By Sabelo J. Ndlovu-Gatsheni
Celebrating Indigenous peoples without their knowledge systems is not celebration. Taking the UN’s 2026 theme on Indigenous midwives as its departure point, this piece argues that knowledge was a colonial technology, that the epistemic line is being redrawn in AI training data, and that cognitive sovereignty is urgent policy.
Every 9 August the world observes the International Day of the World’s Indigenous Peoples. The United Nations theme for 2026 is “Honouring Indigenous Midwives: Safeguarding Life and Well-being.” A midwife is a knowledge holder. To honor her is to make a claim about what counts as knowledge and about who is authorized to produce it. The theme is not cultural. It is epistemological.
The distinction matters, because the celebration of Indigenous peoples has for decades been permitted on one condition: that it stay cultural. Colors, textiles, ceremonies, and cuisines are welcomed. Cosmologies, sciences, medicines, jurisprudence, and philosophies are not or are welcomed only as folklore, that is, as material for anthropology rather than as knowledge with a claim on the present. This arrangement is what I call the cognitive empire, and it is the last empire standing.
How coloniality of knowledge was introduced and globalised
The modern world was not a natural growth. It was designed and assembled through identifiable colonial matrices of power. Race sorted humanity into those who could own and those who could be owned. The line, from the Treaty of Tordesillas of 1494 onward, divided the planet with a stroke of a pen. The map placed Europe at the center and ran the meridian through a London suburb, as though longitude began in England. International law and the nation-state form withheld sovereignty from those judged to lack civilization. And then there was knowledge.
The fifth matrix did the work of naturalizing the other four. The Westernized university, the discipline, the archive, the canon, and the peer-reviewed journal were built to license certain ways of knowing and to disqualify others. African cosmologies, Indigenous sciences, Asian philosophies, and women’s knowledges were reclassified as myth, superstition, or folklore. Boaventura de Sousa Santos calls the resulting structure abyssal thinking: an invisible line running through the world, on one side of which law and reason obtain, on the other side of which they are suspended, and beyond which knowledge is not merely judged wrong but produced as non-existent.
This is why political independence so rarely delivered epistemic independence. A country can hold elections, fly a flag, and sit in the General Assembly while its universities retain inherited curricula, disciplinary boundaries, languages of instruction, and criteria of excellence, so that each new generation of its intelligentsia continues to be certified by metropolitan standards. That is administrative independence. It is not liberation.
What was disqualified was rarely what failed.
Midwifery is the clearest case. Systems of birth attendance developed over centuries, combining physical care with kinship, ritual, botanical pharmacology, and community authority, were displaced by colonial and later developmental medical regimes that classified the midwife as an obstacle to progress. Some of that displacement saved lives, and intellectual honesty requires saying so. But much of it dismantled accumulated knowledge in places where the promised replacement never arrived. What the United Nations now recognizes is that these were systems of care with a rational structure and an intergenerational method of transmission, not superstitions to be tolerated until modernity caught up.
Sri Lanka offers a second and unusually precise illustration. The cascaded tank-village system of the dry zone, the ellanga gammana, is a connected series of tanks organized within a micro-catchment, storing, conveying, and reusing water from an ephemeral rivulet. It evolved over nearly two millennia; the island’s first large reservoir dates to the third century BCE. For most of the colonial and post-independence era, this was treated as picturesque antiquity while planners built elsewhere on other principles. In 2017, the Food and Agriculture Organization designated it a Globally Important Agricultural Heritage System, citing its extent, resilience to natural disasters, and biodiversity, and conferred the award in Rome the following year.
The knowledge did not improve in any phenomenal way until now. What changed was the ruler used to measure it. That is the entire argument in a single example. The question was never whether Indigenous and traditional knowledge works. The question is who holds the authority to declare that it does.
The Vedda make the same point about a people rather than a technology. Sri Lankans recognize the Vedda, who call themselves Wanniyala-Aetto, as the island’s original inhabitants, and the community is not a persecuted minority in the ordinary sense of that phrase. What was placed under suspicion was not their standing but their knowledge. Across three successive colonial dispensations, Portuguese from 1505, Dutch thereafter, and British until independence in 1948, their forest-based ecological knowledge, their healing practices, and their cosmologies were filed under the primitive. The scholarship that followed did not overturn that judgment so much as professionalize it: for the better part of a century the Vedda were studied as ethnographic objects, curiosities to be recorded before they disappeared, rather than as holders of a living science with any bearing on policy. Yet the livelihoods they sustain today rest on a detailed working knowledge of microclimates and on rules of extraction calibrated to what particular territories can bear, which is to say on precisely the competencies that environmental management now spends heavily to reconstruct. The colonial archive preserved the knowledge and denied, in the same gesture, that it was knowledge.
None of this is a new complaint, and it is important that it not be mistaken for one. The counter-archive is old, and it is substantially Asian as well as African. When the Ibadan historians established oral tradition as legitimate historical evidence, they were not adding African chapters to a European book but dismantling the premise that Africa stood outside history. When the Subaltern Studies collective recovered the autonomous political domain of peasants whom colonial and nationalist historiography alike had treated as raw material, and when Dipesh Chakrabarty drew the epistemological conclusion by provincializing Europe, they were doing the same work in another register. What we are being asked to do now is not to invent an argument but to finish one.
Neither reversal nor romance
The first is reversal: declaring European knowledge false and Indigenous knowledge true. That merely reinstalls monologue with a new speaker, and it will lose, because it concedes the premise that only one way of knowing may occupy the room.
The second is romance: pretending that precolonial worlds were free of hierarchy, patriarchy, and cruelty. They were not, and Indigenous and African scholars are usually the first to say so. A decolonial argument that requires an innocent past is a weak argument resting on a false one.
What is proposed instead is an ecology of knowledges. This does not mean abolishing European thought, which would be both impoverishing and impossible. It means ending its monopoly, placing it in conversation rather than in command. It means that knowledge produced in Sinhala, Tamil, isiNdebele, Urdu, or Kiswahili carries the same presumption of truth as knowledge produced in English. It means African philosophy speaking with German philosophy, Indigenous science with laboratory science, and the midwife with the obstetrician as a colleague rather than as an informant.
The line is now being redrawn in code
The epistemic line is being re-inscribed in our decade at planetary scale and at machine speed. The infrastructure through which billions of people now think through search, social platforms, and large language models is concentrated in a small number of corporations headquartered in a single country. These systems are trained on corpora that are overwhelmingly English, overwhelmingly Northern, and overwhelmingly recent. What is missing from a training corpus is not refuted; it is simply absent. And absence is a far more efficient technology of disqualification than anything a nineteenth-century missionary school achieved. A model that has never encountered a tradition of traditional medicine or customary law will not argue against it. It will generate a world in which it was never said.
For Sri Lanka, and for the Asia-Pacific more broadly, this converts an ethical question into a procurement question with a short deadline. Cognitive sovereignty, the capacity of a society to govern the systems in which its own citizens think, requires compute and data under domestic or regional jurisdiction, corpora and models in national languages, digitization of archives on terms set by the communities that hold them, and clear rules on the extraction of traditional knowledge into commercial training sets. Governments that file this under cultural affairs will find it decided for them by vendors.
Three measures
Standing. Give Indigenous and traditional knowledge systems institutional standing rather than ceremonial recognition: accreditation pathways for practitioners, budget lines in health and agriculture ministries, and a place in the curricula of the universities that certify the country’s professionals.
Jurisdiction. Free, prior, and informed consent should govern data as it already governs land. A community’s knowledge should not enter a commercial corpus on terms it did not set and from which it does not benefit.
Language. The International Decade of Indigenous Languages, running to 2032, will be judged by whether it produced usable digital infrastructure, corpora, models, keyboards, translation tools or another shelf of reports.
The stake
Epistemic freedom is not a supplement to political freedom. It is its condition. Honoring Indigenous midwives, if the phrase is to mean anything beyond a hashtag, means conceding that the world holds more than one legitimate way of knowing how a life begins, and therefore more than one legitimate way of knowing how a world should be organized.
That concession is small in cost and enormous in consequence. It is the beginning of what I have called reworlding: the recognition that a world which was made can be remade, and that those assigned to its basement have designs of their own. Reworlding is not, in the end, an academic question. It concerns knowing in order to live, rather than living in order to know.
Sabelo J. Ndlovu-Gatsheni is Professor of History and Tier 1 Canada Research Chair in Pluralistic Societies: Epistemic Pluralism and Ecologies of Knowledges at the University of Calgary, Alberta, Canada. His books include Epistemic Freedom in Africa (Routledge) and Beyond the Coloniality of Internationalism: Reworlding the World from the Global South (CODESRIA).
