English only · Odia translation in progress

Chapter 7: Between Two Worlds

On a January morning in 2014, three months after twelve gram sabhas in the Niyamgiri Hills had unanimously rejected Vedanta’s bauxite mining proposal, a young Dongria Kondh man named Ladho Sikaka stood before television cameras in Delhi and did something that would have been structurally impossible a generation earlier. He spoke in Hindi — halting, accented, but comprehensible — about why the mountain could not be mined. Not because the ore was worthless, but because Niyam Raja lived there. The supreme deity. The hill was not a sacred site in the temple sense, where you build a structure and worship inside it. The entire mountain range was the deity. Mining the mountain was not demolishing a temple. It was killing a god.

Sikaka had learned Hindi at a residential school in Muniguda. He had learned the language of press conferences from NGO workers who had spent years in the Niyamgiri campaign. He wore a shirt and trousers in Delhi, not the traditional rings and headbands he would wear in his village. He carried a mobile phone. And he was articulating, in the vocabulary of environmental rights and indigenous sovereignty that the international press could parse, a cosmology in which a hill is alive. He was, in that moment, running two operating systems simultaneously — the cosmology of the Dongria Kondh, where Niyam Raja governs the hills and the streams flow because the mountain permits it, and the cosmology of the Indian republic, where rights are constitutional, victories are judicial, and a gram sabha is a legal instrument under the Panchayats Extension to Scheduled Areas Act of 1996. Neither system fully comprehended the other. But Sikaka was the bridge. He had to be. No one else could run both.

This is the backwards compatibility problem. In software engineering, backwards compatibility is the property of a system that allows it to accept input from — and produce output for — older versions of itself. When you upgrade an operating system, the new version must still run applications that were built for the old one. If it does not, the upgrade breaks the user’s world. Files become unreadable. Software crashes. Workflows that took years to build are destroyed overnight. The new system may be objectively better — faster, more secure, better designed — but if it breaks the applications that users depend on, the upgrade is a catastrophe. The entire history of successful modernization, in software and in societies, is a history of managing backwards compatibility: how to move forward without destroying what already works.

Tribal identity in Odisha is caught in a backwards compatibility crisis that no one has designed a solution for. Modernity is the upgrade — education, healthcare, roads, mobile phones, legal rights, economic opportunity. The original system is a forest-based civilization with its own governance, its own knowledge, its own relationship to land, its own languages, its own gods. The applications that run on this original system — shifting cultivation, clan governance, oral tradition, sacred geography, community decision-making through councils of elders — cannot run on the new one. And the new one, for all its power, does not offer replacements for what it destroys. It offers schools but not schools that teach in Kui. It offers healthcare but not healthcare that understands why a Saora woman paints ikons on her wall to heal the sick. It offers legal rights but not legal categories that recognize a mountain as a living deity.

The people caught in this incompatibility are not abstractions. They are the IAS officer from a Kondh village who returns for Chaitra Parba and sits in the same circle as the Jani, the ritual specialist whose authority predates the Indian state. They are the college student who speaks Odia in class, Hindi with friends, and Kui when she calls home. They are the Dongria Kondh activist who won the Niyamgiri campaign and now watches Vedanta’s refinery at Lanjigarh continue to pollute the streams below the mountain he saved. They are the 9.6 million people — nearly one in four Odias — who are formally classified as Scheduled Tribes, navigating between a constitutional framework that promises protection and a development model that demands assimilation.

The Original Architecture

To understand what modernity is breaking, you have to understand what was built.

The pre-colonial tribal governance systems of Odisha were not primitive arrangements awaiting the gift of civilization. They were functioning architectures — sophisticated, adapted, internally coherent — that solved problems their builders understood better than any outsider could.

Consider the Santhal Manjhi-Pargana system, operational across Mayurbhanj and Keonjhar. Seven officials governed each village: the Manjhi (headman), Paranik (deputy), Jog Manjhi (youth supervisor), Jog Paranik (his assistant), Godet (messenger), Naeke (priest), and Kudam Naeke (forest priest). Above the village, fifteen to twenty villages formed a Pargana under a Parganat. Above that, a four-tier judicial system — Manjhi Baisi, Mapanjhi Baisi, Pargana Baisi, and Lo Bir — resolved disputes at escalating levels of severity. The Lo Bir was the highest authority: a democratic assembly of all males convened during the annual communal hunt, where any person, including the poorest, could bring complaints against even the Manjhi and Parganat themselves. This was not an approximation of democracy. It was a functioning accountability mechanism that predated British parliamentary procedure in the subcontinent.

Or consider the Kondh mutha system. Fifty named muthas, each simultaneously a clan unit and a territorial unit. Marriage prohibited within the mutha (enforcing exogamy and genetic diversity). Village governance through the Gudi council of elders, the mukhiyas, with a structured dispute resolution process: one party approaches a mukhiya, the baarika (messenger) summons the community, both sides present their case, a collective decision is reached. The Dongria Kondh variant had no overarching political or religious leader at all — a decentralized structure so resilient that when the Supreme Court directed gram sabha votes on Niyamgiri mining, the system generated twelve unanimous decisions without any centralized coordination. The decentralization was not a bug. It was a design feature that made the system resistant to capture by any single authority.

The Bhuiyan pirha system went further. The Bhuiyans of Keonjhar were not merely governed — they governed. They held the right to install the Raja of Keonjhar through the tilak ceremony, and they reserved the right to dethrone him if he became oppressive. Chiefs could not exercise authority until they had received the Bhuiyan token of investiture. The Bhuiyan body functioned as an organized militia, holding lands on conditions of military service. When the British attempted to impose their own candidate as Raja, the Bhuiyans rose in revolt — the 1867 uprising, the largest tribal rebellion in nineteenth-century Odisha. This was not a primitive community resisting modernity. This was a community with a functioning constitutional mechanism defending its sovereignty.

The Juang majang — the rectangular community house at the center of every Juang village — was simultaneously a youth dormitory, a courthouse for elders, a guest house, a cooperative grain store, a musical instrument repository, a ritual venue, and a cultural center. It served the functions that, in modern institutional design, require a school, a courthouse, a community center, a library, and a museum — all in one building, all managed by the community, all operating without external funding or bureaucratic oversight.

Each of these systems shared structural features that any software architect would recognize as sound design. Separation of concerns: secular and sacred authority were vested in different people — the Saora had the Gomango for administration and the Buyya for religious matters, co-equal and complementary. Community ownership of resources: land belonged to the clan, the birinda, the village — never to individuals in the alienable sense, never transferable to outsiders. Federated governance: autonomous villages connected through higher-level councils (pirha, pargana, mutha) for disputes and coordination, but retaining local control over daily affairs. And a built-in accountability mechanism: the Lo Bir, the pirha panchayat, the king-making function of the Bhuiyans — all ensured that power was checked by the community that granted it.

These were not artifacts of a distant past. The Paudi Bhuyan pirha system was still actively practiced when researchers documented it in 2025. The Kondh dispute resolution through mukhiyas still operates in villages across Kandhamal. The Santhal Manjhi-Pargana system, however weakened, still provides a parallel governance layer in Mayurbhanj. The original architecture is not dead. It is running alongside the new one, often in conflict, sometimes in uneasy coexistence, occasionally — as at Niyamgiri — producing outcomes that the new system could not have achieved alone.

The Upgrade That Breaks Things

The modernization of tribal Odisha follows a pattern that software engineers call a “breaking change” — an update to a system that causes previously functioning components to fail.

Start with language. Odisha has twenty-one tribal languages and seventy-four dialects. Several are endangered — Manda, Parji, and Pengo have very few speakers left. Saura is threatened despite having a script. Only six tribal languages have written forms at all: Santali, Ho, Soura, Munda, Kui, and arguably Gondi. Santali alone has constitutional recognition in the Eighth Schedule, making it the only tribal language in Odisha with institutional backing.

The Multi-Lingual Education programme, launched in 2006, was a genuine attempt at backwards compatibility. The state appointed 3,385 tribal language teachers to teach in tribal mother tongues at the primary level. This was real. It mattered. But MLE covers only primary school. The moment a tribal student moves to secondary education, instruction shifts entirely to Odia and English. The bridge language is available for the first few years; then the bridge disappears, and the student must swim.

The result is predictable. The secondary dropout rate among tribal students is 33 percent — more than double the state average of 15 percent. The bottleneck is not lack of intelligence. It is a language incompatibility. A Kui-speaking child who has been learning in Kui for five years is suddenly expected to process everything in Odia, a language she may speak conversationally but has never used for abstract reasoning or scientific concepts. This is the equivalent of porting an application from one framework to another with no migration tool. Some students manage the transition. Many crash.

The students who do make it through carry what researchers call “assimilation anxiety” — the persistent fear that success requires abandoning their cultural identity. The educated tribal professional faces a peculiar bind. In the urban or professional world, she is expected to speak Odia, English, and Hindi fluently. Cultural markers of tribal identity — dress, dietary habits, ritual practices — may be read as markers of backwardness. Reservation creates its own stigma: the assumption that she “got in through quota,” regardless of merit. In the home community, the returning educated tribal may face the inverse suspicion: “you have become like them.” Traditional authority structures — village elders, ritual specialists — may see educated youth as challenges to established hierarchies. The educated tribal is perpetually code-switching between two identity protocols, and neither protocol fully trusts her.

Now consider what happens to forest-based identity when the forest is mined. Tribal identity in Odisha is not cultural decoration layered on top of a universal human substrate. It is functionally tied to ecology. Shifting cultivators — the Bonda with their dangar chas, the Juang with their toila chasa, the Dongria Kondh with their dongar chasa — define their calendar, social organization, and spiritual practice around forest cycles. Remove the forest, and the entire knowledge system becomes irrelevant. Hunter-gatherers like the Birhor (population: 593) and the Mankirdia (population: 2,222) literally cannot exist as communities without forest access. Their population numbers are not demographics. They are extinction warnings.

The Bonda grow over twenty crop varieties through their blend of settled and shifting cultivation while preserving biodiversity — a system that contemporary agroecologists recognize as a sophisticated indigenous response to climate variability. But this system requires forest. When forests are cleared for mining, the traditional food system collapses: mahua, kendu, sal seeds — the minor forest produce that provides twenty to forty percent of tribal household income — disappears. Medicinal plant knowledge becomes useless. Ritual sites are destroyed. The economic basis of barter and seasonal exchange evaporates. Young people have no reason to learn traditional skills. In a single generation, a community shifts from forest-dependent self-sufficiency to wage-labor dependence on the very companies that destroyed the forest.

This is not a trade-off that anyone negotiated. Development arrives as a unilateral imposition — “we are giving you a road” — rather than as a choice made by the community about what it wants to preserve and what it is willing to change. The absence of negotiation is the core problem. In software, you never push a breaking change to production without a migration plan. In tribal Odisha, the breaking changes arrive without warning, without migration tools, and without rollback capability.

The False Binary

The discourse around tribal development in India is organized around a binary that does not survive contact with reality: development-as-integration versus rights-as-autonomy.

Framework A — the development view — holds that tribals are poor because they lack access to modern infrastructure, education, health, and markets. The solution is roads, schools, hospitals, electricity, connectivity, welfare schemes. Integrate tribal areas into the mainstream economy. This framework has measurable appeal. Governments can count roads built, schools opened, families covered by BSKY. Its logic is straightforward: poverty is the problem, growth is the answer.

Framework B — the rights view — holds that tribals are marginalized because their rights over land, forests, and self-governance are systematically violated. The solution is enforcement: implement PESA, enforce the Forest Rights Act, protect tribal land from alienation, respect gram sabha authority. This framework has moral clarity. It points to the 43 percent of land acquisitions that proceed without gram sabha consent, the 27 years of unfinished PESA rules, the 94 percent re-rejection rate of FRA claims under review.

Both frameworks, stated in isolation, are inadequate. The development view treats tribal culture as an obstacle to progress — something to be overcome by integration, like a legacy system that must eventually be retired. The rights view can freeze tribal communities in amber — defending autonomy so rigidly that it becomes a justification for continued neglect. “They want to live traditionally” becomes an excuse for not building hospitals.

The reality is that every tribal community in Odisha wants some version of both. The Dongria Kondh who rejected Vedanta’s mining do not reject healthcare or education. Their literacy rate is 6.35 percent — the second lowest of any PVTG in Odisha. They are not choosing ignorance. They are choosing their mountain. They want the mountain and the school. The current system makes them choose.

The false binary maps directly onto a backwards compatibility problem. Framework A says: upgrade the operating system, and the old applications will adapt or become irrelevant. Framework B says: keep running the old system, because the applications are irreplaceable. Neither approach works. The upgrade-everything approach destroys applications that contain irreplaceable knowledge. The change-nothing approach leaves the system vulnerable to threats — malaria, malnutrition, displacement — that the old architecture cannot address. The actual solution, in software as in development, is the one that requires the most engineering effort: build the compatibility layer. Design the upgrade so that it does not break the critical applications. This requires understanding what the critical applications are, which requires actually talking to the people who use them.

Consider what this would look like in practice. A health system that operates in Kui and understands that a Kondh patient’s relationship to illness involves the Jani and the Dishari — the ritual specialist and the diviner — alongside whatever biomedical intervention is appropriate. An education system where MLE does not stop at primary school but extends through secondary education, where the Santali and Kui curricula include the ecological knowledge embedded in traditional agriculture alongside the periodic table. A legal system where gram sabha consent is not a procedural formality that the sub-collector overrides eighteen months later, but an actual veto that requires the state to negotiate — to find a development path that the community agrees to, not one imposed upon it.

None of this exists. The compatibility layer has never been built. And so the binary persists: integrate or preserve, develop or protect, modernity or tradition. The false choice is not just intellectually lazy. It is structurally convenient for every actor who benefits from the current arrangement. Mining companies prefer the development framework because it justifies displacement. NGOs sometimes prefer the rights framework because it justifies intervention. The state government toggles between the two depending on which is useful: development rhetoric when opening a mine, rights rhetoric when accepting international awards for Niyamgiri. The tribal community is the only party that is never asked what version of the upgrade it would prefer.

Paraja’s Mirror: Tribal Identity in Literature

In 1945, Gopinath Mohanty published Dadi Budha, the first Odia novel centered on a tribal community. Mohanty was not a tribal. He was an IAS officer, the son of a Brahmin family from Cuttack, posted to Koraput district — the deep tribal interior. He would spend decades living among the Paraja, Kondh, and other communities, producing twenty-four novels, twelve short story collections, and five books on tribal language and culture. His 1945 work Amrutara Santana (The Immortal Children) became the first Odia novel to win the Sahitya Akademi Award, in 1955.

But it is Paraja — his novel about the Paraja tribal community of Koraput — that functions as a mirror for the identity crisis this chapter describes. Paraja is not a romantic novel about noble savages. It is a meticulous, devastating account of how the forest bureaucracy and the moneylending system — the mechanisms of the modern state — systematically dispossess a tribal family. The protagonist does not fall because of ignorance or backwardness. He falls because the new system’s rules are designed in a way that makes his old knowledge — his competence with the forest, his understanding of seasonal cycles, his standing in his community — irrelevant.

Mohanty wrote from within the system. He was the administrator. He knew how the paperwork moved, how the forest guard operated, how the moneylender’s ledger translated into lost land. And he understood what most outsiders writing about tribal communities miss: the internal coherence of what was being destroyed. The Paraja were not a broken people awaiting repair. They were a functioning society being dismantled by an external system that could not see them as they were.

This is the literary expression of the backwards compatibility problem. Mohanty documented, in fiction, the moment when the old applications stopped working — not because they were defective, but because the new operating system refused to support them.

The literary record raises an uncomfortable fact. The dominant voices speaking about tribal Odisha have been non-tribal. Mohanty, an upper-caste administrator. Verrier Elwin, a British anthropologist who became an Indian citizen and wrote the definitive ethnographies of the Juang, the Saora, and the Muria Gond. The absence of published tribal voices from within is not an absence of voice — oral literary traditions among Odisha’s tribal communities are deep, containing epics, ritual songs, ecological knowledge, and cosmological narratives transmitted across generations. But the published, documented literary tradition is almost entirely the work of outsiders looking in.

This matters because representation shapes policy. When the tribal experience is narrated by administrators and anthropologists, however sympathetic, the narrative inevitably centers the outsider’s framework. The tribal community appears as the object of study, the beneficiary of policy, the victim of injustice — never as the author of its own story. The backwards compatibility problem extends to narrative itself: the system through which India processes “tribal issues” does not run tribal applications. It runs applications built by people who study tribal communities, not by tribal communities themselves.

[Confidence note: The claim that tribal literary self-representation in published Odia literature is minimal is based on available search results and the research document’s own acknowledgment that “the absence of documented tribal writers from Odisha in search results may reflect a research gap rather than an actual absence.” Oral literary traditions are rich and not captured in publication metrics. Confidence: ~60%.]

The Art That Lives and the Art That Is Displayed

The Saora murals — called ikons, ekons, or idital — are painted on the interior walls of houses in villages across Rayagada, Ganjam, Gajapati, and Koraput districts. Their visual similarity to Warli paintings of Maharashtra is coincidental; the traditions developed independently. But the function is what matters. A Saora painting is not decorative art. It is a house for spirits. Each painting invites a specific deity to inhabit the wall. The painting is created under the direction of or in consultation with the Kudan, the shaman who combines the roles of priest, prophet, and medicine man. The Kudan’s position is achieved, not inherited — you become a Kudan through demonstrated spiritual gifts, not through birth. Verrier Elwin documented this tradition extensively in The Religion of an Indian Tribe (1955), based on years of ethnographic work among the Lanjia Saora.

Today, Saora paintings are available on canvas, on cloth, on e-commerce platforms. They hang in galleries in Bhubaneswar and Delhi. The Odisha State Tribal Museum — which UNESCO hailed as “the finest among India’s 21 tribal museums” in 2010 — displays artifacts alongside live demonstrations by tribal artists on all working days. Government craft promotion initiatives include Saora art in exhibition circuits.

The question is what has been preserved and what has been lost in the transition. When a Saora painting moves from the interior wall of a mud house, where it functions as a portal for spirits, to a canvas sold at a craft fair, what persists is the visual form — the geometric arrangements of humans, horses, elephants, trees, sun and moon. What is lost is the function. The painting is no longer a house for spirits. It is a commodity. The artist may benefit economically. The painting’s meaning has fundamentally changed.

This is the craft-versus-culture distinction, and it maps precisely onto the backwards compatibility problem. The visual form is forward-compatible — it runs on the new system (the market economy, the gallery, the e-commerce platform) without modification. The ritual function is not forward-compatible — it requires the original operating system (the village, the mud wall, the Kudan’s direction, the spiritual cosmology in which walls can house gods) to run. When the form is preserved but the function is lost, what you have is not preservation. It is taxidermy.

The same dynamic plays out across tribal cultural expression. Chhau dance of Mayurbhanj received UNESCO Intangible Cultural Heritage status in 2010. Mayurbhanj Chhau is the maskless variant — unlike the Purulia and Seraikella versions, it relies on facial expression added to martial body movement, enacting episodes from the Mahabharata and Ramayana. The UNESCO recognition brings global attention and preservation funding. It also accelerates the transformation from community ritual to staged performance. The dance moves from the village square to the auditorium, from the seasonal festival to the cultural program, from a martial art with spiritual dimensions to a performing art with aesthetic dimensions. The body of the dance persists. The soul is negotiable.

Kondh women’s facial tattoos — triangles, dots, parallel lines impressed in honor of the deity Bhimul, possibly rooted in chalcolithic cultures — are declining among younger generations. Dalkhai, the women’s dance of western Odisha performed during Dussehra, is increasingly staged at government-organized cultural events stripped of its ritual context. Ghumura, the war-dance-turned-festival-dance of Kalahandi performed with the ghumura drum, seeks a GI tag — a legal instrument of the new system designed to protect the commercial value of traditional knowledge. The GI tag protects the brand. It does not protect the practice.

The tribal festivals themselves face a double pressure. Tourism promotion presents them as exotic spectacles — “witness the ancient rituals of Odisha’s indigenous peoples.” Meanwhile, religious conversion (to Christianity or mainstream Hinduism) erodes the spiritual foundations on which the festivals rest. The economic incentive to perform “tribal culture” for tourists creates a paradox that any software engineer would recognize: the system is being maintained in production long after its original users have migrated away. The festival continues because tourists pay to see it, not because the community practices the faith that gave rise to it.

[Confidence note: The claim about festivals being sustained primarily by tourism economics rather than living practice is stated as a dynamic, not as a universal. Many tribal festivals retain genuine community participation and spiritual significance. The commodification pressure is real but its extent varies by community and geography. Confidence: ~65%.]

Niyamgiri After the Victory

On August 19, 2013, the twelfth and final gram sabha of the Dongria Kondh voted unanimously against Vedanta’s proposed bauxite mine on the Niyamgiri Hills. In January 2014, the Ministry of Environment, Forest and Climate Change completely rejected the project. Amnesty International called it “a great victory for indigenous rights.” It was. The constitutional framework — specifically the Forest Rights Act and PESA provisions on gram sabha consent — had actually worked. For once, the applications running on the original system had been recognized by the new one.

But victory is a point in time. What followed is a process, and the process has been less triumphant.

The Vedanta alumina refinery at Lanjigarh — built on the plains below Niyamgiri before the mining was stopped — continues to operate. The hilltop is protected. The refinery’s environmental impact is not. Amnesty International documented water contamination from red mud and other pollutants. Carbon dioxide emissions cause respiratory problems among local tribal communities. Vedanta annexed sixty hectares of village forest despite approval conditions prohibiting forest use. Kinari village was completely destroyed by refinery operations, displacing over a hundred Majhi Kondh families. The mountain was saved. The people at the foot of the mountain were not.

In February 2016, the Odisha Mining Corporation filed with the Supreme Court challenging the gram sabha resolutions, alleging “technical errors.” The Court rejected the petition. In 2023, the Forest Conservation Amendment Bill raised fears that the protections enabling the Niyamgiri victory might be systematically weakened. Survival International continues to document harassment of Dongria leaders by police, including imprisonment under false charges. The community believes the government is “trying to destroy their community in order to allow mining.”

The mineral remains. Seventy million tonnes of bauxite sit under the Niyamgiri Hills, valued at approximately two billion dollars. As long as the bauxite is there, the economic incentive for mining persists. The current legal protections are strong, and the precedent is powerful. But Indian environmental law has been progressively weakened through amendments, and political will to enforce protections fluctuates with government changes. The 2024 change of government in Odisha from BJD to BJP adds uncertainty.

[Confidence level on future mining threats to Niyamgiri: ~60%. The legal protections are real but not permanent. The economic incentive is permanent.]

This is the aftermath that the backwards compatibility metaphor illuminates most sharply. The Dongria Kondh won a battle to keep the old system running. But the new system has not stopped trying to deprecate it. The refinery operates. The police harass. The amendments chip away at protections. The victory preserved the mountain but did not change the underlying architecture of extraction that targeted it. In software terms, the community successfully blocked a forced migration. But the vendor has not abandoned the migration plan. It is pursuing it through other channels — through the refinery, through legal challenges, through the slow erosion of environmental protections — while the community’s attention is focused on the mountain itself.

The question Niyamgiri poses is not whether indigenous rights can win. They can, under extraordinary circumstances: a Supreme Court willing to enforce the law, international NGO campaigns, sustained media attention, and a community with the social cohesion to generate twelve unanimous decisions. The question is whether this is a scalable model. If every tribal community needs to fight to the Supreme Court to exercise rights that are already guaranteed by law, the law is not functioning. That is not backwards compatibility. That is a system that has been nominally upgraded but where every old application must be manually patched to run.

The God Who Walks in Front

The Sabara Dalapati walks in front of the Jagannath chariot during the Rath Yatra in Puri. This is not a minor ritual gesture. The Sabara — the tribal — leads the procession of the most important deity in Odia collective life. The tradition explicitly links Jagannath to tribal origins: the Nila Madhava legend, in which a tribal chieftain named Biswabasu worships the deity in the forest before the Brahminical tradition absorbs it into the temple at Puri. The incomplete wooden idols of Jagannath, Balabhadra, and Subhadra — replaced every twelve or nineteen years through the Nabakalebara ceremony — are themselves read by scholars as markers of a tribal aesthetic absorbed into the Hindu mainstream.

But the Sabara Dalapati’s walk in front of the chariot is a ritual acknowledgment, not an operational reality. The Dongria Kondh do not worship Jagannath. The Bonda do not worship Jagannath. The Juang, the Saora, the Birhor, the Mankirdia — none of them participate in the Jagannath system. Their deities are Niyam Raja, Dharani Penu (the earth goddess), the forest spirits propitiated through the Kudan’s intercession, the ancestors worshipped at the Sindibor stone platform, the totemic deities of clan and birinda. The Jagannath framework, for all its syncretic absorptive power — its capacity to incorporate Buddhist, Jain, Vaishnavite, Shaivite, and tribal elements into a single devotional system — reaches its limit at the actual tribal communities it claims to honor.

This matters for the identity question because the Jagannath system is the dominant framework through which Odia identity is organized. The temple at Puri is the centripetal force of Odia collective consciousness. The Sabara Dalapati’s walk is the Jagannath system’s way of acknowledging the tribal foundation. But acknowledgment is not inclusion. The tribal communities of Odisha are not part of the Jagannath system in any functional sense. They have their own gods, their own priests, their own sacred geographies. When a Kondh student moves to Bhubaneswar for college and encounters the overwhelming gravitational pull of Jagannath in Odia public life — in calendars, festivals, conversation, identity — she faces a choice that is never stated explicitly but is always present: align with the dominant framework and gain cultural legibility, or maintain the distinctness of Niyam Raja and Dharani Penu and remain, in some fundamental sense, outside the circle of Odia-ness as it is popularly understood.

The parallel to the tribal diaspora experience is direct. An Odia who moves to Bangalore for an IT job faces identity fragmentation — speaking Odia at home and English at work, celebrating Raja festival in an apartment complex that does not know what Raja is, maintaining an identity that the surrounding system does not recognize or value. The tribal who moves to Bhubaneswar faces the same fragmentation, compounded. She is outside the Bangalore mainstream because she is Odia. She is outside the Odia mainstream because she is tribal. Each migration strips a layer of context. Each layer lost is an application that no longer has a system to run on.

The Kandhamal Fault Line

On August 23, 2008 — Janmashtami Day — Swami Lakshmanananda Saraswati and four disciples were murdered at a residential girls’ school in Jalespeta, Kandhamal district. CPI(Maoist) leader Sabyasachi Panda claimed responsibility, stating the killing was for Lakshmanananda’s “anti-Maoist hate speech.” The violence that followed was anti-Christian: thirty-eight to a hundred people killed (the range reflecting the gap between government and independent counts), 395 churches destroyed or damaged, 5,600 to 6,500 houses plundered or burned, 50,000 to 75,000 people displaced, and — a detail that reveals the violence’s true character — thousands of forced conversions to Hinduism.

The standard narrative frames this as Kandha tribals (Hindu) attacking Pano Dalits (Christian). The narrative is partially accurate but critically incomplete. The Pano community — Dalit, Kui-speaking, occupying a historically subordinate social position — had a significant Christian population. Approximately 60 percent of Kandhamal’s 100,000-plus Christians are Pano converts. Swami Lakshmanananda had spent decades running programs to “reconvert” Christian tribals and Dalits to Hinduism. His assassination triggered a wave of retribution that was organized not by spontaneous tribal rage but by Sangh Parivar organizations — VHP, Bajrang Dal, RSS — who mobilized the Kandha community against Christians.

But underneath the communal violence lies the identity question. Why do tribals convert to Christianity? Not because missionaries are uniquely persuasive, but because Christianity offers something the Hindu caste system structurally withholds: dignity. The tribal’s position within Hinduism is ambiguous at best and degrading at worst. Formally outside the caste system, tribals are treated in practice as equivalent to or below Dalits. Conversion to Christianity provides access to mission schools, hospitals, and social networks — material benefits inseparable from the spiritual dimension. It also provides a theological framework in which the tribal is a child of God, full stop, without the qualifying asterisk of caste position.

The counterargument from the Hindu right is that conversion is cultural destruction — that it severs the tribal from traditional practices, festivals, and community ties. This is not entirely wrong. A tribal community that converts to Christianity may indeed abandon the festivals tied to traditional religion, the ritual specialists whose authority derives from the old cosmology, the sacred groves maintained by spiritual taboo rather than legal prohibition. Conversion can be its own breaking change.

But the argument that tribals must be “protected” from conversion by being kept within a Hindu framework that ranks them below upper castes is a protection racket, not protection. It preserves the tribal within a system that does not grant equal dignity. The tribal caught between these forces — the material benefits and theological dignity of Christianity versus the cultural continuity and community belonging of traditional practice, with Hinduism offering a third option that absorbs tribal identity into a caste hierarchy that places it near the bottom — is navigating a trilemma with no good exits.

The Kandhamal violence illuminated this trilemma with fire. It showed that the tribal identity question in Odisha is not merely an academic discussion about culture and modernity. It is a question that produces bodies. The fault line between tribal identity, caste hierarchy, and religious conversion is not an interesting intellectual tension. It is a combustion zone. And the people standing on the fault line — the Kui-speaking families who are simultaneously tribal, Christian, and living in a district where these identities became weaponized — are the most exposed population in Odisha’s identity landscape.

[Confidence note: The communal violence dynamics of Kandhamal are well-documented but the relative weight of organized mobilization versus community-level resentment is debated across sources. The framing here emphasizes organized Hindutva mobilization based on multiple independent reports. However, economic competition between Pano Christians and Kandha tribals over reservation benefits was a significant underlying factor that pure communal-violence framing can obscure. Confidence on overall framing: ~75%.]

The Urban Equation

Over 60,000 families from Bolangir, Nuapada, Kalahandi, Boudh, Sonepur, and Bargarh migrate to neighboring states annually. An additional 40,000 to 50,000 move to brick kilns near Cuttack and Bhubaneswar. The migration is mediated by sardars — labor contractors who advance money to families, creating a debt that binds the worker to the contractor for the season. This is functionally bonded labor, though no one calls it that.

The tribal migrant in the city faces the backwards compatibility problem at its most brutal. In the village, she was a member of a community with standing — a position within the mutha, the birinda, the village council’s jurisdiction. In the brick kiln, she is a unit of labor. The applications that defined her identity — clan membership, ritual participation, forest knowledge, community governance — do not run in the urban environment. The new system offers no equivalent. It offers wages (Rs 8,000 to 12,000 per month for 12-hour days), a washroom that one in five workers has access to, and the legal fiction that she is a free laborer rather than a debt-bonded one.

The urban tribal experience divides into two streams that rarely overlap. The distress stream — the brick kiln workers, the construction laborers, the domestic workers recruited from KBK districts — dominates by numbers. These migrants follow a circular pattern: leave after harvest, work for five to six months, return before the sowing season. They come back with little or no savings. The immediate needs — house repair, medical bills — consume whatever was earned. By June, they borrow again. The cycle repeats. This is not poverty as a static condition. It is poverty as a system with identifiable structural components: absence of non-farm employment in tribal districts, middleman-controlled migration, inadequate wages, debt-based recruitment. Each component reinforces the others.

The other stream is the educated tribal — the college graduate, the civil servant, the professional. Mohan Charan Majhi, a Santal from Keonjhar and four-time MLA, became Odisha’s Chief Minister in 2024. He is the first Santal CM and the first full-time tribal CM in fifty-two years (previous tribal CMs served briefly and were seen as caretakers). His appointment is historic. It is also an open question whether it represents genuine empowerment or strategic representation. In the BJD era, thirty-three reserved tribal Assembly seats were held by tribal MLAs who largely followed the party line set by Naveen Patnaik, a non-tribal from a political dynasty. Whether the BJP’s appointment of a tribal CM changes the dynamic or merely changes the party receiving tribal loyalty remains to be seen.

The educated tribal in the city practices what sociologists call cultural code-switching — constantly moving between tribal identity and mainstream professional presentation. The switching is situational: tribal identity may be emphasized when accessing reservation benefits but suppressed in social and professional settings where it carries stigma. Some tribal migrants minimize identity markers — adopting mainstream dress, suppressing tribal language use, avoiding discussion of village origins. The cost of upward mobility is cultural self-erasure.

This mirrors, at the individual level, what the diaspora chapter of The Leaving documents at the community level. The Odia who leaves for Bangalore faces identity fragmentation — speaking Odia at home and Kannada or English outside, maintaining cultural practices in isolation from the community that gives them meaning. The tribal who leaves for Bhubaneswar faces a doubled version of the same fragmentation. She is diaspora within diaspora, minority within minority, carrying an identity that the dominant system — both the national mainstream and the Odia mainstream — does not have the vocabulary to fully recognize.

What “Development” Means for 66,000 People

The thirteen Particularly Vulnerable Tribal Groups of Odisha number approximately 773,000 people. But the extreme cases — the Birhor (596 people), the Mankirdia (2,222), the Hill Kharia (2,800), the Chuktia Bhunjia (3,086) — total roughly 66,000 human beings whose communities are small enough to disappear within a single generation if the current trajectory continues.

Consider the Bonda of Malkangiri — approximately 12,000 people inhabiting 32 villages in 130 square kilometers of hill terrain. Genetically linked to the first wave of human migration out of Africa approximately 60,000 years ago. Their Bonda Development Agency has operated since 1976. Nearly fifty years of dedicated development programming. The result: 33 percent literacy (against a state average of 73 percent), 79 percent undernutrition among Upper Bonda children, 7.8 percent malaria prevalence, 14.18 percent hepatitis B prevalence, and villages like Barguda where not a single resident is educated because there is no school or anganwadi center.

Or the Kutia Kondh of Belghar, Kandhamal — 4.98 percent literacy, the lowest of any PVTG in Odisha. 25 percent malaria prevalence. This is a community where 95 percent of the people cannot read, where one in four has malaria, and where the Kutia Kondh Development Agency has been operating for decades.

Or the Dongria Kondh who won the Niyamgiri campaign — 6.35 percent literacy. The community that demonstrated extraordinary collective agency in a constitutional process is a community where more than 93 percent of the people cannot read. Their governance system generated twelve unanimous democratic decisions under intense external pressure. Their literacy rate is lower than many countries that are classified as failed states.

These numbers force a question that the development discourse prefers to avoid. If fifty years of dedicated development agencies, multiple constitutional protections, legislative frameworks, institutional bodies, and funding schemes have produced 4.98 percent literacy and 79 percent child undernutrition, the model itself has failed. This is not a matter of insufficient funding or poor implementation. It is a structural incompatibility between the development model and the communities it claims to serve.

The development model assumes a population that can interface with its delivery mechanisms: forms to fill, offices to visit, documents to present, languages to speak, calendars to follow. PVTGs interface with the forest, the season, the clan, the ritual specialist. The development model’s operating system and the PVTG’s operating system are fundamentally incompatible, and no compatibility layer has ever been built. The Bonda Development Agency delivers programs. The Bonda live in the hills. Between the program and the person, there is a gap that fifty years of institutional architecture has not bridged.

The recent progress on PVTG habitat rights recognition is a genuine positive signal. Six of Odisha’s thirteen PVTGs — Paudi Bhuyan, Juang, Saora, Chuktia Bhunjia, Hill Kharia, and Mankirdia — received habitat rights recognition under FRA in 2024. Habitat rights grant communities the right to their traditional territory, socio-cultural practices, livelihoods, and the protection of their natural and cultural heritage. This is, conceptually, the closest thing to a compatibility layer that the Indian legal system has produced: it recognizes that the territory, the practice, and the identity are inseparable, and it grants legal protection to the integrated whole rather than to disconnected components.

But seven PVTGs — including the Bonda, the Dongria Kondh, the Kutia Kondh, the Didayi, and the Birhor — still await habitat rights recognition. And recognition without enforcement is paper without forest. The track record of other legal protections — PESA rules unfinished for twenty-nine years, FRA claims rejected without recorded reasons, gram sabha votes overridden by administrative fiat — suggests that even the most promising legal instruments can be rendered inoperative by the system that is supposed to implement them.

The Compatibility Layer That Was Never Built

Every backwards compatibility problem in software has three possible solutions. You can force-migrate: retire the old system entirely and require all applications to be rewritten for the new one. You can freeze: keep the old system running unchanged and refuse the upgrade. Or you can build a compatibility layer: a translation interface that allows old applications to run on the new system while gradually enabling them to be upgraded on their own terms and at their own pace.

India’s tribal policy has oscillated between the first two approaches while never seriously attempting the third.

The force-migration approach is visible in every displacement project that treats tribal communities as populations to be “resettled” — moved from their territory, given cash compensation, and expected to integrate into the mainstream economy. Hirakud Dam displaced 150,000 people with 35 percent compensation paid. Kalinganagar acquired tribal land at Rs 15,000 to 30,000 per acre and transferred it to Tata Steel at Rs 3.5 lakh per acre. POSCO proposed a $12 billion steel plant in Jagatsinghpur. The pattern is consistent: acquire the land, compensate the people, expect them to adapt. The force-migration approach does not work because it destroys the applications (community, identity, livelihood, ecological knowledge) that cannot be reconstructed with cash.

The freeze approach is visible in the “rights-as-museum” tendency: the romanticization of tribal life that treats any change as loss, any modernization as corruption, any engagement with the market as selling out. This approach does not work either, because it condemns tribal communities to conditions — 4.98 percent literacy, 79 percent child undernutrition, 25 percent malaria prevalence — that no community would voluntarily accept if alternatives were available on their own terms.

The compatibility layer — the third approach — would look something like this. It would begin with the recognition that tribal governance systems are not relics to be studied but functioning institutions to be strengthened. The gram sabha, as it operated in Niyamgiri, demonstrated that traditional governance can interface with constitutional mechanisms to produce legitimate decisions. The challenge is to make this the norm rather than the exception. This means PESA rules that give gram sabhas genuine veto power — not the “consultation” that the Odisha draft rules substitute for consent. It means FRA implementation that processes claims with the evidence support that 87.5 percent of Forest Rights Committees report never receiving. It means development agencies led by tribal professionals, not by non-tribal bureaucrats posted from Bhubaneswar for two-year tours.

It would continue with education that does not force a choice between tribal language and academic success. The Multi-Lingual Education program is a start, but MLE that stops at primary school is a bridge that ends in mid-air. Extending MLE through secondary education, developing curricula in Kui, Santali, and Sora that include traditional ecological knowledge alongside standard science, and training tribal-language teachers at scale — these are the engineering investments that backwards compatibility requires.

It would include economic models that build on forest-based livelihoods rather than replacing them. The Bonda grow twenty crop varieties while maintaining biodiversity. This is not a system to be retired. It is a system to be augmented — with market access for minor forest produce at fair prices (not through the middlemen who currently capture most of the value), with value-addition facilities in tribal areas rather than in coastal processing centers, with cooperative structures modeled on the community ownership traditions that already exist.

And it would require something that no government in India has attempted: a genuine negotiation with tribal communities about what they want to change and what they want to preserve. Not consultation as a procedural formality. Not the gram sabha where the sub-collector reads the government’s plan and asks for a show of hands. An actual process in which the community’s own institutions — the Gudi, the pirha panchayat, the Manjhi-Pargana council, the Sindibor assembly — are recognized as legitimate counterparties with the authority to accept, reject, or modify the terms of modernization.

This is expensive. It is slow. It requires political will that no state government has demonstrated. It requires the mining industry to accept that gram sabha consent might mean “no.” It requires the forest bureaucracy to share jurisdiction with communities it has historically treated as encroachers. It requires the education system to invest in languages spoken by communities of a few thousand people. It requires, in short, the kind of engineering effort that software companies invest when they have no choice — when the old applications are too valuable to break, when the users are too important to lose.

Whether tribal communities in Odisha qualify as “too important to lose” in the political calculus of the Indian state is the question that everything else depends on. The constitutional answer is yes. The operational answer, measured in PESA rules unfinished for twenty-nine years, in FRA claims rejected without reasons, in PVTG literacy rates below 5 percent after fifty years of development agencies, is something else entirely.

What Persists

On an evening in a Kondh village in Kandhamal, a Jani prepares for Chaitra Parba. The spring festival requires specific rituals, specific songs, specific preparations that have been transmitted orally across more generations than anyone can count. The Jani’s authority derives not from a government appointment or a legal framework but from the community’s recognition that he holds knowledge no one else holds. In the same village, a young woman is studying for exams that will determine whether she gets into a college in Bhubaneswar. She speaks Kui with her grandmother and Odia with her textbooks. She has a smartphone. She follows accounts on Instagram that post about tribal rights.

These two facts exist simultaneously. The Jani’s knowledge and the smartphone’s connectivity. The oral tradition and the Instagram feed. The ritual calendar of Chaitra Parba and the academic calendar of exam season. They coexist not because someone designed a compatibility layer, but because human beings are better at running multiple operating systems than any machine. The young woman does not experience a “backwards compatibility crisis.” She experiences her life. She speaks Kui because that is what her grandmother speaks. She studies in Odia because that is what the exam requires. She watches Instagram because that is what connects her to a world beyond the village.

The crisis is not in her. It is in the system that forces her to choose. The system that funds MLE in primary school but not in secondary school. The system that recognizes the gram sabha’s authority in law but overrides it in practice. The system that celebrates Chhau dance at UNESCO while the Mayurbhanj communities that created it lack adequate healthcare. The system that appoints a tribal Chief Minister while leaving PESA rules unfinished.

The backwards compatibility problem is not about technology. It is about power. Who decides which applications get to keep running? In software, the vendor decides — Microsoft, Apple, Google. In the tribal identity crisis, the state decides. And the state has decided, through decades of action and inaction, that the old applications are expendable. Not explicitly. Not through policy declaration. But through the cumulative weight of unfinished rules, rejected claims, diverted funds, unbuilt schools, and the quiet assumption that modernity will solve the problem by making the question irrelevant — that the Kui language will fade, the sacred groves will be cleared, the mutha system will dissolve, and the tribals will become, in the census and in practice, indistinguishable from everyone else.

What persists is the refusal. Twelve gram sabhas voting unanimously against a two-billion-dollar mine. A Jani preparing for Chaitra Parba as he has always done. A young woman speaking Kui to her grandmother and Odia to her textbooks without experiencing this as a contradiction. The Saora Kudan painting ikons on a wall that a market would prefer to see on canvas. The Bonda growing twenty crop varieties on a hillside that a mining company would prefer to blast.

The backwards compatibility problem has a solution that the vendor rarely chooses: let the users decide. Let the communities that built the original system determine what gets upgraded, what gets preserved, and what gets translated into the new architecture. This is not romanticism. It is engineering. The best systems are the ones that respect the knowledge embedded in their predecessors. The worst are the ones that assume the new version knows better than the old one about everything, including things the new version has never encountered.

Tribal Odisha has encountered things the new version has not. How to grow twenty crops while maintaining biodiversity. How to govern a community without centralized authority. How to make a mountain into a god and defend it against a multinational corporation. How to resolve disputes through a council of elders and a collective decision. How to maintain a forest through spiritual taboo more effectively than through legal prohibition.

These are not quaint survivals from a primitive past. They are applications that work. The question is whether the new operating system will be designed to run them — or whether it will, as it has for seven decades, continue treating them as incompatible software to be deprecated, one breaking change at a time.

Sources

Tribal Identity, Literature, and Cultural Expression:

Niyamgiri and Dongria Kondh:

Pre-Colonial Tribal Governance:

Constitutional and Legal Framework:

PVTGs and Development:

Kandhamal:

Migration and Urban Tribal Experience:

Tribal Languages and Education:

Cross-References:

  • The Leaving, Chapter 7 — diaspora identity: /full_read/the-leaving/07-the-diaspora-mind.md
  • The Lord of the Blue Mountain, Chapter 6 — the promise and the wound (Sabara Dalapati, caste within Jagannath): /full_read/the-lord-of-the-blue-mountain/06-the-promise-and-the-wound.md
  • The Long Arc, Chapter 3 — the cathedral in the village (Hirakud, Rourkela): /full_read/the-long-arc/03-the-cathedral-in-the-village.md

Ethnographic Works Cited:

  • Verrier Elwin, The Religion of an Indian Tribe (Oxford University Press, 1955)
  • Verrier Elwin, The Murias and Their Ghotul (Oxford University Press, 1947)
  • Gopinath Mohanty, Paraja (translated into English by Bikram K. Das)
  • Gopinath Mohanty, Dadi Budha (1944)
  • Gopinath Mohanty, Amrutara Santana (Sahitya Akademi Award, 1955)

Source Research

The raw research that informs this series.