Mandatory titles clash with Luhmann
An issue with org-roam and many other digital slipboxes: titles are mandatory.
Maybe this suits Jethro and the other developers, and it suited me when I first started. Coming from a world where all files we ever create must be given a title, it wouldn't have occurred to me as something I'd want to live without. Besides, the title was how I used to find those files in the first place. But I found advice by a zettelkasten blogger (maybe it was Alex Keyhayias or Andy Matuschak) to try the original pen-and-paper Luhmann system for at least a while.
With the Luhmann system, most notes lack titles.
Freedom from titles means:
- You avoid the mental work of titling things
- Notes really do stand and fall with their connections
I used to worry about whether to title notes as nouns or as active statements. Matuschak even has ideas about titles as a kind of API…
But suppose that all new notes just got a random number as title unless you deign to re-title it later. And suppose we modify all our commands such as org-roam-node-insert to use ripgrep for matches in body texts instead of a title search. In a way, we'd treat the bodies as titles to themselves (kind of Logseq-ish!). Any titles matter less since we'd not depend on them to find the right note.
That boils down to a couple of lessons we can learn right now, if we still use mandatory titles:
- Don't sweat the titles – write the first title that comes to mind, and think of them as near-random slugs that only maybe incidentally say something about the content
- Rely first of all on the presence of interlinks to find any note – so refer to topic notes when you're looking for a note you remember. Rely second on ripgrep or filesystem order (both date created and date modified have analogues in the pen-and-paper system), and last (or never) on
org-roam-node-find.