On this page
English only · Odia translation in progress
Chapter 1: The Archaeological Plate
In a village house in Koraput district, on the Eastern Ghats plateau, a Paroja tribal woman sets out the evening meal on a sal-leaf plate. There is a mound of finger millet porridge — jau — grey-brown and dense, the staple her grandmother’s grandmother knew. Beside it, a thin dal of horsegram, simmered with turmeric. A heap of foraged greens — saag from three species she can name in Parji but which have no Odia name, let alone an English one. A smear of red chili chutney. And a small portion of rice, white and polished, from the government ration shop.
She is eating four thousand years at once. The millet and horsegram descend from the same crop complex that Neolithic farmers cultivated at Golbai Sasan around 2300 BCE. The greens come from a foraging knowledge system that predates agriculture itself. The turmeric is indigenous, present in Indian food since before any text recorded its use. The chili arrived from Mexico via Portuguese traders sometime in the sixteenth or seventeenth century. And the polished rice — the newest layer, ironically also the most dominant — is a high-yield variety developed at the International Rice Research Institute in the Philippines in the 1960s, distributed through a Public Distribution System designed in Delhi, and delivered to this plateau village by a supply chain that treats it as the default grain of Indian poverty alleviation.
One plate. Five layers. Each deposited by a different historical force, each still edible, each still carrying the political charge of the moment that created it. The millet is there because Paroja farmers were too marginal to be fully captured by the Green Revolution’s seed-and-fertiliser package. The chili is there because the Bay of Bengal connected Kalinga to the same Portuguese trading network that connected Brazil to Goa. The ration rice is there because a national food policy decided, decades ago, that rice and wheat were the only grains worth distributing — a decision that nearly erased the millet from the plate entirely. To read this meal correctly is to read the compressed history of a state. The plate is a document. The question is what method of reading it requires.
The Cross-Domain Lens: Stratigraphy
In geology, stratigraphy is the science of reading rock layers. The principle is deceptively simple: in any undisturbed sequence of sedimentary rock, the oldest layer is at the bottom and the youngest at the top. Each stratum was deposited by a specific force — a river flood, a volcanic eruption, a glacial retreat, a slow accumulation of marine organisms on an ocean floor — and each preserves evidence of the conditions under which it formed. The geologist reads downward through time, peeling back layer after layer to reconstruct a history that no one witnessed and no one recorded. The rock remembers what human memory cannot.
The power of stratigraphy lies in three features. First, superposition: later layers bury earlier ones but do not erase them. The Jurassic limestone beneath a Cretaceous chalk did not disappear when the chalk was deposited; it was covered, compressed, and sometimes altered by the weight above, but it persists. Second, unconformities: gaps in the record where a layer was eroded or never deposited, marking periods of disruption that are as informative as the layers themselves. Third, intrusions: moments when a force from outside — a magma plume, a tectonic shift — cuts across existing layers, depositing new material that cross-cuts the orderly sequence and rewrites the local geology.
The modern Odia plate is a stratigraphic column. Each historical disruption deposited a layer of ingredients, techniques, and food rules onto the plate, and each layer is still present — buried, compressed, sometimes altered by the weight of what came after, but recoverable. The base layer is the millet-tuber-fish-forest complex of the Neolithic and tribal period, the oldest stratum, deposited by the earliest settled and pre-settled communities on Odia soil. Above it lies the Buddhist-Jain sediment, a vegetarian discipline that partially dissolved over centuries but left structural traces in institutional food rules and a level of vegetarianism higher than a coastal fish-eating culture would otherwise predict. The temple deposit — the rice-ghee-chhena layer of the Ganga dynasty and the Jagannath Rosaghara — is the thickest and most culturally dominant stratum, the one that codified what “Odia food” means in the popular imagination. The Columbian intrusion is precisely that: an intrusion in the geological sense, a force from outside that cut across existing layers and deposited chili, tomato, potato, and cashew into every stratum it touched, rewriting the flavour profile while leaving the structural grammar intact. The colonial fracture — the Na Anka famine, the commercial rice economy, the trauma of mass starvation in a surplus region — is an unconformity: a violent disruption that altered everything above it by making rice surplus a political obsession for the next hundred and fifty years. And the Green Revolution overlay is the most recent and thickest deposit, a homogenising layer that buried almost everything beneath it under a monoculture of high-yield rice varieties.
The metaphor is not decorative. It structures the inquiry. To understand why an Odia plate looks the way it does today, the stratigraphic method says: identify the layers, date the deposits, read the sequence, and pay attention to the unconformities — the gaps and disruptions that shaped what survived and what was lost.
Each layer was deposited by a different force, operates on a different timescale, and carries a different political charge. And unlike geological strata, which can only be read by specialists with rock hammers and mass spectrometers, these food layers can be read by anyone who sits down to eat in Odisha with their eyes open. The plate is the core sample. The meal is the reading.
The Base Layer: Before Rice Dominated
The archaeological food record of Odisha begins at a low mound on the north bank of the Malaguni river, roughly ten kilometres south of Chilika Lake in Khurda district. Golbai Sasan was systematically excavated by B.K. Sinha of the Archaeological Survey of India across four seasons between 1991 and 1996, with follow-up trench work by Deccan College researchers in 2002-2003. The site yielded a stratified sequence approximately 3.5 metres deep, spanning a Neolithic occupation into a Chalcolithic and an early Iron Age phase. Radiocarbon dates from organic samples converge on a calibrated range of roughly 2300 BCE to 800 BCE [Sinha, 2000, Puratattva; Mohanty & Mishra, 2002; Harvey et al., 2006, Antiquity].
What the people of Golbai Sasan ate is preserved in carbonised grain and animal bone. Archaeobotanical recovery yielded charred grains of rice (Oryza sativa), horsegram (Macrotyloma uniflorum), mungbean (Vigna radiata), and browntop millet (Brachiaria ramosa). The rice grains were morphometrically analysed as belonging to the cultivated complex, not a wild variant — meaning that by the mid-third millennium BCE, Kalinga’s coastal hinterland was already practising deliberate rice cultivation. The presence of horsegram and mungbean alongside rice indicates a nitrogen-fixing pulse rotation that improved soil nutrition without any explicit knowledge of nitrogen chemistry [Harvey et al., 2006; Fuller, 2011, Rice journal; Kingwell-Banham, 2018, Quaternary Science Reviews].
The faunal remains are equally revealing. The bone assemblage documents cattle (Bos indicus), buffalo, sheep, goat, pig (both domesticated and wild), chital (Axis axis), nilgai, wild boar, and riverine species. Bone tools — polished celts, fishing hooks, and harpoons — suggest an integrated hunting-fishing-herding economy supplementing the agricultural base [Sinha, 2000; Mohanty & Mishra, 2002; Badam & Prakash, 2008].
Golbai Sasan is not isolated. It sits within a cluster of Neolithic-Chalcolithic sites along the lower Mahanadi and in the Chilika hinterland. At Suabarei in Puri district, excavations in 2015-16 under Jeeban Kumar Patnaik revealed rice dominating the plant assemblage at over ninety percent of recovered macro-remains, with horsegram, mungbean, and browntop millet filling out the remainder. A single rice grain from Suabarei was directly AMS-dated to 3370-3210 calibrated years before present — roughly 1400-1260 BCE — the tightest chronological anchor available for cultivated rice on the Odisha coast [Kingwell-Banham, 2018; ASI, 2015-16]. At Gopalpur in Nayagarh district, a similar assemblage emerged: rice, pulses, millets, and a cattle-dominated bone assemblage with ground stone tools, dated to roughly 2000-1200 BCE [Mohanty, Deccan College, 2007-08]. Sankerjang in Angul district yielded megalithic burials with rice offerings around 1000-800 BCE. Kuchai in Mayurbhanj — excavated by B.K. Thapar in 1961-62 as part of the ASI’s Neolithic survey — is often cited as one of the earliest Neolithic sites in eastern India, with ground stone celts suggesting forest-clearance agriculture, though its bone preservation was extremely poor and no systematic archaeobotany has been published [Thapar, 1963, Ancient India; Basa, 2000, Man and Environment].
The synthesis is clear but must be stated with appropriate caution. The Neolithic and Chalcolithic populations of coastal and near-coastal Odisha cultivated a mixed cereal-pulse-millet farming system, supplemented by domesticated cattle, opportunistic hunting, and riverine fishing. Rice was the dominant cereal in coastal sites by 1500 BCE, but the monoculture relationship between rice and identity that later defines Odia food culture had not yet emerged. Millets and pulses were still substantial components of the cultivated repertoire. The grinding stones recovered at Golbai Sasan, Gopalpur, and Harirajpur indicate cereals and pulses being processed into flour or coarse meal, implying flatbreads, porridges, or fermented gruels [Harvey et al., 2006; Fuller, 2011; Mohanty & Mishra, 2002].
Certain important foods are under-represented in the record because of preservation bias. Fish bones rarely survive in tropical soils, and the Odia coast’s modern dependence on fish is almost certainly a deep continuity that the archaeological record can only indirectly hint at through the bone harpoons and fishhooks at Golbai Sasan. Leafy vegetables, tubers, roots, and forest greens leave almost no macro-remains. The Neolithic diet was therefore likely more diverse than the carbonised grain assemblage suggests [Fuller & Qin, 2009, World Archaeology; Kingwell-Banham, 2018].
The critical gap is the interior. The published sites cluster along the coastal plains and the lower Mahanadi valley. The Koraput-Rayagada-Malkangiri highlands — which today contain the living remnants of Odisha’s indigenous millet and pulse diversity — have almost no published Neolithic archaeobotany. The tribal food systems of southern Odisha, systems that almost certainly descend from pre-agricultural or early-agricultural communities predating the coastal rice cultures, exist as a living ethnographic reality with no corresponding dirt record. Any claim about “three thousand years of continuous tribal food tradition” in Koraput is an ethnographic and linguistic claim, not an archaeological one [Basa, 2011; Kingwell-Banham, 2018; MSSRF Koraput reports, 2002-2012]. This silence matters. As the Tribal Odisha series documents, the tribal communities of the interior highlands — the Kondh, the Saora, the Gadaba, the Bonda, the Paroja — maintain food systems built on millets, tubers, wild greens, forest game, and an encyclopaedic knowledge of the forest’s seasonal offerings. These systems represent what is almost certainly the oldest continuous layer of Odia food culture, a base layer deposited before rice dominated the coastal plains, before any temple was built, before any empire arrived. The base layer never disappeared. It was buried under later deposits. In the highlands, it is still the surface.
The Buddhist-Jain Sediment
In 261 BCE, the Mauryan emperor Ashoka invaded Kalinga. His own edicts record the devastation: a hundred thousand killed, a hundred and fifty thousand deported, and many times that number dead from aftermath. What followed was one of the most consequential dietary shifts in South Asian history, and it began on Odia soil.
Rock Edict I records that in the royal kitchen of Ashoka, many hundreds of thousands of animals had formerly been killed daily for curry, and that this figure had been reduced to two peafowl and one deer, and that even this would eventually stop [Thapar, 1961/2012, Asoka and the Decline of the Mauryas; Sircar, 1979, Asokan Studies]. The edict is a window into what pre-Buddhist elite consumption looked like in an eastern Indian political centre: meat-heavy, ceremonial, extravagant. The conversion to Buddhism after the Kalinga war imposed vegetarian discipline on the royal kitchen and, through the empire’s administrative reach, on institutional food practice across Kalinga.
The Jain presence was even older in some respects. The Udayagiri-Khandagiri caves near Bhubaneswar, carved during the reign of King Kharavela in the first or second century BCE, were residential quarters for Jain monks. The Hathigumpha inscription — seventeen lines of Prakrit carved into rock — describes Kharavela’s restoration of canals that had not been used for three hundred years, a reference to Nanda-period irrigation that implies organised wet-rice agriculture from at least the fourth century BCE [Jayaswal & Banerji, 1929-30, Epigraphia Indica; Sahu, 1984]. Jain monastic diet was governed by the ahimsa code: no meat, no fish, no eggs, no root vegetables after certain observances, no food after sunset. These restrictions, practised by a community that held political and cultural influence in Kalinga for centuries, established one of the two great vegetarian food traditions that would later shape Odia institutional cuisine [Mitra, 1960, Jaina Monuments of Udayagiri and Khandagiri; Sahu, 1964, History of Buddhism in Orissa].
The Buddhist monastic complex of Ratnagiri-Lalitgiri-Udayagiri — the “diamond triangle” of Odishan Buddhism in Jajpur district — flourished from the fifth or sixth century CE into the twelfth or thirteenth century CE. Excavations at Ratnagiri between 1958 and 1961 under Debala Mitra revealed a mahavihara with multiple stupas, two quadrangular monasteries with refectories, and extensive kitchen and storage facilities. At Lalitgiri, ASI excavations exposed grain storage rooms and water cisterns. Udayagiri, the largest of the three, contained clear evidence of large-scale food preparation: stone-paved kitchens, ash deposits, grinding stones, and ceramic storage vessels [Mitra, 1981-83, ASI Memoir 80; Patnaik, 2006, Lalitgiri: A Buddhist Centre; Chauley, 2000, ASI].
The material evidence indicates monasteries operating at a scale that required organised food procurement. Ratnagiri in its mature phase supported several hundred monks. Tibetan pilgrim accounts mention the monastery’s sustenance through lay donation of rice, ghee, and pulses. The Buddhist refectory — the bhattagara — was a standardised institution across South Asian monastic culture: a midday meal of rice, pulses, vegetables, ghee, and seasonal fruit, with no evening solids, and strict prohibitions on meat and alcohol [Sahu, 1964; Ray, 1994, The Winds of Change; Mitra, 1981-83; Beal, 1884, Si-yu-ki].
In stratigraphic terms, the Buddhist-Jain sediment is a layer deposited over roughly a millennium — from Ashoka’s conversion through the decline of the Ratnagiri monastery complex in the thirteenth century CE. It is not a thick layer in the geological sense: it did not replace the base-layer diet of the general population, which remained firmly fish-and-rice. What it deposited was a parallel institutional cuisine, a repertoire of techniques for making vegetarian food satisfying and complex without the flavour shortcuts of meat. The slow cooking of pulses with vegetables. The spice tempering — blooming whole spices in hot oil — to create aromatic depth. The emphasis on textural contrast: crisp and soft, smooth and chunky, in a single meal.
The paradox that the archaeological-institutional record establishes is this: the same geography that sustained an intensely fish-eating coastal population also sustained some of the most austere vegetarian institutions in the ancient Indian world. The two cuisines coexisted for over a thousand years without merging and without displacing each other. The later temple kitchen at Puri inherits the institutional vegetarian tradition. The household kitchen inherits the fish-rice tradition. The unresolved tension between them is a structural feature of Odia food culture rather than a recent development [Eschmann, Kulke & Tripathi, 1978, The Cult of Jagannath and the Regional Tradition of Orissa].
The Buddhist-Jain sediment partially dissolved over centuries as Buddhism declined in Odisha and the Brahmanic-Vaishnava order consolidated. But it left traces that are still measurable. The Dalma that every Odia household cooks — lentils simmered with seasonal vegetables, tempered with panch phutana, without onion or garlic — is, in all likelihood, a distant descendant of monastery cooking. The comparatively high rate of vegetarianism in coastal Odisha, surprising in a region where fish is abundant and culturally central, is a residual signal of this layer. The sediment did not survive as a distinct stratum. It leached into the layers above and below it, altering their chemistry.
The Temple Deposit: Rice, Ghee, and the Fifty-Six Bhogas
The Eastern Ganga dynasty, ruling from roughly 1038 CE under Chodaganga Deva, transformed Odisha’s religious economy by redirecting state revenue into temple construction, Brahmanic ritual, and the institutionalisation of the Jagannath cult at Puri. The Jagannath temple in its current form was constructed under Chodaganga Deva and completed around 1174 CE. The Ganga-era copperplate grants document a pattern of royal land grants to Brahmana settlements and temple establishments — including daily food offerings to the deities [Kulke, 1993, Kings and Cults; Rajguru, 1955-66, Inscriptions of Orissa; Sahu, 1952, New Light on the History of Kalinga].
The mechanism of the temple kitchen’s emergence at Puri is best understood as the superimposition of institutional vegetarian food norms — inherited from Buddhist and Jain monasticism — onto a pre-existing substrate of tribal, Shaiva, and Shakta ritual offerings. Before the Jagannath cult stabilised, the Puri shrine was connected to the tribal deity Nila Madhava, and the earlier ritual offerings likely included non-vegetarian items. The Vaishnava capture of the shrine under the Gangas gradually imposed vegetarian ritual norms, but the absorption carried elements of the earlier substrate — most notably the intercaste commensality of Mahaprasad, which is hard to derive from mainstream Brahmanism and easier to read as a tribal-Buddhist inheritance [Eschmann, Kulke & Tripathi, 1978; Kulke, 1993; Mohanty, 2011, Jagannath Consciousness]. The Lord of the Blue Mountain series treats this absorption at length, documenting how the Jagannath cult incorporated tribal elements — the Savara connection to the deity, the forest foods entering the Bhoga list — into what became a Brahmanised temple institution. For the stratigraphy of the plate, the point is that the temple deposit did not simply overwrite the base layer. It absorbed parts of it, the way a new geological stratum can incorporate fragments of the rock it buries.
The Rosaghara — the temple kitchen — is the deposit’s defining institution. Its scale is the first thing to register:
| Metric | Value |
|---|---|
| Kitchen floor area | ~1,400 sq m across 32 rooms |
| Earthen hearths (chulas) | ~240 active |
| Hereditary cook families (Suaras) | ~600 cooks + ~400 assistants |
| Offerings per day (Chhappan Bhoga) | 56 dishes across 6 services |
| Daily rice consumption | 50-60 quintals (5-6 tonnes) |
| Daily pilgrim footfall (non-festival) | 20,000-25,000 |
| Festival day capacity | Up to 100,000 |
| Ingredients prohibited | Onion, garlic, chilli, tomato, potato, cabbage, cauliflower, beetroot |
[Sources: Mohanty, 2011; Patnaik, 2006; Starza, 1993; Mohapatra, 2006; temple administration]
Daily rice demand of five to six tonnes, sustained year-round, required procurement from dozens of villages on a regular rotation. The Ganga-era land grants allocated entire villages to the temple — sasana villages whose revenue flowed directly into the temple kitchen. This created a de facto food procurement system in which agricultural output bypassed ordinary market mechanisms and was channelled into institutional food production [Kulke, 1993; Patnaik, 2006; Mukherjee, 1981, The History of the Gajapati Kings of Orissa].
What the Chhappan Bhoga codifies is not just a menu but a complete culinary programme encompassing every category of taste, texture, and preparation that the tradition recognises: nine rice preparations, fourteen vegetable preparations, nine milk preparations, eleven sweets, and thirteen pithas. The list rotates with seasonal availability but the structural template has been maintained for approximately eight hundred years [Mohanty, 2011; Patnaik, 2006].
The temple deposit did three things to the Odia plate that no subsequent layer has undone.
First, it placed rice at the absolute centre. Before the temple period, rice was dominant on the coast but coexisted with millets and pulses as substantial components of the diet. The temple’s massive, daily, ritual demand for rice — and the prestige that flowed from the rice-offering to Jagannath — elevated rice from the most important grain to the only important grain. The word for a meal in Odia became, effectively, the word for rice. The gravitational centrality of rice on the Odia plate is a temple-period deposit.
Second, it codified the panch phutana grammar. The temple kitchen’s prohibition on onion and garlic, combined with its reliance on mustard, cumin, fenugreek, nigella, and fennel as the primary aromatic base, standardised a spice profile that radiates outward from Puri into every kitchen in coastal Odisha. Before cookbooks, before food media, before any modern system of culinary standardisation, the Rosaghara functioned as a broadcast tower: pilgrims from Ganjam and Koraput, from Sambalpur and Mayurbhanj, ate Mahaprasad at Puri and carried the taste memory back to their home kitchens.
Third, it established chhena — fresh cottage cheese — as a sacred material. The temple’s demand for daily milk products, channelled through specialised dairy villages, created the institutional basis for the chhena-based sweet tradition that later becomes central to Odia identity. Rasagola, Chhena Poda, Rasabali, Chhena Jhili — all of these trace their institutional origin to the temple-dairy economy of the Ganga and Gajapati periods [Achaya, 1994, Indian Food: A Historical Companion; Ray, 2015, Culinary Culture in Colonial India].
Sarala Das’s Odia Mahabharata, composed in the mid-fifteenth century during the reign of Gajapati Kapilendra Deva, provides the richest literary evidence for the state of the Odia plate at the temple deposit’s peak. Sarala Das localised the Sanskrit epic, inserting fifteenth-century Odia food practices into the narrative: rice varieties, fish preparations, pitha-making, dairy processing, the distinction between habisha (austerity) rice and celebration rice. Scholarly work on the text has catalogued over a hundred food references [Mishra, 1980, Sarala Sahityara Adhyana; Mansinha, 1962, History of Oriya Literature].
The temple deposit is the thickest and most culturally dominant layer on the Odia plate. It is also the layer that most clearly illustrates the stratigraphic principle of superposition: it buried the Buddhist-Jain sediment and the tribal base layer beneath it, but it did not destroy them. The Buddhist legacy persists in the institutional vegetarianism the temple inherited. The tribal legacy persists in the Mahaprasad’s intercaste commensality and in the incorporation of forest foods into the Bhoga list. The temple absorbed what came before it, incorporated fragments, and created a new stratum that became, in the popular imagination, the foundational layer — even though it is, in geological terms, a relatively recent deposit.
The Columbian Intrusion
In geological stratigraphy, an intrusion is a body of rock that has been emplaced by a force from outside the local sequence — a magma plume that cuts across existing strata, depositing new material that cross-cuts the orderly layering. The Columbian Exchange — the transfer of New World plants to the Old World after 1492 — was precisely such an intrusion into the Odia plate. It cut across every existing layer, depositing new ingredients into both temple and household traditions, and rewrote the flavour profile of the cuisine while leaving its structural grammar intact.
The channel was the Portuguese trading-post network along the Indian coast. Vasco da Gama arrived at Calicut in 1498; within decades, Portuguese traders operating out of Chittagong, Hughli, Pipili, and Masulipatnam on the Coromandel coast were carrying New World crops into the Bay of Bengal trading system. The crops that reached Odisha through this channel transformed the plate:
| Ingredient | Origin | Approx. introduction to India | Approx. mainstream use in Odisha |
|---|---|---|---|
| Chili pepper | Mexico/Central America | c. 1500-1550 | c. 1650-1750 |
| Tomato | Mesoamerica | c. 1550-1600 | c. 1800-1900 (late) |
| Potato | Andes | c. 1600-1700 | c. 1800-1900 (late) |
| Cashew | Brazil | c. 1560-1600 (Goa) | c. 1700-1800 (Ganjam) |
| Papaya | Mesoamerica | c. 1600 | c. 1700-1800 |
| Pineapple | Brazil | c. 1550 (Goa) | c. 1800 |
| Guava | Caribbean/Central America | c. 1600 | c. 1800 |
| Maize | Mesoamerica | c. 1600 | c. 1800 (tribal uplands) |
[Sources: Achaya, 1994; Collingham, 2006, Curry: A Tale of Cooks and Conquerors; Katz & Weaver, 2003]
The Odisha-specific dating is approximate because local records are thin. The evidence is largely indirect: temple prohibitions (what the Rosaghara still refuses), inscriptional absence, and literary non-mention. This uncertainty should be stated plainly. The Columbian arrivals reached Odisha’s household kitchen through the Bay of Bengal trade networks centred on Balasore (a Danish and later English trading post in the seventeenth century) and Pipili (a major Portuguese trading post), which distributed the new ingredients along the coastal rice belt [Achaya, 1994; Subrahmanyam, 2012, The Portuguese Empire in Asia; Collingham, 2006].
The chili pepper is the single most consequential intrusion. Before it arrived, the heat in Odia food came from a different family of compounds entirely. Long pepper (pipali), native to eastern India, was the dominant heat source — a warm, aromatic pungency qualitatively different from the sharp capsaicin burn of the chili. Black pepper (gola marichi), imported from the Malabar coast, was used in elite and temple cooking but was expensive. Mustard seed and mustard oil provided pungency. Ginger, fresh and dried, provided warmth. The pre-chili Odia plate was not a mild plate. It was a differently hot plate — a heat built from black pepper, long pepper, mustard, and ginger rather than from capsaicin.
The Rosaghara at Puri preserves this pre-Columbian spicing profile more or less intact. The temple kitchen’s list of prohibited ingredients — chili, tomato, potato, cabbage, cauliflower — is not an arbitrary orthodoxy. It is a reasonably accurate fossil of the fifteenth-century Odia elite plate, preserved because institutional ritual systems update more slowly than household kitchens. This makes the Rosaghara a living archive in the stratigraphic sense: a layer that was deposited before the Columbian intrusion and that has remained sealed against it. When visitors note that temple food at Puri tastes “different” from ordinary Odia food — milder, more aromatic, more reliant on the warmth of black pepper and the complexity of panch phutana — they are tasting the pre-intrusion plate.
The intrusion rewrote household cooking between roughly 1600 and 1900. By the time the British consolidated control of coastal Odisha in 1803, the chili, papaya, cashew, and guava had been absorbed as though they had always been present. In Odia, the chili is simply lanka — a word used without any sense of its foreign origin. In Malayalam, by contrast, the chili is still called kappal mulagu — “ship pepper” — preserving the memory of its maritime arrival [Collingham, 2006]. The potato and tomato spread more slowly and were not fully naturalised in rural Odia cooking until the late nineteenth century. Early-twentieth-century ethnographic surveys in Odisha recorded village elders still considering these “new” or “English” vegetables [Senapati & Sahu, 1966, Orissa District Gazetteers].
The critical stratigraphic point is that the Columbian intrusion rewrote the flavour layer while leaving the structural grammar untouched. The template of the Odia meal — rice at the centre, dal alongside, bhaja (fried vegetable), tarkari (curry), saga (greens), chutney, sweet — did not change. What changed was what filled each slot. The tarkari now included potato and tomato. The chutney now burned with capsaicin instead of peppercorn warmth. The bhaja might be cauliflower or cabbage. But the grammar — the arrangement of elements around rice, the sequential logic of the meal — remained the grammar that the temple deposit had codified. The intrusion deposited new material into existing structures without collapsing them.
The Colonial Fracture: Famine and Commerce
In stratigraphy, an unconformity is a gap in the geological record — a surface that represents a period of erosion, non-deposition, or disruption. Unconformities mark moments when the orderly accumulation of layers was interrupted by a force powerful enough to alter the sequence. The British colonial period in Odisha (1803-1947) was such an unconformity. It did not deposit a new cuisine on the Odia plate the way the temple deposit or the Columbian intrusion did. What it did was fracture the existing sequence — disrupting the relationship between the Odia farmer and the Odia plate by restructuring agriculture around colonial revenue and rice export.
The East India Company’s annexation of coastal Odisha in 1803 integrated the region into the colonial rice economy centred on Calcutta. The Bengal Presidency restructured Odia agriculture around rice export. Cuttack became the primary rice collection and milling node; rice was transported by river and road to Calcutta and from there exported to Mauritius, Ceylon, the Caribbean, and other destinations. Colonial revenue records from 1820-1860 show a progressive rise in rice exports from Odisha ports [Bhattacharya, 1957; Das, 1978, Studies in the Economic History of Orissa].
The structural logic was extraction. Rice left the province before local consumption needs were met. Rain-fed millet and pulse cultivation on upland fields — which had historically provided dietary diversity and a buffer against rice-crop failure — contracted as farmers shifted to paddy monoculture under market pressure and colonial revenue demands payable in cash from rice sales. The buffer against failure thinned [Das, 1978; Hunter, 1872, Orissa; Baker, 1984].
Then the monsoon failed.
The 1866 famine — known in Odia cultural memory as Na Anka Durbhiksha, the famine of the ninth regnal year — is the unconformity’s sharpest expression. The 1865 south-west monsoon delivered the kharif paddy crop in significantly reduced volumes in a region by then structurally dependent on a single rice harvest. Rice continued to be exported from Odisha through Calcutta even as local prices rose and local stocks depleted. The colonial administration had no policy mechanism to halt this export and no political will to create one until mortality was catastrophic [Hunter, 1872; Government of India, 1867, Report of the Commissioners Appointed to Enquire into the Famine in Bengal and Orissa; Sen, 1981, Poverty and Famines].
The death toll is reported as a range because the underlying data is uncertain. The consensus academic estimate is roughly 800,000 to 1.3 million deaths out of a pre-famine Odisha Division population of approximately 3 million — a mortality rate between 25 and 45 percent. Hunter’s widely cited figure is approximately one million. The Famine Commission’s initial estimate of 300,000-500,000 was later revised upward. Recent Odia scholarship places the range at 800,000 to 1.3 million [Hunter, 1872; Bhatia, 1963, Famines in India; Sen, 1981; O Grada, 2009, Famine: A Short History; Mishra, 2015, Na Anka Durbhiksha; Mohanty, 2019, Famines of Colonial Odisha].
The structural causes map onto the extraction-equilibrium pattern that the Long Arc series documents across ninety years of Odisha’s post-1936 history — but Na Anka predates that timeline by seventy years and provides the earliest well-documented instance of the pattern. Rice exported during a local crisis. Revenue demands forcing grain sales when stocks were low. Transport infrastructure too thin for relief. Administrative inaction rooted in ideological commitment to non-intervention in grain markets. Differential mortality across class groups, with landless labourers and artisans dying at far higher rates than landholding cultivators. Sen’s entitlement-failure framework maps onto Na Anka almost exactly: aggregate food availability in the Bengal Presidency was adequate, but Odisha’s residents lacked the purchasing power, political voice, or administrative protection to claim their share [Sen, 1981; Hunter, 1872; Ambirajan, 1976, Population Studies].
Na Anka permanently scarred Odia food culture. The phrase atita kala — the time of scarcity — entered common usage. The concern with grain storage, the cultural taboo against wasting rice, the intensity of the attachment to rice surplus as political guarantee — all of these sharpened in the post-1866 generations. Fakir Mohan Senapati’s Chha Mana Atha Guntha (1902), the foundational novel of modern Odia literature, reflects the post-famine consciousness of a population that had experienced mass starvation within living memory [Senapati, 1902/2005; Mishra, 2015].
For the stratigraphy of the plate, the colonial fracture deposited something unusual: not a new ingredient or technique, but a new psychology. After Na Anka, rice surplus became the non-negotiable priority of Odisha’s food system. The multi-crop diversity that had characterised the pre-colonial plate — millets alongside rice, pulses as substantial components, forest foods as supplements — was further compressed in favour of rice monoculture, because rice scarcity had killed a third of the population. The colonial fracture explains why Odisha’s twentieth-century food policy was rice-centric to the exclusion of almost everything else. The trauma was the deposit. The monoculture was the scar tissue.
The Green Revolution Overlay
Between 1950 and 1955, the Central Rice Research Institute at Cuttack conducted a botanical survey of rice landraces in the undivided Koraput district — the Jeypore tract. Under the leadership of R.H. Richharia and associated researchers, the survey collected seed samples and catalogued approximately 1,745 indigenous paddy varieties from this single district: pigmented landraces (red, black, purple), aromatic varieties, flood-tolerant varieties, drought-tolerant varieties, and varieties adapted to specific altitudes within the district’s 500-1,600 metre elevation range [Richharia & Govindasamy, 1990, Rices of India; CRRI annual reports, 1952-1960; MSSRF Jeypore Biodiversity Initiative reports].
The figure should be cited with care. It refers specifically to the Jeypore tract, specifically to landraces collected during the 1950-55 surveys, and represents what was in active cultivation at that moment. The broader Indian pre-Green Revolution estimate of “over 100,000 rice varieties” is a national figure aggregated across many regional surveys and should not be confused with the Jeypore-specific count [Richharia, 1979; CRRI archives; Deb, 2005, Seeds of Tradition, Seeds of Future].
Then the Green Revolution arrived. IR8 — the “miracle rice” from the International Rice Research Institute, released in 1966 — was followed by IR36, Swarna (MTU 7029), and dozens of other high-yield varieties that progressively displaced indigenous landraces in coastal Odisha’s irrigated zones. The HYVs delivered two to three times the yield of landraces but required irrigation, synthetic fertiliser, and pesticide inputs. The economic logic of a subsistence cultivator with access to irrigation strongly favoured HYV adoption. The ecological and cultural cost accrued to the broader food system rather than to the individual farmer.
When M.S. Swaminathan Research Foundation conducted a follow-up survey of the Jeypore tract in the 1990s and early 2000s, the number of landraces still in active cultivation had collapsed to approximately 350 varieties. From 1,745 to 350 in roughly forty years — an eighty percent decline. The ecologist Debal Deb, who has documented and conserved indigenous rice landraces through his Basudha seed bank, has called this one of the largest destructions of genetic biodiversity in the history of agriculture [Deb, 2005; Deb, 2009, Beyond Developmentality; MSSRF Koraput reports, 1998-2015].
The Green Revolution overlay is the most recent and thickest layer on the Odia plate. In stratigraphic terms, it is the deposit that buried almost everything beneath it. The 1,745 varieties of Koraput represented not just genetic diversity but a knowledge system: each variety embodied the accumulated observation of generations of farmers about how a specific plant interacted with a specific environment — when to plant it, how much water it needed, what pest it resisted, what it tasted like, what ritual purpose it served. When the variety disappeared, the knowledge disappeared with it. A rice variety is not just a genome. It is a relationship between a plant and a community. No gene bank can preserve a relationship.
The parallel millet displacement followed the same trajectory. Finger millet (mandia), little millet (suan), foxtail millet (kangu), kodo millet, and proso millet had been staple grains in the tribal highlands for millennia. The Public Distribution System’s exclusive focus on rice and wheat, the absence of procurement prices for millets, the social stigma attached to millets as “poor people’s food,” and the aggressive promotion of rice as the modern and prestigious staple combined to push millets from the centre of the highland plate to its margins [Seetharaman & Chakraborty, 2020, Millets in India; MSSRF millets reports, 2010-2020; Odisha Millets Mission baseline studies, 2017].
The nutritional cost was measurable. Finger millet contains 300-350 mg of calcium per 100g, compared to 10 mg for rice. It is higher in iron, higher in fibre, and has a lower glycemic index. The substitution of PDS rice for traditional millet consumption in the tribal highlands coincided with rising rates of anaemia, diabetes, and diet-related conditions in populations that had been nutritionally better served by their traditional grains [MSSRF reports; NIN dietary surveys; Odisha Millets Mission baseline, 2017-2019]. The extraction-equilibrium pattern that the Value Chain series traces through Odisha’s mineral economy operates here too, but with a twist: in the mineral case, raw ore leaves the state and value is added elsewhere. In the food case, the extraction was of diversity itself — the 1,745 varieties were not physically exported, but the policy architecture that imposed HYV monoculture structurally rendered them worthless, extracting their place in the farming system as surely as a mine extracts ore from a mountain.
The Green Revolution overlay created a false abundance. Rice production in Odisha increased. Yields per hectare rose. The state moved from chronic food deficit to near self-sufficiency in aggregate rice production. But the nutritional profile of the plate deteriorated. The multi-grain, multi-vegetable, seasonally diverse diet that the pre-Green Revolution plate had preserved — even if imperfectly, even if under pressure from the colonial fracture — was replaced by a rice-dominated monoculture plate supplemented by increasingly homogeneous vegetable and pulse production. More calories, less nutrition. More rice, less resilience. The thickest layer on the plate is also, nutritionally, the thinnest.
The Revival Excavation
In archaeology, excavation is the controlled removal of overlying material to expose buried layers. The past two decades have seen several attempts to excavate buried layers of the Odia plate — to bring millet, heirloom rice, and traditional preparations back to the surface from beneath the Green Revolution overlay. Whether these excavations are succeeding is a question the evidence answers with qualified ambiguity.
The Odisha Millets Mission, launched by the state government in 2017, is the most structurally significant excavation attempt. The programme’s core interventions: revival of traditional millet cultivation through seed distribution and agronomic support, procurement at minimum support prices, integration of millets into the ICDS and mid-day meal programmes, and market development through branded retail channels. The scale is measurable:
| Metric | 2017 baseline | 2023 | 2024 |
|---|---|---|---|
| Blocks covered | 30 | 84 | 142 (across 19 districts) |
| Farmers enrolled | ~15,000 | ~110,000 | ~170,000+ |
| Area under ragi (OMM districts) | ~3,100 ha | ~44,000 ha | ~60,000 ha |
| Millet procurement (quintals/year) | ~0 | ~180,000 | ~260,000+ |
| Millet Shakti retail outlets | 0 | 200+ | 400+ |
[Sources: OMM annual reports; WASSAN; Odisha Agriculture Department; Down to Earth, 2023; The Hindu, 2024]
The programme has made Odisha a global reference case for millet revival. The 2023 International Year of Millets, declared by the United Nations on an Indian proposal, brought international visibility, and OMM officials have presented the Odisha model at FAO and other international fora [OMM reports; FAO, 2023; Down to Earth, 2023; Scroll.in, 2023].
Debal Deb’s Basudha seed bank represents a different kind of excavation — one individual’s decades-long effort to physically preserve what the Green Revolution buried. Deb’s collection, initially based in West Bengal and later relocated to Odisha, has grown to over 1,400 rice varieties from across eastern India, many traced to specific Koraput origins. The Basudha model is conservation through cultivation: each variety is grown annually in the seed bank’s plots to maintain viability, creating a living archive rather than a frozen one [Deb, 2005; Deb, 2009; MSSRF reports].
The Koraput region’s partial preservation of indigenous diversity — maintained by Paroja, Kondh, Bhumia, and Gadaba tribal farmers who kept landrace cultivation alive partly through economic necessity (HYV packages were inaccessible in remote highland hamlets) and partly through cultural commitment to varieties tied to ritual use — was recognised by the FAO in 2012 as a Globally Important Agricultural Heritage System, one of only three in India. The designation is honest in ways the celebration sometimes is not: the preservation happened because these communities were too marginal for the Green Revolution to reach, not because anyone planned for conservation [FAO GIAHS, 2012; MSSRF, 2012; Ramprasad, 2002, Hidden Harvests; Deb, 2005].
The culinary nationalism layer is more recent and more ambiguous. Pakhala Divas — observed on March 20, driven by grassroots social media campaigns since roughly 2011 — represents a diaspora-anchored assertion of food identity. In 2024, the Odisha Tourism Department organised a Guinness World Record for the largest serving of pakhala: 1,174 kilograms. The rasagola GI tag, granted in 2019, distinguishes the Odia version from the Bengali one. GI tags have been filed or granted for Koraput Kalajeera rice (2017), Kandhamal haldi (2019), and others [GI Registry India; Down to Earth, 2019; Odisha Tourism Department, 2024].
These cultural assertions are genuine, but they are excavations of a particular kind: they bring specific items from buried layers to the surface while leaving the structural conditions of burial largely intact. Pakhala Divas celebrates fermented rice while the PDS continues to distribute polished white rice. The Kalajeera GI tag protects a landrace while the procurement system still incentivises HYV monoculture. The millet revival is real, but it remains concentrated in rain-fed uplands where rice is anyway marginal; farmers with access to irrigation continue to prefer HYV rice because the economics still favour it. The structural limits are several: the procurement-price gap between millet and rice, consumer habituation to polished rice after two generations of PDS supply, inadequate millet processing infrastructure, and policy dependence — the entire revival rests on continued state-level political commitment that a change in administrative priorities could withdraw.
The excavation is happening. But the overburden — the Green Revolution overlay — remains thick, and the tools being used to dig through it are, in several cases, the same government instruments that deposited the overlay in the first place.
Honest Limitations
The stratigraphic metaphor has limits that should be named.
Unlike geological layers, food layers interact. The chili combined with the temple recipe. The temple recipe incorporated the tribal offering. A woman in Cuttack cooking dalma with chili-tempered mustard oil is not eating two separate historical layers; she is eating a compound that the Columbian intrusion and the temple deposit created together, a substance that exists only because two layers that arrived at different times reacted chemically to produce something neither contained alone. Geological strata do not do this. Limestone does not combine with sandstone to produce a third rock. Food does. The metaphor illuminates sequence and deposit; it understates recombination.
The metaphor also risks implying that earlier layers are “more authentic” than later ones — that the millet base layer is the “real” Odia food and the Green Revolution overlay is a corruption. This is analytically unhelpful. The Green Revolution layer, for all its costs, is not a foreign imposition in the simple sense. It was adopted by Odia farmers making rational economic choices within the incentive structures available to them. The chili, a sixteenth-century foreign arrival, is now so thoroughly Odia that no one experiences it as foreign. Authenticity is not determined by age. A layer that has been on the plate for four hundred years is not less Odia than a layer that has been there for four thousand.
There are silences in the stratigraphic column that the evidence cannot fill. The household kitchen between 500 and 1500 CE — what an ordinary coastal Odia family actually ate during the Buddhist-to-Gajapati transition — is reconstructable only indirectly. Women’s food knowledge — the seasonal variations, the sick-day preparations, the fermentation and pickling techniques, the rice-sorting and pulse-cleaning skills transmitted through women’s labour — is almost entirely absent from the formal historical record. The fish and riverine protein history, central to Odia food identity, has almost no archaeological record because fish bones do not survive in tropical soils. The Muslim culinary layer in Cuttack, Banki, and Bhadrak is poorly documented in academic scholarship. The precise dating of when Columbian ingredients reached Odisha depends on inference from temple prohibitions and literary non-mention rather than on direct Odisha-specific evidence [Achaya, 1994; research document internal assessment].
These are not failures of the analysis. They are boundaries of the evidence, and they should be visible so that claims do not outrun what the record can support.
The Compressed Archive
An Odia person sitting down to a plate of pakhala bhata with saga bhaja, machha tarkari, a chili-tempered dalma, and a piece of chhena poda is eating a compressed archive. The pakhala — fermented rice water, the food of poor farmers, now a cultural movement and a Guinness record — carries the base layer’s fermentation technology, the colonial fracture’s rice obsession, and the modern revival’s identity politics, all in a single bowl. The saga bhaja — greens fried in mustard oil with a flick of dried red chili — is simultaneously a Neolithic foraging inheritance (the greens), a temple-period aromatic grammar (the mustard oil), and a Columbian intrusion (the chili). The machha tarkari — fish curry — carries the coastal baseline that predates any written record, cooked with turmeric that is indigenous, mustard that the temple codified, and tomato that arrived from Mesoamerica three centuries ago. The dalma — lentils and vegetables tempered with panch phutana — is a probable descendant of Buddhist monastery cooking, using the pre-Columbian spice profile the temple preserved, augmented by raw papaya that came from the Americas. And the chhena poda — burnt cottage cheese — is a temple-period dairy innovation whose sugar caramelisation technique has no recorded origin but whose institutional context is the Ganga-dynasty dairy supply chain.
The plate is more accurate than any textbook. The textbook tells you that Ashoka conquered Kalinga in 261 BCE; the plate shows you what the monastery cooks did with the vegetarian discipline that followed. The textbook tells you that the Portuguese reached India in 1498; the plate shows you how completely a Mexican berry displaced a Subcontinental peppercorn as the heat source of a civilisation’s food. The textbook tells you that a million people starved in 1866; the plate shows you the rice-centric obsession that the trauma deposited. The textbook tells you that the Green Revolution increased yields; the plate shows you the thousand varieties that disappeared to make room for the three that remained.
Reading the plate requires the same discipline that reading a geological core sample requires: attention to sequence, respect for the evidence, honesty about the gaps, and the recognition that what is visible on the surface is always the most recent and least interesting layer. The deep history is below. It is in the millet that a Paroja woman cooks in a highland kitchen. It is in the long pepper that survives in certain Puri temple preparations. It is in the horsegram dal that appears in rural households where the base layer was never fully buried. It is in the fermentation that turns leftover rice into a probiotic food whose nutritional logic modern science is only now confirming.
The Environmental Odisha series documents what happens when the land’s deep structure is mined without accounting for its complexity. The Odia plate has been mined too — its diversity extracted, its layers compressed, its complexity reduced to a monoculture surface. But unlike a mountain that has been hollowed out, the plate can be restored. The layers are still there. The millet still grows in Koraput. The landraces still survive in Deb’s seed bank and in the fields of tribal farmers who never abandoned them. The monastery cooking lives on in every pot of dalma. The temple archive operates daily, cooking without chili, without tomato, without potato, as if the Columbian intrusion never happened — which, within its walls, it has not.
The question is whether the excavation underway — the Millets Mission, the seed banks, the GI tags, the cultural assertions — will succeed in exposing these buried layers before the knowledge needed to read them is lost. The layers will outlast the knowledge. The millet will grow whether or not anyone remembers how to cook it. The landrace will germinate whether or not anyone remembers what it tastes like. But a layer that no one can read is, for all practical purposes, a layer that no longer exists. The urgency is not in the seeds. It is in the hands — the grandmother’s hands that know, without a recipe, how to turn finger millet into jau, how to temper dalma with five spices in the right proportion, how to ferment rice overnight so that it cools the body in the summer heat. Those hands are the real archive. And they are not being preserved in any seed bank.
Next: Chapter 2 — The structural grammar of the Odia plate: why rice sits at the centre, why sweet and savoury share the same course, and what the meal’s architecture reveals about the culture’s organising principles.