Note-taking methods compared

What notes are actually for

Two things, and most students optimise for the wrong one. Notes are a record you can return to, and — more importantly — the act of making them is how you process the material.

The second is where the learning happens. Deciding what matters, how it connects and how to put it in your own words is cognitive work, and it is the work that transcription avoids.

That is why a complete, verbatim record is usually worse than a partial, processed one. If you can take the note without thinking, you have not learned anything by taking it.

It is also why notes you never return to are half-wasted. The processing happened, and the retrieval that would consolidate it did not.

The five methods

Each is good at something specific. Choosing by the material rather than by preference is what makes them work.

Cornell, and why it is the default recommendation

Divide the page: a narrow left column, a wide right column, and a strip across the bottom. Notes go in the wide column during the lecture.

Afterwards — and this is the part people skip — write questions in the left column that the notes answer, and a two or three sentence summary at the bottom.

That afterwards step is the whole point. It forces you to reprocess the material within a day, and it converts your notes into a self-test: cover the right column, read a question, answer it from memory.

So Cornell is not really a layout. It is retrieval practice built into a page format, which is why it outperforms methods that produce better-looking notes.

The time cost is about ten minutes per lecture, done the same day.

Handwriting versus typing

For lectures, handwriting generally produces better recall, and the reason is the constraint rather than the medium.

You cannot write as fast as you can type, so handwriting forces selection and paraphrase. Typing allows near-verbatim transcription, which feels productive and skips the processing.

The research on this is not unanimous and the mechanism is well supported: the benefit comes from the summarising, not from the pen. Typing while deliberately paraphrasing recovers most of it.

Typing wins on searchability, on legibility, on integration with your reference manager, and for anyone for whom handwriting is slow or painful. Those are real advantages.

A reasonable position: handwrite in lectures where you can, type for reading notes where searchability matters, and in both cases paraphrase rather than transcribe.

Reading notes are a different problem

Lecture notes capture something transient; reading notes organise something you can return to. They should be structured differently.

Do not summarise the source. A summary of a paper is a smaller version of a document you already have, and it is filed under the source, which is the wrong axis.

Instead, note what the source claims, what evidence it rests on, and — this is the part that matters — how it relates to what you have already read. That third field is what makes a folder of notes into a literature review.

File by idea rather than by source where you can, or tag both ways. Notes filed only by author cannot answer the question 'what does the literature say about X', which is the question you will actually ask.

Record the page number with anything you might quote. Going back for it later is one of the most common avoidable time costs in a dissertation.

Zettelkasten, in the version that is worth the setup

A system of atomic notes — one idea per note, in your own words — linked to each other rather than filed in a hierarchy.

The claim is that the links produce connections you would not otherwise see, and for a long project drawing on hundreds of sources over years, that is genuinely true.

The overhead is real. Setting it up and maintaining it takes time that a single module does not repay, and a great deal of what is written about it is about the system rather than about the work.

The transferable part, without the apparatus: write one note per idea rather than one per source, put it in your own words, and record how it relates to something else you have read. That is most of the benefit.

For a dissertation or a doctorate, the full version is worth considering. For an essay, it is not.

Digital tools

The tool matters less than the habit, and the time spent choosing one is frequently procrastination.

What a tool needs: fast capture, reliable search, links between notes, and export in a format you will still be able to open in ten years. Most of the popular options do all four.

Integrate it with your reference manager. A note attached to the record it came from, searchable alongside your library, is worth substantially more than the same note in a separate application.

Plain text and markdown survive software changes; proprietary formats do not necessarily. For a doctorate spanning four years that is a real consideration.

Whatever you use, back it up somewhere that is not the same service. Note systems are among the things people lose.

What makes notes useful later

Five properties, and they matter more than the method.

• In your own words. A copied sentence has been recorded rather than understood, and it is a plagiarism risk when it later appears in a draft.
• Marked for what is a quotation and what is your paraphrase, unambiguously. This single habit prevents the commonest route into accidental plagiarism.
• Carrying the source and the page, every time.
• Containing your own reaction — agreement, objection, connection — because that is what you cannot reconstruct later.
• Findable. A note you cannot locate in eight months is a note you did not take.

The quotation-marking habit

Worth its own section because it is the highest-consequence note-taking habit and takes no extra time.

Every time you copy an exact phrase into your notes, put it in quotation marks and record the page immediately. Every time.

The alternative — copying a phrase intending to paraphrase it later — is how unattributed text gets into drafts, and it happens to careful students who had no intention of doing anything wrong.

Months later, a note without quotation marks is indistinguishable from your own words, and you will treat it as such.

Use a consistent visual marker: quotation marks plus a page number, or a different colour, or a prefix. Whatever it is, never vary it.

Returning to them

Notes you never reopen have done half their job. The processing happened; the retrieval did not.

Review within a day, briefly. Ten minutes completing a Cornell cue column or tidying reading notes while the material is fresh is worth more than an hour a fortnight later.

Then space the reviews — a few days, a week, a month. Spaced retrieval is the second best-evidenced study technique after retrieval itself.

Use the notes to test yourself rather than to reread. Covering the answers and recalling them is what builds memory; rereading produces the feeling of knowing without the knowledge.

Before an exam or a chapter, condense: reduce a term of notes to two pages. The condensing is itself retrieval, and the two pages are what you actually revise from.