memo.d.foundation / topics / writing / ahrens-slip-box-methodConnect

Research notes/Knowledge work

Ahrens on the slip box, from atomic notes to writing from the bottom up

Reading notes on the book How to Take Smart Notes by Sönke Ahrens, covering atomic notes, connection, rewriting in your own words, flexible systems and writing that starts from the notes you already have.

By tieubao18 Oct 202613 min readEssay

Tags
writing, zettelkasten, note, method
Words
2,602
Contents
Use with AIOpen in ClaudeOpen in ChatGPT

These notes are adapted from David Drysdale's Moby Diction digital garden (publish.obsidian.md/mobydiction, now offline). They summarize Sönke Ahrens's book How to Take Smart Notes (2017). Han curated and edited them for the Dwarves research vault. Ahrens describes the zettelkasten, or slip box, as a way to take notes. These reading notes argue it is closer to a system for generating knowledge. You build a database of focused notes, link each one to the others, and let the growing web produce ideas you did not plan to have. This page gathers the notes on that book in one place, in the order the argument builds: the principles, the habits of note-taking, the systems around the notes, and the writing that comes out of them.

1Principles

1.1Notes are atomic

The principle of atomicity is a governing idea of the zettelkasten method. Each note holds one unit of thought. To paraphrase Andy Matuschak, a note is about one thing but captures the entirety of that thing.

Splitting a text or a concept this way helps you decontextualize each idea and consider it on its own merits. It also makes links easier. If a document packs several ideas together, connecting any one of them to another idea is awkward. Atomic notes get reused, and reuse creates more chances for connection.

The limit is fragmentation. A note that is too granular loses its context. Be kind to your future self and write each note so it is fully understandable without other references.

1.2Decontextualize ideas to make them new

To engage with an idea, you have to decontextualize it. Our brains fit ideas into comfortable boxes, and the impulse is to make everything familiar. Familiarity fools us into thinking we understand an idea and like it, which short-circuits critical evaluation.

New knowledge needs a break from those patterns. Look at things from a perspective free of ego and open to possibility and experiment. A changed perspective makes ideas more fluid, adaptable and changeable. Divergent thinking helps, as long as you do not censor or restrain the ideas.

Cognitive contradictions promote creativity. Ideas that do not make sense, which psychologists call "meaning threats", disorient us and sharpen our ability to spot unusual patterns. Long exposure to that kind of volatility can provoke maladaptive behaviours. Objectifying our models or ideas can help break them down and rebuild them productively (the strategy notes cover this in Strategy as sensemaking).

Four examples show the move in practice:

  • In Modernist theatre, playwrights like Bertolt Brecht rendered strange the patterns, structures and conventions we take for granted, so that the audience could not identify with the characters. Brecht exposed the fictive quality of his characters and their stories. In the same way, our own ideas are never natural. They are products of past thinking and its paradigms.
  • Tiago Forte (2018) argues that to expose and disrupt our assumptions we must keep disassociating ideas from their context. For Forte, overcoming our own paradigms of thought is the final bottleneck in critical thinking.
  • Design thinking offers "reframing the problem" and "depatterning". Both rest on the belief that processes and structures which force us to treat situations as novel lead to better decisions, and they overturn the brain's reliance on heuristics.
  • In Drawing on the Right Side of the Brain, one exercise turns an image upside down. That breaks the system of symbols we use to draw familiar things, such as the features of a face, and makes you draw the shape and contours as they appear rather than as you have internalized them.

1.3Store information where you would want to encounter it again

The question that matters is where you would like to come across the information again.

When we archive notes we usually file them in a box that makes logical sense to the archivist. That suits a librarian, whose business is archiving. Someone who wants inspiration and new insight should ask instead when they would most like to meet the idea again. This is a central tenet of how a zettelkasten is organized. Its heavily linked structure confronts us with ideas from our past and surprises us with thoughts we forgot we had, at the right moment.

1.4Knowledge is a compounding investment

Knowledge is a long-term investment. We rarely get a big score or a sudden windfall. Insight comes when we connect ideas and process them into something new. That takes effort: thinking about the ideas, writing about them, and linking them to other information we have collected. The effort compounds. The more quality information we gather, the more connections we can make, and the more value we can deliver in our work.

More knowledge also creates the conditions for serendipity. With more ideas, more collisions happen between unrelated ideas, and that promotes innovation and creativity.

2Taking notes

2.1Make notes in your own words

When we read, it is hard to tell what matters most and to remember it. With theoretical texts in particular, we must balance short-term and long-term memory to focus on what is important.

Reformulating what we read in our own words helps. It forces us to consider what the text means, what it emphasizes and what it leaves out. Verbatim copying and quotation marks strip a text of its context and therefore its meaning. They skip the step of interpreting and understanding, and what remains is a random assemblage of sentences that do not cohere. Better to engage with an idea, think it through and rephrase it in our own language.

Reading alone is not understanding. Copying notes in another person's words tricks us into thinking we understand, because the idea feels familiar, and that familiarity can make us think we like an idea we have not examined critically.

The zettelkasten method builds in several acts of translation. You translate the source into fleeting notes. Those become your own thoughts and positions in the slip box. They are translated again when you write for publication. Each translation is a feedback loop that confronts you with gaps in your understanding and forces you to close them.

2.2Take notes with connection in mind

Write every note with the notes you already have in view. A note is far more powerful when it joins a larger network of thought, and when discrete ideas that you have recontextualized form constellations.

Writing with those connections in mind from the start makes you likelier to reveal contradictions and inconsistencies in your thinking. Links between a new idea and an old one also deepen understanding of the material.

2.3Read broadly with purpose

Approach a text with intention. Intention focuses attention on the most important information. Read looking for connections to information you have already processed, and look explicitly for limitations, elisions and contradictions.

Also read broadly, to create connections across contexts and to let ideas evolve over time. Watch for confirmation bias, and do not be afraid to seek out disconfirming evidence.

3Systems

3.1Rigid systems are fragile systems

Systems that are unnecessarily rigid fail. What they did not anticipate, or what does not match the environment they were designed for, threatens them. That makes rigid systems poor tools in uncertain or volatile situations. Good system design accepts that change is inevitable. Instead of resisting friction, it channels friction in productive directions.

3.2Hierarchical systems limit connections across their horizontal layers

A hierarchy sends information and connection up and down, and it inhibits horizontal connection. In a hierarchical organization, information and leadership usually move top down. In an information system, a hierarchy builds knowledge silos. Facts that do not roll up to the same parent folder or unit have walls between them, even when they are related. Whether in an organization or a folder tree, the number of relationships a node can have is effectively limited to its parent and the nodes that share that parent.

Related notes tie this to openness: open societies tend to be more innovative than rigidly hierarchical ones, and hierarchical systems may be more rigid than flat ones. The zettelkasten avoids hierarchy and emphasizes connections between notes instead.

3.3Design environments to facilitate behaviours

Willpower is finite and hard to sustain. Designing an environment that supports the behaviour you want works much better in the long run. A well-designed environment asks less of our limited mental resources, so we can spend them where they have the most impact, and it can nudge us toward good activity without making us virtuous. Reduce the friction around behaviours you want to encourage, and add friction to those you want to discourage.

3.4Writing by hand is a forcing function

Writing by hand has several benefits. Some studies find it activates more areas of the brain than typing, which leads some scientists to argue it promotes "deep encoding". Students who took notes by hand remembered more of what they recorded. Students who typed tended to transcribe instead of taking notes, and they processed the information less.

Handwriting slows you down. You take notes more mindfully and think harder about what you want to say. The friction is productive.

4Writing

4.1Brainstorming surfaces available ideas, not new ideas

Brainstorming is the usual advice for finding ideas. It does not generate new thinking. It only brings forward ideas that are already at hand, in raw and unprocessed form. Those are the ideas on top of our heads, not necessarily our best or most original, and not necessarily the best fit for the problem.

That is dangerous because we get attached to our first ideas and are loath to move past them. They anchor us and block critical examination. After a brainstorm we may invest in an available idea and close ourselves off from disconfirming evidence.

Researchers have also found consistently that brainstorming does not produce an abundance of ideas. Groups come up with fewer ideas and cover a narrower range of topics than individuals working alone. Steven Johnson suggests a weakness may be that brainstorms are time-limited. They run in short sessions and force participants to stay in sync inside a defined window, while the connections they are meant to foster usually develop asynchronously over a longer time.

Ideally you never have to brainstorm. Collect and process ideas continuously into an always-available repository.

4.2Write from the bottom up

Much writing advice, especially for students, says to write from the top down: pick a topic, form a hypothesis, then research. It is better to write from the bottom up. Start with the notes and ideas you have already accumulated and let clusters of ideas guide your focus. Those ideas must interest you, or you would not have gathered a cluster of notes around them.

Bottom-up writing builds on information you have already thought about. You do not have to guess your claim, because you stand on your earlier thinking and the ideas sit in a latticework of context. It eases the anxiety of the blank page, because there is no blank page: you always have something to start from. It also tends to surface disconfirming ideas, because you revisit earlier material and must look again at paradoxes and contradictions.

4.3Writing with a zettelkasten

Ahrens describes his method for writing an article with a zettelkasten. Writing does not start from a blank sheet of paper. It emerges from the bottom up, in conversation with your notes.

The process runs in eight steps:

  1. Make fleeting notes on what you read.
  2. Put them in one place (an inbox) and process them.
  3. Make literature notes that document briefly what you do not want to forget or might use later.
  4. Make permanent notes that outline how the literature notes are relevant to your own thinking and research, one note per idea.
  5. File the permanent notes and link them to other relevant notes.
  6. Develop your ideas based on what emerges in the zettelkasten.
  7. Turn your notes into a rough draft, translating them into a coherent manuscript.
  8. Edit the manuscript.

The translation step is critical. An essay assembled from modular notes, however well written, will probably sound pieced together from fragments and lack unity of form and content. Pasting zettels together and adding transitions is not enough. Flaherty points to some of Emerson's essays, which were put together that way: the voice is flat and the arguments suffer without unity.

Two related notes sit beside this one. The sequence in which a zettelkasten forms ideas may not make a good reading experience, so narrative sequencing matters. And a good research question starts from the zettelkasten: look there first to see how it informs your answer and guides further research.

4.4Writing is thinking

The note's claim is blunt: writing is thinking. Putting someone else's ideas in your own words forces deep consideration of what was said, and you must work through contradictions, gaps in logic and elisions in the text.

Writing also gives a productive feedback loop: to write something clearly in your own words, you must understand it.

4.5Knowledge work means transforming old ideas into new ones

Knowledge work transforms existing ideas into something new. "Every intellectual effort starts with an already existing preconception," Ahrens writes, "that can be transformed during further inquiries and can serve as a starting point for following endeavours."

No idea comes from nowhere. The more information at your fingertips, the more receptive you are to flashes of insight. Knowledge compounds: each idea makes several connections with other ideas, and the constellation of thought grows exponentially large and complex.

Seeing connections in unexpected places raises the odds that an insight is unique and strong. A good strategist looks at old things in new ways, and consuming ideas from diverse sources opens new vantage points. The near-futurist Rohit Bhargava reads magazines written for audiences and demographics completely remote from his own experience. He argues that this helps him spot trends and patterns that would otherwise be blind spots.

Note-taking is therefore critical for knowledge work. Capture ideas as notes, ideally atomic ones, and you can decontextualize and recontextualize them and recombine them into new insights.

These notes link out to others in the research vault. Related notes that exist only in the source garden, which is now offline, are left out.

6References

Dwarves Foundation memo. Written by the team and guest writers.Press [ or ] to move between headings