Note-Taking Systems Compared: Zettelkasten, Cornell, and What Actually Sticks

Niklas Luhmann, a German sociologist who died in 1998, produced roughly seventy books and four hundred scholarly articles over his career. When asked how he managed this, he was matter-of-fact. He did not, he said, really do the writing. His notes did. What he called his Zettelkasten, a slip-box containing around 90,000 index cards accumulated over several decades, was a kind of thinking partner. He would consult it, find unexpected juxtapositions among cards he had written years apart, and those juxtapositions would become the next paper. After his death, researchers at the University of Bielefeld began cataloguing the boxes and discovered that the system was stranger and more precise than Luhmann had ever fully explained in print.

The Zettelkasten has since become a minor obsession in certain corners of the internet, particularly among users of note-taking apps like Obsidian and Roam Research. Most of what is written about it online is wrong, or at least misses the point. The slip-box is not a filing system. It is not a place to store information for later retrieval. It is a method for forcing yourself to think about what you read while you read it, and to keep that thinking in a form that can surprise you later.

The mechanics are specific. When Luhmann encountered an idea in a book, he did not underline the passage or copy it down. He wrote a new note, in his own words, expressing the idea as he now understood it. The note was atomic: one idea per card, short enough to be a single thought. He then gave the card an address, a string of numbers and letters locating it in the physical box, and, crucially, he linked it to other cards already in the system. The links were not categorical. They were associative. This card about Parsons reminded him of that card about systems theory, so he wrote the reference down. Over years, the cards accumulated into a network whose topology reflected the actual structure of his thinking, not the structure of any external taxonomy.

Compare this with the Cornell method, which is what most American students learn in high school. The Cornell sheet is divided into three regions. The largest, on the right, holds the notes you take during a lecture or reading. A narrow column on the left, the cue column, is filled in afterward with questions or keywords that index the notes. A strip across the bottom holds a summary of the page, also written after the fact. The system was developed by Walter Pauk at Cornell in the 1950s, and it was explicitly designed around what cognitive psychology already suggested about memory: that the act of reformulating material, and of practicing retrieval against cues, would produce durable learning.

Both systems, you will notice, share a commitment. Neither treats note-taking as transcription. The Zettelkasten forces you to restate the idea in your own words before you are allowed to save it. The Cornell sheet forces you to return to your notes and generate questions against them, which is retrieval practice in disguise. In both cases the friction is the feature.

This is what most student note-taking lacks, and why most of it fails. The typical undergraduate strategy is verbatim transcription, often now delivered by laptop, in which the lecturer’s sentences are copied as closely as possible into a document that will be opened again only during exam week. A series of studies by Pam Mueller and Daniel Oppenheimer around 2014 examined this directly and found that longhand note-takers, who could not physically keep up with a lecturer and therefore had to summarize in real time, scored better on conceptual questions than laptop note-takers who had captured more words. The mechanism, the authors suggested, was the summarization itself. The longhand students were doing something cognitive. The laptop students were doing something closer to dictation.

Verbatim notes also tend to go unread. The volume is too large, the material feels already-captured, and the student arrives at the exam with twenty pages of lecture transcription and no particular relationship to any of it. The notes are a kind of talisman. Their existence reassures, but it does not teach.

The honest diagnosis, then, is that the choice of system matters less than people think. The Zettelkasten is elegant but elaborate. The Cornell method is simpler but no less effective for students willing to do the back-end work. What they share is what actually matters: they force the student to rewrite, to link, and to retrieve. Any system that does those three things will work. Any system that does not will fail, regardless of how well-designed the templates are or how many plugins the app supports.

This is also why drawing can function as a form of note-taking more powerful than either method. Sketching a process diagram from a textbook passage forces every one of those same operations. You cannot draw what you do not understand. The drawing is the retrieval.

A related point, less often made, is that note-taking should be understood as a way of managing cognitive load rather than accumulating information. When you summarize a reading in your own words, you are offloading the structure from working memory into an external artifact. That artifact then becomes something you can manipulate, consult, and revise without having to hold the whole thing in your head at once. The slip-box is, in this sense, a prosthetic for thinking. So is a Cornell sheet, when it is used properly.

There is a kind of student, and a kind of knowledge worker, who spends more time optimizing the note-taking setup than taking notes. Obsidian plugins, custom CSS themes, color-coded tags, elaborate graph views. These activities feel like work. They are not work. They are furniture-moving, and the temptation to do them is strongest precisely when the actual cognitive task, reading the source and figuring out what you think about it, feels difficult. The test of a note-taking system is not how beautiful the graph looks or how clever the tagging scheme is. The test is whether, six months from now, you can retrieve an idea and do something new with it.

Luhmann’s boxes are kept now in a climate-controlled room in Bielefeld. Researchers have been digitizing them for years, and the work is slow, because each card is short but dense, and each link has to be traced. What emerges from the archive is not a filing cabinet. It is a portrait of a mind that spent forty years arguing with itself on paper. The system was the argument. Most note-taking is not.

Photo via Unsplash.