What this covers
- A note that needs the source open is not finished
- Copying is not note-taking; rewriting is
- One place for notes, chosen once
- Volume is a warning sign, not a measure of effort
There is an enormous amount of advice about note-taking systems and very little about the single property that determines whether notes work: whether you can read them without the source.
Everything else - the app, the format, the colour coding - is preference. This one is not.
The standalone test
Read one of your notes a week after writing it. If you understand it fully, it is a note. If you find yourself wanting to open the document it came from to remember what it referred to, it is a pointer, and pointers fail exactly when you need them.
The practical version of the test: could you revise from these on a train, with no access to the source material? If the answer is no, the notes are not finished yet, and finishing them is a smaller job now than it will be in April.
Why this matters more with PDFs than with books
A textbook is one object you can flip through. A course's PDFs are forty objects with coded filenames. The cost of going back to the source is much higher, so notes that depend on the source fail much more often.
Copying is not note-taking
Selecting a passage and pasting it into a document feels like capture and is almost worthless. The material passes through the clipboard rather than through you, and nothing about it is any more available to you afterwards than it was before.
Rewriting is different in kind. Putting an idea in your own words requires understanding it well enough to restate it, which is a retrieval operation, which is the thing that produces learning. It is slower per page and dramatically faster per mark.
A method that produces usable notes
Works for a lecture PDF, a textbook chapter or a mark scheme.
- Read a whole section first
Without writing anything. Notes taken while first reading are transcription, because you do not yet know what matters.
- Close the document
Genuinely close it, or turn the tablet over. This is the step people skip and it is the step that does the work.
- Write what you remember
In your own words, in sentences. Where you cannot remember something, leave a gap rather than guessing.
- Reopen and fill the gaps
The gaps are the valuable part - they are a map of what did not stick. Fill them, and mark them as things to come back to.
- Write one line of why it matters
The link to the rest of the subject. This is what makes a note findable in six weeks, when you have forgotten the heading it came under.
Getting material out of the document
When to extract the text instead
Some documents are worth converting wholesale: mark schemes, specifications, definition lists, anything you will want to search rather than read.
Extract those to plain text and keep them beside your notes. You then have a searchable copy of the reference material and a written copy of your understanding, which are two different things that people frequently try to make one file do.
What belongs where
Three destinations, and the mistake of merging them.
| Material | Goes where | Why |
|---|---|---|
| Your understanding, in your words | Your notes | The only thing that records what you can produce |
| Reference material you will search | Extracted plain text | Searchable, editable, does not need reading |
| Questions to practise | A topic pack PDF | Needs the original layout and diagrams |
| A figure you will need again | Extracted image, in the notes | Redrawing a graph from memory is a different exercise |
How much to write
Less than feels right. A page of notes per lecture is usually too much; a third of a page is usually about correct.
Volume is the most reliable warning sign there is. Long notes are almost always the result of transcribing rather than understanding, because understanding compresses. If your notes are nearly as long as the source, you have made a slightly worse copy of the source.
Before you close the session
- Every note readable without the source document
- Written in your own words, not pasted
- Gaps marked where you could not remember something
- One line on why each idea matters
- Filed in the one place you always use
- Shorter than the material it came from, by a lot
Questions and answers
Should I take notes on paper or on a computer?
Whichever you will actually reread. Paper wins for diagrams and for not being next to a browser; typing wins for searchability and for volume of material. The decision that matters is having one place rather than several.
Is it worth annotating the PDF as well?
Rarely. An annotation lives in a file you will not reopen, and it duplicates the note without the benefit of having been written from memory.
How do I take notes from a scanned document?
The same way, with the added constraint that you cannot copy anything from it - which, given that copying is the trap, is not much of a loss.
What about diagrams?
Redraw them, roughly, from memory. Extracting the original is useful for reference and does nothing for recall; redrawing badly does a surprising amount.
