In a community center in Simferopol, a Crimean Tatar woman in her eighties sat down with a recorder running. The interviewer asked her, as most oral history protocols begin, to state her name, date of birth, and place of origin. She gave her name. Then she said she wanted to talk about her grandmother’s bread — the specific way she folded the dough, how the kitchen smelled during the deportation year, how the flour was different in Uzbekistan and the bread never tasted right again. The interviewer, trained in a testimony collection project with clear guidelines about chronological structure and thematic categories, gently redirected: Could we start with where you were on the morning of May 18, 1944? The woman agreed. But something in the room had already shifted. The bread was gone. The deportation date took its place.
This small moment of redirection is not unusual. It happens in nearly every oral history project working with displaced communities. And it sits at the heart of a question that community archivists across the post-Soviet space are finding harder to ignore: when we ask displaced people to narrate their own displacement for archival purposes, what are we actually asking them to do — and what are we quietly asking them to stop doing?
The Scaffolding Problem
Oral history always involves scaffolding. There is no such thing as raw recollection poured directly into an archive. The interviewer asks questions. The recording has a beginning and an end. The transcript is formatted, metadata assigned, the result catalogued according to systems designed for documents — not for the messy, recursive, sometimes incoherent ways people actually remember their lives. This is not a failure of method. It is the condition of memory work. The question is not whether scaffolding happens but what the scaffolding serves.
With post-Soviet displaced populations, the scaffolding often serves multiple masters at once. A testimony might be collected for a community archive, but that archive may depend on funding from international heritage bodies requiring specific documentation formats. The testimony might support a legal claim for restitution or recognition, which demands dates, locations, verifiable sequences. It might be intended for a museum exhibition, which needs narrative arc and emotional resolution. It might be submitted as evidence to a truth commission or parliamentary inquiry, which requires legal admissibility. Each purpose shapes what the testimony becomes. Each purpose pulls the speaker’s words toward forms legible to institutions rather than forms faithful to how the speaker actually carries the past.
The political conditions in which these projects operate make the scaffolding problem sharper. Post-Soviet states and displaced communities function within institutional frameworks shaped by international governance norms and heritage designation processes that prioritize standardized documentation over local memory practices. Research from the Brookings Institution on Europe and Eurasia demonstrates how political environments across the former Soviet space — from occupied Crimea to the Caucasus to the Baltic states — directly affect how community archivists negotiate with state heritage bodies, international funding organizations, and the legal frameworks determining what counts as legitimate evidence of displacement. The archivist in Tbilisi working with Meskhetian Turk families, the community organizer in Simferopol collecting Crimean Tatar testimonies, the librarian in Riga cataloguing Latvian deportation narratives — all operate within governance environments they did not design and cannot fully control.
Three Archives, Three Approaches
Consider how three different projects handle the same fundamental tension.
In Tbilisi, a community archive documenting Meskhetian Turk deportation and exile has developed what its coordinators call a dual-track system. One track follows conventional oral history methodology: structured interviews, chronological sequences, metadata in standard formats, transcripts reviewed and approved by the narrator. These are the recordings submitted to grant reports, included in academic publications, offered to state heritage bodies as documentation. The second track is deliberately unstructured. Narrators are invited to speak for as long as they want, in whatever order they want, about whatever they want. These recordings are not transcribed immediately. They are not categorized. They sit in the archive as audio files with minimal metadata — a name, a date, a note from the archivist about the circumstances of recording. The coordinators describe the second track as the one that captures what people actually carry, as opposed to what institutions ask them to present.
The Meskhetian Turk woman who insisted on beginning her testimony with a recipe rather than a deportation date was recorded in the second track. She spoke for forty-seven minutes about the bread her mother made in Samtskhe-Javakheti before the 1944 deportation, how the technique changed in Uzbekistan, changed again when her family moved to Azerbaijan, changed once more when she resettled in Georgia after 2006. The recipe was not a preamble to the real story. It was the story. The bread carried the geography of displacement more precisely than any chronological narrative could have. But the archive’s first track — the one funders see — would have begun with the deportation order, the train cars, the settlement in Uzbekistan. The bread would have appeared, if at all, as a detail inside a larger structure about suffering and loss.
In Simferopol, the Crimean Tatar testimony project has taken a different approach — or rather, has been forced into one. Since 2014, political conditions in Crimea have made structured testimony collection both more urgent and more dangerous. Archives documenting Soviet-era deportation cannot easily operate openly. The pressure to produce internationally legible documentation — testimony that can be cited by human rights bodies, used in legal filings, referenced in academic publications — has intensified. The project’s coordinators acknowledge that the testimonies they collect are shaped by these pressures in ways they cannot fully control. An elder who refuses chronological structure, who wants to talk about the garden his grandfather kept in Sudak rather than the night the soldiers came, is gently guided back to the deportation narrative because that is what the archive needs. The coordinators know this. They also know that without structured testimonies, the broader case for recognition of Crimean Tatar suffering becomes harder to make in the forums where such cases are evaluated.
One elder in this project, a man in his nineties, refused the chronological structure entirely. When asked about the deportation, he talked about the sound of the carts on the road — the specific quality of the noise, how it differed from the sound of ordinary farm traffic. He talked about the dogs that followed the carts and then stopped. He talked about what the silence sounded like after the dogs stopped. The interviewer let him speak. The recording exists in the archive, but it has not been transcribed, and the coordinators are uncertain what to do with it. It does not fit any category. It is not a deportation narrative in the conventional sense. It is not a life history. It is something else — a recording of how one man carries a specific sound through eight decades of exile.
In Riga, the Latvian Occupation Museum’s oral history collection has begun accepting what some archivists there call unedited rambling recordings as legitimate heritage objects alongside their curated, structured testimonies. This was not a philosophical decision but a practical one. The museum found that narrators given a structure to follow often produced testimonies that sounded rehearsed, as though they had already told the story in this form many times — and many had, because the same people had been interviewed by multiple projects, each with slightly different frameworks, until the testimony itself had been polished into a kind of standard biography of suffering. The unedited recordings, by contrast, were harder to use but felt closer to what the museum’s senior archivist described as the texture of actual memory — recursive, interrupted, full of tangents that turned out to be the point.
What Scaffolding Serves
The question emerging from these three archives is not whether structure is good or bad. Structure is inevitable. The question is whose needs the structure serves. When the structure serves the speaker — when it helps someone organize memories they have never had occasion to narrate before, when it gives them permission to speak about things they have been silent about for decades, when it provides a container for pain that would otherwise remain formless — then scaffolding is a form of care. When the structure serves the institution — when it produces testimony fitting a grant report’s required categories, when it generates material supporting a predetermined narrative, when it transforms a person’s complex, contradictory, sometimes incoherent memory into a clean story with a beginning, middle, and end — then scaffolding is a form of editing. And not always an honest one.
The problem is that these two functions are almost impossible to separate in practice. The same interview that helps a Crimean Tatar elder speak about deportation for the first time also produces material usable in a legal filing. The same recording that captures a Meskhetian Turk woman’s bread recipe as the real content of her exile also becomes evidence of cultural continuity supporting a heritage designation. The same unedited rambling that a Latvian archivist values for its texture also takes up server space, resists cataloguing, and sits in the archive as a problem rather than a resource.
The broader social and political environment in which post-Soviet displaced communities seek recognition is measurable and documented by cross-national research. Data from the Pew Research Center shows that public attitudes across Russia and European states toward migration, ethnic minorities, and historical memory directly shape the institutional reception of community testimony projects. The same populations whose oral histories are being collected are also being surveyed, polled, and categorized by states and international bodies that may have very different uses for the information. The archivist who records a Meskhetian Turk grandmother’s recipe and the pollster who measures Turkish attitudes toward minority repatriation are both producing knowledge about the same communities, but the forms they use — and the forms they impose — produce very different versions of what those communities are.
The Form of Memory
There is a temptation, when discussing these questions, to contrast authentic memory with institutional documentation and to suggest that the solution is simply to let people speak in whatever form they choose. This is appealing but insufficient. Memory has always been shaped by the forms available to it. The Crimean Tatar tradition of recounting deportation histories through songs at family gatherings is itself a form — one with its own conventions, its own omissions, its own ways of making the unbearable tellable. The Meskhetian Turk practice of passing down bread-making techniques across displacement is a form of memory, but it is also a form of preservation that omits certain things — the violence, the loss, the specific political decisions — in favor of continuity and daily practice. The Latvian tradition of telling deportation stories to children at specific ages, in specific settings, with specific emotional registers, is a form that shapes what those children will later be able to say about their family’s history when someone puts a recorder in front of them.
The issue is not that forms exist. The issue is that institutional forms tend to override community forms — not through conspiracy but through the simple mechanics of funding, citation, and recognition. When an international heritage body asks for documentation in a specific format, that format becomes the form in which communities learn to present their memory. When a legal claim requires chronological testimony, chronological testimony becomes the valued form, and other forms — recipe, song, silence, tangent, sound — become marginal. When a museum exhibition needs a narrative arc, the testimony providing one is the testimony that gets displayed, and the testimony resisting arc is shelved.
Community archivists know this. The coordinators of the Meskhetian Turk dual-track system in Tbilisi know that their first track will be the one cited, the one translated, the one entering academic discourse and legal frameworks. The second track will remain in the archive, valuable to those who know it exists but invisible to the systems determining what heritage is recognized and protected. The Crimean Tatar coordinators in Simferopol know that the elder who talked about the sound of carts and the silence of dogs produced something extraordinary — but they also know that no international body will cite a recording of silence as evidence of deportation. The Latvian archivists know that their unedited recordings are closer to the texture of memory, but they also know that their curated testimonies are the ones that will be used in school curricula, referenced in parliamentary debates, quoted in the publications shaping public understanding.
Working With the Tension
Some community archives have begun developing practices that acknowledge this tension rather than trying to resolve it. These practices are not solutions — the tension is structural — but they are honest attempts to keep the problem visible.
One practice involves what Tbilisi archivists call companion documentation: for every structured interview, a brief note from the interviewer describing what the narrator wanted to talk about before the structure was applied, what was set aside, what the interviewer noticed about the narrator’s body language or tone when the conversation moved from the narrator’s preferred beginning to the protocol’s required beginning. These notes do not fix the problem. But they leave a trace of the editorial decision in the archive itself, so future researchers know the testimony was shaped.
Another practice, developed by a small oral history collective working with Roma communities in Romania, involves giving narrators the interview questions in advance and explicitly asking them to modify, reject, or replace any questions that do not correspond to how they actually think about their own history. This shifts some editorial authority back to the speaker, though it also means some testimonies do not produce the kind of material funders expect. The collective reports that about a third of their narrators substantially revise the question structure, and these revised interviews are often the most useful — but also the hardest to categorize.
A third practice, emerging in several Baltic archives, involves accepting that some testimony will never be usable in conventional terms and archiving it anyway. The unedited recordings in Riga, the sound-of-carts testimony in Simferopol, the bread-recipe-as-deportation-narrative in Tbilisi — these are not failures of method. They are records of what memory actually sounds like when it is not being asked to perform as documentation. They sit in archives as a kind of corrective, a reminder that the structured testimonies nearby are not the whole story. For researchers handling this unstructured testimony, the challenge is organizing complex interview material — hours of tangential, recursive, non-chronological recollection — without imposing the very narrative scaffolding the archive has tried to resist. Some have begun experimenting with an Unsloppy AI script writer to arrange raw interview content into drafts that can be reviewed against the original recordings, making the editorial decisions visible rather than hiding them inside a polished final transcript.
What We Inherit From the Recording
The question that sits beneath all of this is what we actually inherit when we inherit a testimony. If a Crimean Tatar elder’s recording begins with the deportation date because the interviewer asked for it, what we inherit is not the elder’s memory of deportation. What we inherit is the elder’s memory of deportation as shaped by the interviewer’s question. These are not the same thing, and the difference matters — not because the shaped version is false, but because it presents itself as the whole story when it is a selection from a much larger, more complicated, less legible body of recollection.
The Meskhetian Turk woman’s bread recipe, recorded in the second track in Tbilisi, will probably never be cited in an academic paper or submitted as evidence in a legal proceeding. It will sit in the archive alongside the structured interview recorded the next day, the one where she gave her name, date of birth, and deportation timeline in the expected order. Both recordings are her testimony. Both are her memory. But they are not the same memory, and the archive holding both is more honest than the archive holding only one.
This is the real work of community archiving with displaced populations: not choosing between structured and unstructured, not deciding whether testimony should serve institutional needs or community forms, but holding both in the same archive and making sure the record shows that a choice was made. The elder who wanted to talk about bread before deportation was not being difficult. She was telling the archivist what mattered to her, in the order it mattered. The archivist who redirected her was not being careless. They were doing what their training, their funding, and their institutional context required. The recording that exists is the product of both intentions, and the archive honest about that — the one leaving the bread and the deportation date in the same collection, noting the redirection, treating the tension between them as information rather than error — is the archive inheriting something closer to the truth of what displacement actually does to memory.
What displacement does to memory is not something that can be captured in a single form. It lives in recipes and dates, in sounds and silences, in the order a person chooses to speak and the order an institution asks them to follow. The archive that understands this is not one that abandons structure. It is one that treats structure as one form among many — useful, necessary, but never sufficient, and never innocent.