To Curate Is to Govern

This post is part of a series that examines how colonial knowledge systems in libraries, archives, and museums erase Indigenous, oral, and ecological ways of knowing — and explores how they might be dismantled and reimagined from the perspective of the Global South and the margins. Check all the posts in this section's index.

Archival Infrastructure as Territorial Technology

In the age of global conservation, biodiversity archives have become crucial instruments for organizing ecological knowledge. But beneath their technical veneer lies a deeper function: governance. Archives do not merely record species — they enact jurisdiction. They decide which knowledges count, under what terms they circulate, and who gains the authority to speak for life on Earth. In this sense, to curate biodiversity is to govern it.

While this has long been true of colonial-era cabinets and herbaria, today's infrastructure —digital, distributed, and metadata-driven— is more insidious. It operates under the guise of neutrality, often obscuring its entanglement with political economies, funding regimes, and institutional control. Librarians, archivists, and data stewards are not external to this dynamic. We are its facilitators.

Or its saboteurs.

The Long Archive of Extraction

The (colonial) archive of nature (and of everything else, actually) did not begin with spreadsheets. It began with voyages. With pressed specimens, numbered notebooks, and the taxonomic dreams of empire. From Carl Linnaeus to Joseph Banks, classification served a clear purpose: to translate unfamiliar ecologies (and epistemologies) into manageable, nameable, ownable forms. The archive was a way of enclosing the wild — of making it legible for extraction.

Modern biodiversity databases continue this trajectory. Platforms like GBIF (Global Biodiversity Information Facility) or iDigBio aggregate millions of records: names, coordinates, collection dates, institutional codes. The architecture is distributed, but the logic is centralized. Knowledge flows upward. Metadata standards —Darwin Core, ABCD, TDWG vocabularies— become the lingua franca of global interoperability.

What is obscured is how these systems disembed knowledge from its original ecological, cultural, and epistemic contexts. The archive no longer looks like a glass cabinet. But its function remains enclosure. It governs with the same imperatives: stabilize, sort, extract, control, translate.

Standardization and the Suppression of Plural Ontologies

Standardization is the foundation of modern archival practice. It enables integration, cross-referencing, and reuse. But it also imposes a single epistemic frame.

In biodiversity curation, for example, standardization flattens ontological diversity. It converts seasonal, spiritual, or kin-based taxonomies into static morphological descriptors. A plant known locally as a "rain-bringer," used only in ritual after the second full moon of the rainy season, becomes a simple Latin binomial. Its metadata may be specimen number, herbarium code, or GPS point. Nothing of its original meaning remains.

This is not accidental. As Bowker and Star noted in Sorting Things Out, classification systems encode the priorities of the institutions that build them. What cannot be sorted gets left out. And what gets left out cannot be governed — at least not on its own terms.

This is not merely an issue of "cultural sensitivity." It is a form of epistemic suppression. Ontologies that center relation, obligation, or cosmology are not merely different worldviews: they are different worlds. And they cannot be mapped through Darwin Core fields.

Participation Without Power: The Crisis of Co-Curation

Faced with critiques of top-down, vertical science, many institutions now promote "participatory", horizontal approaches: citizen science, community monitoring, or Indigenous observation networks. These are often framed as inclusive innovations: ways of democratizing data collection.

But participation without control is not co-curation. It is epistemic laundering.

In most cases, local actors feed observations into centralized platforms. But the interpretive architecture remains untouched. Categories are fixed. Vocabularies are locked. Communities may collect data, but they do not define what counts as data, or what meanings it is allowed to carry.

This asymmetry is rarely acknowledged. Participation is measured by metrics —number of contributors, volume of data points— not by authority over framing. Yet framing is where meaning is made. And meaning is what determines ecological policy, funding priorities, and legal designations.

In biodiversity governance (but not only), the power to name is the power to act. Without that power, participation becomes a mechanism for extracting local knowledge under the illusion of collaboration.

Curation as a Sovereign Act

In institutional contexts, curation is often defined as a technical practice: metadata assignment, digital preservation, quality control.

But in community contexts, curation is a sovereign act. It is the process through which a group determines how its knowledge is stored, activated, and transformed across generations.

Such curation does not aim for universal legibility. It aims for continuity. It is embedded in responsibility, not interoperability. And it often resists the logics of acceleration, openness, and permanent availability that define platform-based infrastructures.

To curate, in this context, is not to digitize. It is to decide what may be shared, when, with whom, and why. It is to hold knowledge in trust. Sometimes through silence. Sometimes through opacity. Sometimes through refusal.

This mode of curation cannot be always outsourced. It cannot be always captured by controlled vocabularies or metadata schemas. It requires librarians and archivists to relinquish their role as custodians and adopt a different posture: that of relational stewards operating under epistemic consent.

The Archive as Biopolitical Border

The archive is often treated as a neutral space: a storehouse, a repository, a memory aid. But it is, in fact, a biopolitical machine.

The categories it establishes in terms of biodiversity —endangered, invasive, endemic— are not mere descriptions. They are designations that affect land use policy, conservation funding, and bioprospecting permissions. They draw lines between life that must be protected and life that must be managed. Or extinguished.

Same thing with other areas and epistemes beyond biodiversity.

As Achille Mbembe has argued, modern regimes of power operate through the differential valuation of life. Archives actively participate in this process. They determine which species knowledges, and memories matter, which spaces merit preservation, and which information is admissible in policy discourse.

These decisions are rarely framed as political. But they are. They shape social, cultural, and territorial futures. And they do so through the backend architecture of metadata, not the front-facing rhetoric of inclusion.

Those managing knowledge, memory and biodiversity repositories must confront this reality. We are not just facilitating access. We are helping to draw the borders of the ecological and epistemic world — and deciding who gets to cross them.

Toward a Counter-Archival Practice

If curation is governance, then counter-curation is resistance. And it must begin with a rejection of infrastructural inevitability.

There is nothing natural about the current architecture of data (including biodiversity data). It reflects decisions —epistemic, political, and technical— made by institutions with specific agendas. Challenging these structures requires more than critique; it demands the articulation of fundamentally different archival practices.

This means recognizing plural ontologies not as supplemental "local knowledge fields," but as legitimate organizing principles in their own right. It means developing relational indexing models where authority is distributed, and where knowledge is always contextualized by its conditions of use. It requires abandoning the fetish for clean, decontextualized data, and instead embracing forms of documentation that are layered, partial, iterative, and accountable to their origins — even when they resist coherence. And it entails designing archival infrastructures that make space for refusal: where communities can withhold, redact, reframe, or silence their data on their own terms.

None of this is about making existing systems more inclusive. It is about enabling different systems, governed by different logics, serving different ends.

If You Cannot Surrender Control, You Are Not a Steward

The work of librarianship cannot be reduced to information management. It is a political, ethical, and ontological responsibility. If we accept that libraries, archives, and museums (among others) structure the future —that they determine what life is seen, protected, and acted upon— then our metadata fields, taxonomies, and cataloging choices are acts of worldmaking.

And if we are serious about epistemic justice, we must ask ourselves: Whose world are we making?

Curation is not neutral. To curate is to govern. And if we cannot surrender control over the knowledge we hold —to those who live it, carry it, and care for it— then we are not stewards.

We are simple administrators (and accomplices) of extraction.

  This entry mirrors the chronicle "Curating Biodiversity," a narrative reflection on the same theme.