Getting Started with Zettelkasten

by Alex Nelson, 11 July 2026

Is there a good way to get started with a Zettelkasten?

Well, Zettelkastens have existed for 500 years or so: quite literally any usage of index cards stored in a box qualifies as a “Zettelkasten”.

People usually mean more in the spirit of Niklas Luhmann’s Zettelkasten.

I’ll summarize the basic way to get started with a paper-based Zettelkasten. (I’ve had bad luck with electronic Zettelkastens: the data gets corrupted, software updates are not backwards compatible and I need to start over, backups fail, etc. “Move fast and break things” don’t really work where stability is a virtue.)

Essential steps and ingredients

• You need a box (a shoe box suffices)
• You need printer paper and scissors (or a paper cutter)
• You need a pen (preferably one black ink pen and another red ink pen)

That’s all you need to get started. What do you do?

1. Creating physical slips to write on. Take some printer paper, and cut it into quarters (roughly 5.5 inch by 4.25 inch — or A6 for my non-American friends). When you have few slips cut up, you can start.
   • Index cards tend to be “too thick”, so I recommend printer paper (plus it is much cheaper!). But if you want, you can try index cards. In which case, “Step 1: buy index cards”.
   • My non-American friends may have access to A6 paper already formed. Great! Use that. But in the USA, we don’t have access to that cheaply. So I tend to cut up my own printer paper.
2. Writing. Take one slip. I usually write in “landscape” layout (across the 5.5inch width from left to right).
   1. Leave room to write an ID number. In the upper left-hand corner leave enough space to write an ID number (about 3 cm wide and 1cm tall — or 1 inch wide by 0.25 inches tall) — this is my convention, you can choose a different convention.

   2. Write your thoughts down. Write some idea in the “body” of the slip. For me, this is everything below about 0.25inches from the “top” of the slip—again, this “gap” is my convention, but the “writing stuff down on the slip of paper” is essential—and I tend to leave a little bit of space between each line of writing.

      Important: write down thoughts about one idea, don’t cram too many thoughts on a slip.

      Equally important: write in your own words. (First, to avoid plagiarism. Second, to force yourself to think about the material.)

   3. Forming links. While writing stuff down in the preceding “substep” 2.2, if you happen to talk about something already contained in the Zettelkasten on another slip (the “referred slip”), then you can “link” to that other “referred slip” by writing on the new slip in red ink the ID number of the “referred slip” in parentheses.

      This happens during substep 2.2, as you write your thoughts down you can form links. Or you can leave a gap to write the link later. Whichever you prefer (or is easier).

      It’s important to form links (which are recording relations between the slips), even if they are “obvious” and uninteresting. Because “following links” will produce useful results you can act upon.

3. Integrate slips. Write permanent ID numbers on the slips, then place them in your box.

   • How do you determine what ID number to write down? It does not matter but there are two qualities the ID number must have: it must be “fresh” (not used by any other slip in the Zettelkasten) and it must be “sortable” (you can meaningfully place it in a unique location somewhere in the Zettelkasten).

     “Freshness” ensures there is no ambiguity when referring to the ID number—if there were multiple slips with the same ID number X, it is unclear which slip a link to X references. That is bad because we want to use links to record relationships between two ideas, and this ambiguity causes confusion for “future you”.

     “Sortable” ensures you have a way to unambiguously look up the referenced slip from its ID number alone.

     I have discussed my personal ID numbering scheme elsewhere, but you can choose your own method.

     • Usually I write a “train of thought” (a “twitter thread”) which means I just need to figure out the ID for the first slip, then I can number them sequentially. But it does not matter, because if it should have been placed “somewhere else”…then I can just write on other slips a link to the start of the train of thought.

     • There is no “right location” for a slip, but some places are more suitable than others. For example, ID numbers prefixed by 1/ are about my Zettelkasten. If I write notes about my Zettelkasten, I probably should place them in that section of the Zettelkasten.

       Is it “bad” or “wrong” if I place them in the cluster of notes about “Systems Theory” or “Niklas Luhmann”? No, of course not! I can create links in my pre-existing notes about the Zettelkasten (the ones whose IDs are prefixed by 1/) to reference the new notes I placed elsewhere.

       Sometimes this is even good and desirable, because it will provide a bridge between two distant regions of the Zettelkasten, connecting seemingly-unrelated ideas.

   • Once you have written down the ID numbers on your slips, you should also consider creating links on existing slips in your Zettelkasten referring to these new slips. This will help make sure that these new ideas “won’t get lost”: it’s integrated into the network of notes.

…and that’s it. This is the “floor model” of the Zettelkasten workflow, which you can modify according to how you need it changed.

Notice that step 2 has multiple things going on when “writing notes”, the most important are forming links and writing text.

You can modify this workflow as needed. For academics, you might want to add a step for writing down “reading notes” which would be stored in a separate “shoe box” (the Bibliography apparatus).

What do I write about?

Anything!

But if you are stuck for ideas, I would highly recommend start with writing about the idea of a Zettelkasten. This is what I have done.

To get a sense of what it looks like, I tend to also give “titles” to my slips of paper (Luhmann did not always give “titles” to his slips). The first “section” of my Zettelkasten I can summarize as follows — writing <id>. <title> for each slip:

• 1. Zettelkasten The idea of Zettelkasten as some kind of method for taking notes
  • 1/1 Slips The basic constituent of a Zettelkasten is the slip of paper. Without this, there is literally no Zettelkasten.
  • 1/2 Links Referencing other ideas in the Zettelkasten is how we form relations between ideas. Relations are how we can generate new ideas.

    • 1/2a Branches Sometimes I want to expand on something said in a slip. In these cases, I write (a), (b), (c), …, (z) in red ink, and then append (correspondingly) a, b, …, z to the ID number to form a “branch”. For example, 1/2a is a branch off 1/2.

      If I had more to say about branches, I would modify the ID number for branches to be 1/2a1 Branches. The next slip in this train of thought would have ID number 1/2a2, then 1/2a3, then 1/2a4, and so on.

    • 1/2b Red Ink for Links I realized it is more striking visually to use red ink for links — I just write the ID number in parentheses in red ink. (Initially, I tried using blank ink, and I later changed conventions.)

  • 1/3 Bibliography notes I take notes while reading books. These reading notes (or “Bibliography notes”) go in a separate shoe box. I form “permanent notes” which go in the Zettelkasten from the “Bibliography notes”.

    Importantly: Bibliography notes contain all the information necessary for me to cite the article (or book or website or preprint or…). I typically write this on one side of the reading notes (as well as the date I read it).

    My “bibliography shoe box” is organized by author-date-title (and if necessary, the date I read the material—sometimes I re-read the same article multiple times, in different contexts).

  • 1/4 Workflow How do I actually, you know, “do stuff”? This includes (a) populating the Zettelkasten with new slips, and (b) use these slips to…do stuff. (Those two items were actually branches, did you notice?)
    • 1/4a1 Growing a Zettelkasten Adding new notes to a Zettelkasten involves reading material, thinking about stuff, writing it down, assigning an ID number, then placing it in the shoe box.
    • 1/4a2 Lecture notes Thoughts about how to integrate lecture notes into the Zettelkasten. I ultimately decided to treat lecture notes as an unpublished manuscript which I can reference without requiring any corresponding “bibliography note”.
    • 1/4b1 Using a Zettelkasten I am certain there are multiple ways to use a Zettelkasten. For myself (as for Luhmann), I am using it to write articles and books and things.
  • 1/5 Organization Discussion about how to organize the slips in my Zettelkasten.
    • 1/5a ID Numbers The numbering scheme I am using (which came from a misunderstanding of Luhmann)
    • 1/5b1 Sections Clustering discussions into “sections” makes sense for my Zettelkasten. But I create sections as I need them as opposed to “prepopulating” my Zettelkasten with sections.

    • 1/5b2 Subsection I realize that Luhmann is using the really old-school Tractatus-style numbering scheme where 1.1/ is a section and 1.11 is the first subsection of 1.1 — we don’t do this anymore. Instead we would add a period 1.1.1. I’m going to follow this sensible modern convention.

      Furthermore, I will try to resist adding “too many” sub-…-subsections.

Those are the first 13 or so slips I wrote about in my Zettelkasten. As I run into new problems or situations when adding notes to my Zettelkasten, I think about what conventions I am following. Then I compare it to what I have written about in the first section in my Zettelkasten. Sometimes new situations arise, and I need to add new notes to explain what I am doing.

Other times I find that I am following a different convention than the one discussed in my Zettelkasten—which is perfectly OK! I just add slips into the first section explaining what happened. For example, I initially used black ink for everything (including links, branches, etc.). This turned out to be a really bad idea in practice. So I changed conventions. Later I added a branch off 1/2 Links to make note of this in the slip 1/2b Red Ink for Links.

I recommend following this convention: write about your Zettelkasten as the first section of your Zettelkasten. Sit down and seriously try to answer the questions: what am I trying to do with my Zettelkasten? Why do I need it?

This is a surprisingly nontrivial thing to answer. In general, “Why am I doing [action]?” is a hard question to answer except in the most trivial situations (“I am eating food to keep living”).

Why do you use a Zettelkasten?

tl;dr Writing, I use a Zettelkasten for writing.

I’m writing articles and manuscripts about Mathematics and proof assistants. I use my Zettelkasten to help me in that process.

Most emphatically, I do not use my Zettelkasten as a “personal knowledge management system” (or a “knowledge management system” as the term is used).

So what do I do with it? Well, I suppose one way to describe what I’m doing is performing some kind of “rational reconstruction” of various historical proof assistants, and examining how well these formalize different fields of Mathematics. What do I mean by “rational reconstruction”? I mean: I think about presenting a “simplified toy model” of the proof assistant (or foundations of Mathematics or…) to focus on how it works (while neglecting “kludges” needed to make it work on older hardware).

“Thankfully” we have AI now, which means my life’s passion is irrelevant and without a viable future.

But I have a lot of interests: pure Mathematics, heterodox economics, computer science, physics, philosophy of Mathematics, Linguistics, Proof Assistants. They all have a place in my Zettelkasten. They have feedback to each other, in surprising and productive ways.

I was hoping to write a book about proof assistants. But that doesn’t seem viable (no one cares about them, and no one cares about preserving knowledge). Instead a lot of the material I have in my Zettelkasten has found publication elsewhere.

How do I use this thing?

Method 1: The most direct approach is:

• Think about something you want to write about
• Look up slips about these topics in your Zettelkasten
• Pull out the slips from your Zettelkasten
• Use them as “raw material” for your manuscript.

Method 2: However, it is useful to “follow the links” on the slips you have withdrawn from your Zettelkasten. This is how you make surprising connections between unrelated subjects. (You can then use this to write new slips for your Zettelkasten.)

This gives you ideas to write articles (or blogposts or…) about.

Method 3: Doubtless there are other ways to use a Zettelkasten. You could take a page from Smalltalk, and try “living in” the Zettelkasten (c.f., Alan F. Blackwell’s Moral Codes: Designing Alternatives to AI, especially Chapter 7).

Can I use my Zettelkasten for something other than writing?

Maybe. I don’t know what that would be, though.

It’s not especially useful for studying (if you’re a student), or for learning (if you’re a life-long learner). Sometimes it can give the illusion of knowledge, which is detrimental.

A lot of people tend to think of a Zettelkasten as a way to organize “index cards used for rote memorization”. That is not how Luhmann used his Zettelkasten, it’s not how I use mine.

The central premise behind Luhmann’s Zettelkasten is “There is no writing without thinking.” It’s very specifically geared towards writing.

It’s great for combining ideas in new and surprising ways, which you communicate through writing articles, books, blog posts, etc.

Do I need to understand Luhmann’s writings to use a Zettelkasten?

No, you don’t need to understand Luhmann’s ideas or writings to start or use a Zettelkasten.

Luhmann’s writings are interesting. But so are “systems theory”, “cybernetics”, Principia Cybernetica, and so many other things. If you find them interesting, then read them. Enjoy them!

If you don’t, then you don’t have to read them.

Will these things help you with your Zettelkasten? Maybe?

Some people have speculated that Luhmann’s Zettelkasten system is a “system” (in the sense of “systems theory”), or that it exhibits autopoiesis. It’s worth remembering Luhmann began his first Zettelkasten before learning about systems, and he began his second Zettelkasten a couple decades before “autopoiesis” even existed.

So, strictly speaking: no, you don’t need to understand systems theory or autopoiesis before you can get a working Zettelkasten.

Is this going to help me in [field of study]?

Maybe?

Different fields have different demands, so I cannot say “This is a panacea for all academics.” (That’s why I’m a horrible sales-person.)

Case 1. In my experience, it is not good for Physics. Instead you should try creating an “Encyclopedia” or “Collection of 1-page articles” which summarizes a topic suitable for a graduate student. Avoid over-reliance on calculations because that can hinder you explaining the physics of the subject.

Case 2. For Mathematics, I have mixed results. I suspect this ultimately boils down to how you think about Mathematics.

Case 3. For computer programming, this is not the best approach. You can adopt a form a “literate programming” to have computer programs in your Zettelkasten. The major disadvantage is that most programming languages change quite rapidly, which will make most of your notes “stale”.

If you are willing to work in a “stable” programming language (like: Standard ML, Common Lisp, C99, Smalltalk—languages which are no longer under active development), then you can capture some things usefully in a Zettelkasten. I suspect you will get more mileage out of building your own tool for literate programming and note-taking; the bootstrapping process is a good “test case” for your methodology and tool.

Case 4. For history, biographies, and long-form journalism, the Zettelkasten system outlined could be adapted to your situation. Historians in the late 1800s through the middle of the 1900s (and maybe later) were taught to use index cards to organize information.

For example, David E. Johnson’s Douglas Southall Freeman discusses Freeman’s workflow using index cards as newspaper editor (p.110) and as Historian (pp.329–331).

Edward Gibbon used blank playing cards (the predecessor to the index card) when organizing his sources when writing The Decline and Fall of the Roman Empire.

I have written a long-form profile of a public figure using a modified version of the Zettelkasten to track all the contradictory things this public figure said and did. It was useful for keeping track of a very contradictory figure. (Unfortunately, libel lawyers said there was no way to publish this article because even the most impartial framing of things made the public figure look dangerously unhinged, so I cannot really show it to you.)

Help! My Zettelkasten is too big, I can’t remember where anything is anymore

Yeah, that happens once you get enough stuff in a Zettelkasten.

Let us consider several ideas and approaches. Some might address your particular problem exactly. Others might not. The hope is to inspire you to explore different techniques.

There are several ways to address the issue:

• Organizing slips as you integrate them is one way to delay this problem. Use sections, subsections, and maybe sub-subsections.

• Use a keyword index to keep track of ‘jumping off points’. If you have added links to your slips as I urged you to, then links will help transport you around the Zettelkasten…and so it’s just a matter of “Where do I find material about [topic]?” Writing down an “index” (like an index in a book) and up to 4 ID numbers for each entry helps answer this question. It is not exhaustive.

  What’s a ‘keyword index’ look like? Well, it’s like a book index, except on index cards. Each entry is written on its own line and has up to four ID numbers for you to “jump off into”. The entries can be arranged alphabetically in your index card to help you find what you’re looking for.

  Where do you store a ‘keyword index’? Luhmann kept it separate from his Zettelkasten (literally in a separate “shoe box”).

• Bob Doto uses hub notes. These are structured like the ‘keyword index’ (i.e., it’s a list of entries where each entry has a very finite list of links), except (1) they focus on the location of a single concept as it appears in wildly different locations in the Zettelkasten (for example a hub note about “Time” with entries like: in economics, in physics, in logic, etc.), and (2) the hub notes are placed in the Zettelkasten.

  Why is this useful? Well, if the same concept appears in very different regions of the Zettelkasten, it is useful to have a “hub note” for the concept which points to where it appears in the Zettelkasten.

• Sascha Fast uses something he calls structure notes which resembles the table of contents for a book with links to where the entries appear in the Zettelkasten. (Luhmann did the same thing.) Bob Doto extends this to make it more of what I would call an “annotated table of contents”, adding some explanatory connective text between entries in the structure note.

  I personally have not found this terribly useful in either form.

It’s also good to bear in mind: what are you using this Zettelkasten for? Are your notes you add contributing towards that goal?

Some people try to stuff as much random material into their Zettelkasten hoping it will produce insights and manuscripts…somehow. My experience has been that this is the wrong way to build a Zettelkasten.

Worse: it feels like you are doing something productive, when in reality you are just polluting your Zettelkasten. If you find yourself doing that, then it might be wise to start over. Luhmann had two Zettelkastens, and we do not know how many “false starts” he had with his first one.