01The two audiences.
Notes are written by somebody who has just read the material and is read by somebody who has not thought about it for two months. Those are different people with different needs, and almost all bad notes come from writing for the first one.
In the moment, an arrow and three words genuinely capture the idea, because your head is full of the context that makes them meaningful. In week eleven the context is gone and the three words are archaeology. You end up rereading the source, which means the notes cost time twice and saved it never.
The correction is a single discipline: write the thing the arrow was standing in for. It feels redundant while you are doing it and it is the entire difference between notes that work and notes that get abandoned.
02Five habits that make notes survive.
None takes longer than the version that fails; they simply feel less efficient at the time.
- Full sentences for anything conceptual. Fragments are fine for facts and useless for reasoning, because reasoning is exactly the part your later self cannot reconstruct.
- The source and the page beside anything you might quote or need to check, written at the time. Finding it again later costs ten times what writing it costs now.
- Your own words rather than the source's, because a copied sentence tells you nothing about whether you understood it and a paraphrase tells you immediately.
- A mark for the things you did not follow. A question mark in the margin is the most valuable symbol in any set of notes, and it is the one people avoid because it feels like an admission.
- And a line at the end of each session on what you were about to do next, which costs one sentence and restores twenty minutes at the start of the following one.
03Notes for something that is examined.
Where a paper is coming, notes have a second job: they have to be revisable, which is different from being complete. A perfect transcription of a lecture is a poor revision document, because revising means moving through the material repeatedly and volume prevents that.
So the useful shape is layered. The full notes, then a shorter derived version made from them, and the shorter version is where revision actually happens. Making it is itself the revision, which is why the derived layer is worth more than the original.
In clinical subjects there is a third layer worth having: the mistakes. A running list of things you got wrong in practice questions, phrased as the correction rather than the error, is the single most efficient document going into a paper. That is the same principle described in the exam page, where reading your wrong answers is the actual work.
04Where to keep them.
In one place, and the choice between paper and screen matters far less than that constraint. Notes distributed across a notebook, three documents, a phone and the margins of a printout are functionally lost, because retrieval cost rises fast enough that you stop attempting it.
For a fragmented week the practical answer is usually digital and searchable, since it can be reached from a break room or a parked car, and the setup-and-teardown problem described in the piece about having nowhere to work applies to notes as much as to books.
For material you will be examined on by hand, there is a real argument for writing by hand at least at the derived layer, because the slower channel forces the compression that makes notes useful. Neither answer is universal. Having one place is.
05The test, applied honestly.
Once a term, open something you wrote six or eight weeks earlier and try to state its point aloud without reopening the source. It takes two minutes and it is unpleasant the first time.
If you can, your system works and nothing on this page needs adopting. If you cannot, the fault is nearly always the same one: the notes recorded what the material said rather than what you understood, and the difference between those two is the whole subject.
Everything here is cheap to change mid-term, which is worth saying because note-taking feels like a habit fixed years ago. It is not. It is a decision made freshly at the top of every page.
Questions at this desk.
Is handwriting genuinely better?
It has a real advantage where compression matters, because writing by hand is slow enough to force selection rather than transcription. It has real disadvantages for searching and for studying in fragments. Many people end up typing the first layer and writing the derived one by hand, which takes both benefits.
How much should I be writing?
Less than most people do, and more slowly than feels efficient. The failure in almost every set of student notes is volume without selection, because transcribing looks like working while deciding what matters looks like hesitating. Deciding is the work, and it is what makes the page usable later.
What about recording lectures instead?
Useful as a safety net and poor as a primary method, because a recording defers the thinking rather than doing it, and nobody has time to relisten to a term. Where recording is permitted, use it to check something specific afterwards rather than as a reason to stop taking notes.
Should notes be shared with classmates?
Comparing is genuinely useful and reveals what you missed, particularly the things everybody else marked as important. What does not work is relying on somebody else's, because their notes were written for their understanding, and the whole point of this issue is that notes are addressed to a specific reader.
Notes in the margin, from the people at the desks.
READ THE LETTER