The Liminal Web

The spaces the algorithm skips

There are parts of the internet that exist between genres. Not underground, exactly — they are publicly accessible, indexed, findable. But they do not belong to any platform's category system. They are not blogs, not wikis, not shops, not forums. They are something the internet produced by accident and has not yet named.

These are the liminal zones: the personal sites of artists who update irregularly, the archives of defunct publications that remain online by inertia, the niche aggregators built by one person for their own reference. You arrive at them by following a chain of links rather than a search query. They are discovered, not surfaced.

What liminality means here

The word liminal, borrowed from anthropology, describes thresholds — spaces between defined states. In architectural terms, a liminal space is a corridor, a waiting room, a stairwell. It is a space you pass through rather than inhabit. The internet's liminal zones work differently. They are spaces you can choose to inhabit, but that offer no social infrastructure for doing so. There is no comment section, no like button, no community. There is only the content and the reader.

Why they persist

Liminal web spaces persist because their creators are not optimising for growth. They are maintained because they are useful to the person who made them, or because the person who made them simply has not turned them off. The economics of static hosting are forgiving enough that a site can sit for years at minimal cost. The content stays. The links rot slowly.

For readers, these spaces offer something the algorithmic web cannot: a sense of discovery that has not been engineered. Finding a genuinely useful personal site through a chain of links feels different from finding a recommended article. It feels like the internet working the way it was supposed to work before recommendation systems decided what you should see next.