ARCHAEOLOGY OF FOOD AND EVERYDAY LIFE

LIFE 4.1 Scope: food archaeology is not a prehistoric cookbook This chapter asks what people cultivated, collected, herded, caught, processed, cooked, stored and discarded in the early agrarian landscapes introduced in Chapters 2 and 3. The archaeological record can answer parts of those questions, but it cannot recover a menu in the modern sense. Carbonised grains identify plants that entered fires or deposits; bones and shells identify animals present in the archaeological assemblage; querns, hearths and vessels reveal processing possibilities. None of these, by itself, establishes the proportion of a food in the diet, a named recipe, a ritual rule, or the social identity of the person who prepared it. The method therefore moves from direct residue to bounded inference. This is particularly important for Mithila, Vajji and Anga, because later textual cuisines and modern regional food identities are rich enough to tempt retrospective projection. They belong to later chapters and must not be used to fill prehistoric silence. 4.2 The preservation problem: most food disappears Food is among the least evenly preserved categories of material culture. Leaves, milk, oils, soft fruits, cooked gruels and many roots normally leave little macroscopic trace unless they are charred, mineralised, waterlogged or preserved as microremains. Bone survival varies with soil chemistry and burning; fish bones can be lost through coarse excavation screens; tiny seeds may remain invisible unless flotation or fine sieving is used. The result is a strong preservation bias toward durable shells, bones and accidentally charred plant parts. The absence of a species from an excavation report may therefore mean that it was not present, that it decayed, that the relevant area of the settlement was not excavated, or that recovery methods did not capture it. Archaeobotanical and zooarchaeological evidence are strongest when sampling procedures, stratigraphy and identification criteria are published. Figure 15 formalises this filtering problem and is a standing caution for every quantitative-looking food list in this chapter. Figure 12 — The preservation filter: archaeological food remains are a selective sample. Absence from the recovered assemblage is not automatically absence from diet. GAJENDRA THAKUR 4.3 Chirand: the strongest local conjunction of plant and animal evidence Chirand is unusually valuable because plant remains, animal bones, aquatic fauna, tools, pottery and domestic features can be considered within the same long archaeological sequence. Vishnu-Mittre's study of the Neolithic plant economy reported rice, barley, lentil and field pea from the lower levels, while later summaries also record wheat and pulses such as mung. Nath and Biswas identified a faunal assemblage including humped cattle, buffalo, sheep, pig, dog, deer, elephant, rhinoceros, carp and turtles. Excavation reports add hearths, ovens, bone tools and large ceramic assemblages. Taken together, these materials show a food-producing economy that was neither narrowly cereal-based nor simply a continuation of mobile hunting. Cultivation, animal husbandry, hunting, fishing and the use of wetland resources were concurrent. The evidence is nevertheless an assemblage of different recovery histories; it should not be converted into percentages of diet unless a specific quantified study supports such calculation (Vishnu-Mittre 1972; Nath and Biswas 1980; Verma 2007). 4.4 Rice was central, but 'rice economy' should not erase crop diversity Rice is the most persistent crop in discussions of the Middle Ganga Neolithic because the ecology of the plain favoured wild and cultivated Oryza and because rice has a long later cultural importance. At Chirand, however, the archaeobotanical record also includes barley, lentil and field pea, with wheat and other pulses reported in the wider excavation literature. Senuwar, used here as a comparator rather than as a substitute for evidence from Mithila or Anga, shows even more clearly how rice could coexist with winter cereals, pulses and millets. This diversification matters economically. A household cultivating crops with different sowing and harvesting seasons could spread labour demands and reduce the danger of dependence on a single harvest. It also complicates simple stories in which rice agriculture alone explains sedentism. The archaeological record instead points toward combinations of crops adapted to monsoon, winter and local moisture regimes (ASI 1986-87; ASI 1990-91; Saraswat 2004). 4.5 Cultivation is not the same thing as domestication The distinction between cultivation and biological domestication is essential when discussing early rice. Cultivation refers to human practices such as sowing, tending, water management and harvesting; domestication refers to inherited biological changes in the plant population. Archaeobotanical debates over shattering, rachis morphology and grain form show that these processes need not coincide. Lahuradewa, outside the core study region but within the wider Middle Ganga environmental zone, has been central to debate over very early rice use and cultivation. Later reassessments emphasise that the emergence of fully domesticated rice was protracted. For this volume, Lahuradewa is therefore comparative evidence for the antiquity and complexity of human-rice relationships, not proof that the same chronology applies to every site in Mithila, Vajji or Anga. At Chirand and Senuwar, the safer historical claim is that rice cultivation formed part of established village food economies by their excavated Neolithic phases (Fuller 2011; Tewari et al. 2008). 4.6 Winter crops altered the annual labour calendar Barley, wheat, lentil, field pea and grass pea are not merely additions to a species list. When securely associated with winter cultivation, they imply a second agricultural season alongside monsoon cropping. The Senuwar palaeobotanical reports are particularly informative: Period IA included rice, barley, dwarf wheat, sorghum and ragi millets, lentil, grass pea and field pea, with weeds associated with cultivated fields. The coexistence of monsoon and winter crops implies repeated cycles of land preparation, sowing, weeding, HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II harvesting, threshing, drying and storage across much of the year. Such scheduling can support more stable sedentary households, but it also increases the need for labour coordination and the management of seed reserves. Because Senuwar lies in the Kaimur zone, this evidence is used as a regional model of what diversified farming could involve, while Chirand's local crop list provides the firmer anchor north of the Ganga (ASI 1990-91). 4.7 Field weeds are historical evidence too Archaeobotanical reports sometimes list weeds beside crops, and those minor taxa are easy to overlook. Yet weeds can indicate the disturbed soils and seasonal fields in which crops grew, harvesting height, crop cleaning and the movement of seed contaminants. At Senuwar, taxa such as Vicia, Setaria, Rumex and wild rice occur with food grains. They remind us that a cultivated field was not a modern monoculture but an ecological community. Some weeds may have been tolerated, gathered, grazed by animals or accidentally brought into the settlement with harvested crops. Their presence also cautions against treating every recovered seed as a deliberately cultivated food. Archaeobotanical interpretation therefore depends on assemblage context rather than a binary division between 'crop' and 'wild plant' (ASI 1990-91; Saraswat 2004). 4.8 From harvest to meal: processing created the archaeological signature A cereal does not move directly from field to pot. Harvesting can leave straw and husk; threshing separates grain; winnowing and cleaning remove chaff and weeds; pounding or grinding changes grain size and texture; soaking, parching, boiling or roasting changes the probability that food survives archaeologically. Querns, mullers, pestles and grinding stones therefore belong to the history of food just as much as charred grain. The food economy was partly an economy of repetitive processing labour. The archaeological challenge is that a quern can grind more than one substance and a pestle can serve multiple tasks. Consequently, the presence of grinding equipment supports intensive plant processing but does not identify a specific dish. Figure 12 sets out this evidence chain from residue and tool to appropriately limited historical claim. Figure 13 — From archaeological residue to a defensible history of food. Different evidence classes support different levels of inference. GAJENDRA THAKUR 4.9 Husk in clay: by-product, temper and household waste The earthen oven mixed with husk excavated at Panda in Samastipur is a small but unusually concrete piece of evidence for the proximity of food processing and domestic construction. Husk can be incorporated into clay intentionally as temper, swept into building material as processing waste, or burned as fuel. Each pathway begins with the same underlying fact: cereal by-products circulated through the household rather than disappearing at harvest. Similar husk impressions in burnt clay and pottery at other Middle Ganga sites show how food production entered architecture and craft. Panda therefore provides a direct Mithila-plain anchor for the integration of crop processing, fuel, construction and cooking technology, even though the find does not by itself identify the cereal species used in the oven fabric (Choudhary 2005; ASI 2001-02; ASI 2002-03). 4.10 Cattle were more than meat Nath and Biswas found humped cattle to be the most prominent domestic animal in the Chirand faunal collection, with buffalo, sheep and pig also represented. A food history must resist reducing these animals to carcasses. Cattle and buffalo could contribute traction, manure, mobility, hides and—where dairying occurred—milk, although direct evidence for each use must be demonstrated independently. Bone assemblages can show slaughter and species representation more securely than they can establish secondary products. The historical importance of herding lies partly in its complementarity with cultivation: animals could graze fallow or marginal land, consume crop residues and produce manure, while households retained a movable store of value and food. This mixed agro-pastoral logic is better supported than a specialised pastoral or specialised arable economy for Chirand (Nath and Biswas 1980). 4.11 Sheep, goats and pigs diversified household risk Smaller livestock occupy different ecological and labour niches from cattle. Sheep and goats can exploit browse and grazing that would not support field crops, while pigs can convert household refuse and wetland- edge resources into animal biomass. Their bones at early farming sites therefore indicate a broader strategy of resource use. At Chirand, sheep and pig were well represented in the analysed assemblage; goat is also reported in archaeological syntheses of the site. The precise balance among these taxa remains dependent on identification and sample size, but their presence demonstrates that domestic animal management was not restricted to cattle. Mixed herds and pigs could buffer harvest failure, although this chapter avoids estimating meat availability or herd size without age profiles and quantified faunal data (Nath and Biswas 1980; Verma 2007). 4.12 Hunting remained economically rational Wild deer, rhinoceros and elephant in the Chirand faunal record show that domestication did not eliminate engagement with uncultivated landscapes. Deer are especially important because they supplied meat and antler; antler and long bones could be transformed into points, awls and other implements. The presence of large wild mammals also speaks indirectly to habitats that no longer survive in the same form around the site. Hunting may have protected fields, supplied food, provided raw material or had social meanings that the bones alone cannot recover. What can be said securely is that early farmers did not draw a hard economic boundary between 'agriculture' and 'foraging'. Their subsistence system remained broad- spectrum and ecologically opportunistic (Nath and Biswas 1980). HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 4.13 Fish, turtles and molluscs place wetlands inside the economy Aquatic remains are among the most important correctives to land-centred histories of early farming. Chirand yielded fish, turtles and shell; Oriup in Bhagalpur produced bone fishing hooks, possible net-related terracotta pieces and utilised tortoise shell; broader Ganges Neolithic syntheses also emphasise fish, molluscs and turtles. These finds show that rivers, oxbows, ponds and floodplain depressions were food-producing spaces in their own right. Fishing could be intensified seasonally as water receded and fish became concentrated, while monsoon expansion created different habitats and transport conditions. The evidence does not show that every settlement used the same techniques, but it makes clear that an agrarian village in this landscape was not economically confined to its fields (ASI 1966-67; Nath and Biswas 1980). Figure 14 — A riverine foodscape: fields, herds, wetlands, forest edge and river routes fed into household processing and consumption. 4.14 Oriup: fishing technology in the Anga/Bhagalpur zone Oriup provides the strongest excavated food-related evidence used here for the eastern zone later associated with Anga. In Period I, the excavators recorded an oven, a circular fireplace, numerous bone fishing hooks, bone points, terracotta objects interpreted as possibly associated with fishing nets, and pieces of tortoise shell used as tools. The settlement stood close enough to a river channel for erosion and silt deposition to affect its stratigraphy. Food procurement, cooking and river geomorphology can therefore be read together at the site. This is a more defensible early history of the Anga landscape than projecting the later commercial or literary identity of Champa backwards into the Chalcolithic horizon (ASI 1966-67). 4.15 Wild plants and forest-edge foods are structurally undercounted Wild fruits, leafy plants, tubers, mushrooms, honey and other gathered foods are much less visible than cereal grains, especially in ordinary oxidised alluvial deposits. Archaeobotanical records from the Middle Ganga region nevertheless include wild jujube and other taxa, while the ecology reconstructed from faunal remains implies woodland, grassland and wetland edges accessible to settlements. It is therefore reasonable to include gathering as one component of the food economy, but not to invent a detailed species list for sites that have not produced one. Forest and scrub zones also supplied fuel, timber, fibres and fodder. These non- food resources mattered to diet indirectly because cooking, storage, house construction and animal keeping all depended on them. GAJENDRA THAKUR 4.16 Fuel was part of food technology Every cooked meal required energy, yet fuel is rarely treated as part of economic history. Wood charcoal, burnt clay, ash and hearths show repeated combustion, but identifying the exact fuel species requires anthracological study. Households could draw on wood, brush, dung, crop straw and husk in different combinations. The choice would affect labour, smoke, cooking temperature and pressure on surrounding vegetation. Panda's husk-tempered oven and the repeated hearths or ovens at Chirand and Oriup demonstrate that crop by-products and local biomass were integrated into domestic technology. Without direct fuel analysis, however, dung or particular wood species should remain possibilities rather than claims. 4.17 Hearths and ovens: repeated fire creates social space At Chirand, hearths and oblong ovens were associated with domestic structures; Oriup Period I contained an oven and circular fireplace; Panda yielded an earthen oven. These installations mark places where heat was repeatedly controlled. They could be used for cooking, roasting, parching, boiling with vessels, heating water or craft processes, and a single installation may have served several functions through time. Earlier interpretations sometimes called larger or multiple ovens evidence of 'community cooking'. That is possible, but size alone does not prove communal consumption. The more secure inference is that food preparation was spatially organised inside or beside houses and that fire installations were maintained features rather than accidental burns (Verma 2007; ASI 1966-67; Choudhary 2005). 4.18 Pottery multiplied ways of transforming food The spread of pottery changed the technical possibilities of food preparation even when exact recipes remain unknown. Bowls, jars, basins, dishes and other vessel forms could support boiling, soaking, serving, fermenting, carrying or storage. Sooting, residues, abrasion and use-wear can sometimes narrow these possibilities, but vessel shape by itself is not a recipe. At Chirand and the Chalcolithic sites discussed in Chapter 3, pottery repertoires become more varied and increasingly wheel-made in later phases. This diversification suggests growing differentiation among containment, cooking and serving tasks. It also created new surfaces on which food could be preserved archaeologically as carbonised crusts or absorbed residues, although systematic residue analysis remains limited for the regional sequence used in this volume. 4.19 Boiling, roasting and gruel are possibilities, not named dishes A combination of cereals, pulses, pottery and hearths makes several cooking methods technologically possible. Grain may be boiled whole, cracked, ground into meal, parched or incorporated into mixed preparations; meat and fish may be roasted, boiled or dried. Archaeology can identify the equipment and ingredients more readily than the final culinary form. For this reason the chapter does not translate the assemblage into modern terms such as khichdi, roti, rice-beer or regional fish dishes unless later period- specific evidence independently supports them. Similarity between an ancient technical possibility and a later cuisine is not evidence of continuity. The archaeological contribution is to define the range of feasible transformations and the labour invested in them. 4.20 Storage: one of the hardest parts of the food cycle to measure Storage is economically crucial because it turns a seasonal harvest into year-round food and seed, yet storage facilities are inconsistently preserved and identified. Large jars, pits, bins and raised containers can all store grain, but the same features can have other uses. Regional Neolithic and Chalcolithic sites provide comparative evidence for pits and storage vessels, while the Chirand record is clearer for pottery and domestic HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II structures than for a single, well-quantified storage system. It would therefore be unsafe to estimate surplus from the number of jars. The strongest inference is functional: diversified cropping and sedentary households require some means of holding seed and food between harvests, but the scale, ownership and duration of storage must be established site by site. 4.21 Food security came from diversity rather than guaranteed surplus A household combining rice, winter crops, livestock, fish and gathered resources had several ways to respond when one source failed. Flood damage to a field might coincide with productive fishing; a poor rice harvest could make pulses, livestock or stored grain more important; seasonal wild foods could bridge lean periods. This ecological diversity is a plausible form of risk buffering, not evidence that famine was absent. Early farmers remained exposed to flood, drought, pests, animal disease and storage loss. Figure 14 therefore presents a seasonal scheduling model rather than a prosperity narrative. The historical point is that food security depended on coordinating resources with different seasonal rhythms. Figure 15 — Seasonal scheduling and risk buffering in an early agrarian food system. Timings are schematic and varied locally. 4.22 Food waste is evidence of labour and choice Ash lenses, charred seeds, broken vessels, cut or burnt bones and shell concentrations are often classified as refuse, but refuse is the endpoint of economic decisions. A charred grain may represent accidental burning during drying, a cooking loss or discarded spoiled food; a butchered bone records carcass processing; broken pots mark repeated replacement of household equipment. Middens and floor sweepings can also reveal where tasks took place. The challenge is that deposits may be redeposited by flooding, trampling, pits or later construction. Context is therefore as important as the object. The archaeology of food is strongest when it reconstructs sequences of action—acquisition, processing, cooking, consumption and discard—rather than simply listing species. 4.23 Household labour can be reconstructed more safely than gendered labour Grinding grain, carrying water, tending animals, fishing, collecting fuel, cleaning crops, maintaining hearths and making or repairing containers all required recurring labour. These activities can be inferred from tools and installations, but assigning each task to women, men, children or particular status groups requires independent evidence. Later ethnography cannot be projected mechanically into the third or second millennium BCE. Archaeology may eventually use skeletal stress, activity markers, spatial patterning or burial GAJENDRA THAKUR evidence to refine labour histories, but the sites discussed here do not justify a fixed gender division of food work. The most defensible conclusion is that household subsistence was labour-intensive and coordinated across several activities. 4.24 Feasting and inequality: attractive hypotheses need stronger proof Large hearths, decorated vessels or concentrations of animal bone can invite interpretations of feasting and social competition. Such practices are well documented in many archaeological societies, but they must be demonstrated through context: unusually large serving sets, patterned deposits, high-value foods, repeated event layers or spatial segregation. The evidence reviewed for Chirand, Panda, Chechar and Oriup does not yet support a detailed hierarchy of elite and common diets. Differences in access are possible, especially as settlement complexity increases, but this chapter keeps them as research questions rather than narrative facts. The social archaeology of food begins with variation in deposits; it should not assume stratification before that variation is securely shown. 4.25 Floodplain food systems were mobile even when villages were sedentary Sedentary settlement did not mean that the productive landscape was fixed. River channels migrated, low-lying areas flooded, oxbows formed and silt renewed or buried fields. People could shift cultivation plots, move livestock seasonally, fish at different water levels and collect resources from changing wetland margins. The erosion observed at Oriup and the geomorphological problems discussed in Chapter 1 show why archaeological sites preserve only fragments of these wider catchments. A village mound is therefore the residential node of a larger and mobile foodscape. Figure 13 represents this relationship among fields, herds, wetlands, forest edge and river routes. 4.26 Nepal and the northern Tarai: a boundary of evidence, not a blank history The title of this volume includes India and Nepal, but the prehistoric and early agrarian food record is not equally documented on both sides of the modern border. The Patu evidence discussed in Chapter 2 concerns lithic occupation in the eastern Nepal Siwalik/foothill environment and does not provide an archaeobotanical food assemblage comparable to Chirand or Senuwar. It would be methodologically wrong to transfer the Bihar crop and livestock package northward merely because later Mithila spans the international boundary. For the Nepal Tarai, Chapter 4 therefore records an evidence gap. Later chapters, when textual, archaeological and historical documentation becomes stronger, will reconstruct Nepal-side agrarian and food histories on their own evidence. 4.27 The eastern Anga zone: direct fishing evidence, limited plant evidence For the Bhagalpur/Oriup zone, Period I provides unusually vivid fishing and hearth evidence but not the same published archaeobotanical resolution available for Chirand or Senuwar. This asymmetry should shape the narrative. It is possible to say that riverine resources, tortoise shell, fishing hooks and controlled fire were part of everyday life at Oriup. It is not equally secure to reproduce the full Chirand crop list as though excavated at Oriup. Regional comparison can suggest questions—whether rice, pulses or winter cereals were grown—but those questions remain open until local plant remains are securely reported. This is how a cumulative history avoids turning geographic proximity into fabricated evidence. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 4.28 Conclusion: an agrarian economy remained a mixed food economy The archaeology of food changes the scale of early history. Instead of treating the Neolithic and Chalcolithic mainly as ceramic or technological labels, it reveals daily systems of cultivation, herding, fishing, hunting, processing, cooking and storage. Chirand anchors a diversified local food economy; Panda connects cereal by-products with domestic technology in the Mithila plain; Oriup places fishing and hearths firmly inside the early Anga/Bhagalpur landscape; Senuwar demonstrates, comparatively, the potential complexity of mixed seasonal agriculture. None of these sites justifies a reconstructed cuisine in the modern cultural sense. What they establish is more fundamental: early agrarian communities survived by coordinating multiple ecological zones, multiple species and repeated household labour. Chapter 5 moves outward from these food practices to the larger question of environment, mobility and settlement. Table 4.1 — Food evidence and limits of inference Evidence Principal anchor Strongest Do not infer supported automatically inference Charred grain / husk Chirand; Panda; Plant presence, crop Dietary percentage, Senuwar processing; named dish, (comparator) sometimes ownership cultivation Animal bone Chirand Species present; Exact meat husbandry/hunting consumption, dairy when context allows or traction without evidence Fish / turtle / shell Chirand; Oriup Aquatic-resource use Year-round intensity and fishing or social status of technology fishers Hearth / oven / ash Chirand; Panda; Repeated controlled Specific recipe or Oriup heating communal feast Querns / mullers / Chirand and regional Plant/mineral Which cereal or pestles comparators processing preparation without residues Pottery forms Chirand; Chechar; Containment, Vessel contents from Oriup; Panda heating, serving, shape alone storage possibilities Pits / jars / bins Regional Possible storage Surplus size, taxation comparators; variable where context or household wealth local evidence supports it