Architecture and Settlement

Architecture in Mithila, Vajji and Anga survives unevenly. Massive brick ramparts, stupas, monasteries, rock-cut shrines and temple fragments are visible; timber houses, thatch, fences, workshops and periodic markets are far less so. This chapter therefore joins architecture to settlement archaeology. It asks how water, rivers, routes, fortification, religious institutions, craft production and political authority organized built space, and it keeps excavated sequence separate from legendary or retrospective identification. The result is a social history of built landscapes rather than a catalogue of monuments. 44.1 Architecture as social history, not a catalogue of monuments Architecture records labour, resources, political authority, religious practice and everyday habits in material form. Yet the surviving record of Mithila, Vajji and Anga is uneven. Monumental brickwork, fortifications and religious structures are more likely to survive than timber houses, thatched roofs, fences or temporary market sheds. A history based only on standing monuments would therefore exaggerate the architectural weight of elites and institutions. This chapter treats architecture together with settlement: where people lived, how water and routes shaped habitation, how defensive works reorganized space, how monasteries and temples concentrated resources, and how ordinary building materials disappeared through rebuilding and floodplain erosion. The central method is archaeological rather than typological. A plan, wall, mound, tank or sculptural fragment is interpreted first in its excavated or documented context. Later tradition may preserve valuable memory, but a legendary name does not automatically establish the original function or date of a structure. 44.2 The alluvial plain as an architectural environment North Bihar and the adjoining Tarai are difficult landscapes for durable settlement. Rivers migrate, floodplains accrete and erode, high water tables complicate excavation, and brick can be robbed repeatedly for later construction. These processes affect both how people built and what archaeologists can recover. Raised habitation, embankment-like edges, drainage, tanks and the selection of relatively stable ground were practical responses to water risk. At the same time, rivers were not simply hazards. They supplied transport corridors, fisheries, silted agricultural land and connections to markets. The architectural history of the region therefore cannot be separated from hydrology. A settlement that appears isolated on a modern road map may originally have been oriented toward a river channel that later shifted. Conversely, a large mound beside a present river cannot be assumed to have had the same relationship to water in antiquity. Reconstructing settlement requires geomorphology as well as masonry. 44.3 Preservation bias: earth, timber and thatch versus brick and stone The everyday built environment was probably dominated for long periods by materials that survive poorly. Earth walls can slump into indistinct layers; timber posts leave only stains or post-holes; bamboo and thatch usually vanish; floors may be renewed until earlier surfaces are difficult to isolate. Fired brick, stone thresholds, terracotta plaques and substantial ramparts survive more visibly and therefore attract disproportionate historical attention. This preservation bias is especially severe in wet alluvial conditions. It means that absence of a stone house is not evidence for architectural poverty, and the survival of a temple wall is not proof that religious buildings occupied most of the built landscape. Archaeological interpretation must work from traces of surfaces, hearths, drains, storage, craft debris and spatial organization as well as HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II from walls. Figure 172 summarizes this filter: what survives most clearly is not necessarily what most people inhabited most of the time. Figure 172 — Architecture survives through a preservation filter 44.4 Settlement archaeology and the problem of historical names Ancient and medieval place-names often outlive the physical arrangements to which they once referred, while modern villages may preserve, relocate or reinvent older names. The problem is acute in famous landscapes such as Mithila and Vaishali, where literary and religious memory is exceptionally strong. Historical geography therefore requires several steps: identify the archaeological mound or structural complex; establish its occupational sequence; compare inscriptions, coins and datable artefacts; and only then evaluate textual or traditional names. The procedure protects the history from circular reasoning. A place is not dated because a story names it, and a story is not proved ancient because a modern mound carries the same name. This does not diminish local tradition. On the contrary, it makes tradition itself a historical source whose chronology can be studied. The distinction becomes essential at sites associated with palaces, assemblies, royal capitals or epic figures, where later commemoration can easily be mistaken for excavation. 44.5 Vaishali as a long-lived settlement landscape Vaishali provides one of the clearest examples of a settlement whose archaeological life extends across many political and religious phases. Excavations and protected remains include fortification, stupas, monasteries and foundations, while artefacts range across early historic and later periods. The importance of this evidence lies less in assigning every structure to a famous event than in demonstrating repeated occupation and spatial investment. Defensive works were repaired; religious monuments accumulated; water and circulation routes helped sustain habitation; later communities reused earlier spaces. Such continuity complicates any attempt to draw a single “plan of ancient Vaishali.” What archaeology reveals is a palimpsest. The settlement was not frozen in the republican age, and later Gupta, post-Gupta and medieval activity must be allowed its own history. This long duration also explains why structural remains from different centuries can stand close together without belonging to one architectural programme. 469469 GAJENDRA THAKUR 44.6 Raja Vishal ka Garh: fortification before political identification The mound known as Raja Vishal ka Garh is frequently connected in popular description with an ancient assembly. Archaeologically, the more secure observation concerns fortification. Recent excavation across its ramparts, as summarized by the Archaeological Survey of India, has identified multiple defence-wall phases assigned to the Śuṅga, Kuṣāṇa and early Gupta periods, while finds from the broader site extend from the early historic era into much later centuries. This is substantial evidence for a defended and repeatedly used enclosure. It is not, by itself, proof of the precise constitutional function of every enclosed structure. The distinction is important. Fortification tells us about labour mobilization, boundary-making and investment in protected space; political texts may suggest forms of assembly; but the two evidence streams must converge before a specific building can be named an assembly hall. Architectural history is strongest when it resists the temptation to turn a secure wall into an insecure institution. 44.7 Stupas, monasteries and the making of sacred settlement Buddhist monuments around Vaishali show how sacred architecture could reorganize settlement space. A relic stupa drew memory, pilgrimage and maintenance to a fixed point; monasteries required cells, water, storage, access and support from surrounding producers. Such complexes were not architecturally isolated from the economy. Construction demanded brick-making, transport and specialized labour; continued occupation required food and repair; pilgrimage created routes and periodic concentrations of people. Over time, sacred sites could survive even when surrounding settlement patterns changed. Their visibility today can therefore distort the ancient balance between religious and domestic architecture. The appropriate question is not merely where the stupa stood, but what network allowed it to function. Religious building becomes social history when connected to donors, craftsmen, agricultural surplus, roads, river crossings and the practical routines of residents and visitors. 44.8 Balirajgarh: a fortified site and the value of construction sequence Balirajgarh in Madhubani district is one of the region’s most important fortified archaeological sites. The Archaeological Survey of India’s 1962–63 excavation is especially useful because it documented construction rather than relying on the site’s traditional name. High groundwater prevented excavators from reaching natural soil, but sections through the defence wall revealed a mud-brick core with brick encasement and several phases of reinforcement and repair. The wall was battered rather than vertical, a practical choice for a massive earthen-brick defence. This sequence demonstrates sustained investment in fortification and changing engineering responses. It does not settle every question about the settlement’s identity or political status. Proposals equating Balirajgarh with an ancient capital remain hypotheses to be tested against chronology, material culture and regional settlement patterns. Its strongest contribution to this chapter is architectural: earth and fired brick were combined to create large defensive works on the alluvial plain. 44.9 Fortification as labour, authority and risk management A rampart is not simply a military object. Building and maintaining a large wall requires organized labour, access to earth or brick, planning of gates and drainage, and decisions about what space is worth enclosing. Fortification can therefore indicate political concentration even when the name of the ruling authority is uncertain. It also reflects risk: warfare, raiding, control of trade, prestige competition or the need to manage movement through gateways. In flood-prone country, raised defensive works could interact with drainage and topography, although such hydraulic functions must be demonstrated site by site rather than assumed. Comparing Balirajgarh, Vaishali and Simraongarh shows why one label cannot explain all walls. Their dates, HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II materials and settlement contexts differ. The meaningful comparison concerns construction sequence, enclosed area, gate organization and relation to occupation, not a generic category of “fort.” 44.10 Village architecture: the largest missing archive Most people did not live inside monumental complexes. They lived in villages and dispersed settlements whose architecture was frequently rebuilt with perishable materials. The ordinary house compound may have included sleeping rooms, kitchens, cattle spaces, grain storage, courtyards, fences and work areas whose archaeological signatures are subtle. Seasonal repair could erase earlier wall lines while preserving the continuity of a household plot. Because such remains are less spectacular, they are often underrepresented in published architectural histories. Ethnographic observation of later Mithila compounds can suggest questions—about courtyards, gendered work spaces, thresholds or cattle—but it cannot be projected unchanged into the medieval period. Archaeology must establish each feature independently. The most responsible reconstruction therefore accepts uncertainty: the monumental record is detailed, while the domestic record is fragmentary, even though domestic architecture formed the numerical majority of the built environment. Figure 173 — Settlement as a connected ecological system 44.11 Brick economies and standardization without uniformity Fired brick is one of the most persistent architectural materials of the middle and lower Ganga plains. Its use implies clay extraction, moulding, fuel, firing, transport and skilled laying, and therefore links buildings to an economy of production. Brick dimensions can contribute to chronology when measured in secure contexts, but they should not be used mechanically: reused bricks, local variation and repairs complicate typological dating. Large religious and defensive projects could stimulate production on a scale far beyond household building. At the same time, brick architecture was not uniform. Mud mortar, lime plaster, terracotta facing and stone architectural members could be combined differently according to function and period. The social question is therefore not whether a society “knew brick,” but who could mobilize enough labour and fuel to use it extensively, and how the resulting structures were maintained or cannibalized by later settlements. 471471 GAJENDRA THAKUR 44.12 Anga’s riverine and rocky architectural zone The eastern zone of Anga differs materially from the flatter northern plains. Around Bhagalpur and Kahalgaon, the Ganga interacts with rocky outcrops and hills, creating architectural possibilities not available everywhere in Mithila. The Archaeological Survey of India records a monolithic rock-cut shrine in Bhagalpur district, tentatively placed no later than the ninth century, as well as rock sculptures and cave features in the wider district. Such remains demonstrate that regional architecture was not exclusively brick-built. Stone could be carved in situ where geology permitted, while the river connected these sites to wider routes. This variation cautions against imagining a single “Mithila style” extending mechanically across Mithila, Vajji and Anga. Material culture follows geology as well as political history. A social history of architecture must therefore compare subregions before synthesizing them. 44.13 Vikramshila: a planned monastic complex in Anga The excavated remains at Vikramshila near Antichak provide the clearest large-scale architectural plan in the Anga zone. The Archaeological Survey of India describes a square monastery with 208 cells arranged around a courtyard and a central cruciform stupa; excavation also revealed gateways, votive structures and a building interpreted as a library. This plan is important because it spatially organizes a resident scholarly community. Cells, circulation, worship and teaching were integrated within a single complex, while the surrounding landscape supplied food, water, labour and transport. The monument therefore joins architectural and intellectual history: Chapter 17 treated learned communities, while this chapter asks how architecture made such communities physically possible. Vikramshila also warns against treating institutions as disembodied texts. Manuscripts, teachers and ritual required walls, roofs, storage, pathways and continual maintenance. 44.14 The central stupa and the choreography of movement The cruciform stupa at Vikramshila structured movement as well as vision. Its central position within the monastic square made it an architectural focus from multiple directions. Circumambulatory paths, projecting chambers and terracotta decoration transformed religious practice into patterned movement through built space. Such architecture did not merely contain ritual; it choreographed it. The relationship between cell and stupa also created a daily spatial rhythm between residence and devotion. Similar questions should be asked of other religious sites even when plans are less complete: where did worshippers enter, what route did they follow, what could they see, and which spaces were restricted? Architecture becomes historically meaningful when plan is translated into action. The spatial discipline of a monastery differed from that of a market, fort or household, and those differences reveal distinct forms of community. 44.15 Terracotta as architecture, image and social archive Terracotta plaques at Vikramshila remind us that architectural surfaces can preserve social imagery. Decorative programmes may include religious figures, animals, hunting scenes and other motifs that bring elite, devotional and everyday themes into the same visual field. Terracotta is especially important in brick- building regions because it allows moulded or carved surface articulation without requiring abundant building stone. Yet a plaque is not a photograph of ordinary life. Its scenes are selected, stylized and placed within a religious architectural programme. Their value lies in the conjunction of material, imagery and location. Clay, firing and attachment techniques belong to craft history; iconography belongs to religious and artistic history; placement on a stupa belongs to architectural history. Treating these together avoids separating “art” from the built environment that gave it meaning. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II 44.16 Temple architecture and the danger of survivor typologies Temples in the region survive in radically different states: standing buildings, rebuilt shrines, isolated stone images, architectural fragments and remembered sacred places. A modern temple may occupy an old sacred site while incorporating little or none of the earlier fabric. Conversely, a medieval image may have been moved from its original building. Typological study of śikhara, doorway, moulding or plan can be useful, but only when the surviving element is securely contextualized. Vijayakanta Mishra’s survey of Mithila art and architecture collected many regional examples and remains valuable as a descriptive starting point, yet archaeological dating and fabric analysis must control historical claims. The chapter therefore treats temples as layered sites. Ritual continuity, architectural continuity and sculptural continuity may coincide, but they are not the same thing. 44.17 Stone sculpture without standing temples The abundance of stone religious sculpture in parts of Bihar and Mithila does not imply that equally numerous stone temples once stood intact. Images could be housed in brick shrines, transferred during rebuilding, recovered from fields or tanks, or reused in later structures. This distinction is essential because art history often preserves objects more successfully than architecture. A sculpture’s style may provide a relative date, but its present shrine may be much later. The architectural historian must therefore record provenance carefully: excavated location, reported findspot, current location and any evidence of reuse. The same rule applies to doorjambs, pillars and carved blocks. Spolia can reveal the existence of earlier building traditions while simultaneously obscuring their original plans. Material survival is a chain of movements, not a frozen site inventory. 44.18 Water architecture: tanks, ponds, wells and settlement durability Water structures are among the most socially consequential forms of architecture in the region. Tanks and ponds supplied drinking water, livestock, ritual needs, fisheries, bathing and sometimes construction material from excavated clay. Their embankments also reorganized neighbourhood space. In a monsoon climate, the challenge was not simply obtaining water but storing, draining and separating uses. Archaeological attention to tanks can reveal settlement planning that monumental surveys miss. At Vikramshila, water was integrated with the monastic complex; at other settlements, ponds became anchors of habitation and sacred geography. Later maintenance can preserve a water body while transforming its banks, so present form cannot be assumed ancient. Nevertheless, the persistence of water features makes them important clues for long-term settlement stability. A history of architecture that excludes ponds and drains would omit some of the most heavily used constructed spaces. 44.19 Roads, river crossings and the architecture of movement Settlements depend on movement. Gates, ghats, ferries, raised paths and market streets connect architecture to regional exchange. On the Ganga and its tributaries, river crossings could be more important than land roads; seasonal changes in water level altered routes and landing places. The archaeological signature of a road is often faint compared with a wall, yet the orientation of gates, mounds and religious sites can suggest circulation patterns. Historical texts and later route surveys may help, but route continuity must be tested rather than presumed. Figure 173 treats settlement as a node within flows of water, food, labour, ritual and exchange. This network approach prevents the architectural plan from being read as a self- contained object. A monastery without supply routes, a market without access, or a fort without controlled movement is only half explained. 473473 GAJENDRA THAKUR 44.20 Markets and periodic public space Not all important spaces were permanently monumental. Weekly markets, fairs and festival grounds could concentrate thousands of people while leaving limited architectural residue. Temporary stalls, bamboo structures and open grounds are archaeologically elusive, yet they were central to exchange and public culture. Their locations were shaped by route intersections, river crossings, shrines and administrative centres. Later chapters will treat fairs and markets directly, but here the architectural point is methodological: permanence is not equivalent to social importance. Periodic public space challenges a monument-centred history because its infrastructure may consist of paths, shade, water access and customary plot arrangements rather than walls. When a settlement is reconstructed, open space must therefore be treated as positive evidence, not as an empty remainder between buildings. 44.21 Simraongarh: fortified urbanism in the India–Nepal borderland Simraongarh in the Nepali Tarai is indispensable to the medieval architectural history of Mithila. Archaeological investigations by the Department of Archaeology of Nepal and an Italian mission documented a large fortified urban site through surface survey, stratigraphic sections, ceramics and scientific sampling. The work is especially valuable because the investigators emphasized that chronology remained open to revision rather than forcing all deposits into a single dynastic narrative. Simraongarh’s scale, defensive works and dense archaeological landscape nevertheless demonstrate substantial urban investment in the borderland associated with Karnata power. Its position also makes a wider point: political and architectural history cannot be divided neatly by the modern India–Nepal boundary. Materials, artisans, courtly practices and religious networks moved across the Tarai long before the present frontier. Figure 174 — Four archaeological windows on architecture and settlement 44.22 Simraongarh chronology and the discipline of uncertainty A famous capital attracts precise stories, but archaeology often produces ranges, phases and unresolved sequences. The Simraongarh investigations explicitly noted disturbance, re-excavation and the preliminary nature of early radiocarbon and ceramic evidence. That caution should be retained. Dynastic history can provide a chronological frame for Karnata rule; archaeological layers can date occupation and rebuilding; neither should be made to say more than it can support. The appropriate question is how the fortified urban HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II landscape developed through time, which sectors were occupied, how the ramparts changed, and how the settlement related to surrounding routes and rural production. This discipline of uncertainty is a strength, not a weakness. It keeps future excavation capable of changing the narrative rather than forcing new evidence into an inherited story. 44.23 Courts, palaces and the problem of architectural attribution Courtly centres must have included residences, audience spaces, storage, service quarters and defensive or ceremonial boundaries, yet palaces are often among the most confidently named and least securely demonstrated structures in later local tradition. A mound called a rājā’s palace may indeed preserve elite occupation, but the identification requires excavation or documentary convergence. Timber superstructures and repeated rebuilding further complicate recovery. Rather than inventing palace plans, social history can proceed from firmer evidence: concentration of prestige goods, unusually substantial foundations, defensive placement, inscriptions, or texts that securely name a courtly location. The later Vidyāpati Ṭhakkura’s documentary and courtly world, discussed in Chapters 35 and 38, presupposes material settings of administration and patronage, but those texts do not automatically identify surviving walls. Literary court and excavated palace must remain separate until evidence joins them. 44.24 Karnata and Oiniwar settlement without dynastic determinism The Karnata and Oiniwar periods altered political centres, patronage networks and learned institutions, but settlement change cannot be inferred mechanically from changes of dynasty. Villages continued across political ruptures; markets could grow for ecological or commercial reasons; religious sites could outlast their patrons. Chapter 33 reconstructed social economy from multiple evidence streams, and the same principle applies here. Architecture should be tied to dated building episodes and settlement phases, not assigned to a dynasty merely because the dates overlap. Simraongarh provides strong material evidence for a major Karnata-associated urban centre; elsewhere the record may be thinner. Oiniwar courtly culture can be studied through texts and patronage, but precise architectural attribution remains a separate question. This protects the volume from converting political chronology into an architectural template. 44.25 Sacred clusters and settlement nucleation Temples, monasteries, shrines, tanks and pilgrimage routes can attract settlement around them, but the direction of causation varies. A shrine may be founded at an existing settlement, or settlement may intensify around an established sacred centre. Sacred clusters also accumulate chronologically: an old image, a rebuilt temple, a new pavilion and a modern road may coexist in one precinct. The historian must disentangle these layers. In Mithila, where sacred geography is dense and living traditions are strong, the temptation to treat a present cluster as a single ancient complex is particularly risky. Architectural survey should therefore document fabric, repairs, inscriptions, reused pieces and local narratives separately. The result is richer than a simple date because it reveals how communities continually remake sacred space. 44.26 Craft production and the built environment Architecture depends on craft sectors that often leave their own settlement signatures. Brick kilns, lime preparation, carpentry, metalworking, pottery and terracotta production require fuel, water, raw material and work space. Large building projects can concentrate artisans temporarily or create durable craft quarters. Conversely, craft production may occur at settlement edges because of smoke, fire risk or access to clay. Archaeological waste—wasters, slag, kiln debris, unfinished pieces—can therefore be as informative as 475475 GAJENDRA THAKUR finished monuments. The artistic surfaces of Vikramshila presuppose a technical infrastructure of clay preparation, moulding and firing; fortified walls presuppose mass earthmoving and brick supply. Material culture becomes social history when these invisible labour systems are restored to the narrative. 44.27 Housing, gender and household work: what architecture can and cannot show Domestic space structured cooking, childcare, storage, craft, animal care and privacy, all of which had gendered dimensions. Yet the archaeological record rarely labels a room “women’s space” or “men’s space.” Functional interpretation must therefore proceed from installations, artefact concentrations, access patterns and comparative evidence rather than stereotype. Later ethnographic house forms can generate hypotheses about courtyards or secluded zones, but they cannot be projected unchanged into early or medieval periods. This is particularly important in a region where modern social classification has often been read backward. Architecture may reveal differential access and work areas, but social identity must be independently established. The chapter therefore uses domestic space to ask questions rather than to manufacture certainty. 44.28 Settlement hierarchy: hamlet, village, market, monastery, town and capital Historical landscapes contain settlements of different scales and functions. A large fortified site should not make surrounding villages disappear from analysis, and a monastery may function as a major economic node without being a city in the same sense as a political capital. Settlement hierarchy can be reconstructed through site size, density, craft specialization, imported goods, fortification, institutional architecture and route position. The categories are analytical, not rigid. A village can acquire a market; a shrine can generate a town; a political centre can shrink while remaining sacred. Figure 174 compares four sites precisely to resist a single model. Vaishali, Balirajgarh, Vikramshila and Simraongarh preserve different combinations of defence, residence, ritual and institutional planning. Their differences are historically productive. 44.29 Archaeological method: survey before reconstruction Future research on Mithila–Vajji–Anga architecture will benefit most from systematic landscape survey tied to excavation, remote sensing, ceramic study, archaeobotany and geomorphology. The region contains many mounds and reused architectural fragments whose relationships remain poorly understood. Mapping site distributions can clarify route corridors and settlement hierarchies before spectacular identifications are proposed. Scientific dating should be prioritized where secure samples exist; brick and ceramic typologies should remain anchored to stratigraphy; oral history should be recorded as evidence of memory and site use. The key methodological principle is convergence. Figure 175 formalizes the evidence firewall used throughout this volume: excavated sequence, standing fabric, texts or inscriptions, and living tradition must be distinguished before they are combined. This approach leaves room for discovery and protects the narrative against inherited certainty. HISTORY OF MITHILA, VAJJI & ANGA — VOLUME II Figure 175 — Evidence firewall for architectural history 44.30 Conclusion: built landscapes as changing social systems Architecture and settlement in Mithila, Vajji and Anga were products of ecology, labour, authority, religious practice and exchange. The alluvial plains favoured continual rebuilding and erased much domestic fabric; fortified sites such as Vaishali and Balirajgarh reveal large investments in protected space; Vikramshila demonstrates how a planned monastic complex organized residence, worship and learning; Simraongarh places medieval urbanism across the modern India–Nepal border; the Bhagalpur zone adds rock-cut and stone traditions to a predominantly brick-built history. No single architectural style or settlement type defines the region. What unites the evidence is adaptation: to rivers, materials, routes, institutions and changing political economies. The most reliable history therefore begins from construction sequence and spatial function, then connects them to social life. It does not begin from a famous name and search for walls to fit it. Table 44.1 — Evidence domains for architecture and settlement Evidence domain What it can establish Principal limit Stratified excavation Construction sequence, Limited exposure may not rebuilding, occupation and represent the whole settlement abandonment Ramparts and walls Material, engineering, enclosed Function and political space and labour investment attribution require separate evidence Brick dimensions and Technical practice and relative Reuse and local variation can masonry chronology when stratified mislead typological dating Domestic traces Floors, hearths, storage, posts and Perishable construction is easily activity areas erased or under-excavated Religious complexes Spatial organization of ritual, Monument survival can residence and patronage exaggerate religious share of built space Water structures Storage, drainage, ritual use and Present form may reflect neighbourhood organization repeated later repair Routes, gates and ghats Movement, market access and River migration can shift original controlled circulation relationships Texts and inscriptions Names, patrons, functions and A literary place-name does not dates when explicit identify a surviving wall by itself 477477 GAJENDRA THAKUR Evidence domain What it can establish Principal limit Local tradition Memory, sacred continuity and Tradition is not self-dating clues for field research archaeological proof Regional survey Settlement hierarchy, clustering Surface visibility varies with and route corridors cultivation and geomorphology