I have a habit: whenever I come across an interesting passage, half-form a thought, or suddenly figure something out during a walk, I jot it down on a “card.” It doesn’t have to be physical — a phone memo, a node in Obsidian, or even a message sent to myself on WeChat all work.
The medium doesn’t matter. What matters is: one card, one idea.
It sounds almost stupidly simple, but this constraint has changed the way I process information.
Why Break Ideas Apart?#
When we take notes, we instinctively organize by topic. Reading a book? List key points under the title. Attending a class? Write notes under the course heading. It’s intuitive, but it has a problem: the notes get locked into the context in which they were created.
Come back three months later and you’ll find those notes only make sense within their original framework. Economics notes belong to economics; psychology notes belong to psychology. They never talk to each other.
Card-based thinking is different. You strip each idea out and let it stand on its own. A card might be just two or three sentences, but it’s self-contained — understandable without any surrounding context.
Three months later, you can place the card “sunk cost fallacy” next to the card “why people stay in terrible relationships for too long,” and suddenly realize they’re saying the same thing.
How to Write a Card#
A few principles I’ve figured out through trial and error — they may not work for everyone.
Use your own words. Copying the original text verbatim means you haven’t actually thought about it. When you transcribe, you feel like you understand, but really it’s just your hand moving. Rephrasing in your own words — even if less elegant than the source — at least means the idea has passed through your brain.
Make it standalone. Imagine your future self picking up this card three months from now, with no memory of what book you were reading or what problem you were mulling over. It should still make sense. If it doesn’t, you’ve written something too dependent on context.
Note the source, but don’t elaborate. A single line of attribution at the bottom is enough. Don’t paste in the entire original passage — that’s making a scrapbook, not a thinking tool.
One card, one idea. If you find yourself drifting in two directions while writing, split it into two cards. This is the hardest rule to follow and the most worth sticking to.
Collision Is the Point#
Once you’ve accumulated enough cards, interesting things start to happen. Pull out a few at random, lay them side by side, and you’ll spot connections you never would have imagined.
Sociologist Niklas Luhmann used this method to produce over 70 books and 400-plus academic papers. He called his card box a “conversation partner,” saying it constantly surprised him. The ideas he put in came out transformed, because the proximity between cards created combinations he’d never planned.
My own experience isn’t quite that dramatic, but I’ve definitely had similar moments. Once, while sorting through cards on “habit formation,” I happened to have a card on “path dependence in urban planning” sitting nearby. The similarity between them sparked an insight: the way personal habits form is a lot like how city roads develop. Once a path is worn in, everyone who comes after reinforces it — even if it’s not the optimal route. That idea eventually became an article.
If I’d filed things the traditional way, “habit formation” would be in the psychology folder and “path dependence” in the urban studies folder. They would never have met.
Practical Ways to Do It#
You don’t need to buy a wooden card box to get started — though that is admittedly pretty cool.
Paper index cards. The most primitive and most direct. The upside is that limited physical space forces you to be concise. The downside is that retrieval is painful once you have a lot of cards.
Obsidian / Logseq / Roam. The advantage of digital tools is bidirectional linking. You can reference one card from another, and the software automatically establishes a connection on both sides. Over time, your notes grow into a web rather than a tree.
Notion databases. If you’re already a Notion user, managing cards as database entries works too. Each record is a card; use tags and relations to build connections. Be warned, though: Notion’s structure is fairly rigid, so you’ll need to consciously resist the urge to over-organize.
Apple Notes / messaging yourself. Good for capture. The key is to periodically process these scattered jottings into proper cards. If you only capture and never process, your notes app turns into a junkyard within a month.
The tool doesn’t matter. Doing it consistently does.
Common Pitfalls#
Cards that are too long. If a card exceeds 200 words, be suspicious. Length usually means you’ve crammed more than one idea into a single card.
Collecting without reviewing. A card’s value isn’t realized the moment you write it — it’s realized when it gets recombined later. If you never revisit old cards, this method degrades into ordinary note-taking. My rule: spend 20 minutes each week randomly flipping through old cards with no agenda, just to see what new connections surface.
Obsessing over categorization. Trying to design the perfect tagging system from day one is a trap. Write first, categorize later — let the structure emerge on its own. Luhmann’s card box had no strict classification system; he simply built reference links between cards and let the structure reveal itself.
Chasing perfection. Cards are drafts — intermediate artifacts of thought. They don’t need to be polished sentence by sentence. Rough is fine, as long as your future self can understand them. Perfectionism only makes you afraid to start writing.
From Cards to Output#
Accumulating cards isn’t the end goal. When it’s time to write an article, prepare a talk, or make a decision, try this:
Spread all the relevant cards out and arrange them like puzzle pieces. Move them around and see what story different arrangements tell. You’ll find that the hardest part of writing — figuring out what to say — has already been done through your everyday card practice. All that’s left is stringing them together and polishing the prose.
That’s exactly how I wrote this article. I didn’t sit down and “plan” what the essay should look like. I pulled out the relevant cards I’d accumulated over the past six months, laid them on the table, and the skeleton of the piece was already there.
Thinking isn’t a one-time event. It’s a process that’s always happening — we just usually don’t write it down. Card-based thinking simply gives that process a container: one small enough, flexible enough to capture ideas, preserve them, and reactivate them at moments you’d never expect.
Thanks for reading! Feel free to share this article or drop me a line to chat.