Labour, Service and Occupational Communities

Communities Early medieval social history is often reconstructed from landholders, rulers and institutions because they dominate the written record. This chapter reverses the angle of vision. It asks who cultivated fields, carried messages, guarded roads, supervised ferries, made tools, carved images, copied documents, supplied monasteries and performed the countless services that turned rights on copper plates into functioning society. The evidence does not permit a complete occupational census. It does permit a controlled reconstruction of differentiated labour, service and skill—and of the gradual processes through which some livelihoods became more tightly organised as hereditary social groups. 28.1 Labour as a historical problem Labour is easy to see in the results of production and difficult to see in the surviving texts. Fields were cleared and cultivated, embankments maintained, cattle tended, boats moved, pottery fired, metal repaired, images carved, documents copied and institutions fed. Yet royal charters usually record rights, beneficiaries and officials rather than the names and conditions of the people who performed most of this work. A social history of labour must therefore reconstruct relationships from several kinds of evidence instead of treating a list of occupational labels as a census. For Mithila, Vajji and Aṅga in the early medieval centuries, the best regional controls are Pāla-period charters from Munger and Bhagalpur, the archaeology of Antichak/Vikramaśīla, and material evidence from Mudgagiri. Wider eastern Indian inscriptions help identify possible patterns, but they cannot be imported wholesale into north Bihar. 28.2 Occupation, service, status and caste are not synonyms A person’s work, mode of dependence, social rank and later caste identity need not coincide. Kṣetrakara identifies cultivation more directly than it identifies a bounded caste. Sevaka denotes service but does not by itself tell us whether the service was domestic, military, administrative, temporary or hereditary. Terms such as mahattara and kuṭumbin point to rural standing and household position rather than a single occupation. Conversely, later normative texts may classify communities by inherited status even when earlier records reveal more fluid economic roles. This distinction is essential because early medieval history has often been written backwards from later caste ethnographies. The safer method is chronological: establish what a contemporary record actually names, identify the social relation implied by that term, and only then ask whether later evidence shows increasing heredity or closure. 28.3 Four evidence streams and their limits Figure 108 separates four evidence streams. Copper-plate charters are strongest for administrative visibility: they list officials, residents, cultivators and groups whose compliance with a grant mattered. Archaeology reveals production without necessarily naming producers. Normative and literary texts preserve social categories, ideals and stereotypes, but they are not demographic surveys. Comparative work on eastern India can illuminate processes such as the gradual systematisation of occupational groups, yet regional timing varied. No single stream is sufficient. An iron object proves metal use, not the caste of the smith; an official title proves an office, not its lifetime tenure; a ranked social list proves a classificatory language, not the exact everyday relations among all people named. Labour history becomes reliable when these evidentiary limits are built into the argument rather than added as a final disclaimer. 303303 GAJENDRA THAKUR 28.4 The Bhagalpur plate as a social document The Bhagalpur plate of Nārāyaṇapāla, issued in year 17, is especially valuable because its notification formula extends far beyond a royal elite. It addresses a long sequence of officers and servants and then turns to residents, describing a social field from Brahmans and leading persons down to Meda, Andhra and Caṇḍāla. The same grant requires resident cultivators to obey the new fiscal arrangement and deliver the appropriate shares and dues. This does not make the plate a complete map of society: the list is formulaic, status-laden and shaped by the legal purpose of transferring a village’s revenues. Even so, it demonstrates that the grant was imagined as affecting people who cultivated, administered, policed, transported, served and inhabited the locality. A land grant was therefore also a reordering of obligations. 28.5 The Munger plate and the labour of administration Devapāla’s Munger plate of year 33 contains an equally revealing roster. Alongside high officers appear functionaries connected with punishment, policing, tolls, forts, frontiers, messaging, travel, ferries, animals and local administration. It then addresses residents led by mahattaras and kuṭumbins and extends the social range to groups including Meda, Andhraka and Caṇḍāla. The importance of the passage lies less in translating every title with false precision than in recognising the density of specialised service around royal government. Administrative power depended on labour: messages had to travel, revenue had to be assessed and moved, animals maintained, crossings supervised and coercion organised. The court and the countryside were connected by people performing these functions. Many of them may have been mobile; others were embedded in local society. 28.6 Cultivators and the base of the agrarian economy The resident kṣetrakaras named in Pāla charters were the people whose cultivation made a revenue grant economically meaningful. They were not necessarily a homogeneous peasantry. Some may have held substantial household plots; others worked smaller holdings or land controlled through superior rights. The charter tells us that cultivators had obligations to render specified dues after a transfer, but it rarely tells us how labour was recruited within the household, whether hired workers were employed, or how harvest risk was shared. Ryosuke Furui’s reconstruction of early medieval eastern India stresses increasing differentiation in land relations and the growth of larger holdings above actual cultivators. For the regional history of Mithila and Aṅga, this suggests a stratified agrarian world without allowing us to assign a single tenure status to every kṣetrakara. 28.7 Household labour and invisible cooperation Much agricultural work took place within households. Ploughing, transplanting, weeding, harvesting, threshing, storage, animal care, fuel collection and food processing demanded coordinated labour across seasons. Charters normally name household heads or collective categories and therefore hide much of the work of women, children, dependants and junior kin. This invisibility should not be mistaken for absence. Agricultural manuals and comparative rural evidence show that successful cultivation required timed cooperation rather than the isolated activity of a male landholder. Household labour also linked production to reproduction: children learned skills, food was processed, tools maintained and animals cared for. The historical household was thus both a social and productive institution. Because the surviving sources distribute recognition unequally, a labour history must distinguish who worked from who appeared in legal documents. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 28.8 Mahattaras, kuṭumbins and rural differentiation Terms such as mahattara and kuṭumbin appear in early medieval eastern Indian records as indicators of rural standing. They should not be flattened into modern equivalents such as landlord, peasant proprietor or village elder in every context. Their importance here is relational. They occupied positions from which they could be notified, witness transactions, mediate boundaries or represent local interests, while poorer cultivators and labourers were less visible. This differentiation affected access to labour. Better-resourced households could command family labour, exchange work with neighbours, hire assistance or extend credit during agricultural bottlenecks. Less secure households might sell labour or accept unfavourable advances. The rural economy therefore contained unequal capacities to mobilise work. Those inequalities are more historically defensible than a simple division between free peasants and serfs. 28.9 Agrarian labour beyond the named cultivator Actual field work could exceed the labour capacity of the household that held cultivation rights. Peak seasons created demand for additional hands, and larger holdings could only be managed through dependants, wage-like arrangements, share relations, attached labour or combinations of these. The evidence from Pāla north Bihar does not permit a numerical estimate of landless labourers, and the language of later agrarian systems should not be projected backwards. Nevertheless, the gap between legal control over land and the physical work required to cultivate it is historically significant. Where landholding became more stratified, the possibility of separating the manager or superior right-holder from the actual tiller increased. Furui’s wider eastern Indian model places agrarian labourers at the lower end of increasingly layered land relations by the later early medieval period. 28.10 Figure 108: reconstructing labour without a false census Figure 108 summarises the evidentiary method. Charters illuminate people made legible by fiscal and administrative transactions; archaeology illuminates the products and spaces of work; normative texts illuminate ranking and social ideology; comparative history helps identify processes whose regional presence must still be demonstrated. The convergence of these streams allows strong statements about specialised labour and differentiated social relations. It does not justify claims that every occupational group was already a closed caste or that every named status corresponded to one economic function. This distinction matters most for communities at the lower end of textual hierarchies, whose voices are rarely preserved in documents produced for courts, monasteries or Brahman beneficiaries. 305305 GAJENDRA THAKUR Figure 108 — Labour history: distinct evidence streams 28.11 State service as an occupational world The official rosters of the Munger and Bhagalpur plates reveal government as a large service complex. Beyond ministers and commanders were officers concerned with accounts, justice, policing, tolls, territorial supervision, forts, messengers, transport and animal resources. Some titles probably represented substantial offices; others may have denoted operational personnel or categories of royal retainers. Together they show that state formation generated work that was not directly agricultural. Administrative labour depended on literacy, numeracy, mobility, record keeping and coercive capacity. Mudgagiri/Munger, a Pāla jayaskandhāvāra, would have concentrated many such services when the royal camp or headquarters was present. The town’s significance therefore included a labour market created by government itself—clerks, guards, suppliers, transporters, artisans and attendants serving an administrative centre. 28.12 Cāṭa, bhaṭa and sevaka: coercion and service Pāla charters repeatedly refer to cāṭas, bhaṭas and sevakas, often in formulas that prohibit their entry into donated villages. The terms have been interpreted in relation to soldiers, police-like retainers and servants, but their precise institutional meanings could vary. The exemption formula is nevertheless revealing. Villagers experienced royal personnel not only as abstract authority but as bodies that could enter settlements, demand supplies, enforce orders or impose burdens. Excluding such entry enhanced the autonomy or privilege of the donee. From the standpoint of labour history, the formula also reminds us that military and service personnel had to be housed, fed and moved. State service redistributed demands onto rural producers even when the surviving charter presents the matter as a legal privilege. 28.13 Ferries, boats, messengers and animal handlers The Pāla official lists include titles connected with ferries and with elephants, horses, camels, cattle and other animals, as well as messengers and personnel concerned with movement. In the riverine geography of Mithila, Vajji and Aṅga, these were not peripheral occupations. Crossing the Ganga or its tributaries required boats, landing places, skilled handling and local knowledge of currents and seasons. Royal communication depended on relays of people and animals. Armies and courts depended on fodder, stable work and HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II veterinary knowledge even when the texts do not individualise those workers. Transport labour linked agrarian villages to markets, administrative centres and religious institutions. It also carried risk: flood, channel change and seasonal scarcity could transform the value of local expertise. 28.14 Figure 109: the charter’s social field Figure 109 places the main categories visible in Pāla grants into three analytical zones: administrative service, rural society and ranked social classification. The zones overlap. A local notable might also serve the state; a cultivator might transport produce; a person placed low in normative hierarchy might possess specialised ecological or craft knowledge. The point of the diagram is therefore not to freeze people into boxes but to show how the charter assembled different groups because a grant needed their recognition or obedience. The administrative document makes society visible through its own purpose. Historians must reconstruct the wider life behind that selective visibility. Figure 109 — A Pāla charter’s social and service field 28.15 Craft labour at Mudgagiri/Munger Recent study of Pāla sculptural material from Munger is important for labour history because unfinished sculptures preserve stages of carving and polishing. Such objects demonstrate production, not merely consumption. A finished image in a temple can travel far from its workshop; an unfinished image is stronger evidence for local or nearby working activity. Stone carving required quarry access or imported blocks, transport, specialised tools, design knowledge, roughing out, finishing and often inscription. Mudgagiri’s administrative and religious importance created demand for such work. We should not assume one permanent guild or hereditary workshop without direct evidence, but the material record supports skilled śilpakāras operating within a patronage environment where court, temple and urban demand could sustain specialised production. 28.16 Antichak/Vikramaśīla and institutional demand The excavated complex at Antichak demonstrates the labour demands of a large monastic institution. Brick architecture, drains, cells, shrines, terracotta plaques, pottery, seals and sealings, metal objects, stone 307307 GAJENDRA THAKUR sculpture and bone, shell and ivory objects represent many kinds of work. The monastery required builders during major construction phases and continuing labour for repair, cooking, cleaning, water management, storage and provisioning. Some craft production may have occurred on site; other goods were brought through surrounding markets and villages. Archaeology cannot assign every object to a named community, but it establishes an economy of skilled and unskilled labour around the institution. Vikramaśīla was thus not only a centre of learning and ritual but also a sustained consumer of regional labour. 28.17 Potters and the archaeology of everyday production Ceramics are among the most abundant archaeological traces of labour because vessels break and are discarded. Pottery at Antichak and other sites points to repeated production for cooking, storage, serving and ritual use. Potters needed clay, water, fuel, forming skills, firing knowledge and access to consumers. Production could be household based, village based or concentrated near larger settlements; the archaeological record rarely tells us the social identity of the potter directly. The historical importance of pottery lies precisely in its ordinariness. Monumental patronage highlights elite art, but daily institutional and household life depended on large quantities of replaceable craft goods. A social history of occupations must therefore place ordinary ceramic labour alongside prestigious stone and bronze work. 28.18 Metalworkers, repair and the economy of tools Iron tools and fittings, copper objects and bronzes reveal different sectors of metalworking. Agricultural expansion required cutting implements, plough components, knives and repair. Buildings required clamps, hinges, nails and tools. Religious institutions demanded lamps, vessels and images. Military establishments required weapons and equipment. The same smith could serve a mixed local clientele rather than a single specialised market. Metalworking also linked workers to supply chains for ore or scrap, charcoal and fuel. Because finished metal was frequently recycled, archaeology underrepresents the scale of production. The presence of iron and copper objects at Antichak therefore signals technical labour while leaving workshop organisation open. We should distinguish evidence for metal use from evidence for a resident hereditary smith community. 28.19 Stone carvers, image makers and artistic skill Pāla-period eastern India is famous for stone sculpture, but the social history of the people who made it is less visible than the art itself. Munger’s unfinished pieces help recover sequences of manufacture, while wider eastern Indian evidence preserves occasional references to artists and engravers. By the Sena period, the Deopara inscription famously praises Śūlapāṇi as a leading figure among the artists of Varendra; whether the phrase proves a formal guild remains debated. For this volume the safer conclusion is that skilled artistic communities and reputational networks existed, with mobility between patronage centres. Stylistic similarity can suggest shared training or models, but it should not be converted automatically into a named workshop unless material or inscriptional evidence supports that step. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II Figure 110 — Occupation depended on place and demand 28.20 Scribes and engravers as documentary professionals Copper-plate grants were material products of specialised literacy. A royal decision had to be formulated in Sanskrit, copied, checked, engraved on copper and authenticated with a seal. The Bhagalpur plate names Maṅghadāsa, son of Śubhadāsa and associated with Samataṭa, as its engraver. Such notices are valuable because they reveal named technical specialists and, in some cases, geographic mobility. Sanjukta Datta’s work on Pāla copper plates shows that engraving and charter production formed a documentary culture with identifiable practitioners. Scribes and engravers occupied an intermediate world between administration and craft: their work required linguistic training, graphic skill and institutional access. Their visibility contrasts sharply with the anonymity of most cultivators and manual workers. 28.21 Ritual specialists and the labour of institutions Temples and monasteries created regular work beyond the roles of monks, teachers or Brahman recipients named in grants. Worship required flowers, lamps, food offerings, water, cleaning, music or recitation in some contexts, repair of buildings and management of supplies. The Bhagalpur grant explicitly supports worship, offerings, charitable feeding, construction and care for the sick at a Śaiva-Pāśupata establishment. These purposes imply cooks, suppliers, builders, attendants and medical provisioning even though the workers are not individually named. Institutional endowments thus transformed agricultural revenue into recurring service demand. The economic significance of religious patronage lay partly in this conversion of landed income into labour, materials and maintenance. 28.22 Merchants, brokers and market-facing occupations Chapter 27 showed that markets linked villages, towns and institutions. Labour history adds the people who made those connections work: merchants, brokers, transporters, weighers, packers, boat operators and petty vendors. Direct evidence for each role in Mithila and Aṅga is uneven, so the Rajbhita inscription from the wider Pāla world remains a comparative control rather than a local template. It records an organised vaṇiggrāma associated with rural markets and demonstrates that commercial specialisation could develop outside large cities. Merchant organisation also created demand for credit, storage, security and information. 309309 GAJENDRA THAKUR The boundary between occupation and social group could therefore harden through repeated cooperation and shared interests, but this was a process rather than an automatic consequence of trade. 28.23 Kaivartas and the danger of a single occupational translation Some Pāla-period eastern Indian charters include Kaivartas in ranked social lists. The term has long associations with water, fishing, boating and later social identities, yet it should not be given one timeless occupational meaning. Riverine skills were economically important across the Ganga plain, but a label could broaden, change status or acquire political significance over centuries. The later Kaivarta mobilisation in Varendra belongs to a different regional and chronological context and cannot be used to define every earlier Kaivarta in Bihar. The historical value of the term lies in showing that groups associated in later evidence with specialised livelihoods were becoming socially legible. The exact relation between livelihood, status and heredity must be established case by case. 28.24 Meda, Andhra, Caṇḍāla and the rhetoric of social range The Munger and Bhagalpur charter formulas extend their address lists to communities such as Meda, Andhra or Andhraka and Caṇḍāla. The phraseology communicates comprehensiveness and hierarchy: the king’s order reaches from prestigious Brahmans and leading residents to socially subordinated groups. It would be hazardous to assign each label a fixed local occupation. Some may reflect ethnic, regional, occupational or status histories that had already been reinterpreted by Brahmanical classificatory language. What the formula establishes with confidence is unequal social recognition. People placed at the lower end of these lists were part of the inhabited and productive landscape even when elite texts represented them through stigma. Their inclusion in a grant notification is therefore evidence of presence and subordination, not a complete ethnography. 28.25 Women’s labour and the documentary silence Women are dramatically underrepresented in the occupational vocabulary of royal charters, yet rural and institutional economies could not function without their work. Food processing, water collection, animal care, textile production, childcare, transplanting, weeding and post-harvest tasks are among activities that comparative agrarian history consistently identifies as gendered labour, though the precise division varied by place and household. Women also appear as donors and patrons in some eastern Indian records, demonstrating that legal visibility was possible under particular conditions. The methodological rule is to avoid replacing documentary silence with invented detail while refusing to equate silence with inactivity. Where a task is necessary to a documented production process, the historian may identify the labour requirement without pretending to know the exact gender distribution in every locality. 28.26 Burdens, exactions and dependent labour Land grants frequently describe donated property as exempt from various burdens and from the entry of certain royal personnel. The phrase parihr̥ ta-sarva-pīḍa—freed from all burdens—signals that rural communities could be subject to claims beyond ordinary revenue. It does not enumerate those claims, and it should not automatically be translated as a precise system of forced labour. Elsewhere in early medieval India, viṣṭi is well attested, but its occurrence and meaning vary by region and period. For Pāla north Bihar the cautious conclusion is that state and service demands could impose labour-like or provisioning burdens on settlements, while the exact balance among compulsory work, supplies, cash dues and customary service HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II remains uncertain. Exemption formulas reveal the pressure of extraction even when they do not describe every mechanism. 28.27 From repeated livelihood to hereditary occupational group Ryosuke Furui’s regional synthesis argues that early medieval eastern India witnessed increasing organisation of hereditary occupational groups and their later systematisation within a jāti order. Figure 111 presents this as a process rather than a starting assumption. Skills can be transmitted within households because training, tools and contacts are valuable. Repeated service to the same patrons can deepen dependence. Marriage networks and normative classification can then reinforce group boundaries. Yet these stages need not occur everywhere or at the same speed. A potter family might transmit craft skills for generations without belonging to a fully closed regional caste system in an earlier century. The chronology of social closure is therefore as important as the identification of occupation itself. Figure 111 — From work to social group: a historical process 28.28 Regional variation across Mithila, Vajji and Aṅga The three regional zones did not offer identical labour environments. North Bihar’s alluvial and riverine landscape privileged cultivation, cattle, fisheries, ferries and seasonal transport, while its charter record emphasises villages and administrative units. Aṅga’s Ganga corridor, Mudgagiri and the large institution at Antichak created stronger archaeological visibility for urban, artisanal and institutional labour. Older centres in the Vajji landscape continued to carry religious and route significance, but early medieval occupational documentation is thinner. These differences are partly historical and partly archival. Where inscriptions survive, officials and rural categories become visible; where excavations are rich, crafts and material work come forward. Regional comparison must therefore distinguish differences in past society from differences in what has been preserved and studied. 28.29 Methodological rules for occupational history Six rules guide reconstruction. First, do not equate occupation, title, status and caste. Second, treat charter lists as administrative classifications, not censuses. Third, use archaeology to demonstrate production 311311 GAJENDRA THAKUR before assigning producer identity. Fourth, distinguish necessary labour from documented labour: a monastery required cooks and cleaners even if their names are absent, but their precise organisation may remain unknown. Fifth, keep comparative eastern Indian evidence explicitly comparative unless a regional anchor exists. Sixth, date processes of heredity and jāti consolidation rather than projecting later social formations backward. These rules produce a less dramatic but more reliable account. They also recover workers as historical actors without pretending that the sources preserve their voices equally. 28.30 Conclusion: work connected the regional economy Early medieval Mithila, Vajji and Aṅga were held together not only by kings, grants and institutions but by work. Cultivators produced the agrarian base; transport workers connected river and road; officials converted royal orders into local practice; artisans built, repaired and supplied towns and monasteries; scribes and engravers materialised documentary power; servants and military retainers sustained coercive institutions. Social hierarchy shaped who controlled resources and who remained anonymous. Over time, repeated livelihoods, household transmission, patronage and Brahmanical classification could contribute to more hereditary occupational identities, but the process was uneven. The strongest regional evidence therefore supports a differentiated labour society in motion—not a timeless caste chart. Chapter 29 turns from occupational organisation to the changing relations of household, kinship and social reproduction that sustained these communities. Table 28.1 — Evidence for labour and occupational communities, with limits of inference Evidence What it shows Historical use Limit of inference Bhagalpur plate of Long roster of officials Administrative Not a census; Nārāyaṇapāla and servants; residents; visibility of specialised formulaic hierarchy kṣetrakaras; ranked service and cultivators does not define each social range group’s occupation Munger plate of Officials for tolls, Specialised state Titles do not reveal Devapāla policing, territorial service and wages, recruitment or control, movement, differentiated rural hereditary tenure ferries and animals; society mahattaras and kuṭumbins Grant exemption parihr̥ ta-sarva-pīḍa and Villages could face Does not by itself formulas exclusion of cāṭa-bhaṭa burdens and intrusive prove a specific local entry service demands system of viṣṭi Antichak/Vikramaśīla Pottery, terracotta, Large institutional Objects rarely identify archaeology stone, metals, seals, demand for skilled the social group of the architecture and drains and routine labour producer Unfinished Pāla Visible stages of carving Local or nearby Does not establish sculptures at Munger and polishing skilled stone-working one permanent guild activity or caste workshop Bhagalpur named Named specialist Professional literacy, One named artisan engraver Maṅghadāsa associated with charter engraving skill and does not define all manufacture mobility scribal organisation Wider eastern Indian Occupational labels and Comparative model Cannot be projected Kaivarta / later processes of group for increasing unchanged into every occupational evidence consolidation hereditary locality or earlier organisation century Furui’s regional Stratified land relations Long-term process of A model to test synthesis and occupational social differentiation regionally, not a groups increasingly substitute for local systematised toward jāti evidence