The first time I saw the map, I felt sick. Arcs of light connecting the temples of Egypt to the glass cases of London, the shrines of Benin to the drawers of Berlin. One hundred thousand artifacts—stolen, bought, bartered, or simply ‘collected’—and every single one of them now tethered to a Western institution by a digital thread. I built Ex Situ, an open-source spatial index that does something simple and radical: it refuses to host the data.
Most cultural heritage projects reproduce the problem they’re trying to solve. They pull images, descriptions, and provenance into a new database, clean it up, and present it as neutral. But neutrality is a myth. Every museum’s taxonomy is a political artifact. Categories like ‘Islamic Art’ or ‘Ethnological Collection’ were invented to separate, to other, to justify removal. If you duplicate that taxonomy, you sanitize the crime.
So I made a different choice. Every museum is a crime scene. The code is the evidence. Ex Situ doesn’t store a single image. It only indexes the relationships—the arc from the origin site to the current institution. The images stay on the museum’s own servers. The data stays flat, deliberately avoiding the hierarchical taxonomies that encode bias. This is not a technical limitation; it’s a political statement.
You’ve probably wondered why museums hold so much of the world’s heritage. You’ve probably seen the empty pedestals in Athens, the missing statues in Easter Island, the looted bronzes in Nigeria. We all feel the weight of it. But what if the architecture of the software itself could hold the institutions accountable? By refusing to host, I forced the museums to own their own labels. Every time a user clicks on an artifact, they are routed directly to the source. The museum’s own categorization stares back at them. The responsibility stays where it belongs.
The twist is this: the project is an indexer, not a hoster. It’s connective tissue between archives that were never designed to speak to each other. The Met, the V&A, the SMB Berlin—eight collections, one thousand years of displacement, all mapped as arcs. But the arcs don’t lie. They show the asymmetry. Lines from every continent converging on London, Paris, New York. This is not a map of culture. It’s a map of theft.
I started this as my MA thesis in 2022. I’ve been building solo ever since—Next.js, Deck.gl, Python ETL pulling from open-access APIs. It’s open source, AGPL-3.0. I recently added a Markdown export so researchers can download provenance data for a filtered set of artifacts. But scaling alone is overwhelming. I need the community to help: performance tuning, data pipelines for missing museum APIs, code contributions. Because this isn’t just a project about the past. It’s a blueprint for how to build ethical data systems in a world where every database is a political act.
When you open the map, you see the scale. The thousands of Korean artifacts in Japanese museums that aren’t even indexed yet. The Crimean treasures now in Moscow. The diaries of colonial officers who bragged about ‘bringing back so much stuff.’ The map is incomplete, but it’s already damning. Every missing line is a silence that refuses to be broken.
Ex Situ doesn’t make demands. It doesn’t call for repatriation. It simply shows the arcs. And once you see them, you can’t unsee them. The code is the statement. The rest is up to us.
FAQ
Q: Why not just host the images and data to make it easier for users?
A: Because hosting duplicates the museum's taxonomy and sanitizes the problem. By keeping images on the museum's servers, the user is forced to confront the source institution's own categorization—which is often colonial and biased. The technical choice is a political one.
Q: What practical action can someone take after seeing the map?
A: Researchers can download provenance data to analyze patterns of displacement. Developers can contribute code to add missing museum APIs. Activists can use the map as evidence in repatriation arguments. The project is a tool, not a finished product—it needs community to grow.
Q: Isn't this just a simple database of links? What's so revolutionary?
A: The revolution is in the architecture. Most cultural heritage projects aggregate data and present it as neutral, which accidentally reinforces the status quo. Ex Situ's flat, non-hosting model is a deliberate refusal to sanitize. It's a system designed to hold museums accountable by making them own their own data—and their own history.