Neolithic record most relevant to the northern alluvial plain is concentrated at Chirand in Saran and Chechar–Kutubpur in Vaishali. They are not interchangeable. Chirand has a long excavation history, a thick Neolithic deposit and unusually rich bone and antler industries together with plant and animal remains. Chechar–Kutubpur is geographically closer to the later Vaishali–Vajji core and HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II therefore crucial for this book, but the excavated exposure was smaller and the published evidence is correspondingly more limited. The strength of the chapter lies in keeping those evidentiary differences visible. Other Bihar sites such as Senuwar and Taradih are indispensable for understanding the broader Neolithic of Bihar, but they lie outside the principal north-bank Mithila–Vajji focus. They are used here to test regional models, not to fill local gaps. In the Anga/Bhagalpur zone, excavated sequences at Oriup and Champa become especially informative from Chalcolithic and later horizons; those belong to Chapter 3 and are not pulled backward to manufacture a Neolithic Anga story. Figure 5 — Selected prehistoric anchors used in Chapter 2: direct study-zone evidence and explicitly labelled regional comparators (colour schematic; not to scale). 2.7 Chirand: a durable settlement on a riverine margin Chirand lies on a major riverine corridor in north Bihar, near the Ganga–Ghaghara system. Excavation revealed a substantial Neolithic occupation underlying later cultural periods. B. S. Verma’s early reports emphasized the thickness of the Neolithic stratum and the variety of bone, antler, stone and ceramic materials. The site therefore provides something rare in the alluvial plain: not an isolated implement but a repeated occupation sequence in which architecture, tools and subsistence remains can be read together (Verma 1971; Verma 2007). The settlement evidence includes post-holes, reed or bamboo impressions in burnt clay, rammed floors and circular or semicircular wattle-and-daub structures. These are lightweight technologies, but lightweight does not mean temporary. Mud, reed and bamboo are rational building materials in an alluvial environment where timber and grasses are available and heavy masonry is unnecessary for domestic structures. Rebuilding may have been frequent; the archaeological signature of repeated rebuilding is part of settled life rather than evidence against it. GAJENDRA THAKUR 2.8 Chirand’s bone, antler and stone technologies One of Chirand’s most distinctive archaeological contributions is the scale and variety of its osseous industry. Excavators reported celts, scrapers, chisels, hammers, needles, points, borers, awls, diggers, pins, arrowheads and ornaments made from bone, antler, ivory and shell. The significance lies not simply in the list of objects. Manufacture required selection of suitable skeletal elements, cutting, scraping, grinding, perforation and finishing. Unfinished pieces indicate that at least some production took place within the settlement rather than all objects arriving ready-made (Verma 1971). Stone technology remained important alongside bone. Ground and polished implements, pestles, querns and microlithic components indicate several chains of work: clearing or woodworking, processing plant foods, hunting and maintaining composite tools. The archaeological lesson is technological plurality. A ‘Neolithic’ community did not abandon small chipped stone simply because polished axes existed, nor did bone tools represent an inferior substitute for stone. Different raw materials were selected for different mechanical properties. 2.9 Plant economy at Chirand: what the seeds actually show Palaeobotanical work at Chirand is unusually important for reconstructing subsistence. Vishnu-Mittre identified food-plant remains from Neolithic levels including rice, barley, lentil and field pea; later syntheses also discuss wheat and other pulses in the Bihar Neolithic record. These finds establish cultivation and consumption of multiple crops rather than a rice-only economy (Vishnu-Mittre 1972; Pokharia et al. 2017). The chronology of the oldest Chirand plant remains has long been difficult because some early radiocarbon determinations were internally inconsistent. That problem should not be hidden by quoting one spectacular early number. Modern regional syntheses commonly place the Bihar Neolithic broadly in the late third to second millennium BCE, with Chirand often discussed around c. 2000–1500 BCE or somewhat wider depending on calibration and phase definition. The safe historical conclusion is the presence of a settled food-producing community in later prehistory, not an artificially precise founding year. The combination of rice with winter cereals and pulses is socially more informative than a single crop label. It implies knowledge of different growing seasons and a household calendar that could distribute labour across the monsoon and cool seasons. Whether every listed crop was produced locally in every phase cannot be assumed; nevertheless, the assemblage supports a diversified plant economy. 2.10 Animals, hunting and aquatic resources at Chirand The Neolithic economy was not a replacement of wild foods by domestic foods. Archaeozoological summaries from Chirand include domestic cattle, buffalo, sheep, goat and pig together with hunted deer, wild boar and other wild fauna. Fish, tortoise, molluscs and other aquatic resources fit the site’s riverine setting. This is better described as a mixed agro-pastoral-foraging economy than as an agricultural economy with hunting as a marginal survival (Sinha 1994; eHRAF Ganges Neolithic synthesis). The presence of wild fauna cannot by itself tell us the social organization of hunting, and cut-marked bones do not reveal whether particular species carried ritual meanings. Those questions require evidence not presently available. What can be said is that cultivation, animal keeping, fishing, hunting and gathering were compatible activities within one settlement system. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 2.11 Chechar–Kutubpur: direct prehistoric evidence in modern Vaishali district Chechar–Kutubpur is central to Volume II because it provides excavated prehistoric evidence inside present Vaishali district, close to the later landscape of Vaishali and Vajji. The Archaeological Survey of India’s 1977–78 excavation report describes Period I as a pre-metal occupation over natural soil. Post-holes and burnt clay with reed impressions indicate wattle-and-daub houses plastered with husk-tempered earth; the available evidence suggested that at least some structures were circular. Most importantly for subsistence, the report characterized the economy as developed and rural, with agriculture accompanied by hunting and angling (Archaeological Survey of India 1977–78: 17–18). The report’s wording must be used carefully. ‘Developed rural economy’ is an excavator’s interpretation, not a census of occupations. The direct evidence is architectural traces, tools, ceramics and subsistence indicators. The absence of metal in Period I helps distinguish the horizon technologically, but it does not by itself supply an absolute date. Nor does the site prove that the later city of Vaishali grew directly from this settlement without interruption. What it does prove is that food-producing communities occupied the Ganga-side alluvial zone of modern Vaishali before the urban and republican histories treated in later chapters. 2.12 Architecture and settlement permanence at Chechar–Kutubpur Wattle-and-daub architecture is often archaeologically modest because walls decay and only post-holes, floors and burnt clay survive. At Chechar–Kutubpur, however, the combination of post-hole patterning, reed impressions and plastering material is enough to move interpretation beyond a transient camp. It indicates constructed domestic space and investment in maintaining a dwelling. The reported circular plan also places the site within a wider range of Neolithic house forms known from the Middle Ganga Plain. Permanence should not be understood as immobility. Households could occupy a village for many years while members moved seasonally for fishing, grazing, hunting, fuel, raw materials or marriage exchange. Archaeology of a settlement mound records where durable activities accumulated, not the full annual movement of its inhabitants. 2.13 A subsistence mosaic, not a staircase of progress The older language of ‘food gathering’ followed by ‘food production’ can imply a clean evolutionary threshold. The Bihar evidence is more usefully understood as a mosaic. Cultivated crops, domestic animals, wild game, fish and gathered plants coexisted. Different resources reduced different risks: cereal cultivation could generate storable calories; herds converted vegetation into meat and milk; fishing exploited predictable aquatic niches; hunting and gathering widened the seasonal food base. This mixed system is exactly what should be expected in floodplain settings with strong seasonal contrasts. Households did not need to choose once and for all between farming and foraging. Their economic resilience may have depended on retaining several strategies. The proportion of each strategy almost certainly varied by site, year and season, and the surviving archaeological record does not permit a single percentage allocation. GAJENDRA THAKUR Figure 6 — Prehistoric subsistence as a seasonal mosaic rather than a one-way transition from foraging to farming (colour analytical diagram). 2.14 Rice and the wider Middle Ganga problem Rice occupies a special place in debates over the Neolithic of the Middle Ganga Plain, but the history of rice should not be made simpler than the archaeobotanical evidence allows. Sites such as Lahuradewa in eastern Uttar Pradesh have produced very early rice-related evidence and have transformed discussion of cultivation in the region. At the same time, scholars have debated whether the earliest rice remains represent fully domesticated rice, managed wild stands, cultivation before full domestication, or later intrusive material. Fuller and later archaeobotanical syntheses emphasize that domestication is a biological and cultural process whose chronology cannot be established from the mere presence of rice alone (Fuller 2006; Pokharia et al. 2017). For Mithila–Vajji–Anga history, Lahuradewa is therefore comparative evidence, not local proof. It demonstrates that early cultivation and rice management in the wider Middle Ganga world may have deep roots. It does not prove that the same practices began at the same date in Vaishali, Saran, Bhagalpur or the Nepal Tarai. Local chronology must come from local contexts. 2.15 Crop diversity and seasonal scheduling The combination of rice, barley, wheat and pulses documented in the wider Neolithic–early historic Ganga archaeobotanical record points toward increasingly complex seasonal scheduling. Rice generally belongs to the monsoon cycle, while barley and wheat are winter crops in the modern region. Pulses can occupy several niches in rotation or mixed cultivation. Even without assuming modern agronomy, the archaeological presence of crops with different ecological requirements indicates accumulated practical knowledge of soils, rainfall and season. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II Such scheduling has social implications. Sowing, guarding, harvesting, threshing, processing and storage concentrate labour at particular times. Households must coordinate labour and cope with sickness, flood or crop failure. Yet archaeology rarely reveals the rules by which labour was allocated. It would be unjustified to assign particular tasks to men, women or named social groups simply because later ethnographies do so. 2.16 Domesticated animals and the limits of identification Cattle and other domestic animals broadened the economic base of Neolithic communities. They supplied meat and potentially milk, dung, traction or transport, although each use requires independent evidence. A cattle bone identified as domestic does not prove plough agriculture. Likewise, sheep and goats indicate managed herds but do not reveal ownership arrangements. The responsible reconstruction separates species identification from the social uses that may or may not be archaeologically visible. Faunal assemblages are also filtered by preservation and recovery technique. Small fish bones, bird bones and molluscs are easily undercounted in excavations that did not use fine sieving or flotation. Older excavation reports can therefore make terrestrial mammals appear disproportionately important. This is another reason to treat subsistence percentages with caution. 2.17 Wetlands, fish and floodplain knowledge The riverine setting of Chirand and the reported angling at Chechar–Kutubpur underline a form of expertise that is easy to obscure in narratives centred on crops. Fishing requires knowledge of channel seasonality, spawning, shallow-water habitats and appropriate gear. Tortoise, molluscs and wetland plants extend the resource zone beyond the cultivated field. Oxbows, abandoned channels and marsh edges were productive environments even when they were inconvenient for permanent architecture. This ecological knowledge also links Chapter 2 back to Chapter 1. Floods were not only disasters. Seasonal inundation renewed soils, created fisheries and rearranged grazing and gathering zones. The same process could also destroy houses or bury fields. Prehistoric settlement in an alluvial plain was therefore adaptation to variability, not conquest of a stable landscape. 2.18 Archaeological visibility: burial, erosion and false absences The distribution map of known prehistoric sites is not a transparent map of where prehistoric people lived. A mound on a stable levee can accumulate visible deposits, while a low flood-basin camp can disappear beneath younger alluvium. A river cut may expose artefacts far from their original depositional context, and modern embankments, brick cutting, road building and agriculture can remove the upper part of a site. Survey intensity is uneven as well: some districts have received repeated archaeological attention, others very little. For this reason, the absence of a reported Neolithic mound in a particular part of Mithila or Anga is not evidence that the area was empty. It is evidence that secure sites have not yet been demonstrated there. That distinction will be maintained throughout Volume II: archaeological silence is a research condition, not a licence to invent continuity. GAJENDRA THAKUR Figure 7 — Archaeological visibility on an active alluvial plain: burial, exposure and reworking can distort the apparent distribution of prehistoric settlement (colour schematic). 2.19 Mobility within sedentary life Sedentism is best treated as a spectrum. Circular huts, floors and repeated occupation indicate residential investment, but a village-based household can still be highly mobile in its use of the landscape. Fishing trips, herd movements, hunting, raw-material procurement and visits to related communities create networks extending beyond the mound. The settlement is a node in those movements, not their boundary. This perspective helps reconcile two kinds of evidence that are sometimes made artificially contradictory: durable domestic architecture and continued exploitation of wild resources. A household need not be a permanently enclosed farming unit to be sedentary in archaeological terms. 2.20 Material circulation without invented trade routes Beads, stone raw materials, shell, ivory and other objects at Neolithic sites invite questions about circulation. Some materials may have been obtained locally; others may have moved through exchange or seasonal travel. Without geochemical sourcing, however, it is unsafe to draw precise trade routes. The presence of a bead made from a visually distinctive stone is not enough to identify a quarry, a merchant class or a long-distance market. The correct historical inference is more modest: prehistoric households were connected to material landscapes wider than the immediate dwelling area, and some objects probably travelled between communities. Chapter 12 addresses trade only when stronger evidence for organized exchange becomes available. 2.21 What cannot yet be reconstructed: language, caste and political form Nothing in the Patu, Chirand or Chechar–Kutubpur evidence permits identification of a spoken language. Archaeology does not allow these communities to be labelled Maithili-, Angika- or Bajjika- HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II speaking. Nor does it demonstrate the caste categories documented in much later periods. Burial practice, house size and artefact variation can sometimes indicate social differentiation, but the present evidence is insufficient to reconstruct a named hierarchy. The same applies to political organization. A cluster of houses does not prove a chiefdom; the absence of a palace does not prove egalitarianism. The social scale of Neolithic communities can be explored through settlement size, storage, production and differential access to resources, but the present north-Bihar data are not rich enough to support a detailed constitutional model. 2.22 The Anga zone: a deliberate boundary of this chapter For the eastern Bihar zone later associated with Anga, the archaeological sequence becomes substantially clearer in Chalcolithic and Black-and-Red-Ware horizons at sites such as Oriup and Champa. Excavations at Oriup, near Antichak in Bhagalpur district, identified a Chalcolithic Period I beneath later NBPW, Pala and medieval occupations. Champa likewise produced lower black-and-red-ware horizons below its early historic development. These are important, but they belong chronologically and analytically to Chapter 3. This deliberate boundary prevents a common form of narrative inflation. Because Anga is historically prominent in the first millennium BCE, there is a temptation to supply it with an equally detailed Neolithic prehistory whether or not excavated evidence exists. Volume II will instead allow different regions to have different levels of prehistoric visibility. 2.23 Nepal–India comparison without a single origin story Patu and the north-Bihar Neolithic sites belong to different archaeological traditions and ecological settings. Patu represents an early-Holocene lithic adaptation in the Siwalik foothills; Chirand and Chechar– Kutubpur represent later food-producing settlement in the Gangetic plain. Their inclusion in one chapter does not imply descent of one from the other. It demonstrates that the later trans-border region contained several prehistoric trajectories. This plurality is historically valuable. The later cultural connectedness of Mithila across India and Nepal need not be justified by inventing a single prehistoric people. Long-term connectivity can emerge repeatedly through routes, marriage, markets, pilgrimage, language and political institutions in much later periods. Prehistory supplies a deep human landscape, not a charter of modern identity. 2.24 Conclusions: what Chapter 2 establishes Four conclusions are secure enough to carry forward. First, prehistoric occupation is documented both in the eastern Nepal foothill zone and in the north-Bihar alluvial plain, but the archaeological traditions are not identical and should not be collapsed. Second, Neolithic communities at Chirand and Chechar–Kutubpur combined built domestic space with mixed subsistence: cultivation and animal keeping existed alongside hunting, fishing and gathering. Third, alluvial geomorphology strongly filters what survives and what is discovered, so archaeological absence must remain an open research question. Fourth, none of this evidence licenses the projection of later language, caste, political or regional identities into prehistory. Chapter 3 begins where the material record becomes broader: Chalcolithic and early agrarian communities, increasing settlement numbers, copper alongside stone and bone, changing ceramic traditions, and clearer evidence in the eastern Bihar/Anga zone. The transition will be treated as a regional process with local differences, not as a universal stage that arrived everywhere at once. GAJENDRA THAKUR Table 2.1 — Principal archaeological anchors and limits of inference Site / zone Secure signal Subsistence / settlement Main caution Patu / Rato Contextualized cobble- Human occupation and stone- Not evidence for Khola, Mahottari tool industry on older working in a forested farming or later terraces; early-Holocene foothill/stream environment Mithila/Madhesh dating framework identity Chirand, Saran Neolithic settlement Mixed cultivation, animal Early dates are with huts, stone and keeping, hunting, fishing and debated; do not rich bone/antler gathering force one exact industries, plant and founding date animal remains Chechar– Pre-metal Period I; post- Food-producing rural Small excavation; Kutubpur, holes, reed-impressed settlement in present Vaishali no direct line may Vaishali daub, circular huts; district be assumed to later agricultural and Vaishali city or hunting/angling Vajji polity evidence Senuwar / Well-studied regional Help frame crop and Comparative Lahuradewa comparators for chronology questions in the evidence only; not farming and wider Middle Ganga region local proof for archaeobotany Mithila/Vajji/Anga