At what point does adapting a Zettelkasten stop being a Zettelkasten?
Hello Folks,
I’m still quite new here, so I’m genuinely curious about this. What Are We Actually Trying to Preserve in a Zettelkasten?
When we try to build a Zettelkasten today, are we trying to preserve Luhmann’s method, or are we trying to preserve what made his system so effective?
Those might not be exactly the same thing.
Luhmann had a physical box, sequential numbering, Folgezettel, links, and a very particular way of writing and developing notes.
The more I read about Zettelkasten, the more I notice that people seem to disagree about some of the details: Folgezettel, numbering, structure, digital tools, tags, links, etc.
And I find that disagreement interesting.
I’m wondering where people here draw the line.
If I keep the idea of writing small, focused notes, connecting them, developing ideas through those connections, and using the system to think rather than simply store information — but I change the structure to fit the way I work digitally...
Am I still doing Zettelkasten?
Or are some of the old constraints actually part of what makes the method work?
For example, is Folgezettel something we should try to preserve because it provides a particular cognitive benefit?
Or was it simply a very clever solution to the limitations of a paper-based system?
I don't have a strong opinion yet. I'm actually hoping the people who have been doing this for years can change my mind.
What is one part of Luhmann's system that you think is genuinely essential — and one part you think we can safely leave behind?
Howdy, Stranger!
Categories
- 3K All Categories
- 153 Research & Reading
- 699 The Zettelkasten Method
- 11 Knowledge Work
- 101 Writing
- 470 Software & Gadgets
- 155 Workflows
- 732 The Archive
- 15 Plug-In Showcase
- 88 Resolved Issues
- 225 Projects Logs and Journals
- 85 Project: Zettelkasten.de
- 53 Critique my Zettel
- 175 Random
- 375 Introduce Yourselves!
Comments
For me (and I stress for me, since there is no shared definition of Zettelkasten and there cannot be one: no authority has ever produced such a definition...), a practice stops being a Zettelkasten when it begins to lose the three essential characteristics that, over time, I have come to believe a Zettelkasten should have.
I don't remember whether I came up with them myself, whether Sascha formulated them somewhere, or whether it's somewhere in between, but I have made these three points my own, and I consider them to be the essence of Zettelkasten:
So, essentially: if the connections and relations start to disappear, if the system stops enabling me to develop thoughts and knowledge, if over time it ceases to become the tool through which I continuously confront my accumulated reflections, then I no longer have a Zettelkasten.
There is, however, one very important caveat: if I stop having a Zettelkasten and the system continues to meet my needs, that's perfectly fine.
It may actually be better to have a system that can no longer be called a Zettelkasten but that I continue to use and that still works for me, rather than forcing it to function in a way that doesn't suit me, introducing friction that might eventually lead me to stop using it because I no longer feel at home with it.
How Luhmann did it remains, for me, just one of many possible ways of doing a Zettelkasten. At the same time, however, I think it is equally true that, for a system to be called a Zettelkasten, it must somehow reimplement some of the dynamics that Luhmann's system had in relation to the development of thought and knowledge.
The difference is in implementation details, rather than in the underlying principles.
In the practice, however, I don't think a system can "suddenly" cease to be a Zettelkasten.
It is probably more useful to think of these characteristics as existing along a continuum.
A system can be "more or less Zettelkasten-like": it can gradually lose some of the characteristics that make it a Zettelkasten, becoming something that is increasingly distant from the original idea, rather than crossing a clear line at which it suddenly stops being one.
In that sense, there may be systems that are "almost" Zettelkastens, or systems that remain Zettelkastens while gradually losing some of the dynamics that made them so in the first place.
For example, a progressive use of AI that increasingly replaces the author's own exercise of thinking with the automated generation of thought would progressively "de-Zettelkasten-ize" the system.
Likewise, if a system gradually accumulates so much complexity that the friction involved in adding new reflections and making new connections becomes too great, its dynamics can gradually "freeze". The network may still exist, but it is no longer actively developing. In other words, the Zettelkasten gradually shifts from being a living system to becoming an inert one.
@andang76 put it very well. I only have an anecdote.
I have practiced taking notes and reflecting on them using a collection of plugins for the Vim text editor, the main one being vim-zettel for about six years. I think it's Zettelkasten-like, but the only book I read on the topic was How To Take Smart Notes, so I don't know. Still, it works for me and I've published many posts thanks to it. I call it Zettelkasten in my mind because it reminds me of Luhmann and his remarkable corpus.
That's all to say, discussions on definitions are interesting, and it's good to have a well defined set of rules to follow when implementing a system, but at the end of the day the only thing that matters is whether it works for you.
I think the essential part of his personal technique was the "Folgezettel". (That's not Luhmann's term, but a term frequently used in this forum.)
Somewhat ironically, it's the Folgezettel that many people leave behind, myself included. ;-) (However, I do have a small section in my zettelkasten that implements Folgezettel structure with links in YAML frontmatter instead of Luhmann numbers. My notes live in Obsidian.)
For me the key feature of a "Zettelkasten" is collecting. That's the part you can't leave behind.
My favorite Zettelkasten tutorial was published 250 years ago by Johann Jacob Moser.1 It uses the word collectanea for the things collected in the wooden boxes.
In my opinion, many of the features that we love to debate are secondary to curating a collection of notes. Such a collection might be unordered (as in Eminem's box of song notes) or it can have some sophisticated system. The collection might serve different purposes.
Luhmann's ordering system was, well, unique. One reason, why so many people leave it behind, is that they are not Luhmann. :-) Their brains work differently. They have other interests. Fortunately plenty other systems have been tried out and documented. I'd guess that my digital system has more in common with Moser's century old technique than with Luhmann's.
Note: My argument is based on how the term "Zettelkasten" has been used in the German language for a long time.
https://archive.org/details/10554451bsb/ ;↩︎
This is a better version of the discussion I opened a few days ago, thanks for it.
They aren't. I don't think those are quite the best statements of the alternatives, though. His method was the way he worked with his system (in the way I use the word); the system was the collection of components he devised: notes, a container for them, a way to link them, a way to find them, and the way that cards that were near each other physically were likely to be near each other conceptually as well. One might include his index cards that contain entry points, structure cards, and so on but I think of them as auxiliary practices he needed because of the constraints imposed by his system being paper-based.
I don't think there is one thing. If you focus only on the linking you end up with a wiki-like system. That misses an important aspect, which is that notes that were physically close were usually conceptually close. So once you got an entry point, you were likely to find an apt note nearby. Or you could notice that you didn't have an apt note and decide to fill in the gap. Or the difference between a note and what you thought the area was about triggered new thoughts and led to a cascade of new notes.
So I consider that his system supported both the wiki-like aspect and the relatedness-by-propinquity aspect. I think both are important. If you only have the wiki part you are missing an seriously important component.
Exactly how he personally made use of these capabilities was his method.
I'm trying to discover and use what aspects of Luhmann's system cause it to be effective. I have no real interest in hanging on to the label of "Zettelkasten" for its own sake.
Try to complete my previous post.
For me, Folgezettel and other things like that is not "essential" to a Zettelkasten.
First of all, if I completely do without it, I don't break any of the properties I described in my previous post.
Second, I can choose not to use Folgezettel as an implementation detail, but if I can understand why it was introduced, the underlying principles behind its use, and the benefits it provides, I can look for another construct that implements the same dynamic and use that instead to achieve the same properties, if those properties are important to me. Which is exactly what I did :-)
The same reasoning I applied to Folgezettel can be applied to many of the implementation details used by Luhmann or other authors.
Folgezettel, putting an ID in note titles, writing notes atomically, adopting a bottom-up or top-down approach, "planning" the construction of the network or letting it emerge spontaneously over time, using paper or Obsidian, using an index or not, using folders, using only the "classic" note types or abandoning them and/or introducing other types, putting only "ideas" into the network or including other kinds of content, using categorization or not, and so on.
Every single choice brings certain properties with it, but none of these choices is a real requirement for achieving a network, developing thought, and being able to confront your past own thinking.
There have been other systems of connecting thinking, and Luhmann had a particular way of implementing it for himself. If history was reflected on differently, we might all be talking about the Ashby system and still have similar discussions.
From what I've read about how various scientists took their notes, they all had ways of connecting ideas, but they personalized it to themselves. That's really what the Luhmann system is, a personalization.
I would say that we are actually following these scientists examples by creating our own personal system, instead of trying to copy them exactly. It's not as important what name we give it.
Yes!
Yes!
Counter-argument: proximity in a digital hypertext systems could be determined by the number of links you have to follow in order to get someplace. It's up to the author to add links, so that conceptually close notes are just one click away. (And we have tools like tags and search that make proximity less relevant.)
The problem for us today is that his method is poorly documented. Even if we consume all available sources, we don't really know, what went on in his head. We can study the artefact (the actual zettelkasten) in great detail and still miss out on key elements of his method.
For example, what part of his work happened only in the manuscripts of the many articles and books? How exactly did he consult zettels while working on a manuskript? And most important of all: what happened only in his head?
This might be the wrong question.
Maybe we are missing what made his method so effective?
I think part of his success is that he designed his own method, a method that worked really well for him. He adapted elements from other methods that he found useful. He ignored stuff that didn't work for him. He added own inventions. He knew how to read. He used a particular writing style that trained his mind.1
So maybe the actual lesson is to do your own thing? :-)
Luhmann: Lesen lernen. 2000 https://web.archive.org/web/20241125104449/https://luhmann.surge.sh/learning-how-to-read ↩︎
No, that's not at all what I mean. I mean that there is a way to scan "nearby" cards to get an idea of their conceptual arena and/or their flow of argument. Luhmann could get this by riffling through a stack of cards. My system gets this by having an outline one can scan, and and by the particular way search results are presented.
It's hard to appreciate the value of this ability until you have experienced it. The value comes from being able to grasp the conceptual neighborhood immediately by seeing it, not by some abstract number or by counting the jumps from one individual node to another.
What is in play here is the matter of scale I discussed in another comment yesterday. A focus on details obscures grasping the bigger pattern. In this case, individual cards, and to some extent even individual links, are the fine-grained scale, but you need to back out to see the pattern. This is what scanning does. Structure cards are helpful but no replacement.
Luhmann had local structure, not global structure. Local sections of a card box were isomorphic to parts of an outline, and they generally (with exceptions) were related conceptually.
That is why I tried to understand the deeper underpinnings of his system: cognitive, library, and psychological science among them.
After all, every football player kicks and heads the ball differently but they are all playing "football". I'm playing "Zettelkasten".
The advantage of a feature-rich app like Obsidian is that I have a many options available to filter and display linked notes. For example: VirtFolder https://forum.zettelkasten.de/discussion/3478/obsidian-plug-in
The advantage of capturing lots of structured data in my notes is that I can create a wide range of meaningful queries.
Of course the advantages come at a cost. Obsidian has a steep learning curve and customization takes time.
Wooow, I have been visited the website of Ashby and actually I have never heard about it before.. What is this and why it can be similar? I'm curious about it.
I mean, is this only a way to search indexed pages or is something more structurated?
I love that phrase!
"I'm playing "Zettelkasten".Because actually that is what I wanted to do all this time!That's why I cannot be using Obsidian system. I know that app can help me to build myself some structure about ZK, but actually, there is a huge percentage that I finish creating some personalized system in there, rather than follow real ZK real structure. As I am not a ZK expert, the error gap is huge for me.
That's why I prefer to get a ZK app that really exists to make easy my taking-note day-life. So I will be only preoccupied to enter my notes and the system will be organizing everything by itself, and my only preoccupation will be to study my notes, study my connections and keep entering more notes in time.
I think the
"heart of ZK"is to interior your notes and develop some reasoning about them, "talk to them" like Luhmann used to say and keep developing my own ideas based on my researches and stuff.If I am able to build or find something like this, I will be happy to share!
>
I think this is one key about a group divided in 2 or 3 branches. Because, If you are using some method that mostly cannot be called ZK anymore, but works fine for you, so in that case I would say that is a "personalized system", maybe based on ZK, but is no longer ZK. In that case, we were not being talking about ZK anymore, but a derivated system that can be perfectly better or not. But, there are some people that prefers to maintain the essentials of ZK.
That's why I said that I cannot use Obsidian as my based ZK app, because that requires me to build my own system and there is a huge gap that I can finish building my own way to store my notes, and I am a member of the group part, let's call it #1 haha. I prefer to maintain ZK essentials at all cost.
I will copy-paste my last reponse to tomp here:
That's why I prefer to get a ZK app that really exists to make easy my taking-note day-life. So I will be only preoccupied to enter my notes and the system will be organizing everything by itself, and my only preoccupation will be to study my notes, study my connections and keep entering more notes in time.
I think the "heart of ZK" is to interior your notes and develop some reasoning about them, "talk to them" like Luhmann used to say and keep developing my own ideas based on my researches and stuff.
If I am able to build or find something like this, I will be happy to share!
Luhmann left behind a small paper that is fundamental for me: "Kommunikation mit Zettelkästen."
My experience with the Zettelkasten falls into two distinct phases: the time before I studied that paper and the time after.
In my view, if one wants to discuss Luhmann’s Zettelkasten and how it works, it is impossible to ignore the contents of that paper.
I rarely see the concepts from that paper cited in discussions about what the Zettelkasten (Luhmann’s version) actually is, yet they are far more important than other details.
Regarding the Folgezettel debate.
Searching can't replace the browsing dynamic enabled by Folgezettel. They are two completely different navigation dynamics.
One of the most relevant functions of the Folgezettel is allowing you to implement browsing through ideas that belong to sequences (according to a specific semantics attributed to the physical proximity).
That being said, in a digital system, making of chains/branches of thoughts an their browsing is easily implementable without resorting to the Folgezettel: you can build notes where you arrange links to other notes (Structure Notes) that you can use like browsable maps.
This can be seen as a classic example of aspects that belong to different levels of abstraction.
Folgezettel and Structure notes are different constructs, that is, different implementations, which can be used to achieve certain behaviors within a Zettelkasten, such as creating chains of thought, branching, and enabling browsing.
These behaviors, in turn, have value because they produce certain benefits, such as facilitating orientation within the network, the exploration of "close ideas" and their progressive development through juxtaposition, and the creation of compositions (the sequences) that express meanings greater than those contained in a single note.
We can therefore distinguish at least three levels:
constructs → behaviors → benefits
Confusing these levels can easily lead us to mistake a particular implementation for the behavior itself, or even for the benefit we are trying to achieve. But Folgezettel and Structure notes are not the end goal: they are two different ways of implementing certain behaviors, which in turn serve to achieve something more important.
FYI, these are the texts and interviews I had in mind:
I've seen it mentioned frequently in the forum, eg when talking about ZK as communication partner.
Yes indeed, d'accord. Here is my distillation from several translations of that article, and from some of his zettels:
Research partner
Secondary memory
Alter ego
Constantly communicate
Creativity booster
Dialogs with
I think he wrote little about details and workflow because he wanted to communicate at the level of that list.
Actually, it can, when you implement the structure with attributes inside the files. You need to search the files in order to see the structure. If the search is automated and visualized as an outline, it looks like a static outline. (See the VirtFolder example mentioned earlier.)
If you store the information inside the files, you can even encode multiple sequences. For example in YAML frontmatter you could encode:
The trick is that you only have to store the link. Overviews are being computed as outlines, graphs or whatever visualization the app supports.
Structure notes, aka MOCs, are a rather crude and analog technique. :-) They make it more difficult to write queries. The upside is that you have an overview note that has room for comments.
Actually, you will do better to be loose and flexible about these "sequences". They might be closer to sets than sequences, and their semantics may change from place to place and time to time. It's best to take the attitude that
Of course, you may devise conventions that help you. In my experience, you will always come across some situation that your convention doesn't work for and then you have to come up with a way to handle it. This is inevitable because any way to display the structure of the collection can only be a lower dimensional projection of the complex whole.
Folgezettel is a way to show or collect a related train of notes (or ideas) so that they can be grasped as belonging to a larger unit. It doesn't have to be implemented with a numbering scheme (depending on your system, of course). The goal is valuable, in my experience; there can be more than one way to get there. Whatever the method, it should be very simple to create and view. Preferably it should be highly intuitive. Otherwise it will cause unwanted cognitive effort and only get in the way. Changes will also be harder.
Yes, it’s another way to implement note sequences, one I use myself, without the automatic visualization, and it has its own strengths and weaknesses.
One of the key advantages of the Folgezettel or structure notes is that you can get an overview, a grasp, of multiple notes in a series at the same time, not just the immediate neighbors.
If you picture how a paper card file works when you flip through it with your fingers—moving quickly back and forth or peeking at the cards behind by spreading them apart—the concept of "scanning ideas" becomes clearer.
Luhmann himself wrote in the paper I mentioned—if I recall correctly—that it was important for him to write only on the front of the notes so he could scan them in sequence without even removing them from the slip box. He placed great importance on that physical dynamic.
Crude and analog like writing :-)
But that writing, that effort is thinking itself.
Placing links "in the right point" in the canvas a MOC represent is a relevant thought process, in my experience. And a source of fourther knowledge developing. Contrasts, holes to be filled, patterns, abstractions emerge.
Same dynamics that emerge with folgezettel, I presume.
Having never used the Folgezettel method in the practice, I don't know how "edge cases" or the need to "break convention in a specific instance" are handled in practice.
One of the advantages of structure notes is that, since they are logical rather than physical constructs, you can create as many as you like to map the same set of notes, according to different way of organizing links; furthermore, because they are local rather than global constructs, you are free to design each individual structure note however you see fit.
Yes. "Placing" links on/in a metaphorical canvas works for me too. In my toolkit I consider this a separate tool.
On some level, yes. The difference is that Folgezettel is an incremental process. Every new zettel needs to be attached to an existing zettel. Once a number is assigned, the connection is fixed. You can't rearrange them later. Some clusters and patterns will emerge, but maybe not as intended.
With MOCs or structure notes you are more flexible. You can update them or create multiple versions. And you can arrange and rearrange them.
You may have seen my posts where I experiment with different visual layouts for zettels. The layouts visualize the same folgezettel structure, yet they invite a different kind of reading.
Ah, that explains a lot. :-)
I recommend you try it out with 100 paper index cards. Write real zettels about thoughts/ideas that you care about. Follow the process described in the article. (As inspiration I recommend @emiwid's recently posted images of an analog zettelkasten.)
I tried it out myself and did not like it. But at least I know how this moment felt like, when I had written a few cards and needed to decide where to attach them, so that I could file them in the box. And it was also interesting to experience, how it changed my writing, when I tried to write notes so that they would easily attach to an existing note.
I didn't make it to 100 cards… :-)
I never clicked with the idea of using Folgezettel, even without even trying it: I just sensed too many drawbacks for my personal attitudes.
In particular, I don't want things and decisions to be set in stone, built like concrete that lasts forever once placed. I don't like immutability. I constantly rewrite and move things around; they are dynamic and continuously evolve over time.
However, I studied it quite a bit and tried to understand its benefits. I acknowledge its objective validity, and I'd say I realized pretty early on that I could achieve some of its advantages in other ways.
There are people who use it, and with enough practice, if it fits your way of working, it probably works. Generally speaking, though, it's usually the first thing I see people drop from their Zettelkasten implementations.
I really recommend to try it out. The physical experience matters!
At least try to make the examples in Kommunikation with Zettelkästen tangible:
I consider it just a different approach to linking notes. Unlike Luhmann, he didn't use only index-sized cards, he used both journals and index cards. By assigning unique ID's (in his case hand-written numbers) to each page extended across 25 notebooks, he transformed them into a single large super-notebook. The inserted ID's could then link any idea on any page to a related idea in another page in the same journal, or a different journal, but all within this multi-journal super-notebook system. So it was a network of links connected different ideas across time and space.
I think it's another approach to creating a large network of linked ideas, it's just a different implementation from how Luhmann did it with his Zettelkasten. I don't think either of these are better or worse, they're just different, and we're all creating our own system in one way or the other.
Wooow! I deffinately haven't heard about it before!
I think each style is having its own advantage, maybe they have been build to solve different things, different problems they wanted to solve. However, I has found in ZK some perfect way to store my notes, to study them, to find them, and develop them... However, Ashby sounds interesting! I will review it more often too. Thanks for the explanation!