The Building That Decided Who Had a History
You walk into a colonial-era records office in Nairobi, Kolkata, or Bridgetown, and you understand the hierarchy before you read a single word. Floor-to-ceiling shelves. Acid-free folders. Humidity held at roughly 45 to 55 percent relative moisture, controlled with the seriousness of a bank vault. The administrative correspondence of district officers survives in extraordinary condition, cross-referenced and boxed and waiting. The oral testimonies of the people those officers governed are not there at all. Not damaged. Not misfiled. Absent, because the building was never designed to hold them.
That absence is not accidental. It is architectural.
What Gets a Shelf and What Gets a Floor
The physical logic of a nineteenth-century British colonial archive followed a precise hierarchy. Paper records produced by the colonial administration occupied purpose-built metal or hardwood shelving, often raised off the floor to protect against monsoon flooding and termite damage. The British Records Office model, exported across the empire after roughly the 1870s, specified minimum shelf depth, maximum stack height, and the use of buff-coloured folders that shielded documents from light. These were serious, expensive engineering decisions, and the money shows.
Contrast that with what happened to petitions submitted by local populations. Historians working in the National Archives of India have described finding bundles of vernacular-language petitions, sometimes thousands of them, stored in cloth-tied packets on the floor of outer rooms, or in outbuildings with no climate protection. The packets survived, when they survived, because of luck rather than design. The building's circulation plan placed European administrative correspondence in the inner, climate-stable rooms. Everything else was peripheral. Literally.
This wasn't pure malice. It was something more durable: an unexamined assumption baked into professional practice, the kind that doesn't need a villain because it runs on autopilot. The archivists trained in London learned that a record worth keeping was a record produced by a legitimate administrative actor. In the colonial framework, that meant European officials. The rest was, in the vocabulary of the period, "background material" or "local correspondence," categories that implied lower preservation priority and received lower-quality physical storage as a direct result.
The Folder as a Verdict
Consider the mechanics of how a document gets classified at intake, because this is where physical design and ideological assumption meet most precisely.
When a district officer's report arrived at a colonial records office, it entered a registration system. It received a unique reference number, a subject heading, a cross-reference to related files, and physical housing in a numbered folder within a numbered box on a numbered shelf. Retrieval was possible years later in minutes. The system was genuinely impressive, a feat of bureaucratic engineering.
When a village headman's complaint arrived, addressed in Tamil or Yoruba or Malay, it typically received a summary translation, not a full transcription, attached to the relevant district officer's file as an enclosure. The original document was sometimes retained, sometimes discarded. Even when retained, it existed as an appendage to the European record rather than as a record in its own right. No independent folder. No independent reference number. No independent shelf location. To retrieve it, you had to already know about the European document it was attached to, which is to say you had to approach indigenous voices only through European intermediaries, by design.
That structure persists. Archivists at the Bodleian Libraries and at the British Library have noted that large categories of colonial-era documents from colonised populations remain accessible only by working through the administrative files of the people who governed them, a research path that subordinates one account to another before the historian has read a single line. The folder, in this sense, was a verdict on whose account of events deserved to stand alone.
Two Researchers, One Archive, Different Centuries
Consider two historians working in the same colonial archive fifty years apart. The first, working in the decade after independence, finds the European administrative records immaculate: bound, indexed, cross-referenced, stored at stable temperature. She writes a political history of the colony that is detailed and well-sourced, because the sources are physically available to her in a way that makes research tractable. Her footnotes run deep.
The second wants to write a social history from the perspective of the colonised population. She finds the administrative records still excellent. She also finds that the petitions, the land-use maps drawn by local surveyors, and the records of customary courts were stored in a detached outbuilding that lost its roof in a storm sometime in the 1980s. Roughly thirty percent of that material survived. What remains is water-damaged, partially illegible, and uncatalogued. Her footnotes are thinner. Her argument must hedge where the first historian's could assert. The building's original design has shaped not just what survives but the epistemological confidence with which history can be written, and thirty percent survival is not a rounding error.
This is not a hypothetical scenario. Archivists in former British East Africa have documented precisely this pattern of differential survival.
What People Get Wrong About This
The common objection runs like this: colonial archives preserved more than they destroyed, and criticising their design is anachronistic, judging nineteenth-century institutions by twenty-first-century standards.
This objection misunderstands the argument, and I think it misunderstands it deliberately. The point is not that colonial archivists were uniquely villainous or that they should have anticipated postcolonial historiography. The point is that physical design encodes assumptions, and those assumptions had consequences that stretched far beyond the intentions of anyone who drew the original blueprints. A building that allocates acid-free folders to one category of document and a damp outbuilding to another is making a statement about value, whether or not anyone in that building thought of it that way. Good intentions are not a conservation strategy.
The distinction also matters between the dramatic bonfire of records, which happened, and is well documented, and the quieter, structural disappearance of certain voices through the ordinary operation of a system never designed to keep them. The latter is in some ways harder to remedy, because it leaves no single moment of decision and no single person to blame. Think of limescale building up inside a pipe until the flow stops: the cumulative logic of a thousand small choices about shelf placement, each individually unremarkable, collectively decisive. That kind of loss does not announce itself. It shows up decades later as a gap in the footnotes, a hedge in an argument, a community whose land claims the archive describes as "inconclusive."
Post-independence governments inherited these buildings and their hierarchies. Some invested in retroactive cataloguing and conservation. Many did not have the resources, and the gap between what was spent on European administrative records and what was available for everything else is, where figures survive, striking. The physical infrastructure of forgetting outlasted the political system that built it.
The Archive Is Always an Argument
So ask yourself: when a historian writes that the evidence for a particular community's land claims is "sparse" or "inconclusive," how much of that sparseness is history and how much of it is humidity?
The tendency to treat archives as passive repositories, neutral containers for whatever documents happen to end up inside them, is a comfortable fiction. The physical history of colonial record-keeping makes that position very difficult to hold. The shelf depth, the humidity control, the folder system, the circulation plan of the building itself: each was a decision, and each decision reflected a theory about whose activities constituted history worth preserving.
When researchers today find rich administrative records and thin or absent records from colonised populations, they are encountering the long-term effects of those decisions as much as they are encountering the past itself. The archive is not where history is stored. It is where certain histories were chosen, and the rest were left to the weather.