For a long time, anthropology treated cultural difference as a matter of beliefs about one world. The “ontological turn” asks a sharper question: what if there isn’t just one world awaiting multiple interpretations? What if there are many worlds, partially overlapping, occasionally commensurable, often not?
The ontological turn coalesced in the 2000s across anthropology and STS (science and technology studies), drawing on work by Philippe Descola, Eduardo Viveiros de Castro, Annemarie Mol, Bruno Latour, and later Martin Holbraad and Morten Axel Pedersen. The pivot is from epistemology (how people know the world) to ontology (what the world is). Instead of translating others’ concepts into our categories—“they believe spirits cause illness”—researchers attempt to let those concepts compose the analytical terms: spirits as persons, jaguars as subjects, objects as social actors. The aim isn’t whimsy; it’s method.
Theory Snapshot
Three moves define the turn:
-
Take concepts as world-making, not mere opinions. Viveiros de Castro’s “Amerindian perspectivism” argues that beings (humans, animals, spirits) share a culture but inhabit different natures—a form of multinaturalism. A jaguar sees blood as manioc beer; a human sees beer as beer. The difference isn’t a mistaken belief but a different bodily position in a pluriverse.
-
Map ontological regimes. Descola proposes four modes—naturalism, animism, totemism, analogism—to describe how societies sort interiorities and physicalities. These are not museum labels; they’re living arrangements that organize ethics, property, and politics.
-
Enactment over representation. Mol’s The Body Multiple shows how “atherosclerosis” is done differently in clinic, lab, and ward—multiple realities enacted through practices. Ontologies are not just thought; they are performed in institutions, instruments, and routines.
Method in Practice
In this approach, analysis slows down. Rather than correcting “local beliefs” with scientific truth, ethnographers ask: what realities have to hold for these practices to make sense? If a fisherman in Oceania treats a reef as kin, the question becomes practical: how do kinship obligations allocate harvest, repair damage, or sanction extraction? Likewise, if data engineers say “the model learned,” the ontological turn doesn’t scoff at anthropomorphism; it tracks how workflows, dashboards, and legal fictions materialize a learning entity with rights, liabilities, and budgets.
Case in Point
Consider human–wildlife conflict. A conservation NGO frames elephants as endangered resources in a shared ecosystem. Farmers frame elephants as political agents rerouting through their fields when the state neglects irrigation. Meanwhile, local ritual specialists might treat elephants as emissaries from ancestral domains. Each enactment distributes responsibilities differently: insurance schemes, legal personhood for rivers, ritual compensations, electric fencing. The policy catastrophe begins when one ontology masquerades as neutral “reality” and the others are demoted to “belief.”
Friction and Critique
The ontological turn attracts heat. Critics warn it can slide into anything-goes relativism, evacuate power analysis, or romanticize “others” while ignoring capitalism, gender, and race. The best work counters by coupling ontology with politics: who gets to decide which world is operative in court, clinic, or climate model? Latour’s diplomacy of “composition” and decolonial scholars’ insistence on pluriversality push the turn toward institutions—treaties, standards, audits—where worlds negotiate under asymmetry.
Why It Matters
Climate change, pandemics, and AI governance are ontological battlegrounds. Are viruses agents or objects? Is a dataset property, commons, or kinship archive? Is a forest a carbon sink, a cathedral of beings, or an infrastructure? These are not semantic disputes. They script which futures are fundable, insurable, and livable. An ontological sensibility equips us to design with difference—to build policies and interfaces that don’t erase worlds in the name of efficiency.
The ontological turn’s wager is modest and radical: let analysis be remade by the realities it studies. Not to flatten conflict but to render it intelligible on its own terms. We don’t need consensus about one true world to act together; we need institutions capable of hosting worlds—with procedures for friction, translation, and repair.
