Culture · Essay
Whose memory does a country keep? The multicultural archive problem
A society that holds several pasts at once has to decide whose records get kept, in which language, by whom. The Windrush scandal showed what happens when the state gets it wrong: the documents were destroyed, and then proof of them was demanded.Written to last.
By Confinity Editorial · 2026-04-23 · 6-minute readQuiet tools, kept out of the way.
Every national archive was designed by somebody, for somebody. The parish registers assume a church. The census assumes a fixed address. The naming conventions assume one culture's way of building a name. For most of their history, the great public record systems worked well for the majority they were built around and badly for everyone else, and this was not much noticed, because the everyone else did not run the archives.
A multicultural society inherits those systems and a population they were never designed for. What happens next is not an abstract question about inclusion. It is a practical question about whose grandmother can be found in fifty years, and it has already gone badly wrong at least once in living memory.
Between 1948 and 1973, hundreds of thousands of people arrived in Britain from the Caribbean, legally, as citizens. Many were children on their parents' passports. The clearest record of their arrival was a set of landing cards held by the Home Office.
In 2010, the Home Office destroyed the landing cards. Then, under later immigration rules, the same department demanded that these now-elderly people prove their decades of continuous residence, and treated the absence of proof as grounds for losing jobs, housing, medical care and, in dozens of cases, for detention and removal from the country they had lived in for sixty years.
The Windrush scandal is usually told as an immigration story. It is also the purest archival story of the century: a state destroyed a minority's records and then made those records the price of belonging. People's lives were derailed for want of a document they had never been responsible for keeping.
The lesson generalises uncomfortably well. Whoever controls the record controls, eventually, who is believed. A community whose past sits only in somebody else's filing system has no fallback when the filing system fails, or is weeded, or turns hostile.
The institutional response, over the last few decades, has three recognisable strands.
Community archives: collections built and held by the communities they document, rather than about them. They keep what national institutions never collected, the shop records and society minutes and wedding photographs of populations the mainstream archive did not think to ask.
Oral history done as shared authority: recording people in their own words and treating the speaker as co-owner of the result, rather than raw material. The interview archives built this way now hold, for many migrant communities, the only first-person account of the journey that exists anywhere.
And plain multilingualism: a record kept only in the official language quietly loses everything that was said, named and sung in the others. At Ellis Island the inspectors worked through interpreters and, contrary to the myth, did not rename anybody; the names that changed were mostly changed later, by the families themselves, to fit a country that could not spell them. Both halves of that story, the kept name and the changed one, are family history, and only a record in both languages holds them.
Scale the whole problem down from a nation to a household and it does not change shape.
A family that has crossed a border holds its memory in two languages, at least one of which the grandchildren will read badly. Its documents are split between two countries' systems, and one of those countries may be difficult, dangerous or simply expensive to ask. Its story, the part that explains everything else, exists in the memory of the generation that made the journey, and that generation is going.
The pattern repeats across this whole library: the griot lines thinned by migration, the jiapu copies that survived abroad when the originals burned, the Windrush generation itself, whose best protection would have been copies of their own papers in their own hands.
That last phrase is the practical conclusion. The remedy for depending on somebody else's archive is holding your own: the documents scanned, the story recorded in the language it happened in, the names in both spellings, kept by the family. Confinity is built to be that holding place, private and exportable, so that the record of how a family got here does not sit solely in an institution that once already proved it could lose it.
The Windrush case, which is the whole argument in one story