On this page
English only · Odia translation in progress
Chapter 1: Before the Temples
In 1939, a young anthropologist named Nirmal Kumar Bose walked along the banks of the Burhabalang River in Mayurbhanj district, at the edge of what is now the Similipal Tiger Reserve, and picked up a stone. It was not a remarkable stone to the untrained eye — palm-sized, made of quartzite, with one end crudely flaked to form a cutting edge. But Bose, working alongside D. Sen from the University of Calcutta, recognized it for what it was: a hand axe. A Lower Paleolithic hand axe, shaped by a human hand perhaps three hundred thousand years ago, possibly longer. Over the next three years, Bose and Sen excavated the site at Kuliana and pulled from the earth a collection of heavy-duty tools — choppers, cleavers, flake tools, hand axes — that established Kuliana as the first excavated Paleolithic site in India. Not one of the first. The first.
This fact bears repeating, because it is almost never mentioned in mainstream accounts of Indian civilization. The first scientifically excavated evidence of Stone Age human presence in the subcontinent came not from the Gangetic plain, not from the Indus Valley, not from the Deccan plateau, but from the forested highlands of northern Odisha. And this was only the beginning. Since Bose’s excavation, the Burhabalang River basin and its tributaries have yielded over a thousand stone artifacts spanning the Lower Paleolithic through the Mesolithic, a continuous record of human habitation stretching across tens of thousands of years. The nearby site of Kuchai produced a microlithic horizon — tiny, precisely flaked blades and points — beneath a Neolithic layer, while Baidypur became the first reported Neolithic site in the state, with ground stone celts and cord-impressed pottery suggesting settled agricultural communities by roughly 2500 BCE.
What Bose found at Kuliana was not just a collection of old rocks. It was proof that the land we now call Odisha has been inhabited as long as any land on earth. That the cultures which eventually produced the temples, the scripts, the maritime empires, and the Jagannath synthesis did not arrive from elsewhere into an empty landscape. They grew on top of something already ancient. And that ancient something — the deep substrate of human life in this geography — is what this chapter is about.
The Deep Time
The geological canvas matters. Odisha’s landscape is not an accident; it determined who could live here, how they lived, and what they eventually became. The state sits at a meeting point: the Eastern Ghats running north-south through its western highlands, the coastal plain widening from the Mahanadi delta to the Chilika lagoon, and the Chota Nagpur plateau extending into its northern districts. Each zone offered different resources, different dangers, different possibilities. The highlands had caves, rock shelters, stone for toolmaking, forests dense with game. The coastal plain had alluvial soil, rivers for irrigation and transport, access to the sea. The plateau had laterite soils, mineral deposits, and deciduous forests that would sustain human communities for millennia.
The Paleolithic evidence clusters, unsurprisingly, in the highlands. At Kuliana in Mayurbhanj, the tools are predominantly quartzite — Lower Paleolithic choppers and hand axes that belong to the Acheulean tradition, the same broad technological culture that humans practiced from East Africa to Western Europe. Recent explorations in the area have documented 115 newly discovered stone tools from Kuliana, Kalabadia, and Kendudiha, ranging from the Lower Paleolithic to the Mesolithic, predominantly quartzite with some quartz and sandstone. The toolmakers were selecting specific stone types for specific purposes. This is not random bashing. It is engineering, at the most elemental level.
Move west and south, and the evidence shifts from open-air tool scatters to the intimate spaces of rock shelters. At Ushakothi in Sundargarh district, a cave measuring forty-six metres in height and sixty-one metres in width sits at the foot of Maheswar hill, deep inside the forest. The paintings on its walls range from small geometrical and floral patterns to large animal motifs — deer, cattle, stag, sambar — and human figures shown hunting, domesticating animals, fighting, and dancing. Two grinding holes on the floor of the rock shelter and the collection of microliths from the surrounding slopes confirm that this was not merely a canvas but a habitation site. People lived here, processed food here, and painted the walls with images of their world using red and white ochre pigments.
Further south, in what is now Nuapada district, the Yogimatha rock shelters near Khariar contain paintings assigned by archaeologists to approximately the tenth millennium BCE — about twelve thousand years ago. The most significant image is a bull followed by a cow, a calf, and a human figure, which researchers interpret as an early depiction of animal domestication. If that reading is correct, it places the people of western Odisha among the earliest communities in the Indian subcontinent to make the conceptual leap from hunting wild animals to living alongside them. At Gudahandi in neighbouring Kalahandi, the cave paintings may date to the fifteenth millennium BCE — seventeen thousand years ago — making them among the oldest artistic expressions in all of eastern India. The figures are painted in red, blue, and black: humanoid shapes, geometric designs of squares and rectangles, circular wheels with spokes, horse-shoe forms, and the outlines of deer.
Then there is Vikramkhol, near Belpahar in Jharsuguda district, where a rock face measuring thirty-five feet by seven feet carries incised markings that have generated scholarly argument since their discovery in the 1930s. Dr. K.P. Jayaswal, who first studied them, tentatively dated the inscriptions to around 1500 BCE and argued they represented a transitional script — a bridge between the pictographic writing of Mohenjo-daro and the Brahmi script that would later become the standard writing system of the subcontinent. Other scholars, including Richard Salomon, have dismissed them as non-literate rock carvings. The debate remains unresolved. What is not debated is that someone, very long ago, sat at this rock face in a forest in western Odisha and deliberately inscribed marks that they intended to be meaningful. Whether those marks constitute “writing” in the formal sense is less important than the cognitive act they represent: the impulse to record, to communicate across time, to leave a mark that outlasts the maker.
What emerges from all this scattered evidence — the Kuliana hand axes, the Ushakothi paintings, the Yogimatha domestication scene, the Gudahandi geometric art, the Vikramkhol inscriptions — is a picture of continuous human presence across an enormous span of time. Not a single culture, not a single people, but a succession of communities adapting to the same landscape, inheriting and modifying the technologies and practices of those who came before them. The deep time of Odisha is not empty. It is layered, like geological strata, and the layers are human.
The Austric Foundation
If the Paleolithic and Mesolithic evidence tells us that people have lived in this land for hundreds of thousands of years, the next question is: who were they? Not in the genetic sense — ancient DNA studies from this region are still sparse — but in the cultural and linguistic sense. Whose languages, whose agricultural practices, whose social structures, whose ways of understanding the world formed the foundation layer of what would eventually become Odia civilization?
The most likely answer, supported by linguistic reconstruction, comparative ethnography, and increasingly by population genetics, is: the Munda peoples.
The Munda languages — Mundari, Santali, Ho, Kharia, Juang, Bonda (Remo), Bhumij, and others — belong to the Austroasiatic language family. This is the same family that includes Khmer in Cambodia, Mon in Myanmar, and Vietnamese. The connection is not fanciful; it is established by systematic comparison of vocabulary and grammatical structures across these languages. The linguist Paul Sidwell has proposed that Munda speakers arrived on the coast of modern Odisha approximately four thousand to thirty-five hundred years ago — around 2000 to 1500 BCE — having crossed the Bay of Bengal from Southeast Asia. Genomic analyses of Munda populations broadly support this timeline, revealing admixture with Southeast Asian ancestry components dating to approximately 3,800 years before the present.
This is a startling idea: that the oldest continuous cultural layer in Odisha may have arrived by sea from the east, not by land from the northwest. The Munda homeland, in this reconstruction, is not the Chota Nagpur plateau where the largest Munda-speaking populations now live, but the Mahanadi delta coast, from which they spread inland over centuries.
But here is the crucial nuance. Whether the Munda speakers brought entirely new cultural practices to an uninhabited landscape (they did not — the Paleolithic and Mesolithic evidence proves prior habitation) or whether they arrived into a land already populated and their cultural practices fused with what was already there, the Munda cultural imprint on Odisha is so deep and so pervasive that it constitutes the bedrock upon which everything else was built. Consider what the Munda communities contributed — not as a historical curiosity, but as living practices that persist in forms most Odias no longer recognize as having Munda origins.
Rice. The centrality of rice to Odia life — the culture, the cuisine, the economy, the ritual calendar — needs no elaboration. But the technique of wet-rice paddy cultivation, the system of bunded fields and controlled water management that makes intensive rice farming possible, has deep Austroasiatic roots. The Munda speakers carried rice cultivation knowledge from Southeast Asia, where the domestication of Oryza sativa began. At Golbai Sasan in Khurda district, archaeobotanical analysis of a Neolithic-Chalcolithic settlement dating to approximately 3,500 years before the present confirmed that residents grew rice alongside browntop millet, mung bean, horsegram, and pigeon pea. This is not incidental farming. It is a diversified agricultural system, the kind that develops over generations of accumulated knowledge. The Munda contribution to this system — if the maritime arrival hypothesis is correct — was not the invention of rice farming in Odisha but its intensification, the introduction of techniques that transformed rice from one crop among many into the centre of the agricultural economy.
The sacred grove. The sarna — a patch of forest, often dominated by sal trees, left entirely untouched because it belongs to the spirits — is one of the most distinctive features of Munda religious practice. The sarna is not a conservation area in the modern sense. It is a place where the non-human world is sovereign, where human authority does not apply, where the spirits of the land — Marang Buru, Jaher Era, the ancestral spirits — reside and are propitiated. The village priest, the pahan, conducts rituals at the sarna on festival days, but no one cuts wood there, no one ploughs there, no one builds there. The grove is inviolate.
This practice is so widespread among tribal communities across Odisha — not just Munda speakers but Kondh, Saora, Gadaba, Gond — that it cannot be attributed to a single source. But the Munda form of it, the sarna proper, is the most systematized and the most institutionally embedded. And here is the point that mainstream Odia culture prefers not to examine too closely: the reverence for groves, for specific trees, for the idea that certain patches of forest are sacred and must not be touched, pervades Hindu practice in Odisha in ways that have no Vedic origin. The temple gardens, the sacred peepal trees, the village deity groves found across the coastal plain — these are, in structural terms, sarna practices dressed in Brahmanical clothing. The underlying logic — that the land itself has agency, that certain spaces belong to forces older than human settlement — is Austric, not Sanskritic.
Totemism and clan organization. The Munda communities are organized into totemic clans called kili. The Nag (Snake) kili, the Bagh (Tiger) kili, the Hansda (Eel) kili — each clan traces its identity to a non-human ancestor or a natural feature, and marriage within the clan is prohibited. Among the Dhelki Kharia, the eight clans include Soren (rock), Muru (tortoise), Samad (deer), and others. This system of social organization — defining kinship through connection to the natural world rather than through lineage from a human ancestor — is fundamentally different from the caste system that would later overlay it. But traces of totemism persist in Odia Hindu society in ways that are rarely acknowledged: village deity cults organized around animal symbolism, the snake worship of Naga Panchami that goes far deeper than its Puranic justification, the clan names among many Odia families that derive from natural objects rather than from gotras.
The village council. The parha system of the Mundas is a form of governance so complete that it functioned as a legislative, executive, and judicial body long before the British established their courts and revenue offices in the region. Several villages together formed a cluster called parha, with the head of a group of villages called manaki and the overall chief known as parha raja. Decisions were made by collective deliberation, disputes resolved through customary law, and resources managed communally. The khuntkatti land system — joint clan ownership of land and forests, where no individual could rise above the community and land could not be treated as a commodity — was the economic foundation of this governance.
This matters because the village councils of rural Odisha, the gram sabhas that the Supreme Court invoked in the Niyamgiri case, are not colonial inventions. They are descendants — attenuated, modified, overlaid with statutory authority — of a governance tradition that predates the Indian state by millennia. The instinct for communal decision-making, the suspicion of individual accumulation, the sense that the village is a collective entity with collective rights — these are Munda structural principles that survived Brahmanization, survived colonialism, and remain visible in the political culture of rural Odisha today.
Language. The direct lexical contributions of Munda languages to Odia remain understudied compared to, say, the well-documented Sanskrit borrowings. But the influence is there, particularly in agricultural vocabulary, in names for plants and animals endemic to the region, in kinship terms, and in the phonological texture of spoken Odia itself. Many Munda terms appear in Vedic texts compiled between 1500 and 500 BCE, suggesting that Munda speakers were present in the broader eastern Indian landscape early enough to influence even the language of the Indo-Aryan newcomers. The Odia language’s distinct character — its phonology shaped by centuries of interaction with tribal tongues, its vocabulary drawing from indigenous sources rather than Persian and Arabic — reflects this deep Austroasiatic substrate.
The Juang of Keonjhar deserve particular mention. They are an Austroasiatic people endemic to Odisha — found nowhere else on earth — who speak a language belonging to the Munda family. Their oral tradition identifies the Gonasika Hills as their primordial homeland. Historically known as patuas, “leaf-wearers,” because their women wore girdles made of leaves rather than woven cloth, the Juang represent what may be the oldest continuously inhabited cultural niche in Odisha. Their very existence — a tiny population of perhaps fifty thousand, speaking an unwritten Munda language, practicing forms of shifting cultivation and forest management that predate any historical record — is evidence that the Austric foundation is not merely a theoretical layer to be reconstructed from linguistic analysis. It is alive, fragile, and still present.
The Dravidian Layer
On top of the Austric foundation — or perhaps alongside it, because cultural contact is rarely sequential and neat — lies another stratum: the Dravidian.
The Kondh are the largest tribal group in Odisha. They speak Kui and its southern dialect Kuvi, languages that belong not to the Austroasiatic family but to the Dravidian — the same family as Tamil, Telugu, Kannada, and Malayalam. The Kondh homeland is the highland interior of Odisha: Kandhamal, Kalahandi, Koraput, Rayagada. The terrain is dramatically different from the Munda-speaking lowlands and plateaus of the north. This is Eastern Ghats country — steep valleys, dense forest, laterite ridges, and a landscape that has historically resisted easy access from the coastal plain. The eighty percent forest cover of Kandhamal district is not an accident of modern conservation. It is the legacy of communities who managed these forests as living systems for centuries.
The Gadaba, the Paraja, the Saora (who are linguistically Austroasiatic but culturally share many Dravidian highland practices), the Pengo, the Konda — these communities form a cultural zone in southern and western Odisha that is distinctly different from the Munda heartland of the north. Their languages — Kui, Kuvi, Gondi, Ollari, Konda — have no commonality with Odia, as linguists have noted; they are structurally Dravidian, with grammatical features, phonological patterns, and core vocabulary that link them to the languages of peninsular India rather than to the Indo-Aryan north.
What did the Dravidian communities contribute to the cultural grammar of Odisha? Several things that are so deeply embedded they have become invisible.
Terraced agriculture. The Kondh system of highland farming — cutting steps into steep hillsides to create flat planting surfaces, managing water flow through gravity-fed channels, rotating crops across elevation gradients — is a sophisticated engineering response to difficult terrain. The Dongria Kondh of Niyamgiri maintain dongar orchards on hill slopes growing over a hundred varieties of crops: turmeric, ginger, pineapple, oranges, bananas, jackfruit, and several varieties of millet. These are not “primitive” gardens. They are managed agroforestry systems that mimic the layered structure of the surrounding forest, blending cultivated crops with semi-wild species in arrangements that function simultaneously as food production, soil conservation, water management, and biodiversity preservation. Western agronomy has a name for this approach — it is called “permaculture,” and it was invented in the 1970s by Australian academics who were, in effect, rediscovering what the Kondh had been practicing for millennia.
The earth goddess and the meriah. Of all the Kondh practices that collided with British colonialism, none generated more horror — or more self-serving justification for imperial intervention — than the meriah sacrifice. The meriah was a human being, often purchased from Pano traders, who was sacrificed to Dharni Penu (or Tana Penu), the earth goddess, to ensure the fertility of the soil. The victim was called meriah in Odia and toki or keddi in Kui. The ritual was embedded in the regional political system, performed by specific ritual specialists at the expense of local rajas and patras.
The British discovered the practice in the 1830s and launched three decades of military “meriah campaigns” to suppress it. Samuel Macpherson was sent to lead these campaigns in 1835 and drafted detailed reports on Kondh life, viewing them through the classical literary and evolutionary lenses of his era. The Kondh resistance to British suppression — the Khond Uprisings of 1837 and 1856, led by Chakra Bisoi — was not merely a defence of human sacrifice. It was a defence of political and religious autonomy against a colonial power that used moral outrage as the entering wedge for territorial control.
What the meriah controversy reveals, beneath the colonial drama, is something important about pre-Hindu religious practice in this region: the centrality of blood sacrifice to agricultural fertility, the personhood attributed to the earth, the transactional relationship between human communities and the land they cultivate. The earth is not an object to be worked. She is a being to be propitiated. And propitiation requires giving something precious — the most precious thing of all — back to her. When the British forced the substitution of buffalo for human sacrifice, the structural logic of the ritual remained intact. The buffalo sacrifice that continues among Kondh communities today is the meriah in attenuated form. And the underlying theology — that the earth is alive, that she must be fed, that agricultural abundance is a gift that requires reciprocation — persists throughout rural Odisha in harvest festivals, in the worship of Dharni Mata, in the ritual first-ploughing that precedes the sowing season. These are not Hindu practices in origin. They are Kondh practices absorbed into the Hindu frame.
Linguistic influence. The Dravidian impact on Odia is more extensively documented than the Austroasiatic substrate, partly because the interaction has been longer and more intensive. Retroflex consonants — sounds produced with the tongue curled back against the palate — are more prominent in Odia than in most other Indo-Aryan languages, a feature that linguists attribute to Dravidian influence. Words like poda (burnt), gudi (temple), and badi (a type of dried food) have Dravidian roots. The use of postpositions rather than prepositions in Odia syntax mirrors Dravidian grammatical structure. Certain suffixes used to form nouns and adjectives in Odia — patterns involving -ka and -ta — are thought to be borrowed from or inspired by Dravidian models, specifically from Telugu and Kannada contact. The Ollari Gadaba language, a tribal Dravidian language spoken in the border areas of Andhra Pradesh and Odisha, shows extensive structural influence from Odia, evidence of centuries of two-way linguistic exchange.
The Dravidian layer in Odisha is not a replacement of the Austric layer. It is an addition. In the northern districts — Mayurbhanj, Keonjhar, Sundargarh — the dominant tribal languages are Munda. In the southern and western highlands — Koraput, Rayagada, Kandhamal, Kalahandi — the dominant tribal languages are Dravidian. The two zones interpenetrate at the margins, and in both zones, the later arrival of Indo-Aryan Odia overlaid both substrates without erasing either. What modern Odia is, linguistically and culturally, is a three-layer structure: Austric at the base, Dravidian in the middle, Indo-Aryan on top. Remove any layer and the whole edifice becomes incomprehensible.
Megalithic Odisha
Scattered across the highlands of Odisha — in Mayurbhanj, Keonjhar, Sundargarh, Koraput, Nuapada — stand stones that no one remembers erecting. Dolmens: horizontal slabs resting on vertical uprights, creating chambers that once held the dead. Menhirs: single standing stones, some taller than a person, driven into the earth to mark a presence, a memory, a claim. Stone circles: rings of boulders arranged with deliberate geometry, their purpose debated by archaeologists but understood intuitively by the tribal communities who still live among them.
The megalithic culture of Odisha is one of the least studied and most extraordinary archaeological phenomena in eastern India. The sites are numerous — hundreds of them, scattered across the forested highlands, many unexcavated, some undocumented, most unknown to anyone except the local tribal communities who regard them as the work of ancestors. The dating is uncertain for most sites, but the broad consensus places the megalithic period in eastern India between roughly 1500 BCE and 500 BCE, overlapping with the Iron Age and in many cases directly associated with iron-working communities.
What makes the Odisha megaliths remarkable is not just their age or their number. It is the fact that the tradition is not dead. Among the Gadaba of Koraput, menhirs are still erected during the annual memorial feast called Gota Mela and during the Ongongota (Crab festival). These are not reconstructions or heritage performances. They are living rituals in which a community pools its resources to quarry, transport, and erect a stone pillar in memory of a deceased member. The Gadaba menhirs are located near rice fields, and tribal belief connects them explicitly with agricultural fertility — the dead, marked by stone, continue to participate in the productive life of the community. The Bonda create stone walls and circles for communal assemblies. The Gond of Nuapada maintain megalithic traditions with distinct customs connected to beliefs about life after death.
This is a phenomenon that demands careful attention. In most of the world, megalithic cultures are studied through archaeology alone — the builders are gone, their practices reconstructed from material remains. In Odisha, the builders are still building. The Gadaba erecting a menhir in 2025 is performing an act that is structurally identical to the act of the unknown people who erected the stone circles of Keonjhar three thousand years ago. The continuity is not metaphorical. It is direct.
What do the megaliths tell us about the societies that built them?
First, social organization. Erecting a megalith is a collective enterprise. A menhir of any significant size requires quarrying stone from a source that may be kilometres from the erection site, transporting it without wheeled vehicles across rough terrain, and installing it vertically in a prepared pit. This requires coordination, surplus labour, and communal decision-making. A society that builds megaliths is not a collection of isolated families. It is a structured community with the capacity for collective action — a parha, in effect, before the word existed.
Second, beliefs about death. The megalithic burials found across Odisha — dolmens containing human remains, cist burials surrounded by grave goods — indicate a belief that the dead retain some form of agency or presence. They are not simply gone. They must be housed, marked, remembered. The stone is permanent in a way that wood and flesh are not; it is a commitment to keeping the dead present in the landscape of the living. This is ancestor worship in its most literal form — not the propitiation of vaguely defined “ancestors” in a ritual context, but the physical installation of the dead in the productive landscape of the community. The Gadaba menhir stands near the rice field because the ancestor it marks is understood to participate in the field’s fertility. The dead feed the living. The stone makes the connection visible.
Third, knowledge of iron. The megalithic period in Odisha overlaps precisely with the Iron Age. At Badmal-Asurgarh in Sambalpur district, a fortified Iron Age settlement dating from the early first millennium BCE to the second century BCE has yielded extensive evidence of iron smelting and smithing — furnaces, forges, finished iron objects, ore, slag, and ingots. The site sits on the left bank of the Harihar stream, a tributary of the Mahanadi, and covers about four hectares enclosed by a lozenge-shaped earth rampart with three passageways. The name “Asurgarh” — fort of the Asur — connects the site linguistically to the Asur people, a Munda community historically identified as India’s earliest iron smelters. The Asur smelting technique was distinctive: they built furnaces fueled by charcoal from sal trees and produced iron of reportedly exceptional quality — rust-resistant, according to some accounts. Women sang to the furnace during the smelting process, a practice that linked metallurgy to fertility ritual in ways that the industrial mind finds difficult to process but that makes perfect sense within an animist framework where the transformation of ore into metal is understood as a kind of birth.
The iron smelters of the Mahanadi valley were not peripheral actors in Indian technological history. In eastern India, iron smelting technology with proper furnace management and advanced reduction procedures evolved around 700 BCE. The communities engaged in this work — Asur, Agaria, Lohar, Birjia, Kol, Ho — were the bearers of knowledge that would eventually make the Kalinga kingdom a military power, its iron-tipped weapons and tools the product of a metallurgical tradition stretching back into the unnamed centuries. When the Tata family built their steel empire in Jamshedpur in the early twentieth century, they were, in a sense, industrializing a skill that the Asur had practiced on this same plateau for three thousand years. The Tatas’ industrial smelting practice effectively made the Asur’s traditional process redundant. The Forest Conservation Act of 1980, which restricted the felling of sal trees, severed the last connection between the Asur and their ancestral craft. A metallurgical lineage spanning three millennia ended not with a war or a natural disaster but with a bureaucratic regulation.
Proto-Urban Odisha
Five kilometres southeast of modern Bhubaneswar, buried under agricultural fields and encroached upon by the expanding city, lie the remains of a fortified settlement that challenges conventional narratives about where urban civilization emerged in eastern India. Sisupalgarh — the name means “fort of Sisupala,” connecting it, probably apocryphally, to the Mahabharata — was discovered in 1948 by the archaeologist B.B. Lal, one of the most important figures in Indian archaeology. Excavations between 2005 and 2009 by Monica L. Smith of UCLA and R.K. Mohanty of Deccan College revealed a city far older and more sophisticated than previous estimates had suggested.
The earliest occupation at Sisupalgarh dates to the seventh or sixth century BCE — contemporary with the earliest phases of Taxila in the northwest and the mahajanapada period in the Gangetic plain. The northern rampart has been radiocarbon dated to between 510 and 400 BCE. The city was continuously inhabited for nearly a thousand years, from the fifth century BCE to the fourth century CE.
The physical layout is striking. Sisupalgarh is roughly square, enclosed by a massive earthen rampart with eight monumental gateways — two on each side. Geophysical surveys and excavations have confirmed broad streets connecting each of the eight gateways, forming a grid. The population has been estimated at twenty to twenty-five thousand people — a genuine city, not a village or a military camp. Houses were built of stone and brick, typically with two or three rooms and a verandah in the front. A sophisticated drainage system managed water flow through the settlement.
The artifacts tell a story of connectivity. Punch-marked coins indicate participation in the monetary economy of the broader subcontinent. And then there is the extraordinary find: Roman rouletted ware. Fragments of this distinctive pottery, produced in the Mediterranean world and traded across the Indian Ocean, have been found at Sisupalgarh, confirming that this city was plugged into a trade network that stretched from the coast of Odisha to Rome. Sisupalgarh was not an isolated settlement developing in a vacuum. It was a node in a web of exchange that connected the eastern Indian coast to the wider world.
The same pattern of trade connectivity appears at Manikapatna, an ancient port site on the coast near Chilika Lake. Excavations conducted by the Odishan Institute of Maritime and Southeast Asian Studies between 1989 and 1993 uncovered Indian rouletted ware, knobbed ware, fragments of Roman amphorae, red glazed ware, Puri-Kushana coins, and a Kharosthi inscription on a potsherd. In the upper levels, Chinese celadon ware appeared in huge quantities. This is evidence of maritime trade spanning centuries — with Rome, with Sri Lanka, with China — and it aligns with the literary tradition of the sadhavas, the merchant-mariners of Kalinga who sailed to Southeast Asia and whose departure is still commemorated every year during Kartika Purnima in the festival of Bali Yatra.
At Golbai Sasan, near the northern shore of Chilika Lake, the Neolithic-Chalcolithic settlement that produced the earliest evidence of rice cultivation in Odisha also yielded woodworking adzes and other tools consistent with boat building. The connection between the people who grew rice, the people who built boats, and the people who eventually sailed those boats to Java and Sumatra and Bali may be more direct than the temporal gap between 1500 BCE and the historical period of the sadhavas would suggest. The relationship between Odisha and the sea is not a later development grafted onto an agrarian culture. It is there from the beginning, embedded in the Neolithic toolkit.
Sisupalgarh raises a discomforting question for anyone who thinks of Odisha as historically peripheral. A city of twenty-five thousand people, with eight monumental gateways, a street grid, stone houses, a drainage system, and trade connections to Rome, was thriving near the site of modern Bhubaneswar half a millennium before the Common Era. And yet this city has no literary record. No inscription describes its founding. No dynasty claims it. It exists only in the archaeological record, as a city that was real, sophisticated, and completely forgotten. Today its ramparts are being slowly consumed by Bhubaneswar’s urban sprawl. Satellite imagery shows encroachment from all sides. The Archaeological Survey of India has declared it a protected monument, but protection on paper and protection on the ground are, in Odisha as elsewhere, different things.
The Cultural Grammar
If you step back from the specifics — the stone tools, the cave paintings, the megaliths, the fortified city, the iron furnaces, the rice fields — and ask what patterns connect them, what recurring motifs emerge from ten thousand years of pre-historical evidence, the answer is a set of deep cultural principles that predate any formal religion, any state, any written text. Call it the cultural grammar of Odisha: the underlying syntax of meaning that persists beneath the changing vocabulary of successive civilizations.
The land is sacred. This is the most fundamental principle, and it is shared across every pre-Hindu cultural layer in Odisha — Austric, Dravidian, megalithic. The land is not an inert resource to be owned and exploited. It is alive. It has agency. Specific places — groves, hilltops, river sources, stone circles — are seats of power that human beings must approach with respect and propitiation. The Munda sarna, the Kondh Niyamgiri, the Gadaba menhir field, the Juang forest — all express the same underlying conviction: the earth is not ours. It is we who belong to the earth.
This principle did not disappear when Hinduism arrived. It was absorbed, reframed in Sanskritic vocabulary, and attributed to different deities. But the structure remained. The temple is, in a sense, a sarna with a roof. The sthala purana — the local sacred history that every Odia temple possesses — almost always traces the deity’s presence to a pre-existing sacredness of the site, a svayambhu (self-manifested) quality that predates human construction. The temple was built because the place was already holy. The holiness came first. And in most cases, that holiness was recognized and maintained by tribal communities long before the first Brahmin priest arrived.
Rice is the centre. The centrality of rice to Odia life extends beyond nutrition. It organizes the calendar (the agricultural year structures festival timing), the economy (land value is measured in paddy yield), the social order (rice-giving and rice-receiving mark hierarchical relationships), and the ritual life (no puja is complete without rice, no marriage proceeds without it, no death rite is finished without the final offering of cooked rice). The Nuakhai festival — the celebration of the new rice harvest, predominantly in western Odisha — is perhaps the most explicit surviving expression of a pre-Hindu agricultural ritual, in which the first grains of the new harvest are offered to the deity before anyone in the community may eat. The structure of Nuakhai — first fruits offered to a higher power, then distributed through the social hierarchy — maps precisely onto Munda harvest ritual forms.
Communal organization precedes individual authority. The parha, the gram sabha, the village council that decides by consensus rather than by vote — these are not democratic innovations in the modern sense. They are expressions of a social logic in which the community is the primary unit and individual authority is derivative, temporary, and accountable. This logic runs so deep in Odisha’s political culture that it surfaces even in contexts far removed from tribal governance. The Odia tendency toward collective political movements — the language movement of the early twentieth century, the Naveen Patnaik phenomenon (which functioned less as a cult of personality than as a collective identity project), the persistent strength of cooperative institutions in rural areas — may owe more to the Munda parha than to any modern ideology.
The dead remain. Ancestor worship — the belief that the deceased continue to participate in the life of the community, that they must be fed, remembered, and consulted — is the common thread connecting the megalithic builders to the Gadaba menhir erectors to the Odia household that observes shraddha (annual ancestor rites) and sets aside a portion of festival food for the departed. The Hindu framework reinterpreted this practice through the concept of pitri (ancestors) and embedded it in the Brahmanical ritual calendar. But the underlying logic — the dead are not gone, they are here, they have needs and powers, the boundary between living and dead is permeable — is pre-Hindu, pre-Buddhist, pre-everything. It is the oldest religious idea in this landscape, as old as the megalithic dolmens that were its first architectural expression.
The sea is a threshold. The maritime dimension of Odia identity — the sadhavas, the Bali Yatra, the Boita Bandana — is usually discussed as a feature of the historical period, linked to the Kalinga maritime empires of the early centuries CE. But the boat-building tools at Golbai Sasan, the Neolithic settlement near Chilika, push this maritime consciousness back by at least a millennium. And if the Munda speakers arrived by sea from Southeast Asia, as Sidwell’s hypothesis suggests, then the very oldest cultural layer in Odisha was, literally, a product of ocean crossing. The sea is not a boundary to be feared. It is a threshold to be crossed. This attitude, embedded in the deepest cultural memory, differentiates Odisha from the landlocked cultures of the Indian interior and connects it to the maritime civilizations of Southeast Asia in ways that the temple-and-dynasty narrative of official history tends to obscure.
The Transition
Sometime in the middle of the first millennium BCE — the dating is necessarily imprecise, because the transition was gradual and varied by region — new cultural currents began to reach the land that would become Odisha. They came not as a flood that washed away what existed before, but as rivers that merged with the existing waterways, altering the flow without eliminating the source.
The arrival of Vedic Brahmanical culture in Kalinga is attested indirectly by late Vedic literature. The Baudhayana Sutras, a Dharmashastra text compiled perhaps in the sixth or fifth century BCE, mentions Kalinga as a peripheral region — “impure” by the standards of the Brahmanical heartland in the Gangetic plain. This designation is revealing in both directions. It tells us that Brahmanical culture was aware of Kalinga but considered it culturally alien, which confirms that the pre-Brahmanical cultures described in this chapter — the Austric, the Dravidian, the megalithic — were still dominant when the Vedic gaze first turned eastward. And it tells us that the process of “Brahmanization” — the assimilation of local deities, practices, and communities into the expanding Hindu framework — was not yet complete.
Buddhism arrived early and with force. The Ashoka inscriptions at Dhauli and Jaugada, dating to the third century BCE, are among the earliest written records in Odisha and document the aftermath of the Kalinga War — the bloodbath that supposedly converted the emperor to non-violence. Buddhism thrived in Odisha for centuries, producing major monastic complexes at Ratnagiri, Udayagiri, and Lalitgiri — the “Diamond Triangle” of Buddhist sites that are among the most significant in India. Jainism, too, found fertile ground. The Hathigumpha inscription of King Kharavela, carved in a cave in the Udayagiri hills near Bhubaneswar in the second or first century BCE, opens with salutations to the Jain Arhats and Siddhas and describes a ruler who patronized Jain institutions while also — notably — tolerating other religious traditions.
But here is the critical insight about how all these layers interacted, and it is the insight that makes Odisha’s cultural history different from a simple story of conquest and replacement: nothing was ever fully erased.
When Buddhism came, it did not eliminate the tribal worship of earth spirits and ancestor deities. It layered on top. When Brahmanical Hinduism came, it did not eliminate Buddhism or the tribal practices beneath it. It layered on top. When Shaivism, Shaktism, Vaishnavism, and Tantra contended for dominance in the medieval period, the older practices persisted beneath and within each new tradition. The result is a culture of extraordinary depth — not depth in the sense of a single tradition developed over time, but depth in the geological sense: multiple distinct strata, each preserving the logic of its era, each visible to those who know how to look.
The Jagannath cult — which will be explored in the next chapter — is the supreme expression of this layering. A deity whose origins are almost certainly tribal (the daru devata, the wooden god, the lord of the forest), who was absorbed into Vaishnavism, whose temple rituals contain Tantric elements, whose festival practices echo Buddhist egalitarianism, and whose worship is open to all castes in a way that defies orthodox Brahmanical exclusion. Jagannath is not a Hindu deity who replaced earlier gods. He is a palimpsest — a text written and rewritten on the same surface, where each layer of writing remains faintly visible beneath the latest. The Austric reverence for the sacred tree is there. The Dravidian earth worship is there. The megalithic ancestor consciousness is there. The Buddhist universalism is there. The Shaiva and Vaishnava devotional frameworks are there. They coexist, not in the harmonious way that tourist brochures suggest, but in the tense, productive, contradictory way that living cultures actually work.
This is what makes Odisha’s cultural history feel different from a surface reading of Indian civilization would suggest. Most popular histories present culture as a series of replacements: the Aryans replace the Dravidians, the Hindus replace the Buddhists, the Mughals replace the Hindus, the British replace the Mughals, modernity replaces tradition. Odisha’s actual cultural record — the archaeological, linguistic, and ethnographic evidence examined in this chapter — tells a different story. It is a story of accumulation rather than replacement. Of layering rather than erasure. Of persistence rather than surrender.
The temples that would rise along the coast of Odisha in the centuries to come — Parasuramesvara, Mukteswar, Lingaraj, Rajarani, and finally the incomparable Sun Temple at Konark and the Jagannath Temple at Puri — were built on foundations far older than their stone. The foundations were not physical. They were cultural: a set of convictions about the sacredness of land, the centrality of rice, the authority of the community, the presence of the dead, and the permeability of the boundary between the human and the more-than-human world. These convictions were carried by peoples whose names we do not know, who spoke languages that have no written record, who built in wood and earth and song rather than in stone and script. The temples came later. The culture they stand on was already ancient.
Next: The Sacred Geography — How Buddhism, Jainism, Shaivism, Shaktism, and Tantra collided and fused in Odisha to produce a syncretic culture unlike any other in India, culminating in the Jagannath synthesis.
Source Research
The raw research that informs this series.