Once you predetermine the categories, it gets difficult to connect ideas from one category to another — let alone multiple. Each idea becomes a rigid part of its category. It’s not a problem to have categories (it guides how you tag notes, for example) — I’m saying the rigidity becomes a problem.
How, then, should we manage our knowledge without predetermined categories? It’s hard to imagine, but if you think of categories as “folders” in your computer, it makes sense to create “shortcuts” from one folder to another for better access.
First, the problem is that conventional note-taking is too linear to form new insights. He used notecards instead. Second, as your note collection grows, it gets harder to retrieve ideas. So, like Berners-Lee, (perhaps, the other way around) he organized them using links rather than strict categories. Yes — Luhmann was a badass who used hyperlinks back when it wasn’t even a thing yet.
Anyway, because there were no categories, there were also no hierarchies — or more appropriately, hierarchies didn’t form until they made sense. In other words, hierarchies were emerged instead of established from the start. That allowed new ideas to form — not to mention high-level insights that wouldn’t even cross anyone’s mind.
It’s as if Luhmann was deliberately countering the flow at which specializations emerged from disciplines. Instead of starting from the top — a predetermined specialization — he starts from the bottom, the content, and then builds them back up into new topics.
