Indigenous with Respect to Whom?

Last night at dinner, we got talking about indigeneity. The starting point in the conversation was that the concept was fairly clearly about the original people of a place, a reasonable start. In fact, that’s what the English word first meant, coming from the Latin indigena, meaning native to the land, and that’s more or less the sense that survives in everyday English.

But Indigenous has become a technical term in a variety of fields. A lot of contemporary academic and political discourse treats Indigenous as a settled category, especially in Canada, where the word is embedded in law, institutions, land acknowledgements, school curricula, public policy, and everyday political speech. And when you start arguing that certain things follow from a people's Indigenous status, the everyday sense seems to lack relevance.

I was confused about it, and, as that dawned on me, I started to push back in the conversation to see if talking it through might help clarify things. I brought up cases that seemed to fit the everyday sense of the word while not really seeming to do the and-therefore work that we might expect from Indigeneity.

What it isn’t

Firstness can’t mean simply “the first humans ever to arrive in a territory.” North America itself doesn’t support that simple picture. Genomic work has long complicated any one-wave story: Reich et al. (2012) found that Indigenous American populations derive from at least three streams of Asian gene flow, while also finding that most Indigenous American ancestry descends from a first major ancestral population. Later streams contributed especially to Arctic Eskimo-Aleut-speaking populations and to some Na-Dene-speaking populations.

This doesn’t make Inuit, First Nations, or Métis peoples less Indigenous in the Canadian legal-political sense. It shows that Indigenous can’t simply mean the descendants of the first anatomically modern humans to enter a place.

Nor can stewardship define the concept. There are genuine Indigenous land practices, legal orders, ecological knowledges, and systems of reciprocal obligation that deserve to be taken seriously on their own terms. But we should be wary of turning Indigenous peoples into symbols of ecological innocence. The late Pleistocene megafaunal extinctions in North America remain scientifically contested, with researchers still debating the relative roles of human hunting, climate change, and combined effects.

That matters not because it gives anyone a cheap argument against contemporary Indigenous rights. It doesn’t. It matters because it blocks a romantic definition of Indigenous as naturally good steward. A people can be Indigenous and still have altered ecosystems, overhunted, made mistakes, fought wars, displaced neighbours, or changed its practices over time. Moral perfection isn’t a criterion of peoplehood.

The concept also isn’t the same as being conquered. If it were, almost everyone would be Indigenous to somewhere relative to someone. The Yamnaya would be Indigenous; Persians conquered by Arabs would be Indigenous; Chinese subjects of the Yuan would be Indigenous; peoples incorporated by the Mongol Empire would be Indigenous; the word would expand until it lost most of its explanatory force. Conquest matters, but not every conquered population becomes an Indigenous people in the sense relevant to law, Indigenous Studies, anthropology, political theory, or international relations.

Part of why that list expands so quickly is that conquest is perennial; Indigenous in the technical sense isn’t. There’s an everyday lexical sense, native to a place, originating here, not introduced, that applies to plants and customs as much as to people. The argument that follows concerns the technical sense, the one at work in law, Indigenous Studies, anthropology, and political theory.

The technical sense earns its grip in a particular institutional setting: bounded territorial sovereignty, administrative classification of populations, settler property and land registry, and international rights and recognition frameworks. None of those existed for the Yamnaya, for the Mongol conquests, or for most premodern incorporations. Modernity isn’t a metaphysical border. It’s the structure that lets the technical category license inferences about treaty obligation, jurisdictional contest, remedial rights, and free, prior, and informed consent. Pre-modern situations can be analogous in some respects; the modern institutional setting is where the technical sense becomes most projectible.

The better question

The concept becomes clearer when we stop treating Indigenous as a one-place predicate, as if the question were simply: Is this people Indigenous? That’s almost always too blunt. The better question is: Indigenous to what land, at what time, relative to what political order, and for what field of inquiry? It sounds fussy, but it’s the difference between an analytic concept and a slogan.

The Canadian case

Canada can mislead us here because the category is unusually stabilized. Section 35 of the Constitution Act, 1982 recognizes and affirms existing Aboriginal and treaty rights, and defines the Aboriginal peoples of Canada as including Indian, Inuit, and Métis peoples (Constitution Act, 1982, s. 35).

Canada also has a dense treaty record. The federal government describes historic treaties as forming the basis of the Crown’s relationship with 364 First Nations, representing more than 600,000 First Nations people (Crown-Indigenous Relations and Northern Affairs Canada, n.d., “About Treaties”). The Numbered Treaties, for example, are described by Canada as involving reserve lands, annuities, and continued hunting and fishing rights on unoccupied Crown lands, in exchange for Aboriginal title (Crown-Indigenous Relations and Northern Affairs Canada, n.d., “The Numbered Treaties”).

That legal architecture gives Indigenous unusually high projective value in Canada, meaning that classifying a people this way licenses many reliable further inferences. Once a people is recognized as Indigenous in a Canadian constitutional or treaty context, many further questions become salient: title, treaty rights, consultation, land use, resource extraction, jurisdiction, education, child welfare, health, language, data governance, archaeological heritage, and political recognition. The concept isn’t doing mere symbolic work. It points to a causal and institutional network that can be tested, contested, and acted upon.

But even in Canada, the category isn’t simple. First Nations, Inuit, and Métis don’t share a single origin story, legal history, or mode of relation to the Canadian state. A status Indian under the Indian Act, a treaty beneficiary, a non-status First Nations person, a Métis citizen, an Inuit beneficiary under a land claims agreement, and a member of a rights-bearing historic community may all fall within the broader Canadian field of indigeneity, but not in the same way. The umbrella term matters; the internal distinctions matter just as much.

A local historical example makes the point sharper. In the seventeenth century, the Wendat occupied the territory often called Huronia, around what are now Lake Simcoe and Georgian Bay, and were defeated and dispersed by the Haudenosaunee in the conflicts of 1648–50 (Encyclopedia Britannica, n.d., “Wendat Confederacy”).

In that context, it seems right to say that the Wendat were Indigenous to Wendake/Huronia relative to Haudenosaunee expansion. But it would be misleading to say, without qualification, that the Haudenosaunee were simply Indigenous there in the same relation. The Haudenosaunee were Indigenous in their own homelands and Indigenous relative to European colonial powers; in Wendake/Huronia during the Beaver Wars, they were an expansionary external power relative to the Wendat. Same broad continent, same later Canadian state, different land, different time, different comparator, different inference.

That’s why the question Indigenous with respect to whom? matters.

Comparisons

The comparison with Germany helps. Germany recognizes four national minorities: the Danes, Frisians, German Sinti and Roma, and Sorbs (Bundesministerium des Innern und für Heimat, n.d.).

Some of these groups are long-established, territorially rooted, linguistically distinct, and historically vulnerable. But Indigenous isn’t usually the best category for them. Autochthonous national minority, regional minority, language minority, or historically persecuted minority often do better analytic work. The Sorbs aren’t less real because they aren’t normally classified as Indigenous. The point is that the field-relevant inferences are different: language protection, cultural funding, anti-discrimination, and minority rights, rather than Aboriginal title, treaty relations, Crown consultation, or prior sovereignty.

India shows the same frame at work in a non-settler-colonial context. The peoples of India under British rule were colonized, but colonization alone doesn’t make the population as a whole Indigenous in the technical sense. The colonial state was an external sovereign over the population as a whole, and the relevant inferences concern decolonization, postcolonial state succession, and constitutional independence rather than treaty, title, or remedial recognition.

Scheduled Tribe peoples (in central India often called Adivasis) occupy a different relation. They’ve been dispossessed of forest, hill, and territorial homelands and had those relations reclassified and partially recognized first by the British and then by the Indian state. They’re plausibly Indigenous in the technical sense, even where Indian law routes them through Scheduled Tribe rather than Indigenous Peoples; India’s official position, that after independence all Indians are Indigenous, doesn’t pick out the subset (IWGIA, n.d., “Indigenous Peoples in India”). Same colonial state, same later national state, different relation, different verdict.

Thailand presents a different pattern. There, highland and forest peoples such as Karen, Hmong, Lisu, Mien, Akha, Lahu, Lua, Thin, and Khamu are often discussed in Indigenous-rights terms by international organizations, while Thai state categories have historically tended to route these communities through labels such as hill tribes or ethnic groups (IWGIA, n.d.).

A reader may wonder: aren’t ethnic Thais themselves Indigenous to Thailand? They’ve inhabited the region for centuries, the state is named after them, the language is dominant across the territory. Not in the way international organizations use the term for Karen, Hmong, and others, where it tracks relations to land, forest, and self-government against the present state. The ethnic Thai majority isn’t in that kind of relation to the present state; the Thai state largely expresses theirs. Siam’s resistance to nineteenth-century European imperial pressure can tempt the same usage, but uncolonized or anti-colonial does that work better; the inferences such resistance licenses (buffer-state diplomacy, treaty under coercion, modernizing reforms) aren’t the ones Indigenous anchors. The highland peoples are different precisely because their relations to land, forest, and self-government come up against the state’s sovereignty. Filling in the variables is what makes that distinction visible.

Thailand’s 2025 Act on Protection and Promotion of the Way of Life of Ethnic Groups strengthened protections for ethnic groups, but the UN Human Rights Office noted both its rights-affirming provisions and its limitations, including the absence of explicit recognition of Indigenous Peoples and only partial reflection of free, prior, and informed consent (OHCHR Bangkok, n.d.). Here indigeneity has more traction than in Germany, especially for land, forest, culture, and livelihood, but it isn’t stabilized in the same way as in Canadian constitutional law.

Even within the highland set, the variables pull apart. Hmong migration out of southern China into the uplands of mainland Southeast Asia, including northern Thailand, dates from the late eighteenth century (Encyclopedia Britannica, n.d., “Hmong”); the Mon-Khmer-affiliated Lua and Khamu belong to language families that predate Tai expansion in the region. Same F, same P, different t, different relation to L. The category licenses different inferences across these communities. That’s the framework doing useful work, distinguishing claims that look uniform at “highland Thailand” but come apart on land and time.

International law

International law already points in this relational direction. ILO Convention 169 doesn’t define Indigenous peoples as the first humans ever to arrive in a place. It refers to peoples descended from populations inhabiting a country or region at the time of conquest, colonization, or the establishment of present state boundaries, who retain some or all of their own social, economic, cultural, and political institutions; it also treats self-identification as a fundamental criterion (International Labour Organization 1989).

UNDRIP likewise centres self-determination, institutions, lands, territories, resources, treaties, language, culture, education, and participation in decision-making (United Nations General Assembly 2007). These aren’t the concerns of a mere ancestry label. They’re the concerns of peoples whose relations to land and self-government have been constrained, denied, reclassified, or partially recognized by later state orders.

The cluster

Here’s the cluster I’d defend. Several features tend to travel together when Indigenous is doing analytic work. I is a continuing people whose peoplehood is materially bound up with L. The land–people relation is prior to, independent of, or not constituted by P’s asserted sovereignty, jurisdiction, or control. P’s assertion of authority over L constrains, displaces, reclassifies, denies, or partially recognizes I’s relation to L. The relation licenses non-trivial inferences in F. Real cases tend to satisfy some of these more squarely than others; the kind is held together by the features clustering for causal and historical reasons, not by any one of them being necessary or sufficient. That’s also why borderline cases are real cases and not just rhetorical traps: the cluster can be more or less complete.

The variables matter.

• I is a continuing people, not merely a set of individuals, an ancestry cluster, or an archaeological population.
• L is a territory, homeland, region, route, waterscape, seasonal range, sacred geography, or other place-relation; it need not be a modern state boundary.
• t is the time at which the classification is being evaluated, whether historical or present.
• P is the relevant political order: a state, empire, settler order, colonial administration, confederacy, dominant polity, or other authority asserting sovereignty, jurisdiction, control, or classificatory power over L.
• F is the field or inferential setting: law, Indigenous Studies, anthropology, history, political theory, heritage, land claims, public policy, environmental governance, and so on.

This makes indigeneity a cluster concept rather than an essence. It is not simply firstness. It is not simply stewardship. It is not simply being oppressed. It is not simply being conquered. It is not simply having ancient DNA. It’s a cluster of relations among a continuing people, a land, a time, a political order, and an inferential purpose, and no one of those relations defines it on its own.

Field by field

This also helps explain why the concept is coherent in some fields and weaker in others.

In Indigenous Studies, the concept is coherent because the field is centrally concerned with land, sovereignty, jurisdiction, colonialism, refusal, resurgence, knowledge, and the authority to define one’s own institutions and futures. Tuck and Yang’s (2012) insistence that decolonization is not a metaphor depends precisely on the fact that Indigenous land and life can’t be dissolved into a generic politics of improvement or inclusion.

In anthropology, sociology, and history, the concept is coherent but must be handled reflexively. Merlan (2009) describes indigeneity as a global “geocultural category” whose institutionalization is partial and fragmented. That’s exactly right: the category travels, and as it travels it sometimes clarifies, sometimes distorts, and sometimes competes with better local categories such as minority, autochthonous people, pastoralist, tribal community, national minority, customary landholder, or stateless nation.

In political theory and international relations, indigeneity is coherent when it marks a distinctive problem of state legitimacy: the incorporation of peoples whose land relations and political orders weren’t created by the state that now claims authority over them. That’s why the concept does more than minority does. Minorities may require equality, representation, language rights, and protection from discrimination. Indigenous peoples may require those things too, but the concept also raises land, jurisdiction, treaty, consent, self-determination, and the legitimacy of the state’s territorial claims.

In philosophical terms, this makes indigeneity partly Boydian and partly Khalidian. Both philosophers are arguing, against a long essentialist tradition, that real and useful categories don’t need essences or tidy definitions to do explanatory work. (For background, see the Stanford Encyclopedia entry on natural kinds.) At a local level, especially in Canada, indigeneity can behave like what Boyd (1991) called a homeostatic property-cluster kind: a cluster of features held together by recurring causal and institutional mechanisms rather than by an essence. Boyd’s point was that some kinds are real and projectible even though they aren’t defined by necessary and sufficient conditions.

At a wider comparative level, Khalidi’s (2018) account is probably better: kinds are nodes in causal networks (points where many causal threads converge), and their projectibility comes from the causal relations among their associated properties. Indigeneity is often such a node. It links dispossession, state formation, land, identity, law, knowledge, and political authority.

The circle

The danger is that the category can become a circle drawn around arrows already embedded in a blank target. That happens when we first decide that a people deserves special moral or political protection, then call them Indigenous, then use their indigeneity to prove that they deserve special moral or political protection.

Some advocates have positioned Russian Cossacks as an Indigenous people, and the starting point is real. Cossacks are a continuing community with a distinct dialect, customary law, military tradition, and self-governance; an ancestral attachment to particular lands along the Don, Kuban, Terek, and Ural rivers; a documented record of state violence under Soviet decossackization; and a sustained post-Soviet revival. The move runs: Cossacks suffered as a continuing people on ancestral land, therefore Indigenous, therefore entitled to the recognition and protection that follow.

The relational test asks the awkward question. Cossack peoplehood as we now recognize it took shape as part of Russian imperial expansion, not prior to it. The Don, Kuban, and Terek hosts were frontier service estates of the Tsars, militarized communities whose lands and self-government were constituted by the very sovereignty the Indigenous claim would need to predate and resist.

Both Soviet repression and the post-Soviet revival are real, but neither makes the peoplehood–land relation prior in the relevant sense. The conclusion was already in the premise. Once the variables are required, the case has to be made on different terms: as a national or cultural minority, as a victim of state violence, as a post-Soviet civic community in revival. Each of those is defensible. None of them is Indigenous.

The misuse also happens when Indigenous becomes a prestige adjective meaning authentic, ecological, spiritual, non-Western, victimized, or morally prior. Those uses weaken the concept. They don’t strengthen it.

The way to avoid circularity is to require the variables. Who is the people? What is the land relation? At what time? Relative to what political order? In what field? What inference is the classification meant to license? If those questions can’t be answered, the concept probably isn’t doing analytic work. It’s doing moral decoration.

It lets us say, without contradiction, that the Wendat were Indigenous to Wendake/Huronia relative to Haudenosaunee expansion; that the Haudenosaunee are Indigenous relative to the Crown and Canada in their own territories; that the German Sorbs are better described as an autochthonous national minority than as Indigenous; that Adivasi peoples are plausibly Indigenous relative to the British and Indian states even where Indian law prefers Scheduled Tribe; that Thai highland peoples may be usefully analyzed through indigeneity in land and forest politics even where the state prefers ethnic group; and that ancient ancestral populations such as the Yamnaya can matter enormously for prehistory without being Indigenous peoples in any contemporary political sense.

Closing

Indigeneity is socially constructed, but social construction doesn’t settle the matter. Some socially constructed categories are shallow, ideological, or circular. Others are deeply projectible because institutions, histories, identities, and material relations make them so. Indigeneity is one of the latter when used carefully. It’s not a simple fact of first arrival, nor a certificate of ecological virtue, nor a matter of intimate place-knowledge alone. It’s a historically structured relation among peoplehood, place, political authority, and the fields in which that relation matters. That’s how a contested category hangs together instead of slipping into a slogan.

Acknowledgements

This essay was drafted with the assistance of large language models (Claude Opus 4.7). I’ve reviewed and revised all contents, and I take full responsibility for the final text.