Why highlighting alone fails
- It requires no processing. You can highlight a sentence you don't understand, and people frequently do — see why highlighting doesn't work.
- It marks importance without saying why. In revision week the yellow tells you nothing you can act on.
- It over-selects. By chapter three most students are highlighting a third of the page, at which point the signal is gone.
- It produces recognition, not recall. Re-reading your highlights feels like knowing, which is the precise illusion that costs marks in exams.
Highlighting isn't worthless — it's a fine locator. The mistake is treating it as the study activity rather than as an index you'll annotate on top of.
Four marks that are worth making
| Mark | Means | Why it earns its space |
|---|---|---|
| A question in the margin | "What would happen if…", "why this and not that?" | It's a retrieval cue and a future practice question in one. |
| A one-line summary at a section end | The section's claim in your words | Forces the compression that proves comprehension. |
| A link | "cf. p.212", "contradicts the lecture" | Builds the connections that single-pass reading never makes. |
| A confusion mark | "??" plus what specifically is unclear | A precise to-do; "??" alone is useless in March. |
Notice that all four are things you wrote, not things you selected. That's the whole distinction, and it's why annotation survives as revision material while highlighting doesn't.
The pass structure
- 1
Skim first, mark nothing
Headings, figures, summary, first and last paragraph. Five minutes. You cannot judge what's important on a first read, which is exactly why first-read highlighting over-selects.
- 2
Read a whole section before marking anything in it
Annotating as you go means marking what seems important mid-paragraph rather than what turned out to be. Read, then mark.
- 3
Write the section's claim in one sentence
In the margin or at the section break, in your words. If you can't, you didn't follow it — which is worth discovering now rather than in April.
- 4
Add one question per page, minimum
Questions convert directly into revision material later, which no highlight does.
- 5
Flag confusions specifically
"Why does the sign flip here?" is actionable in office hours. "??" is not.
This is SQ3R with the marking made explicit, and it fits the same time budget once the pointless re-reading is removed.
Physical books, PDFs and library copies
The medium changes the mechanics but not the principle: something you wrote must end up attached to the page.
- Your own physical copy. Margins, pencil, and the inside back cover as a running index of the questions you generated.
- Library or borrowed copies. Sticky notes, or a notebook with page numbers as the left column — slower, and it forces the summarising that makes annotation valuable anyway.
- PDFs. Comment tools rather than the highlighter. If your reader makes a comment as easy as a highlight, you'll write more; if it doesn't, switch readers.
- Tablet handwriting. The best of both for most people — search plus the selection pressure that handwriting imposes. Note-taking apps versus paper covers the trade.
The index page that makes it retrievable
Annotation's failure mode is that it's scattered across four hundred pages you'll never reopen. The fix is an index on the inside cover, built as you go: page number, and three words about what's there.
Two lines per chapter is enough. It turns the book from something you'd have to leaf through into something you can query in seconds, and it takes about ten seconds a session to maintain.
Annotation isn't revision
The last mistake is the largest. Annotation is a comprehension and capture technique; it is not how you learn the material. A perfectly annotated chapter that you then re-read four times produces the same weak result as any other re-reading.
The annotations have to become questions you answer without the book — which is the retrieval practice that actually builds retention. If you wrote a question per page, you've already got most of the way there.
How much to annotate
A rough calibration: one written mark per page, one summary line per section, one question per page. That's substantially less ink than the average highlighted textbook and considerably more thinking.
If you're marking more than that, you're transcribing. If you're marking less, you're reading passively and the book will look untouched by the time it matters. See active reading strategies for the wider habit.