Writing Slowly

From time to time I venture opinionated answers to questions people have about making notes — and right now is one of those times.

NB: A Zettelkasten is a simply the German term for a box with paper slips in, a once-popular way for scholars and writers to make and keep their notes. By extension it’s the name of a contemporary method for making digital notes too; but is there a Zettelkasten method? —

Question: What made it click for you, and what do you wish you knew before starting?

I’ve been reading about Zettelkasten for a while and keep going back and forth. On one hand it seems like the most thoughtful way to build a knowledge base. On the other hand every time I try to set it up I spend more time on the system than on actual thinking. Genuinely curious — what made it click for you? And what do you wish you knew before starting? I’m also doing a small research project on how people manage knowledge and what frustrates them most. If you want to share more structured thoughts. Reddit/Zettelkasten

My answers:

what made it click for you?

Workable modularity. My ADHD makes long-form writing a chore, so it’s been really helpful to find a process to make lots of bits add up to something worthwhile. I’m not sure everyone has this same issue though.

And what do you wish you knew before starting?

That the ‘rules’ are actually very simple and that keeping it simple is the key.

• Plain text (Markdown) notes.
• Each note is a single idea with a unique ID.
• Each note deserves a clear title.
• Notes link meaningfully to other notes.

That’s it, and I wish I hadn’t tried everything else first.

If you want to share more structured thoughts.

Do I ever?

Question: A second brain for my life?

I know that zettelkasten is used mainly for academia and creating written work but could I use it to create a second brain for my life and for learning subjects related to engineering? I’m rather a forgetful person and I often miss what’s important about other people, things I’ve learnt and even things about myself. I believe that this system could force me to think and help to create long-term memory based on system of small connections. My goal is not any measurable output like writing an article or a book, but to not lose myself in the world of people, ideas and information. Is there a better system/solution for my problem or should I try out Zettelkasten for that? Reddit/Zettelkasten

My answer: Yes you can use a Zettelkasten for personal knowledge management, but in this case, a better fit might be a personal wiki.

A Zettelkasten breaks your thoughts apart to build new arguments (analysis and synthesis). In contrast a personal wiki keeps related thoughts together while still preserving the benefits of linked, self-authored thinking.

Strict Zettelkasten modularity, one idea per note, exists to help writers reassemble notes into articles or books. If your aim is simply to retain knowledge and track your life, forcing every thought into an atomic note requires a friction that may be unnecessary for your use case. A personal wiki lets you keep longer, evolving pages while still keeping some of the core habits of the Zettelkasten approach. I ultimately found long wiki-style pages hard to use in my writing process, but not everyone intends to write.

But even if you do go for a wiki there are still some Zettelkasten habits that might be useful:

1. Don’t just collect information, process it too

It’s tempting just to clip text or save links. But you’ll probably forget why you did it. Write notes about concepts or people or whatever in your own words. Translating ideas into your own language encourages active recall and builds long-term retention. Crucially the ideas become yours rather than someone else’s. That’s where a lot of the value lies.

2. Create dense, contextual links

Skip the heavy folder hierarchies. Instead, link pages directly within your sentences so notes connect through natural relationships. This mimics (to some extent) how human memory works and it makes retrieval fairly easy.

3. Let the structure emerge from the bottom up

Don’t bother designing a master template on day one. Some people use classifications like the Dewey Decimal system. I haven’t found this useful because for most of these categories, you ain’t gonna need it. Create individual pages for specific topics, projects, or people first. It’s only worth building broader index pages when you have enough notes to justify them.

4. Decouple your notes from chronological time

There’s a risk that a daily-log-style wiki might hide patterns over time. Dedicated pages for ongoing subjects let you aggregate scattered details about a person or topic into one living document. I’m not saying don’t journal. I like journaling. A lot. But I transclude sections of my journal into separate notes to help me spot the wider patterns.

5. Iterate to improve as you go

Whatever system or app you use, just start, and make gradual improvements to your system as you work with it. There’s nothing like doing it wrongly to clarify what you need to do differently. Few decisions, if any, will be fatal. I favour the simplest possible and least locked down approaches so I can shift without finding I’m locked in to someone’s proprietary system. But even then you can probably still escape if necessary.

Oh, I nearly forgot to say that I’m also a rather forgetful person and I’ve found that almost any form of at least minimally structured writing (rather than just totally random notes) has been really helpful. Another thing that really helps is a regular review of my notes.

Question: Struggling with reference numbers?

Im a PhD student in Literature, Ive recently discovered the Zettelkasten method and fell in love very quickly. One problem, I have so much doubt favour my coding system. At the moment I’m using 1.1.1.1(a/b/c), however this is pointless because I genuinely dont have references for each point. Does anyone have any examples of their own tracking system or keys? I’m stuck before I properly begin. Reddit/Zettelkasten

My answer: The numbering just shows you where you had a clear run of connected ideas, and this is an entirely subjective judgement, which is barely worth getting concerned about. You can’t really ‘do it wrong’.

To start, think of it as ‘running numbers’. The first note is number 1, the second is number 2, the third is number 3, and so on. It couldn’t be easier. This is a very standard approach to library cataloguing, especially for acquisitions, where the lowest number is the oldest acquisition and the highest number is the newest.

So far, so simple. Niklas Luhmann took this a step further, but this next step is also fairly simple.

Let’s say you have six different ideas and you make it all the way to note 6, when you suddenly realise that you’re not done with the subject of note 2. Now you have a choice. You can now:

• just write note 7 and link it back to note 2, or
• you can label it as note 2a instead.

The same goes for note 2a as well. If you want to write more, you can now choose whether to make it:

• note 2b (because it follows on from the idea in note 2), or
• note 2a1 (because it follows directly from note 2a, not just note 2), or else
• to label it note 7, because it’s simply a new note with a new idea.

That’s it. That’s the whole system.

Observe that the numbering just shows you where you had a clear run of connected ideas, and this is an entirely subjective judgement, which is barely worth getting concerned about. You can’t really ‘do it wrong’. Further, the decisions about numbering hardly paint you into a corner, since you can always link notes across this labelling, and despite it. For example, you can always link note 7 to note 2, which is similar but not identical to creating a note 2a instead. Either way you have formed a useful connection between your notes.

People argue about how much use they get from this kind of numbering, which is commonly referred to as ‘Folgezettel’ (follow-on notes). Niklas Luhmann clearly seems to have benefitted from it (his Zettelkasten was pre-digital though). But even so he also noted that the important thing is that each note should have ‘a firm, fixed place’ (Stellordnung). In other words, a fixed ID to allow cross-references to be accurate.

My own approach is to use unique timestamps. Again, these are really just slightly fancy running numbers. On reflection, the timestamp information is marginally less useful than the Folgezettel information would be. I can’t really see ‘chains of thought’ at all. But the key remains: to make useful links between notes, and to ensure each note has a unique and permanent identifier.

Question: Is there a simplified Zettelkasten approach?

I’m trying to find a system where I can organize my thoughts, find patterns and also help with my projects. “Full” Zettelkasten seems a little overkill, tbh. Do you guys know any other method or maybe a simplified Zettelkasten approach? Reddit/Zettelkasten

My answer: Yes, there is a simple approach, which is to start really simply and to add complexity only when you’re comfortable - or never.

I turned to the Zettelkasten approach because I have ADHD and get overwhelmed/bored by needless complexity. This is what has worked for me. I grew my Zettelkasten organically, in stages — and only moved forward when I was fully comfortable with what I was doing and could see in practical terms what benefits the next stage would offer. Stage one is the only essential stage. The rest follow naturally, or not at all.

Stage one: the basic principles

I have a few simple rules that I mostly stick to.

1. Plain text (Markdown) notes.
2. Each note is a single idea with a unique ID.
3. Each note deserves a clear title.
4. Notes link meaningfully to other notes.

That’s it. This effectively is a Zettelkasten, even though it’s very simple. With this you will go far. You don’t really need the other stages. They’re all just embellishments to the basic principles.

Stage two: reference notes and main notes

After you’ve done this for a bit, you’ll probably notice that it’s useful to have two kinds of notes:

• reference notes, that give bibliographic details of a source you’re reading, watching or listening to; and
• main notes, that make a single point.

You might well find you’ve already been doing this, but it won’t hurt to get a little more systematic about it. Each reference note will link to a set of main notes that relate to that reference source. And each main note that refers to a source will link back to a reference note, so you can find all the notes that relate to that source. That’s pretty much it, and you don’t really need the next stages.

Stage three: hub notes and structure notes

This is where you notice that a reference note acts as a kind of hub for your thoughts inspired by that particular source. You realise that you can have other kinds of hub note too if you want. Like a note that identifies a major concept, which links to a set of further notes that all explore that concept in more detail.

And while a hub note is pretty much just an unordered list of notes sharing a common theme, you can also create structure notes, where the list is ordered in some way, and is on its way to becoming an outline of a piece of writing. That’s it. You don’t need the next stage, which is…

Stage four: four simple folders

Bob Doto, author of A System for Writing, has suggested four folders are all you need for a complete Zettelkasten and I’ve found this advice helpful. Here they are:

• in-box (fleeting notes you haven’t yet fully written)
• reference notes
• main notes
• sleeping (the overflow from your in-box, consisting of notes you’ve pretty much abandoned, but that you might get back to some day).

More than just folders, these represent a simple workflow. Basic notes start in the in-box, and move to another folder as you work on them, or else abandon them. The principle is that it’s better to write a super-crude note for the in-box than to write nothing, and it’s better to put a note to sleep than to worry about whether it’s ‘complete’ or not. The notes you work on are the notes that engage your attention.

Now there’s just one more stage, in which you start writing.

Stage five: produce something

Some people treat their Zettelkasten as a knowledge bank, and just keep saving it all up, but I prefer to make regular withdrawals by creating things from my notes. Some say that making notes is valuable in itself, because note-making is part of the thinking process.

Fine, I agree, but now since I’ve done the thinking via my notes, I might as well have something to show for it. The German sociologist Niklas Luhmann and other twentieth century, pre-digital scholars used their notes to write articles and books. It wasn’t just a database without a purpose. As you produce things, you’ll be able to fine-tune your Zettelkasten to serve your unique purpose. You might well stick with the four basic principles of stage one, because in my opinion, the purpose of a Zettelkasten is to do something with it.

Now if you’ve read this far and are thinking ‘why did he make it so complicated?’ just go back to stage one and consider those four basic principles. That really is the core.

Question: How do you bridge the gap between ‘academic’ notes and your Zettelkasten?

I’m a big fan of Sönke Ahrens’ “How to Take Smart Notes” and the Zettelkasten methodology, but I’m curious about how other students handle the intersection between their college coursework and their long-term knowledge base… My question for you: Do you keep your college notes completely separate from your personal Zettelkasten, or do you mix them? If you mix them, how do you prevent your vault from being cluttered with “temporary” exam-specific info versus long-term knowledge? Reddit/Zettelkasten

My answer: My experience is that there’s a big difference between undergraduate style study where you mostly have to remember ‘facts’ and the ‘correct’ answer, and honours or postgrad work, where you’re researching for yourself.

In the case of the former, I found my notes to be really specific and strongly focused on exactly what I needed to know just to pass exams and assignments. So I decided to firewall my Zettelkasten and keep my study notes separately. This was probably the right decision, since I hardly had time to scratch myself, let alone take thoughtful notes. A lot of the material was literally just for one test, with no sense that I’d ever need to know it again.
A down-side to keeping these notes separately is that this knowledge now doesn’t cross-pollinate with the rest of my Zettelkasten, but since I’ve kept these study notes I can always go back to them if I feel the need. But as predicted, I mostly haven’t felt the need to do so. Also:

When a specific topic from a class really interests me or feels relevant to my broader understanding of AI, I “extract” it into an atomic note in my main vault.

This is the way to go.

I expect a few people will have a different perspective, but this is my experience.

Question: I spent a year building my graph. It looks great. It doesn’t move me forward

A question about the usefulness of graph views in tools such as Obsidian. The original post was deleted but the comments are still available at Reddit/Zettelkasten.

My answer: There are useful things to be done with these graph views.

For those not yet put off, I’d recommend Arthur Perret’s article What’s the point of the graph view?, and the response by Hendrik Erz, Quo Vadis PKM - as well as Eleanor Konik’s It’s not just a pretty gimmick, which she mentioned in the comments here.

There are useful things to be done with these graph views, though I must admit I myself have not done them.

Question: How did Luhmann come up with those 11 super-themes around which he organized his entire zettelkasten?

In ZK2, how did Luhmann come up with those 11 super-themes around which he organized his entire zettelkasten? [A comment relating to a video of researcher Johannes Schmidt describing the Zettelkasten of Niklas Luhmann]

My answer: Luhmann’s second Zettelkasten (ZKII) was arranged according to eleven top-level categories, based on subjects or themes associated with his singular major project, a theory of society.

Niklas Luhmann was a German sociologist who wrote about 90,000 notes in two collections. Here are Luhmann’s top-level categories for his second Zettelkasten:

1. Organisation theory
2. Functionalism
3. Decision theory
4. Amt: office, post, job, duty
5. Formal / informal order
6. Sovereignty / State
7. Isolated/individual terms, problems
8. Economy
9. Ad hoc notes
10. Archaic societies
11. High cultures

In his lecture on Luhmann’s Zettelkasten, Johannes Schmidt observes that these headings are far from comprehensive. Instead they strongly echo the progression of Luhmann’s scholarly interests over many years.

“Looking at these you quickly see that this does not describe a certain body of knowledge to work through like in the first Zettelkasten. If you know a bit about the development of Luhmann’s theory you quickly recognize this as a historic record of research interests.” [Source doc]

Schmidt claims the first five sections are an organised summary of Luhmann’s research interests in the 1960s, while the later categories are less organised and more ad hoc.

But even in the supposedly more systematic sections, there is a marked unevenness to the amount of material. For example, section 21 of category 5. — Formal/informal order — is labelled ‘Functionalism’. This section contains between 15,000 and 16,000 notes out of 67,000 total (more than 1/5). In marked contrast, there are other sections that are almost empty.

And the top-level Zettelkasten categories don’t align with the major contours of his theory. These are the subsystems of society that I can clearly identify from Luhmann’s work. They don’t really match his top level Zettelkasten categories. Im not saying this is any kind of problem, just making an observation:

• Economy (Payment/Non-payment),
• Law (Legal/Illegal),
• Science (True/Untrue),
• Politics (Power/Opposition),
• Religion (Immanence/Transcendence),
• Art (Fit/Unfit),
• Education (Better/Worse),
• Mass Media (Information/Non-information),
• Health (Ill/Healthy),
• Intimacy (Love/Non-love)

But Luhmann didn’t have this mapped out in 1963 when he started his second Zettelkasten. He derived it from working in his Zettelkasten for many years. I see this as a fantastic example of someone working on what mathematician Richard Hamming would have called their 10-20 basic problems — but that’s a story for another day.

Question: How long should a note be?

I understand the core of zettelkasten is writing atomic notes so that as they link they build an organic and ever evolving web of your knowledge and so on and so forth. My question however is how small should atomic be? I’ve read that if it’s longer than three paragraphs, you should probably split it into more notes and link them up. Reddit/Zettelkasten

My answer: There’s no fixed length of a note. What matters is the difference between simplexity and complexity.

A simplex note, typically involving a single idea or concept, can be re-used in different contexts with little effort. A complex note tends to involve several ideas braided together, and to reuse that note you have to untangle the threads and separate them out. It’s easier to write that complex note but the complexity leads to more work later.

This concept of ‘separation of concerns’ is key for computer programming. For example, it’s evident on the cover of a well-known book, A Philosophy of Software Design, by John K. Ousterhout, which you can see in my article: I found a way to create order from my jumbled ideas. I’ve found it’s also a really helpful principle for writing useful notes.

But I didn’t and couldn’t appreciate this difference until I’d written quite a few notes and tried to reuse them. Eventually it clicked: the point of ‘atomic notes’ isn’t to write short notes (although that’s OK); it’s to write modular notes.

You work out the right length by noticing what happens with your longer notes when you try to use them.

Question: Can you work with a hybrid digital Zettelkasten/ physical notebook system?

Reddit/Zettelkasten

My answer: Pen and paper is still great, especially while reading, which is why I haven’t gone completely digital.

Here’s what I do with my notebooks:

• Treat each notebook as a cite-able source, exactly like any other source.
• Number each notebook and every page.
  • I number my notebook pages starting from p.1 each new notebook, but if I was to start over I might instead follow cyberneticist W. Ross Ashby and number all pages in one long continuous sequence. He made it to 25 volumes and p.7189.
• Refer to all notebook notes by unique ID, e.g. “Notebook 2, p.30”.
  • All links are made in my digital Zettelkasten, not in the notebooks. If a link is worth making, it’s worth writing a short note about. Otherwise I end up with enigmatic links, the purpose of which I’ve forgotten. I suspect this is a potential problem with references like ‘> (C.08.notebookB: 210 - Stiegler 2)’ - but I might have misunderstood this.
• For drawings and mindmaps that are worthy standalone artifacts, I photograph them and then link to the image in my Zettelkasten (I have a ‘media’ folder for this purpose, as my main Zettelkasten is text files only).

This system requires me to keep all my notebooks, even when they’re mostly full of ephemeral or fleeting notes. But I’ve found this is OK. Old ’temporary’ notes have later proved to be valuable so I tend to keep them all. But also, I try to write my ‘main’ or ‘permanent’ notes in such as way that they make full sense even without access to the original source, so if I did lose my notebooks, it wouldn’t be a disaster for my Zettelkasten.

I tried just having paper notes but ultimately, full-text search was just too useful, which is why I’ve mostly gone digital. I also find semi-automatic backlinks, an ability to find ‘orphans’ (notes that are not directly linked from another note), and an ability to find ‘missing’ notes (notes that don’t yet exist, but have been linked to from other notes) really useful.

But pen and paper is still great, especially while reading, which is why I haven’t gone completely digital.

Question: What exactly becomes a permanent/atomic note?

My problem is simple: Where do I exactly draw the line between what should be made into atomic notes (externalized) and what should be memorized (internalized) or written in some other form (another form of externalization)?
…Niklas Luhmann wrote an average of only 6 atomic notes per day. I just cannot wrap my head around what filters he put down. Everything seems important. Reddit/Zettelkasten

My answer: Personally I don’t use the ‘internalized’ category you mention, because I have a very average memory so I forget almost everything I attempt to remember. That means it’s easier just to write things down.

The filter that’s hiding in plain sight is that all Luhmann’s notes (in his second Zettelkasten, ZKII anyway) related closely to his scholarly project of writing a new ’theory of society’. He either didn’t write any notes that couldn’t directly inform his sociological work, or else he only wrote about what he’d made notes of. Either way he seems to have been filtering rather strongly. This relates to your point number one, which you say doesn’t solve your problem¹. I’d suggest it might. 

But personally I don’t use the ‘internalized’ category you mention, because I have a very average memory so I forget almost everything I attempt to remember. That means it’s easier just to write things down (’externalized’?). 

That leaves two different kinds of notes:

• Things I might write about (these go in my Zettelkasten); and
• Things I almost certainly won’t write about, such as shopping lists (these are not for the Zettelkasten). 

Question: Should I use structure notes?

I have a question about structure notes. If my goal is not writing should i use them in my zettelkasten? Reddit/Zettelkasten

My answer: Even if you’re not writing to publish, this process of working out different levels of abstraction can be very useful – it can really clarify your thinking and understanding.

Ideas work at different levels of abstraction. A series of simple atomic notes, where each note relates to a single idea, can be combined to create a note at a higher level of generalization. A hub note would perhaps just list the component notes, but a structure note provides added structure (yes, really!), to suggest how a small group of simple ideas combines to form a more complex idea. Even if you’re not writing to publish, this process of working out different levels of abstraction can be very useful – it can really clarify your thinking and understanding.

Consider a student writing notes about blood clotting. Their initial note might cover matters such as platelet aggregation, coagulation cascades, anticoagulant proteins, and clot breakdown. So is this one idea (blood clotting) or four (the ones I just mentioned)?

In this example, I’d suggest writing a single structure note, with links to four further notes at a lower level of abstraction. Here’s an example:

How blood balances clotting and flow

Blood must prevent hemorrhage without causing thrombosis. Several mechanisms achieve this balance:

[[Platelets form hemostatic plug]]: platelets aggregate at injury sites and create an initial seal. This triggers [[Thrombin converts fibrinogen to fibrin]], which generates a stable mesh that reinforces the plug. As a safeguard, [[Protein C prevents excessive clotting]] when it inactivates coagulation factors when clotting is no longer needed. Finally, [[Plasmin breaks down clots]] after tissue repair completes and restores normal flow.

When these mechanisms fail, bleeding disorders or thrombosis results.

Notice that this note describes a process (blood clotting), which none of the more specific notes alone could realistically cover. That’s the extra value the structure note delivers. Notice also that the note-writing process is reversible. You could start with a structure note, then break it down into its component parts and write related notes at a lower level of abstraction, or you could start with one of the more basic notes and then work up to the structure note.

For my money, this is one of the major benefits of the Zettelkasten approach: it’s like an ongoing lesson in analysis and synthesis across different levels of abstraction.

Finally, a disclaimer: please don’t rely on this example during your next medical emergency involving blood loss. I have little real idea how blood clotting actually works, so I made it up.

Thanks for reading this far. Amazing! You are a true knight (or dame) of the order of the Zettelkasten. But what if you didn’t find your own question answered here? That’s a shame, but don’t despair. Here’s a previous edition of the answers to note-making questions service:

Some urgent notemaking questions find answers

And you can always hit me up on micro.blog or Mastodon or BlueSky for a geeky conversation. Oh, and you can subscribe to the weekly email too.