We treat notes as evidence of thought — as proof that we engaged, processed, understood. They are not. Notes are the residue of an encounter with an idea, not the encounter itself. The distinction matters more than it might seem, because conflating the two is one of the most reliable ways to feel productive while making no intellectual progress whatsoever.
There is a notebook on my shelf — physical, spiral-bound, nearly full — from a period several years ago when I was working through a difficult set of ideas in philosophy of mind. I filled it over the course of about four months. Page after page of careful notes: definitions copied from textbooks, arguments transcribed from papers, diagrams reproduced from lectures, key terms underlined in the particular emphatic way that feels, in the moment, like it is reinforcing something.
I picked it up recently and read through it. Almost none of it was mine.
Not in the plagiarism sense — I had cited everything. But in a more fundamental sense: the ideas on those pages had passed through me without leaving much behind. I had recorded them, carefully and diligently, and the recording had felt like understanding, and it had not been understanding. The notebook was full and my comprehension was thin. The relationship between the two things — the fullness and the thinness — turned out to be less accidental than I had assumed.
The Confusion and Where It Comes From
The confusion between notes and thinking is not arbitrary. It has a psychological structure, and understanding that structure is more useful than simply criticising the mistake.
Taking notes feels like thinking because it involves many of the same surface behaviours. You are attending to an idea. You are selecting what seems important. You are producing output. The physical or digital act of writing something down has enough in common with the act of processing an idea that the two experiences are difficult to distinguish from the inside.
There is also a phenomenon that cognitive scientists call the fluency illusion — the tendency to mistake the ease with which information can be processed for understanding of that information. When you read a sentence and it makes sense to you, the experience of sense-making feels like comprehension. It is not always comprehension. It is sometimes just parsing — the recognition of familiar words in familiar grammatical structures, producing a signal of coherence that the mind interprets as understanding. Copying or paraphrasing text into notes amplifies this effect: the act of engaging with the material produces a feeling of engagement that is real, but that feeling is not a reliable indicator of how much has actually been understood.
The philosopher Michael Polanyi described a related distinction with his concept of tacit knowledge — the observation that we know more than we can tell, and correspondingly, that we can tell more than we know. A note is a telling. What it captures is the articulable surface of an idea — the words in which the idea can be expressed, the examples used to illustrate it, the argument structure as it appears in a text. What it does not capture, and cannot capture, is the understanding that makes those words mean something — the web of prior knowledge into which the idea must be integrated, the questions the idea raises, the ways in which it conflicts with or confirms what you already believe.
The Generation Effect
There is a body of research in cognitive psychology on what is called the generation effect, first demonstrated by Norman Slamecka and Peter Graf in 1978. The finding is straightforward but has large implications: information that you generate yourself — that you produce, rather than passively receive — is remembered significantly better than information you merely copy or read.
In the original experiments, participants who completed word pairs (HOT–C___) remembered the target words better than participants who read the completed pairs (HOT–COLD). The effort of generation — of producing the answer from partial information — produced better retention than the passivity of reading a complete answer. Subsequent research has extended this finding across many domains: self-generated examples, self-generated explanations, and self-generated connections between ideas all produce better retention and better transfer than equivalent amounts of time spent reading material someone else generated.
The implication for note-taking is uncomfortable: the note that is most comfortable to take — the verbatim transcription, the copied definition, the screenshot of the slide — is the note that does the least work. It requires no generation. It requires only recognition and reproduction. The effort that feels like engagement is, cognitively, passive consumption wearing the costume of production.
This is why the Cornell Note method, or Andy Matuschak's concept of evergreen notes, or the Zettelkasten principle of writing notes in your own words — all of which are often dismissed as unnecessarily effortful — are producing a real effect rather than a stylistic preference. Forcing yourself to articulate an idea in your own words, without looking at the source, requires generation. It requires you to reconstruct the idea from your own understanding rather than copy it from someone else's formulation. The resistance you feel when you try to do this — the difficulty of finding your own words, the uncertainty about whether you have understood correctly — is not an obstacle to learning. It is learning.
What Recording Actually Does
This is not an argument against notes. It is an argument for understanding what notes are actually for, as distinct from what we habitually use them as.
Notes, understood clearly, are not storage. They are not the place where understanding lives while you are not using it. The understanding lives in you — in the neural structures that constitute your grasp of an idea — and notes are a set of external cues that can help you retrieve and reconstruct that understanding.
The distinction matters practically. Storage implies that you can put something in and get the same thing back out. If notes were storage, then the quality of a note would be measured by the fidelity with which it preserved information — by how accurately it reproduced the original source. But notes as retrieval cues are measured differently: by how effectively they allow you to reconstruct understanding that you already have, by how usefully they prompt the connections and applications that constitute actual comprehension.
A note that says "consciousness is the hard problem — Chalmers 1995" is poor storage — it loses almost everything from the original text — but it can be an excellent retrieval cue if, when you encounter it, it triggers a genuine reconstruction of the arguments, the objections, the implications. A note that copies three paragraphs from the original paper is better storage but may be a worse retrieval cue — it gives you the text back without requiring you to reconstruct the understanding, which means the understanding does not get exercised.
Richard Feynman, who was famously good at understanding things deeply, described his method for testing comprehension as the deliberate attempt to explain a concept as simply as possible, without jargon, to someone without background in the field. If you could not do this, he argued, you did not understand the concept — you had the words without the understanding the words were supposed to carry. This test is worth applying to your notes: could you reconstruct the idea from the cue without consulting the source? If not, the note is recording something you have not yet understood.
The Highlighting Illusion
Before reaching the notebook, there is an even earlier stage where this confusion operates — one that more people encounter, and that is worth examining on its own terms.
Highlighting — underlining, starring, colour-coding — is probably the most widespread study strategy in the world, and it is also, by the research, among the least effective. A 2013 meta-analysis by Dunlosky and colleagues, reviewing the evidence for ten common study techniques, found that highlighting produced low utility across virtually every measure: it did not improve comprehension, did not improve retention relative to reading without highlighting, and produced no detectable benefit for transfer to new problems.
The explanation for this is the same as the explanation for the note-taking problem, and it is worth stating clearly: highlighting is the selection of content without the processing of it. It identifies the thing that seems important without requiring you to do anything with the importance. The highlighted passage sits in the text, distinguished by colour, and produces a feeling of having engaged with it — of having marked it as significant, of having done something more than passive reading. The feeling is real. What it is tracking is not comprehension but attention, which is not the same thing.
There is a specific moment in reading that highlighting exploits: the moment when something clicks — when a sentence lands in a way that produces a flash of recognition or insight. This moment feels like understanding, and sometimes it is. But often it is the sense of understanding without its substance: the recognition of a familiar pattern, the resonance of an idea with something vaguely held, the appeal of a well-phrased sentence. Highlighting that moment does not deepen it. It marks the location of something that felt important and returns you to passive reading. The flash does not get followed. The click does not get interrogated.
What would follow the flash is the question: why does this land, and what does it connect to? What would interrogate the click is the attempt to say, without looking at the highlighted passage, what it means — to reconstruct the argument in your own words and find where the reconstruction fails. This is the work that highlighting never does and that the feeling of highlighting prevents, because the feeling of having done something with the passage is sufficient to satisfy the mind's sense of engagement.
The notebook full of notes, the Kindle device full of highlights, the Readwise account full of daily review cards — all of these can be constructed entirely from the surface of reading, without the encounter with ideas that produces genuine understanding. They look like evidence of intellectual engagement because they share properties with such evidence: they are produced by engagement, they are voluminous, they are organised. What they do not share with such evidence is the thing that matters most: the requirement that the ideas pass through genuine comprehension before being recorded.
The Notebook as a Performance
I want to return to my spiral-bound notebook for a moment, because I think its failure was not random.
The period in which I filled it was a period of considerable intellectual anxiety. I was working on something difficult, I was not sure I was understanding it well enough, and I was, I think, using the act of note-taking partly as a performance of diligence — for myself, primarily, but also for the imaginary observer who might have been watching to see whether I was taking this seriously. The fullness of the notebook was evidence I was taking it seriously. The evidence was, it turned out, misleading.
This is a function that notes serve that is almost never discussed in productivity literature: the function of self-reassurance. The growing collection of notes tells a story about a person who is doing the work, who is engaging, who has accumulated something. It is a story that feels good to inhabit. It is, frequently, a story that is somewhat detached from what the notes actually contain — which is, more often than most people would like to admit, an archive of ideas that were encountered without being absorbed.
The problem is not the note-taking itself. The problem is treating the presence of notes as evidence of understanding rather than as evidence of encounter. Encounter is necessary but not sufficient. Understanding requires something more, and that something more is consistently more effortful and more uncomfortable than the note-taking that precedes it.
What Thinking Actually Looks Like
Thinking — genuine intellectual engagement with an idea rather than the passive recording of it — has a phenomenology that is distinctly different from the phenomenology of note-taking, and recognising the difference is the beginning of using notes well.
Thinking is generative. It produces something that was not in the source material: a connection the source did not make, a question the source did not raise, an objection the source did not anticipate, an application the source did not consider. When you are genuinely thinking with an idea rather than about it, you are not reproducing the idea's own structure — you are doing something with the idea that the idea itself does not do.
Thinking is also effortful in a specific way. It involves uncertainty — the recognition that you do not yet know what you think, that the idea is still unstable, that the formulation you have reached is provisional. This uncertainty is uncomfortable, and the discomfort is a reliable signal that thinking is occurring rather than recording. Recording an idea feels relatively smooth. Thinking with an idea feels turbulent.
The philosopher Hegel is supposed to have said that the familiar is not known merely because it is familiar — that the most common conceptual error is mistaking recognition for understanding. Notes, at their worst, are a machine for producing recognition. You encounter the idea, you note it, you encounter the note, you recognise it. Each recognition feels like confirmation of understanding. None of it requires you to do anything with the idea that you could not have done by reading the source.
The corrective is to treat thinking, not recording, as the primary activity — and to treat notes as tools in service of thinking rather than products of it. Notes that record thinking are different from notes that substitute for it. The former ask you to articulate what the idea does for your understanding: where it connects, what it disrupts, what question it answers and what questions it opens. The latter ask you only to preserve what the idea says.
Practices That Close the Gap
The gap between notes and thinking is not closed by taking fewer notes. It is closed by changing the relationship between note-taking and the cognitive work that note-taking is supposed to support.
The most consistently useful shift is the simplest and the most resisted: close the source before you write. Not after you have found the formulation you want to copy — before you have found it. Attempt to reconstruct the idea from memory, in your own words, and then compare your reconstruction with the source to find where you were wrong, where you were imprecise, where you omitted something important. The comparison is the learning; the discrepancy between what you thought you understood and what you actually understood is the information that drives genuine comprehension.
A related practice is what the educator Mark Guzdial calls elaborative interrogation — asking, of each idea you encounter, why it is true. Not what it says, but why it is so. The effort of constructing an answer to this question — even a partial or uncertain answer — produces significantly better retention and better transfer than restating the idea without explanation. The question is uncomfortable because it often reveals that you cannot yet answer it, which is precisely the information you need.
Finally: write notes that have arguments rather than descriptions. A note that says "Kahneman distinguishes System 1 and System 2 thinking" is a description. A note that says "Kahneman's distinction between System 1 and System 2 thinking is useful for understanding why the fluency illusion is hard to correct — if fast thinking is what produces the feeling of comprehension, and fast thinking is also what note-taking triggers, then the feeling is self-reinforcing" is an argument. Arguments require generation. They require you to do something with the idea rather than simply record that the idea exists.
The Honest Assessment
The spiral-bound notebook on my shelf is not a record of thinking. It is a record of reading — careful, attentive reading, but reading nonetheless. The ideas it contains have a relationship to my understanding that is indirect and partial: they passed through me in a way that left traces, and the traces are not nothing, but they are considerably less than understanding.
I have learned to treat this as a starting point rather than an indictment. The encounter the notebook records is real. The reading was not wasted. But the work of converting encounter into understanding — the generative, effortful, uncertain work of doing something with ideas rather than storing them — is work that notebooks cannot do and that their fullness does not indicate has been done.
The question to ask of any note you take is not "does this accurately record the idea?" but "does this require me to think?" If the answer is no — if the note can be produced by copying, transcribing, or paraphrasing without any uncertainty about whether you have understood correctly — then the note is recording an encounter. The encounter may be valuable. But the note itself is not thinking, and its presence on the page is not evidence that thinking has occurred.
Notes are one of the most useful tools for thinking, and one of the most reliable substitutes for it. The difference is not in how many notes you take, or how carefully you organise them, or how sophisticated the system is in which they live. The difference is in whether the act of taking the note requires you to understand — or merely to record that understanding is something you intend to do later.