In 1970, the Rhodesian government enacted censorship preventing official publications from stating as fact that Great Zimbabwe had been built by Africans.
Read that again, because the date is the shocking part. Not 1870. Not the height of the scramble. Nineteen seventy — colour television, moon landings, my parents’ generation already alive.
The site in question is a stone city in the south of what is now Zimbabwe, built in dry-stone masonry over two or three centuries, with walls running eleven metres high in places. It is the largest ancient structure in sub-Saharan Africa. The people who built it were the ancestors of the people living around it.
This was not a mystery in 1970. Gertrude Caton-Thompson had excavated the site in 1929 and concluded it was African work. Peter Garlake, Rhodesia’s own Inspector of Monuments, had spent the second half of the 1960s researching it and reached the same conclusion with considerably more evidence.
What happened to Garlake is the thing I want to sit with.
The Fort Victoria demand
In the Rhodesian Parliament, the member for Fort Victoria District, Colonel George Hartley, argued that the theory of indigenous construction was pure conjecture, and that Garlake should give equal platform to the alternative — that Great Zimbabwe had been built by light-skinned people from somewhere else.
Garlake resigned in 1970 and left the country. He went to the University of Ife in Nigeria, finished his manuscript there, and published Great Zimbabwe in 1973. He did not return to Zimbabwe until 1981, after majority rule.
Meanwhile the Smith government commissioned work from writers willing to argue the alien-builders case. Bruwer had published in 1965. Gayre followed in 1972.
So we have the full apparatus in one place: a state that needed a particular claim about African capacity, an inconvenient archaeological record, a professional pushed out for reporting it, and paid substitutes brought in to muddy the finding.
That is not ignorance. Ignorance is passive. This was production.
What Mudimbe named
V.Y. Mudimbe published The Invention of Africa in 1988, and the concept that has stayed with me from it is the colonial library.
His argument, compressed to the point of unfairness: Africa as an object of knowledge was largely constructed by European discourse. The categories, the questions, the vocabulary of what counts as evidence of civilisation — all of it was assembled elsewhere, by people with reasons, and then applied here. The library is not a building. It is the accumulated body of writing about Africa that later writing has to argue with, including African writing.
The trap Mudimbe identifies is that even refutation happens on the library’s terms. You end up proving that Africa had cities, had writing, had states — using a checklist of civilisational markers you did not write and might not have chosen.
Great Zimbabwe is the clean case because the enforcement was written into law and you can date it. But most of the library was never that explicit. Hugh Trevor-Roper, Regius Professor at Oxford, told a television audience in 1963 that African history amounted to little worth teaching — the past of pre-European Africa being, in his framing, movement without significance. He was not censoring anyone. He was describing the state of a field that had been constructed to have nothing in it.
The claim of absence was never a finding. It was a decision about where to look, hardened over two centuries into a fact.
Two writers who described the machinery from inside
Binyavanga Wainaina published a piece in Granta in 2005 that took the form of instructions for writing about the continent. It is satire, and its method is simply to list the conventions honestly: which landscapes are permitted, which human types recur, what emotional register the reader expects, how Africans may appear and in what supporting capacity. Set out plainly as rules, the conventions look absurd. That was the point, and I would rather you read it than have me summarise it further.
Chimamanda Adichie’s contribution came four years later, and the detail that lodges is autobiographical. As a child in Nigeria she wrote stories in which the characters were white, ate apples, played in snow, and discussed the weather — because those were the only stories she had read, and so those were the things she understood stories to contain.
That is the library working at the level of a seven-year-old’s imagination, before any argument can reach her. By the time you are old enough to contest the frame, you have already been furnished with it.
The mechanism, stated plainly
Strip out the theory and the thing runs in four steps.
First, someone with authority declares an absence — no history, no development, no capacity for stone architecture.
Second, that declaration shapes what gets funded, taught, published and collected. Nobody excavates what is agreed to be empty.
Third, the resulting thinness of the record is cited as confirmation of the original declaration.
Fourth — and this is the step people miss — the loop becomes self-sustaining without anyone needing to lie. By 1963 Trevor-Roper did not have to suppress anything. The field genuinely was thin, because of step two. He was reporting accurately on a situation that the library itself had produced.
Great Zimbabwe only required an explicit law because the evidence was too large and too obviously well-built to ignore. Eleven-metre walls are difficult to leave out of the record. Most of the continent’s past was easier to simply not look for.
Where the library lives now
I would not spend a Monday on 1970 if this were only history.
The library’s present form is quieter. It shows up in what gets treated as the default assumption in a room where a decision is being made.
It shows up in the pricing of risk. African sovereign debt carries spreads that a number of studies have argued exceed what the underlying fundamentals justify — a premium attached to perception rather than to default history. That gap is the library expressed in basis points.
It shows up in datasets. The corpora that train large language models are overwhelmingly English, Chinese and a handful of European languages. Over two thousand languages are spoken on this continent and the vast majority have almost no digital text. A model that has read very little in Hausa or Bamileke will be poor at those languages, and that poverty will be experienced by users as evidence that the technology does not work here, rather than as evidence about who built the training set.
It shows up in citation. African scholarship on African subjects is systematically under-cited relative to work on the same subjects produced in the global North. The library still has editors.
And it shows up in the pitch meeting, where a founder spends the first ten minutes of an investor call establishing that the market is real. That is unpaid labour, performed by the person with the least time, to correct a frame someone else installed.
The counter-archive
Here is the capability side, and it is not rhetorical.
The response to a manufactured absence is a documented presence, and that work is underway. Timbuktu’s manuscripts are being digitised, tens of thousands of pages of them already public. African-language NLP projects — Masakhane’s work across dozens of languages, the various national corpus efforts — are building the text that does not exist yet, deliberately, from scratch. African journals and university presses are publishing scholarship that used to have to travel abroad to count.
None of that is glamorous and all of it is underfunded. But it is the correct response, because it attacks step two rather than step three. Arguing with the conclusions of the library is slow work. Changing what is in it is slower, and permanent.
The constraint is money and attention. Digitisation is recurring operational cost with no revenue model, which makes it the hardest kind of thing to fund and the first thing cut. The Timbuktu manuscripts sit in Bamako’s humidity while the budget for climate control gets discussed.
What I take from Garlake
The detail that stays with me is not the censorship law. It is that Garlake was Rhodesian. His father had commanded the Rhodesian armed forces. He was, by birth and position, inside the settler establishment, employed by the state as its Inspector of Monuments.
And he still resigned, left, and published, because the evidence said what it said.
The library is powerful, but it is not sealed. It is maintained by individuals making individual decisions about what they are willing to certify. Some of them, including some with every incentive to go the other way, decline.
That is worth remembering when the frame you are working against seems too large to move. It is not one thing. It is a very large number of small decisions, and some of them are yours.
Day 6 of 60. This series names one real capability and one real constraint in every piece. Sources: EBSCO, History of Censorship in Zimbabwe, on the 1970 statute; Garlake obituaries in The Standard and The Zimbabwean, 2011–12; Shadreck Chirikure et al., “Peter Garlake (1934–2011), Great Zimbabwe and the politics of the past in Zimbabwe,” Azania, 2012; V.Y. Mudimbe, The Invention of Africa, 1988; Binyavanga Wainaina, Granta 92, 2005; Chimamanda Ngozi Adichie, “The Danger of a Single Story,” 2009.