The Colonial Spine of Classification

This post is part of a series that explores how metadata can be used as a site of resistance, refusal, and poetic subversion. From classification to linked data, the series investigates how cataloging practices can encode oppression, and how they can be reimagined to challenge dominant systems and speak from the margins. Check all the posts in this section's index.

Introduction

If the individual fields of a metadata record represent the surface —the skin— then classification systems form the skeletal infrastructure beneath. They are not merely supportive scaffolding; they are the epistemic spine of the information system, determining how knowledge is anatomically arranged, navigated, and legitimized. In this architecture, systems like Dewey Decimal Classification (DDC), Universal Decimal Classification (UDC), and Library of Congress Subject Headings (LCSH) do not operate as neutral containers. They act as regulatory regimes — taxonomic apparatuses that not only reflect but impose ideological order.

In Foucauldian terms, classification can be seen as a disciplinary technology: one that partitions knowledge, assigns identity, enforces visibility, and polices boundaries. These systems do not just arrange books or records; they perform ontological labor, encoding worldviews into the very act of organization. The assumption that knowledge can be universally structured according to fixed hierarchies is itself a colonial gesture — one that abstracts, isolates, and reifies knowledge forms within a framework designed to serve particular cultural and institutional priorities.

To critique metadata fields without interrogating the classificatory logic that governs them is to miss the deeper architecture of control. It is like protesting a locked door while refusing to examine the blueprint of the house — a house built by and for the epistemic elite. Classification systems determine what can exist within a catalog, what is intelligible within a system, and what is rendered residual, invisible, or illegible. They are not accidental tools; they are intentional acts of epistemic governance.

The Anatomy of Oppression

Classification systems are infrastructures. And like all infrastructures, they carry ideologies — quietly, efficiently, persistently.

Dewey Decimal Classification (DDC)

The Dewey system, born in the 19th century, reflects its Protestant, white, North American origins with painful clarity. Nowhere is this more evident than in class 200, which is devoted to religion: Christianity occupies the vast majority of the range from 200 to 290, while all other world religions are compressed into the final digits, from 295 to 299 — a symbolic relegation that speaks volumes. This isn't just an issue of outdated balance or imperfect coverage; it reveals a hierarchical worldview in which Christianity is not merely a religion, but the religion — the implicit norm against which all others are measured and diminished.

The same structural bias recurs in other major classes. In the 300s, dedicated to social sciences, Western political, legal, and economic systems dominate the conceptual terrain, leaving little room for Indigenous or non-state forms of governance and sociality. In the 900s, which encompass history and geography, the narrative arc is similarly centered on Europe and North America, with the rest of the world appearing as a fragmented afterthought — reduced to numbered subdivisions and colonial categories. The geography of classification mirrors the geography of empire.

Universal Decimal Classification (UDC)

Despite its name, the Universal Decimal Classification (UDC) is neither universal nor ideologically neutral. Constructed atop Dewey's foundational structure and molded by the priorities of European academic institutions, UDC presents itself as a flexible system — primarily through its extensive use of auxiliaries, which allow for modifications and combinations. However, that very flexibility is governed by strict, predefined frameworks. The so-called "common" and "special" auxiliaries offer only the illusion of openness: they permit variation, but only within limits set by the system's original logic.

Notably, "special auxiliaries" often position Indigenous or localized knowledge categories as peripheral modifiers rather than central classifications. They are appended to dominant structures rather than allowed to constitute primary classes of their own. This reinforces a hierarchical structure in which Eurocentric concepts retain primacy, while other epistemologies are filtered through the lens of secondary annotation.

UDC's decimal precision may appear neutral, even scientific, but it functions as a mask — one that conceals deep-seated assumptions about what constitutes legitimate knowledge. A cosmology grounded in Quechua relationality or a taxonomy shaped by ritual ecology cannot be meaningfully expressed through decimal fragments. The system's apparent adaptability breaks down when confronted with worldviews that refuse to be atomized.

Library of Congress Subject Headings (LCSH)

The Library of Congress Subject Headings (LCSH) functions as a thesaurus of violence — not always through overt slurs, but through the framing and filtering of meaning. Its harm lies in its terminology, yes, but more profoundly in its insistence on naming from the perspective of power.

For decades, the system used the term "Illegal aliens" to refer to undocumented migrants — a dehumanizing label that conflated legal status with identity and reinforced carceral, xenophobic narratives. It took years of organized protest by librarians, scholars, and activists to have the term changed, revealing the system's resistance to reform even in the face of sustained critique.

Similarly, Black life and experience were long cataloged under reductive subject strings like "Blacks — Social conditions," a phrase that describes inequality as an ambient feature rather than the outcome of structural racism. The wording subtly but effectively redirects attention away from systems of power and toward vague descriptions of circumstance.

Queer identities fared no better: well into the 1990s, LCSH collapsed non-normative sexualities under terms such as "Sexual deviation," inheriting a clinical, pathologizing vocabulary from medical and legal discourses that criminalized LGBTQ+ existence. These terminologies weren't anomalies — they were systemic expressions of the worldview embedded in the classification.

Each subject term functions as a lens, and every lens has a history. In LCSH, that history is rarely neutral. The system demands that all knowledge objects pass through its predefined portals to be included at all. And those portals are shaped by the same hierarchies, exclusions, and cultural biases that pervade the broader institutional apparatus. To be cataloged, one must first be translated —and often distorted— by the language of authority.

Why Reform Always Fails

Every few years, institutions announce reforms with great ceremony. The Library of Congress updates offensive subject headings — but only after protracted internal debates and political controversy. The Dewey Decimal Classification makes symbolic gestures toward inclusivity by diversifying religious or geographical categories — typically by shifting decimal points or adding minor subdivisions. Committees across the library world propose "inclusive terminology," draft "cultural sensitivity" guidelines, and establish equity task forces to advise on classification policies.

But none of these interventions touch the spine. They rearrange the organs, repaint the skin, and celebrate incremental "progress," all while leaving the foundational architecture intact. The classificatory logic —the epistemological grammar underpinning these systems— remains untouched.

And that is not an accident. The system is not malfunctioning. It is operating exactly as it was designed to: to stabilize a particular worldview, to regulate knowledge according to Western, extractive, and imperial standards of legibility. These are not glitches in an otherwise neutral system; they are features of an ideological machine.

Rupture, Not Repair

If repair is structurally foreclosed, then rupture becomes not only necessary, but urgent. Not a rhetorical or symbolic rupture, but one enacted through daily, deliberate acts of subversion. Across library, archival, and digital platforms, practitioners and communities are already developing strategies that confront the limits of classification by building parallel logics, embedding refusal, and reclaiming space within and beyond the system.

One such strategy is the creation of what I call overlay classifications. Rather than attempting to reform the dominant colonial taxonomy from within, some communities choose to build parallel systems that exist alongside it, occupying the same bibliographic or digital infrastructure but operating according to different ontological premises. A leading example is Ngā Upoko Tukutuku (Māori Subject Headings), developed in Aotearoa New Zealand as a collaborative effort between Māori knowledge holders and national library institutions. These headings articulate categories grounded in whakapapa (genealogy), whenua (land), and wairua (spirituality), reflecting an Indigenous semantic world that resists reduction to LCSH terms. The two systems coexist, but they do not speak the same language — and that difference is the point.

A second approach is the use of what I call dual schemes. Digital platforms such as Mukurtu CMS provide Indigenous communities with tools to describe and organize their collections using their own cultural protocols, vocabularies, and taxonomies, while still maintaining optional compatibility with broader standards like Dublin Core or LCSH. This model does not require communities to choose between sovereignty and interoperability. It respects internal logic while permitting strategic engagement with external systems without epistemic compromise.

A third form of intervention can be more spatial than semantic: what I call fugitive shelving. In this practice, librarians and community archivists may organize materials according to alternative logics —thematic, affective, seasonal, or ritual— rather than according to standardized call numbers. A book might be shelved next to an object, a photograph next to a song, not because they share metadata values, but because they share meaning in a cultural or spiritual sense. These arrangements may be undocumented, oral, ephemeral, and dynamic — resisting fixity and undermining the hegemony of the catalog. They may function as living classification systems that operate outside the grid of institutional logic.

Finally, some resist within the catalog itself through acts of what I call radical relabeling. When harmful subject terms cannot be deleted, they can be annotated. In public-facing systems, librarians add tooltips, footnotes, disclaimers, or contextual notes that signal the violence embedded in the terminology. These paratextual interventions do not erase the harm, but they refuse to normalize it. By publicly naming the politics of classification, they open space for critique. This practice —sometimes described as "cataloging as dissent"— transforms the catalog from a site of compliance into a site of confrontation.

Each of these tactics —overlays, dualities, spatial alternatives, and discursive interruptions— represents a refusal to surrender knowledge to the logics of domination. They are not perfect solutions, but they are cracks in the colonial spine. And from those cracks, other systems can grow.

Tactical Questions

If you intend to confront classification in your own practice —whether as a librarian, archivist, researcher, or developer— the work must begin with critical questioning. But these are not rhetorical prompts for philosophical reflection; they are concrete, diagnostic tools for epistemic audit.

Begin by asking: Who authored this taxonomy, and for whom was it built? No classification system emerges in a vacuum. Every hierarchy, every term, every subdivision reflects decisions made by someone — often in a specific institutional, cultural, and political context. Identifying the authorship of a system is the first step in exposing its assumptions.

Then ask: What kinds of knowledge are excluded, collapsed, or rendered incoherent within this structure? What cannot be expressed? What is forcibly generalized? Often, entire domains —oral tradition, relational epistemologies, non-binary identity, ritual practice— are either missing entirely or awkwardly inserted into ill-fitting categories.

Consider also: Which epistemologies are granted primacy? What counts as legitimate knowledge? What is assumed to be stable, nameable, and universally classifiable? Most traditional systems elevate written, discrete, object-centered, and Western knowledge forms, while others are filtered, fragmented, or domesticated to fit.

Now shift perspective: What would this catalog look like if it had been built from another worldview? One grounded in seasonal change, kinship networks, or cosmological relation? Asking this question not only reveals the limitations of the current system: it gestures toward alternative logics that are not only possible, but already in practice elsewhere.

And finally, ask: Where are the gaps, and who benefits from them? The absence of a term, the rigid structure of a hierarchy, the invisibility of a category — these are not just oversights. They serve someone. Every omission or distortion is a political act, often with real consequences for visibility, access, citation, and legitimacy.

These are not abstract critiques. They are fieldwork tools. To ask them is to begin dismantling the scaffolding of authority that classification pretends to neutralize.

Toward a Post-Classificatory Imagination

What if classification were not a grid, but a story? Not a table, but a ritual? Not a hierarchy, but a relation?

In dominant Western epistemologies, classification is often framed as a matter of control: to know something is to name it, to assign it a place, to fix its coordinates within an abstract and universal schema. The logic is extractive, taxonomic, and fundamentally disembodied. It seeks to isolate the object from its context, flatten its relations, and repackage it for retrieval. But other knowledge systems operate differently.

Across many Indigenous, oral, and ritual epistemologies, classification is not imposed from above — it is emergent, embedded in the rhythms of land, language, and community. Categories do not remain static across contexts; they shift with the seasons, the speakers, the rituals in which they are invoked. A plant, for example, may belong to one category when it is blooming, another when it is harvested, and still another when it is sung into a story. A narrative may be epistemological in one context, ontological in another, and cosmological in yet another. In such systems, classification is not a means of containment: it is a mode of relation, a way of caring for knowledge by locating it within a living web of responsibility.

This is not a cosmetic variation. It is a fundamental ontological difference. Where Western systems demand fixed, mutually exclusive categories, relational systems tolerate ambiguity, embrace multiplicity, and understand knowledge as dynamic and co-constitutive. They refuse the idea that all knowledge must be legible to external standards of order — especially when those standards are rooted in histories of conquest, discipline, and institutional gatekeeping.

We do not need to wait for permission to begin building systems that reflect these other logics. They already exist —embedded in memory, in ceremony, in practice— though they have long been excluded from cataloging standards and information infrastructures. The task now is not to "reform" the old system by adding new categories to old hierarchies. The task is to confront it directly, expose its colonial foundations, and build in its margins and against its logics — to construct what it cannot imagine.