The productivity internet has been telling you to link your notes for a decade. Bidirectional links, backlinks, graph views — the promise was a zettelkasten for everyone: ideas that surface automatically, connections you'd never have seen, thinking that compounds itself. For most people, what actually happened was a more elaborate way to file things they'll never open again.

Here's what I want to be honest about first: the underlying idea isn't wrong. Memory is associative. Linking ideas does mirror how knowledge connects in the brain. Luhmann's zettelkasten genuinely worked — he published 70+ books from it. The problem is how that idea got packaged, marketed, and then adopted at scale. Seven beliefs built up around linking that feel obviously true and are, on inspection, mostly false.

Why this list

Linking became a default practice before anyone had good evidence it worked for average users at scale. The case studies that moved the needle were on expert use of structured note systems — not on someone using Roam for six months and checking their graph view weekly. The beliefs below deserve the same critical look we'd apply to any productivity practice: does this actually work, and for whom?

1. "More links = a richer knowledge network"

The belief: Every link you draw is a new path through your knowledge, so linking aggressively compounds over time into something genuinely useful.

What actually happens: A link you draw and never traverse does nothing. Richness comes from traversal — from following a link when you need to think — and most links in most systems are drawn reflexively during capture and never followed. After a year of linking, the average user finds their notes the same way they always did: search.

Open your graph view right now and pick a random link. When did you last follow it?

2. "The graph view will reveal unexpected connections"

The belief: As your notes accumulate, the graph will surface insight you couldn't have planned for — serendipitous connections between ideas that were sitting there all along.

What actually happens: The graph view is useful for admiring your note system — a reliable dopamine hit — and for almost nothing else. Real unexpected connections surface when you reread a note in a different context, not when you stare at a ball-and-string diagram. The graph shows structure; it doesn't show meaning. The note-taking apps that sell their graph views hardest have the lowest follow-through rates on using them for actual thinking.

3. "Link during capture to keep the cost low"

The belief: Linking at capture time is efficient — you're already in the note, so connecting it to related notes immediately lets the graph grow naturally without a separate step.

What actually happens: Linking during capture adds a small decision tax to every note you take — what does this connect to? Multiply that by a hundred notes a week and you've added real friction to the one step that benefits most from being frictionless. The result is either capture slowing down or linking becoming reflexive and meaningless: connections drawn without thought, because it's what you're supposed to do.

The most productive note-takers I've observed link after the fact, during review, when a connection is genuine and immediately serves a project.

4. "Backlinks surface context automatically"

The belief: Because backlinks appear without effort — every note shows you what links to it — you gain context for free, without maintaining any index.

What actually happens: Backlinks show you which notes link here, not why those notes are relevant right now. A note with thirty backlinks has thirty pointers and no ranking — which of those thirty matters depends on your current project, and the backlink panel can't know that. Relevance requires context; backlinks provide volume.

Semantic search does what backlinks actually promised: it surfaces the notes most relevant to what you're thinking now, without the upfront cost of drawing any links.

5. "Orphan notes mean your system is broken"

The belief: A note with no links is a wasted capture — an island that can't contribute to your network and will be forgotten.

What actually happens: Some notes are perfectly useful as standalone items. A meeting decision, a project spec, a reference table, a book quote — none of these need to be woven into your theory of knowledge to be valuable. Forcing links onto them manufactures structure for its own sake.

The goal is notes that help you think and act. An orphan note you can find when you need it is worth more than a well-linked note you never open.

6. "You're building something like Luhmann's zettelkasten"

The belief: The linking method that worked for Niklas Luhmann — producing decades of original scholarship — can be adapted by anyone willing to put in the effort with a modern app.

What actually happens: Luhmann spent fifty years linking and pruning a physical system with slow, deliberate connections — each note a distilled idea, each link earned through thought. What most people build in digital tools is a fast-growing append-only collection that looks like a zettelkasten in structure and shares almost nothing with one in practice. The friction of the analog system was doing work that the frictionless digital version skips.

Luhmann's results were remarkable. So was the decades of compulsive discipline. The results don't transfer without the discipline.

7. "A good note should be heavily linked"

The belief: The importance of a note is proportional to how many things link to it — dense connectivity signals a central idea worth returning to.

What actually happens: Some of the most valuable notes are atomic ideas that don't need a hundred links. They need to be findable and clear. Over-linking a note to signal its importance is the note-taking equivalent of highlighting every sentence in a textbook — when everything is emphasized, nothing is.

A note earns importance through being used, not through being pointed at.

The one thing to change today

Stop linking during capture. Write the note and don't touch the link field. Then once a week, during review, add one link where a connection is genuinely useful for an active project. That discipline — one earned link per note per review — produces a more navigable network in three months than three months of aggressive linking during capture.

The underlying move is to let relevance drive linking instead of completeness. In JustJot.ai, semantic search already surfaces your most relevant notes at the moment you need them — so the links you do draw can be the ones that genuinely add meaning, not the ones you drew to convince yourself your system was working.