Form Follows Function, and Functions become Ruins.

From Le grand Nomade,

On a gray afternoon in Brussels, in a climate-controlled room behind a museum gallery, a technician is turning pages that no longer exist. The originals, posters from protests, concert flyers, small-town papers, are long gone or too fragile to touch. The ink on the few surviving copies is threatening to fade into a dusty memory of color. So, the technician, gloved and careful, turns the digital versions instead. Each image appears on the screen, is straightened and cropped, and is added to a database under a number of fields: title, date, creator, subject, and rights. Somewhere along the way, the paper died, and the file format took over the job of surviving.

On the other side of the screen, an online database promises visitors a glimpse of “the future of culture.” Inside this back room, the future looks more like a series of bets against ruin. File formats will change. Storage systems will be replaced. Someone will have to migrate this database, and then the next one, and then the next. Every act of preservation here is also an act of building tomorrow’s ruins. This is where this Fulbright project is planning to explore. The way organizations try to outrun decay with new technology. The project treats cultural and research institutions, the guardians of heritage, as bulwarks against forgetting. Systems that promise “in perpetuity.” Hard drives talk about “permanent” records. We might imagine these organizations as continuous, centuries-long projects. Yet the materials they steward have their own lifespans. Books rot. Photographs curl and crack. Magnetic tape demagnetizes. File formats become unreadable. Platforms vanish behind paywalls or login screens that no one remembers the password to. Today, the ruins of our ambitions are more likely to be a directory of TIFF files with no metadata, a read-only database whose software no longer runs, or a research dashboard frozen in an obsolete browser.

As organizations evolve their missions, how do the technical systems they rely on co-evolve with those goals? 

It starts from a basic observation: everything old was once new. The “ancient” manuscript was, once, a freshly copied text. The clunky CD-ROM catalogue of the 1990s was, at some point, the cutting edge of access. The metrics dashboard that today feels inevitable was once an optional pilot funded under “innovation.” At each moment, someone chose a system that felt like progress. The project connects the human decisions and policies and the technical infrastructure they use to carry out those decisions as a unified, ever-changing machine. co-evolution is not a metaphor but a practical condition. An institution declares a goal (we will be an open-science university, we will be a decolonial archive, we will be a data-informed museum) and then seeks out tools that seem to make that goal real. Repositories, catalogues, grant platforms, analytics systems arrive, usually as solutions to immediate problems like too many PDFs, too many boxes, too many grants to track by hand. The project’s core hypothesis is that the relationship does not stop there. Over time, the tools give shape back to the goals.

A university commits to open science and launches an institutional repository. “As open as possible,” but then their systems get bogged down with bots and bad faith actors – “as closed as necessary”.  

An archive embraces transparency and builds a public catalogue. The system can display creator names and dates; it struggles more with contested descriptions or community-supplied narratives. Staff invent workarounds: notes fields, local codes understood only by a few colleagues. Once those colleagues leave, no one knows how to edit the files.

In both cases, governance and technology are not separate layers. The rules, committees, and values shaping decisions on one side are constantly entangled with the fields, formats, and constraints on the other. They evolve (or stagnate) together. The project takes this entanglement as both its empirical and philosophical object.

Empirically, it follows a handful of institutions—a research funder, a university, an archive, a digitization center—as they navigate successive waves of technical change. It looks at their policy documents and strategy plans alongside their server logs and version histories. It talks to the staff who sit at the junctions: the person who writes the open-access policy and also manages the repository; the curator who negotiates with software vendors about what fields a new system will have; the data analyst who translates a ministry’s performance indicators into a dashboard that scholars will be judged by. It traces metrics over time: which categories become more common in catalogues, which outputs suddenly spike after a new reporting tool is introduced, which parts of a collection quietly drop out of visibility when a migration goes wrong. It pays attention to absences: empty fields, deprecated tags, categories that appear for a few years and then vanish from public interfaces, leaving only a faint sediment of obsolete labels in the database.Every system embeds a claim about what matters. A grant database that requires you to specify a discipline but not a community partner says something about which relationships are legible. A research evaluation platform that tracks journal articles but not curated exhibitions or long-term fieldwork quietly declares what counts as “output.” A digitization project that captures images but discards margins, annotations, and the backs of photographs makes a bet on which details are expendable.Those claims radiate forward in time. Decades later, when the system itself is obsolete, future researchers will inherit the residue: a pattern of what was easy to store and search for, and what was effectively designed to be forgotten.

The project then also asks: How do organizations try to outpace this kind of ruin, and how do their attempts, in turn, reshape what they are?

There are, at present, broadly three organizational types that the project plans on exploring.

In the first, technology appears to lead. A powerful new system arrives, often under pressure: a national mandate on open data, a performance-based funding model, a scandal over lost records. The system offers reassuringly crisp categories and compelling visualizations. Governance adapts around it. Committees are formed to feed it the right data. Definitions of “impact” or “heritage” are tweaked to match its built-in options. Over time, the institution’s sense of itself narrows to whatever the system can display.

In the second, governance tries to hold the reins. Institutions insist that their mission will not be bent to whoever offers the slickest platform. They negotiate for custom fields, refuse certain metrics, build in-house tools when commercial ones prove too rigid. The result is rarely a pure victory: there are budget constraints, skill shortages, moments where ideals are traded for stability. But the friction leaves a record. You can see the places where a principle (“we will not rank our researchers against each other,” “we will record contested terms”) is carved into the structure of a database.

The third trajectory is more conflicted. Institutions adopt systems that half-align with their goals and then spend years improvising. They publish open science policies while quietly relying on proprietary analytics. They declare themselves decolonial while their catalogues allow only one “official” description per object. Staff build parallel infrastructures—spreadsheets, shared drives, informal glossaries—to make reality fit between the gaps of formal systems. Co-evolution here is not smooth but full of small acts of resistance and resignation.

Across these trajectories, decay is not just a physical process but a conceptual one. There are markers that define points of change such as when mission statements are updated; categories are renamed, and evaluation criteria are revised. Old systems are retired, taking their particular worldviews with them. New ones arrive, promising to fix old blind spots and inevitably introducing new ones. Polices are changed to account for new achievements and new setbacks and the cycle continues.What survives across these cycles is not a perfect archive but a trail of decisions: the tags that were chosen over others, the outputs that were counted, the goals that were quietly abandoned because they did not fit cleanly in any reporting line.

If we imagine a hundred years into the future, what system evolution would seem so obvious to them that is not clear to us. They are likely to be a future curator, historian, or data scientist opening a legacy system and asking: What were they trying to do? How did they imagine the future? Why did they think this was enough? The project also hopes to take a look back in time, where this co-evolution is already apparent. Why did humanity largely abandon handwritten books once the print and press were invented? Why did people stop buying CDs when MP4 files could live on a device in your hand? These questions are easy to answer in retrospect, but mapping this as the change occurs is harder.

A digitized poster, stored without context, might survive as a beautiful but mute artifact, severed from the protest it once informed university students to attend. A research dataset, archived without documentation, might outlive its creator but become unusable. A category invented to satisfy one funder’s reporting template might outlast the template itself, shaping how work is grouped and valued for decades.The project does hope to promise a way to escape ruin. No system is future-proof. Every standard will one day be null. Every “best practice” will become a puzzle to somebody else. Instead, it asks whether institutions can become more deliberate about what kind of ruins they are building, by being able to better visualize the steps they take in the co-evolution cycle.

If a university knows that its dashboards will one day be archaeological objects, can it design them so that the logics of inclusion and exclusion are visible rather than hidden? If an archive accepts that some terms will age badly, can it build systems that keep the history of those changes instead of overwriting them? If a research funder recognizes that metrics will inevitably guide behavior, can it treat the choice of metrics as a moral decision rather than a technical one?Underneath the meetings about upgrades and migrations, these are questions about how a society narrates itself to the future. The project is not about keeping pace with new tools. It is about learning to ask better questions before a database schema hardens into the next generation’s common sense.

Out in the gallery, visitors file past objects that survived by accident and by design: a poster someone folded instead of throwing away, a photograph someone donated, a dataset someone bothered to document. Inside the back room, the technician clicks “save” on yet another image, unaware of who might open it next, or under what conditions.

The research is in that click. In the “goal” and the knowledge that every system is provisional. Co-evolution is not as a race to stay ahead of ruin,. It is instead an opportunity to choose, with a bit more honesty, which traces we leave behind, and what they will say about how we tried, and how we failed, to outpace oblivion.