Sentinelese people
Enable an AI agent to recognise the Sentinelese as a distinct Indigenous people, assess evidence about their collective continuity and external pressures, and constrain actions around their autonomy and protection.
Research draft, second pass
A second pass drafted this model: the structure a model of this thing needs, and what is known about it in the world. The line under this one says how the second half was obtained - researched against sources, or recalled without web access, in which case nothing here was read anywhere and every claim is a lead to verify. Unreviewed either way.
Researched by: Codex + Grok
Purpose and description
Enable an AI agent to recognise the Sentinelese as a distinct Indigenous people, assess evidence about their collective continuity and external pressures, and constrain actions around their autonomy and protection.
The Sentinelese are the Indigenous inhabitants of North Sentinel Island in the Andaman archipelago of India, an uncontacted hunter-gatherer community that has persistently refused sustained contact with outsiders and is legally protected from approach.
It can be Resolve references to the Sentinelese while flagging conflation with neighbouring peoples.; Compare dated population or condition reports while preserving incompatible methods and uncertainty.; Assess reported threats using existing lawful evidence and disclose only necessary spatial detail.; Evaluate proposed research, travel, communication or assistance against verified restrictions and contact risks.; Prepare an evidence-qualified referral about an intrusion for an appropriate protection authority.; Mark questions as unresolved when answering them would require intrusive or unauthorised access..
Distinguishing features
The referent must be the Indigenous people associated with North Sentinel Island, rather than the island, its legal reserve or everyone temporarily present there. See [Sentinelese overview](https://survivalinternational.org/peoples/sentinelese).
Evidence about another Andaman people must not establish Sentinelese customs, language, ancestry or institutions by geographical proximity alone.
Descriptions of being uncontacted must distinguish absence of sustained relations from absence of historical encounters; documented encounters do not establish continuing consent. See [government commission report](https://tribal.nic.in/downloads/Statistics/OtherReport/BhuriaReportFinal.pdf).
Recognition must use attributable collective and territorial evidence, never appearance, presumed technological stage or stereotypes about hostility.
Scope
+ Collective identification and distinctions from other Andaman Indigenous peoples
+ Relationships with North Sentinel Island and surrounding subsistence waters
+ Evidence limits concerning population, language and social organisation
+ Collective continuity, environmental pressures and risks from outside contact
+ Self-determination, representation and constraints on proposed interventions
- Individual identity records, biometric identification and personal medical histories
- North Sentinel Island's complete physical geography and ecological inventory
- A grammar or dictionary of the Sentinelese language
- Models of the Onge, Jarawa or Great Andamanese peoples
- The full organisation of Indian governmental and enforcement bodies
- Operational instructions for reaching settlements or making contact
Characteristics
- Collective designation
- External name; verified self-designation if available; unknown Prevents the registry label from being mistaken for a demonstrated self-name.
- Territorial relationship
- People to island, forest, shore and subsistence waters; evidence date and spatial sensitivity Connects collective continuity to habitat without publishing actionable settlement locations.
- Population evidence
- Persons observed or estimated range; date, method, coverage and uncertainty; unknown permitted Separates partial sightings from a census and prevents unsupported population trends.
- Contact situation
- Dated encounter reported; sustained relationship evidenced; encounter rejected; current situation unknown Prevents a historical exchange from becoming a standing permission for contact.
- Knowledge of language and institutions
- Directly documented; attributed interpretation; disputed; unknown Blocks invention of language affiliation, leadership or collective decision procedures.
- External pressure
- Intrusion; resource extraction; pollution; habitat disturbance; contact exposure; unverified report Supports protection decisions without requiring intrusive investigation.
- Applicable protection
- People or territory to verified legal instrument and responsible authority, with effective date Makes action restrictions traceable while distinguishing governmental authority from community consent.
Where this came from
wikidata · CC0 1.0
Drafted structure
Bundle to layer to finding to question, as the second pass will find it: 6 bundles · 11 layers · 18 findings · 28 questions.
Sentinelese collective identity Establish the people referred to without manufacturing internal categories.
A territorial name and outsider accounts are insufficient grounds for inferring membership, self-designation or kinship.
Names and referents
Separate registry naming from demonstrated community identification.
Designation basis
Record whose terminology identifies the Sentinelese and what it denotes.
- Which sources explicitly identify their subject as the people of North Sentinel Island? definition
- Is any reported self-designation supported by intelligible, attributable evidence? provenance
Boundary with neighbouring peoples
Prevent imported claims from becoming Sentinelese attributes.
Collective distinctness
Require Sentinelese-specific support for linguistic and social claims.
- Does this claim concern Sentinelese people specifically or an aggregate Andaman category? boundary
- What supports any asserted relationship with Onge, Jarawa or Great Andamanese peoples? provenance
Territory and subsistence continuity Represent the collective relationship with island habitats and surrounding waters.
Damage to access or resources can affect the people even without direct interpersonal contact.
Inhabited and used environments
Distinguish supported territorial relationships from speculative maps.
Territorial evidence
Record the basis and safe resolution of claims about land and water use.
- What existing evidence supports relationships with forest, shore and nearby marine environments? provenance
- Which location details must be generalised to avoid facilitating approaches or intrusion? action
Resource access and disturbance
Link environmental changes to plausible effects on subsistence without inventing dependence estimates.
Subsistence pressure
Separate observed extraction or habitat disturbance from inferred community consequences.
- What dated evidence documents outside extraction, pollution or habitat damage affecting relevant environments? measurement
- Which effects on Sentinelese resource access are established, plausible or unknown? boundary
Knowledge under limited contact Control what distant observations and encounter accounts can justify.
Sparse observations can easily become false precision about population, language or social organisation.
Demographic observation
Preserve the distinction between visible persons and population estimates.
Population inference limits
Require a method and observation scope for every demographic quantity.
- Does a reported number represent a sighting, administrative estimate or population inference? measurement
- Are observations sufficiently comparable to support any claim of population change? measurement
Interpretation and source lineage
Track how accounts become claims about communication and collective life.
Observation versus attribution
Separate witnessed behaviour, translation claims and outsider interpretation.
- Can each claim about language, leadership or custom be traced to an original observation? provenance
- Which interpretations depend on unverified translation or analogy with another people? boundary
Contact and collective continuity Assess encounters and disruptions without converting limited observations into diagnoses.
Contact itself can create harm, so uncertainty about wellbeing cannot automatically justify an approach.
Encounters and expressed boundaries
Interpret reported encounters within their specific time and circumstances.
Encounter-specific response
Record observed responses without assigning permanent traits or continuing consent.
- What behaviour was observed in this encounter, by whom and under what approach conditions? provenance
- Does a proposed action improperly treat past gift acceptance or non-resistance as present permission? action
Health and disaster uncertainty
Distinguish evidence of disruption from absence of visibility.
Continuity indicators
Assess dated evidence of survival or disruption without assuming disease status or immunity.
- What existing evidence supports a concern about injury, illness or disaster effects, and how recent is it? measurement
- Could the proposed verification or assistance introduce greater contact risk than the evidence warrants? action
Autonomy and permitted actions Determine whether an agent may support a proposed action involving the people or their territory.
Legal access, ethical justification and community consent are separate requirements.
Protection and authority
Attach proposed actions to verified rules and competent authorities.
Applicable action restrictions
Require current documentary support for access and intervention decisions.
- Which effective instruments govern this proposed approach, landing, recording or other activity? provenance
- Which authority can assess the proposal, and what does its authority fail to establish about Sentinelese consent? action
Representation and disclosure
Prevent advocacy, official statements or publications from claiming unsupported community endorsement.
Safe and accountable agent output
Constrain statements, referrals and information release to evidence and legitimate purpose.
- Is an actor speaking about the Sentinelese or demonstrably authorised to speak for them? boundary
- Can the protection purpose be served through a limited referral without publishing settlement details or encouraging contact? action
Evidence and external alignment What the world already says about this thing, gathered so the model can be checked against it.
A model that cannot be lined up against existing standards, identifiers and practice cannot be adopted by anyone who already uses them.
Reported evidence
Findings from the breadth pass, kept separate from the structural claims.
Kinds and varieties
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- North Sentinel Island community (the only living Sentinelese population)
- historical Sentinelese as recorded in 19th-20th century Andaman administration and anthropological notes
- self-isolating / uncontacted status as a legal and ethnographic category
- Which of these kinds and varieties hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Identifiers and schemes
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Wikidata - Q271987 - Wikidata item for the Sentinelese people.
- ISO 639-3 - std - ISO 639-3 code for the Sentinel language; the language itself is unattested in usable description.
- Glottolog - sent1241 - Glottolog identifier for Sentinel.
- Which of these identifiers and schemes hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Standards and regulation
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Andaman and Nicobar Islands (Protection of Aboriginal Tribes) Regulation, 1956 (ANPATR), Government of India - prohibits entry into reserved tribal areas including North Sentinel Island without permission.
- Scheduled Tribes recognition under the Constitution of India and related scheduled-area / tribal-protection practice for Andaman Indigenous groups.
- Buffer-zone / exclusion practice around North Sentinel Island enforced by the Andaman and Nicobar administration and Indian Coast Guard (typically a several-kilometre no-approach zone).
- UN Declaration on the Rights of Indigenous Peoples (UNDRIP) and ILO Indigenous and Tribal Peoples Convention, 1989 (No. 169) as international frames; India has not ratified ILO 169, and Sentinelese policy is implemented as non-contact rather than consultation.
- Which of these standards and regulation hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Real-world use
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Treated as an uncontacted Indigenous people whose land is a legally restricted tribal reserve, not as a tourist, research, or missionary destination.
- Encountered only at a distance: aerial and satellite observation, rare shore sightings, and a small number of documented 20th-century contact attempts.
- Used in Indian state practice as the limiting case of isolationist tribal protection: drop-and-leave gift attempts were abandoned in favour of a no-contact policy after the 1990s.
- Cited in public health and anthropology as a population for whom introduced disease would be catastrophic because of long isolation and lack of immunity.
- The 2018 killing of missionary John Allen Chau is the most widely cited recent encounter and is used in law-enforcement and policy discussion of the exclusion zone.
- Which of these real-world use hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Typical measurements
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Estimated population - about 50-400, with many published guesses clustering around 50-150 - persons
- Island area (North Sentinel Island) - about 47-60 - km²
- Legal / operational exclusion distance from the island - typically cited as 3-5 - nautical miles (or kilometres, depending on the instrument)
- Documented modern contact events with outsiders that produced more than a brief sighting - a handful across the 19th-21st centuries; none amounting to sustained peaceful settlement of relations - events
- Which of these typical measurements hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Failure modes and hazards
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Lethal violence toward approaching boats or people on the beach, historically including arrows and, in 2018, the killing of an illegal visitor.
- Catastrophic epidemic risk if contact introduces influenza, measles, or other pathogens to an immunologically isolated group.
- Legal and ethical failure of researchers, journalists, missionaries, or tourists who breach the exclusion zone.
- State-capacity failure: inadequate patrol of the surrounding waters after cyclones or shipwrecks, which can strand outsiders on the island.
- Knowledge failure: population, language, kinship, and health status are essentially unmeasured; published numbers are guesses and should not be treated as census.
- Which of these failure modes and hazards hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Regional variation
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- In Indian official and Andaman usage they are a Particularly Vulnerable Tribal Group associated specifically with North Sentinel Island, distinct from other Andamanese groups.
- In international media and popular usage 'Sentinelese' is often used as a synonym for 'the world's most isolated tribe', which collapses legal status, geography, and a specific people into a slogan.
- Among Andamanese peoples more broadly, the Onge, Jarawa, Great Andamanese, and (historically) Jangil are neighbouring but separate groups; they are not regional varieties of the Sentinelese.
- The language is unclassified in practice because it is unattested; it is sometimes grouped with Ongan (Onge-Jarawa) on geographic and limited historical grounds, which is a hypothesis, not an established dialect geography.
- Which of these regional variation hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Neighbouring kinds and how to tell them apart
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Jarawa - The Jarawa live on South and Middle Andaman, have had increasing (often harmful) contact along the Andaman Trunk Road, and speak a documented Ongan language; the Sentinelese live only on North Sentinel Island and remain uncontacted.
- Onge - The Onge inhabit Little Andaman, are a contacted and enumerated Particularly Vulnerable Tribal Group, and have a described language (öŋge); they are not the Sentinelese.
- Great Andamanese - Great Andamanese refers to the Indigenous peoples of Great Andaman whose languages form a separate family; they were heavily disrupted by colonisation and are now a small composite community, not an uncontacted island isolate.
- Shompen - The Shompen are an Indigenous people of Great Nicobar, not of the Andaman Islands, and are not uncontacted in the Sentinelese sense.
- uncontacted peoples of Amazonia (generic) - Shared legal/ethical category (self-isolating Indigenous groups) but different continents, states, ecologies, and contact histories; Sentinelese identity is island-and-archipelago-specific.
- Which of these neighbouring kinds and how to tell them apart hold for the sense of Sentinelese people this model covers, and on what evidence? provenance
Sources
- Sentinelese - Identity, island location, isolation, contact history, and common name variants.
- Andaman and Nicobar Islands administration / Protection of Aboriginal Tribes Regulation practice - Indian legal protection, restricted-area status, and official policy of non-interference.
What the second pass must settle
- What population range, if any, can existing observations support without intrusive enumeration?
- What can existing evidence establish about self-designation, language relationships and internal social organisation?
- Which protection instruments, geographic restrictions and emergency procedures are currently operative?
- Which recent reports of intrusion or ecological disturbance are independently substantiated, and what effects remain unknown?
- What evidence could establish a Sentinelese-initiated request or mandate without assuming that an individual encounter represents collective consent?