The Complete Guide to Atomic Note-Taking
Atomicity refers to the idea that knowledge is made up of discrete building blocks. The Principle of Atomicity is a processing direction in note-taking, aiming for one knowledge building block per note. It is not a rigid law, but a guiding compass. It needs to be contextualised for each application.
Post edited by ctietze on
Author at Zettelkasten.de • https://christiantietze.de/
Howdy, Stranger!
Comments
Some thoughts:
I feel you are missing a third option where they are not saying you have to rely on heuristics and are willing to engage (not stating this is the case here, just an observation). They hear you out and just believe you are over-complicating a concept because you don't have a strong grasp of what is important in relation to the concept and are adding details /or sub-concepts that are unimportant.
I am not following what the problem of level 3 is in the context of this article. I looked back and was not able to work out what it is. I also don't think I have a strong grasp of what are the details that make up level 3 overall. Could have used 1-2 sentences stating what the procedures of the method are that differentiate from keeping a zettelkasten without the zettelkasten method.
I love level 4 and think about it a lot. I secretly wish I had my own coding skills (don't have the will to try vibe coding) to experiment with creating Obsidian plugins that can help with pushing one towards better thinking.
I really like Super Memo Guru (Piotr Wozniak) definitions for Concept and Model. I think of a model as a set of concepts connected by rules. I also don't quite agree with your knowledge building blocks, as they don't seem discrete enough from each other. As to me models are made up of concepts. Hypotheses are made up of arguments. I have no clue what I would say the building blocks of knowledge are instead.
I was completely lost on this aspect. I understand the zettelkasten iceberg level 2 and 3 but am not following along with what actually is atomicity level 2 and 3.
This is big to me and something I wonder a lot about. It makes me think this Andy Matuschak Patreon article. It unfortunately is locked behind a Patreon wall and I didn't make a copy of it, so I don't have exact quotes of what I'm thinking about. But he discusses how he abandoned his evergreen notes practice as it was accumulating too much technical debt, much better to just create a flat outline of sorts (Cal Newport style) per project. I thought that would be an interesting topic to tackle on the blog and see how you've handled that concern of technical debt over the years in your zettelkasten.
Impulse of creating/notes without directly moving to a later stage of maturity seems dangerous to me. Getting into Collector's Fallacy territory. This reminds me of how the brain generates an overabundance of neurons and synapses then prunes them to become more efficient. I think the worry is if you spend to much time in impulses than you never get around to pruning and aren’t particularly efficient. I’d say a certain level of efficiency is necessary due to our limited time/energy on this planet. Why do I say all this? Because impulsing without pruning or maturation feels good (collectors fallacy).
I got confused because you put your comment before the quote. I am pretty sure that it is the convention to put the quote first and then your comment. The dividers didn't show in the post, which would've helped.
It is based on a specific pattern-language of meaning. The eye is the all-seeing eye that emerges as a phenomonon in schizophrenia as a result of an overactive left hemisphere and a dysfunctional right hemisphere. A feeling of being constantly watched. (Based on Iain McGilchrist)
The puppy is the self, still unformed and full of potential. It can be both the user and the idea aiming up.
It is AI-generated.
This is based on my observation. So, I miss the third option, since I wished people would engage with the level 3 concepts.
Level 3 lacks the justification. The problem of level 3 is that it solves level 2 problems by giving a formal framework on how to make sure that you understand what an atom is (scuba diver directly looking). With this framework, you can actually make sense of heuristics more and use them more effectively. But level 3 still lacks a deeper justification of why this actually should work. Level 4 provides the justification by stating that atomicity is a trait of knowledge itself and the principle of atomicity is a principle of treating knowledge accordingly.
"Level 4 is reached when you see that level 3 is actually about sound thinking and not some methodological gimmick."
I think it is a good idea to learn to code for one's thinking ability. I am not sure how to learn to code to gain general benefits. I think it is the same with strength training for athletics. If you are really weak, any strength training will benefit you. But none of the people who are reading this are the equivalent of weak. So, a more sophisticated approach is needed.
You are mixing two frameworks here. If you go with the definitions of Wozniak, you can't make any consistency claim about the inventory, since the consistency is dependent on the condition that you go with the definitions. It is like exchanging sugar for honey, wheat flour with rice flour and then claim that the recipe is not working.
But I won't defend my inventory here. You can use any inventory of building blocks that you like. Well, the inventory should be working. Meaning: It should not just be a bunch of definitions but a bundle of ideas, methods, thinking tools etc.
To create an inventory, you need to have a reasonable framework of completeness of what you are mapping. If we are talking about knowledge building block, the inventory should map the process of knowledge work.
Atomicity level 2 means that you applied heuristics, level 3 means that you identified the knowledge building block and made sure that the note just contains this one building block.
I had a short exchange with Andy over this very topic. There were quite some issues with his practice, which explain the problems he had, at least in part.
It is indeed an interesting topic which I wouldn't label as technical debt but the problem maintenance. But there is no conscious solution on my part. The reason why I have almost zero technical debt is that my Zettelkasten practice is pretty much just a thinking practice. So, I don't have to clean anything up, just to keep everything tidy and working on the level of the method. If something is messy, it is messy for a good reason: My level of thinking is still messed up. I don't engage with messy areas of my Zettelkasten, unless I want to progress my understanding. If I need to connect a new idea to something that is messy in my Zettelkasten, I will clean that mess up as an act of furthering my understanding.
One aspect, for example, why Andy accumulates technical debt is that his Evergreen Note are public. This creates the need for them to be understood by others. There is a lot of stuff in my Zettelkasten that nobody than me can understand, that violates good practices etc. But it is my personal Zettelkasten and I don't need to care. When I see that something is inconsistent, incoherent or whatever, I can just accept it or write a comment with a quick idea on how to fix this and then move on. Impossible if the notes are available to others.
But I don't want to dissect Andy's notes in a thread in a forum because Andy's work warrants a more careful approach than a spontaneously generated answer.
Your stages seem to be on the level of workflow. I oriented myself by the maturation stages in practical knowledge work.
It is a straightforward condition: If you want to make the Zettelkasten a home for your mind and thinking, you should move the complete thinking process into the Zettelkasten.
You are right that there is a problem of the Collector's Fallacy. However, the root cause of the problem is not to be found in the method, but in one's approach to knowledge. A person that never disciplined his mind to follow through after the first impulse, will not do it with any system and external pressure (e.g. a deadline for an essay) will be the mechanism of making this happen.
If you try to deligate the mind's disciplination to external constraints, you risk of detraining your mind or never developing its discipline in the first place.
But you can watch the video to see that it works if you practice it.
I am a Zettler
I encourage whoever is reading this to check out S.S. Seward’s Note-taking (available free on Google Books). He anticipates the concept of atomicity when he instructs the reader to capture the idea of a curious portion of a text or lecture. That’s all the principle of atomicity is about. Capturing ideas. Not a summary, or an enumeration of facts and the secondary details pertinent to an idea but the special discrete points that the author is trying to communicate throughout the text.
Knowledge building blocks give you a mental bank of the type of ideas you that you find in a text. The stages of an atomic note are just ways to track the development of an idea—from a speculative bit of information into a concrete intimation that affects how you think and behave after its registered in your mind as correct.
Only taking notes on things that you intend to advance according to an implicit timeline sounds dangerous too. Like Sascha, I think this disrupts an inherently internal phenomenon by introducing external stakes. This is something that inhibited my own processes in the past. I reckon as of late that efficient impulses are a matter of taste—being able to distinguish what bits of information are relevant to your broader research and anticipating their usefulness. This is where having a “toolbox of building blocks” comes in handy. Luc B. Beaudoin’s CUP’A method is a terrific heuristic for processing information. I use Notenik to store my knowledge base and it gives you different ways to view your collection of note files in the UI. With the
Statusfield I can store the stages of my notes as numerical values and sort them. This is just an example of ways you can track the progress of notes made on an impulse. Although I must confess, I can’t think of any note of my own that wasn’t conceived like this.I think this is one of the strongest distinctions to make when talking about knowledge work. Productivity is one thing with its own set of practices and interests. “Practical knowledge work” (researching, note-taking, composing, publishing) is another. Atomic note-taking is an epistemic framework or a toolkit for working with knowledge. Something than your workflow can adapt to. I find that Nick’s reservations with impulsive notes are addressed here and in the broader section from where this passage originates:
The Zettelkasten Method as a whole seems impartial to any one productivity method. Bring your own. Sascha’s integration of BASB into his own system gave me some insight into this.
What we’re looking to achieve with our notes is made straightforward in this piece:
When it comes to cultivating knowledge I don’t think empirical data is always an appropriate metric for efficiency.
Resources like “SuperMemo Guru” appear inclined toward productivity. Beaudoin has a term that I’m fond of, “Cognitive productivity”. These kind of methods can help aid your knowledge work but I don’t think that success in this area alone can constitute for excellence when it comes to understanding things—the comprehensiveness of the methods not withstanding. The cultivation of knowledge is the objective of knowledge work and productivity workflows facilitate this effort. However I think that workflows are often confused as the either the means—or worst—the end.
Devising a productive system of learning does not bring one any closer to the discovery and recognition of Truth. But it can help keep you consistent on the journey. This is similar to the path of spritual reformation in Islam. For example, a seeker may be prescribed or adopt practices in addition to the obligatory and customary acts of worship. But these auxiliary practices in and of themselves do not bring them closer to their ultimate objective (viz. salvation in the afterlife).
Nice recommendation! (It is also available via the internet archive: https://archive.org/details/cu31924012997627/page/n3/mode/2up if you want to download it in various formats. It is public domain)
Yes! I made a connection with what you wrote further down: Knowledge Building Blocks give you an external reference point if you got the idea right. (Just an addition to what your wrote)
To add to that: External stake to the internal process could be translated to: You put public pressure on your private sphere. It is absolutely not correct to make this translation, but it echoes the problem of specific restrictions.
This is speculative: If you only pursue for what you already have a plan, you might become less creative because creativity is dependent on a more open-ended, playful attitude.
Yes, exactly!
You will be quoted. This is exactly it!
I am a Zettler
I appreciate the comments.
@Sascha
>
This captures the tension perfect and may additionally result in one pursuing less ambitious work/research. Of course this implies a certain level of control over your work/output that may escape a lot of people. I found Cal Newport’s Slow Productivity a useful read in general but I felt like it makes this assumption. Indeed it may be a privilege overall, to be so good you can afford to work slow and deep.
Ah! Still I dream.
I did it — I have reviewed and processed the essay, the video, and the discussion.
A lot of work, and there would be a lot to say.
The short answer is that this is a very meaningful and relevant work.
In my view, it is essentially the transposition of bottom-up object-oriented development — taken from software development and brought into idea development. The process is especially evident in the video.
It seems to me that this work marks a major evolution in the theme of atomicity.
It addresses some of the limits of the simple model of “one idea per note.” It presents tools for achieving atomicity in a rational way, as well as techniques to avoid fragmentation.
The video, in particular, is a powerful example — not only of how atomization can be done, but of how general Zettelkasten work can be done. A pure organic development, very similar to how, in software development, a cluster of classes, objects, modules emerges from a coding task. That use of refactoring is exactly how I see Zettelkasten (and in fact, thanks to my background as a developer, that’s how the model of atomicity came to me “naturally”).
One observation I can make — a possible completion of the work — would be to better explain why pursuing atomization is worthwhile. That would make the value of the effort clearer. Understanding ideas better, clarifying them, and being able to make connections are only three of many benefits.
I’ll just mention a few (non-exhaustive list):
I see a potential critical point in the block-based model, which should probably be discussed more deeply.
Perhaps there’s too much implicit focus on the structural properties of the blocks for retrieving the architecture of ideas, and less on their ability to express creativity, taste, subjectivity, and personal perspectives. Of course, the article does not exclude this possibility in any way, but it might be worth emphasizing it explicitly. Not only can I “capture the truth,” but I can also “interpret” or have my own perspective.
When we talk about blocks, patterns, and structures, it’s very easy to be distracted by their formal and representative properties — while subjectivity, which distinguishes Zettelkasten from other models, risks being overlooked. For example, if I have a “model block” or an “argument block,” it’s easy to end up expressing only the descriptive part of the model, capturing the idea "as it is" from the source.
In the last minutes of the video, you briefly mention your personal reason for using that dynamic about IQ — it’s a short but very important moment to me. Presented so briefly, it might not be perceived as fundamental.
In general, despite the soundness of the atom-based model, or its block-based evolution as you’ve shown, over time I find it easier to introduce the concept of atomicity using the photography metaphor rather than the atomic one.
In the future, when asked to explain it, I think I’ll rely more on photographic metaphors.
The great advantage of photography is that everyone understands it — you can bring up relatable examples of photos that “work” or don’t work:
At the same time, however, the photographic model works well for showing what works and what doesn’t, but perhaps provides fewer tools for explaining how to do it well.
That might just be my personal limitation, since I’m much more experienced as a developer and system modeler — I master those two mindsets better than that of a photographer. I can recognize a good photo but wouldn’t be able to teach how to take one, while a good photographer could develop a sound photography based mindset for atomization. Nevertheless, simple photographic metaphors have, in the past, illuminated for me very clearly how to pursue atomicity (landscape, portrait, and so on).
I’ve made this side discussion about photography mindset not to diminish the block-based mindset you presented, but to highlight what I believe is a valuable opportunity.
I’m quite sure there are Zettelkasten practitioners who don’t fully feel comfortable thinking in terms of “blocks.”
That’s fine: these people can use the same method as a “template framework,” inserting their preferred mindset (for example, the photographic metaphor) in place of the block vocabulary, while keeping the same skeleton that provides progressive modeling on four levels, organicity, refactoring and so on.
So, if you don’t like blocks, try replacing them — keeping everything else the same — with a vocabulary of models that you resonate with more.
Used as a framework with interchangeable mindsets, this model has a much broader scope than originally intended
I’ve personally found as many as five different mindsets. One of these is a "making cookies" mindset
In a nutshell, learning to master atomicity is like learning to master photography or OOP development.
Same thing, same dynamics, same learning path, same "thinking in objects" once the highest level reached.
And if someone is lucky enough to know how to do either of the latter two, I think they’ve already got a clear path ahead.
Many thanks for your kind words.
Even as a non-programmer (I only learned Pascal and a bit of Delphi), I envision the process very similar to coding. The similarity stems from a specific shared pattern of information mapping that is beneath both processes. I can't summarise it enough to make it part of a post here, but to give you a preview: It has to do with the strange fact that mathematics doesn't care about nature, yet nature seems to speak math perfectly. More: I don't think that the Logical Empirism is as dead as it is proclaimed.
The notion that knowledge is a malleable social construct is as bizarre as stating that 2+2=4 is a social construct.
You are correct. This is already planned in two dimensions:
My focus stems likely from two sources:
I don't understand. You mean the above?
Atomicity is not a mere metaphor. My claim is that knowledge is organised in discrete building blocks that serve a specific function (taken directly from the article). Knowledge is not like atoms. Knowledge is atomic.
The atomic model of knowledge is not a metaphor. The difference between a model and a metaphor is that a model is based on assumption about reality, a metaphor is an epistemic/didactic device to give you a feel and a rough direction to start your thinking process.
The similarity of both introduce two fallacies:
The photography metaphor works great as many level 2 approaches work great. However, as a metaphor it still is subject the the level 2 problems as layed out in the article.
"Showing" would mean that you convince by providing learning experiences. The How depends on understanding the underlying learning reality.
It seems to me that something got mixed up because of the AI-translation.
I have to cite myself: The individual in individualisation is the specific task and not the individual person.
I am not 100% confident in the completeness of the inventory of knowledge building blocks. But the underlying reality of knowledge isn't anything that you can avoid. Take arguments a an example: There are languages that work (first/second order logic, syllogisms, etc.). All of them are annoying if you are not in love with math. But you can use some vivid metaphor about logic and argumentation to understand it. The same is true for any knowledge building block.
Hopefully not.
I am a Zettler
I've split the off-topic detail discussion about photography as a kind of knowledge work:
https://forum.zettelkasten.de/discussion/3341/photography-is-knowledge-work
Hard to make clean cuts with the cross quotes everywhere. But interested readers can read up to here, then follow the link to continue.
Author at Zettelkasten.de • https://christiantietze.de/