Rethinking PKM Part 2: Links, Tags, and Filepaths

A key takeaway from this PKM deep-dive is that file organization has measurable performance limits—and that those limits help explain why many people rely more on folder navigation than search. Citing research discussed in *The Science of Managing Our Digital Stuff*, the creator notes a “magic number” of 21 items per folder: beyond that, it becomes more efficient to create subfolders. The underlying timing claim is stark—about 2.2 seconds to move one step in a folder hierarchy versus roughly 0.1 seconds to scan a file—so folder depth and breadth directly shape how quickly information can be found. The same book also points to a cognitive angle: typing search terms may cost more mental effort than using spatial memory to locate files, which aligns with the creator’s own preference for tools like xCQLI Brain over search.

From there, the focus shifts from performance to structure. The session builds a Venn-diagram framework (based on a circle map and double-bubble style comparisons) to sort common PKM actions into four overlapping “containers”: file paths, links, and tags. The definitions are concrete. File paths are treated as rigid, long-term storage—folder placement plus file names—like library shelves. Links represent relationships between notes; when enriched with metadata fields in tools such as Obsidian, they can function like an ontology or knowledge graph. Tags label notes so they can form “virtual spaces,” prime pattern recognition, and support similarity grouping (by type, project, or status).

The most practical portion is the placement of specific PKM activities into the overlaps. “Tag for action” is placed closer to tags, with a caveat: while it could be done via links (e.g., linking a task note to a to-do list), the creator finds it less practical than encoding actionable meaning in file names. “Reminder to resurface notes” sits between tags and links because both can support later retrieval—either through labeling or through contextual relationships that trigger resurfacing in the right situation.

“Group similar items” lands in the overlap because folders and tags both classify, while links can also create classification by connecting notes to a project or a taxonomy built from notes. “Add keywords to a note” is treated as more tag-like than path-like, while “Add context to notes” is strongly linked to relationships: context includes where information came from, how it will be used, life-cycle status, and even situational details like location or time—details that links are meant to connect.

Finally, workflow tracking is placed on the boundary between tags and file paths. The creator is ambivalent about whether status belongs in naming conventions or labels, but emphasizes a major advantage of file paths: they are platform independent and remain usable for years even if the tool changes. Links and tags can evolve, but folder structure and naming—if kept clean—act as durable anchors for long-term retrieval.

Overall, the framework argues for a deliberate division of labor: file paths for archival stability, links for relationship and context, and tags for flexible labeling—while acknowledging that many PKM tasks naturally spill across categories.