Physical hoarding has a natural brake: you run out of room. Digital hoarding has none. A terabyte costs less than lunch, so the decision that clutter used to force, keep or discard, never has to be made.
The consequence is not a full disk. It is an archive whose contents you have no working model of, which makes it useless in a specific and frustrating way: you cannot rely on it, so you keep the information in your head as well, and get the costs of both systems.
What it looks like in notes
It is rarely dramatic. The usual signs:
- Thousands of notes, and a strong suspicion that most of them are duplicates or fragments.
- An inbox or "unsorted" folder with four figures in it, untouched for years.
- Multiple half-migrated systems, none of them decommissioned, each possibly containing the thing you need.
- Web clippings you have never reopened, saved articles you will not read, screenshots with no context.
- Reluctance to delete anything, justified by "it might be useful."
The last one is the engine. Every individual item genuinely might be useful. The cost is not paid per item, it is paid across the whole collection, in reduced confidence that anything can be found.
Why it is not a storage problem
If it were about space, the fix would be a bigger drive. It is about retrieval confidence, and that degrades as the ratio of signal to noise falls.
A collection you trust behaves like an extension of your memory. A collection you do not trust behaves like a room you avoid. The tipping point is not a number of notes, it is the moment you start searching your own archive and expecting to be disappointed. See cognitive offloading: the whole benefit of writing things down depends on believing you can get them back.
There is also an emotional component that the tidying advice usually skips. People keep material because deleting feels like admitting they will never do the thing, never read the book, never write the piece. The archive is partly a record of intentions, and pruning it means retiring some of them. The antilibrary makes the case for the healthy version of that.
The distinction that actually helps
Not everything in your system is the same kind of thing, and treating it uniformly is what produces the mess.
| Category | Rule |
|---|---|
| Records — receipts, contracts, tax, medical | Keep indefinitely, archive out of the way, never in the working set |
| Reference — things you look up | Keep, but only if findable. An unfindable reference is not reference |
| Thinking — your own notes, ideas, drafts | Keep almost all of it. This is the irreplaceable part and it is usually small |
| Intake — clippings, saved links, screenshots | Aggressively temporary. Most should expire |
The category that causes digital hoarding is almost always the fourth. It is also the one where the material is, by definition, still available elsewhere. A saved article you have not read in three years is not a note, it is a bookmark to something the internet still has.
A pruning approach that survives contact with reality
Wholesale reorganisation projects fail. They take a weekend, produce a burst of satisfaction, and leave the underlying habits untouched, so the mess returns within months. Smaller and duller works better.
Archive, do not delete. Move everything older than a chosen date into a single archive/ folder, excluded from your default view but still searchable. This gets you a clean working set in one action with zero risk, and it is reversible.
Set an expiry on intake. Anything in the clipping or read-later pile that is older than 90 days should go without individual review. If you have not read it in three months, you will not.
Delete duplicates and stubs on sight. Empty notes, "asdf", the same idea captured four times. This is the one category worth deleting outright.
Do not review item by item. Deciding on five thousand things individually is the trap. Decide by rule, apply in bulk.
Fix the intake valve. Pruning without this is bailing without patching. The rule that works: capture a thought in one second, but capture someone else's content only with a sentence saying why. That sentence is the filter, and it takes ten seconds. Most things will not be worth ten seconds, which is the point. See the collector's fallacy.
What not to over-correct into
There is an opposite failure, and it is worse. Aggressive deletion of your own material, elaborate filing rules that make capture expensive, or a minimalism that discourages you from writing anything down.
Your own notes are cheap to keep and impossible to reconstruct. The target is not fewer notes, it is fewer things that are not notes. See when to delete notes and how to declutter your notes.
The test
Once a month, try to find something you know you saved. If you find it in under thirty seconds, your system is working regardless of how many notes it holds. If you do not, the problem is not volume, it is naming, structure, and search.
More in information overload, PKM mistakes, and personal knowledge management.