The invisible systems that sort your world.
Every database, form, and algorithm relies on categories. But who decides what counts? The answer reveals how power works in the information age.
Every time you fill out a form, search a database, or get sorted by an algorithm, you're encountering classification systems at work. Male or female. Urban or rural. High risk or low risk. These categories feel natural, inevitable even. But they're not. They're designed, and that design has consequences.
My research into classification systems started with a simple observation: the categories we use to organize information don't just describe the world, they shape it. A medical database that only recognizes two gender categories creates real problems for transgender patients. A credit scoring system that classifies certain zip codes as 'high risk' perpetuates housing discrimination. The map is not the territory, but sometimes the map rewrites the territory.
What makes this particularly fascinating is how these systems become infrastructure. Like electricity or water pipes, they fade into the background once they're established. Everyone builds on top of them. Everyone depends on them. And that invisibility is precisely what makes them so powerful.
When categories become concrete
Classification systems don't start out as infrastructure. They begin as practical solutions to immediate problems. A hospital needs to organize patient records. A government needs to count citizens for representation. A company needs to sort customers for marketing.
But something interesting happens when these ad hoc solutions succeed. They standardize. They spread. Other organizations adopt the same categories to maintain compatibility. Software gets built around them. Legal frameworks reference them. Training programs teach them.
Eventually, what began as one institution's organizing principle becomes everyone's organizing principle. The classification system transforms from a tool into a foundation that everything else depends on. At this point, changing it becomes incredibly difficult, not because the categories are naturally correct, but because so much has been built on top of them that modification would require rebuilding entire systems.
This infrastructural quality explains why some problematic classifications persist long after we recognize their flaws. It's not just that people are resistant to change. It's that the categories have become load-bearing walls in the architecture of information.
The politics hiding in plain sight
Here's what's particularly insidious about infrastructural classification: it appears technical when it's actually political. When categories become invisible infrastructure, their social dimensions disappear from view. The choice to classify people as 'employed' versus 'unemployed' seems like a straightforward administrative decision, but it determines who qualifies for benefits, who gets counted in economic statistics, and whose work gets recognized as legitimate labor.
Every classification system embeds assumptions about what matters, what exists, and what can be safely ignored. These assumptions reflect the perspectives and interests of whoever had the power to create the system in the first place. But once the system becomes infrastructure, those embedded perspectives start to look like objective facts about the world.
My research found that this invisibility is not accidental. It's a feature of successful infrastructure. The more essential a system becomes, the more it fades from conscious attention. This creates a paradox: the classifications that have the most power over our lives are the ones we're least likely to question, precisely because they've become too fundamental to notice.
The negotiation never ends
Despite their appearance of stability, classification systems are constantly being negotiated. Different communities have different needs, different ways of understanding the world, different priorities. The categories that work perfectly for one group create problems for another.
Consider how medical classification systems handle mental health. Psychiatrists need diagnostic categories that correspond to treatment protocols. Insurance companies need categories that correspond to billing codes. Patients need categories that correspond to their lived experiences. Researchers need categories that correspond to biological mechanisms. These needs don't always align.
What emerges from this ongoing negotiation is not the 'correct' classification, but rather the classification that can accommodate the most powerful stakeholders. The system that becomes infrastructure is the one that manages to serve dominant interests while appearing neutral enough to gain widespread adoption.
This means that changing classification systems requires more than just better categories. It requires changing the balance of power between the communities doing the negotiating. Technical solutions to classification problems are rarely just technical.
Building better worlds through better categories
Understanding classification as infrastructure opens up new possibilities for intervention. If categories shape reality, then designing better categories can help create better realities. But this requires recognizing that classification work is inherently social and political, not just technical.
For database designers, this means considering whose perspectives are embedded in your schema and whose are excluded. For policy makers, it means recognizing that administrative categories don't just organize bureaucracy, they organize people's lives. For technology developers, it means understanding that the taxonomies built into search algorithms and recommendation systems actively shape what knowledge becomes visible and accessible.
The goal isn't to eliminate the politics from classification, which is impossible, but to make those politics visible and accountable. When we treat classification systems as neutral infrastructure, we abdicate responsibility for their social consequences. When we recognize them as designed systems that embody particular values and serve particular interests, we open up space for democratic participation in their creation and modification.
Classification systems will always be with us. The question is whether they'll be designed by accident or by intention, in service of existing power structures or in service of more equitable futures.
Get the next one
An occasional note when something genuinely new ships here — essays, free tools, projects. No schedule, no filler, easy out.
Need something like this built?
I design and ship AI tools, full-stack apps, and data pipelines — end to end, to production. Tell me the problem in a sentence; I'll give you an honest read on fit within a day.
Work with me →