The Iceberg Model and Atomicity
This discussion was created from comments split from: Cardcraft And Atomicity.
Howdy, Stranger!
Categories
- 3K All Categories
- 152 Research & Reading
- 692 The Zettelkasten Method
- 7 Knowledge Work
- 100 Writing
- 464 Software & Gadgets
- 154 Workflows
- 730 The Archive
- 15 Plug-In Showcase
- 88 Resolved Issues
- 225 Projects Logs and Journals
- 83 Project: Zettelkasten.de
- 53 Critique my Zettel
- 171 Random
- 373 Introduce Yourselves!
Comments
I'm confused. In The Complete Guide to Atomic Note-Taking you write in section "Level 3: Atomicity and the Zettelkasten Method":
Level 3 of the iceberg is labeled "Zettelkasten Method".
My brain reads: "Zettelkasten Method" = level 3 = knowledge building blocks = a very specifc theory of knowledge.
What am I missing?
You seem to read "x is y" as only "x is y". However, "x is y" is compatible with "z is y".
In the guide, I don't argue for the position that knowledge building blocks knowledge atoms and only them are ones. I don't argue at all. I just make a claim and continue based on the claim.
So, to move from level 2 to level 3, you have to adhere to a model for atomicity. But, as stated above or somewhere else. The decision for a model doesn't have to be consistent for the total Zettelkasten. I use completely different atoms for story elements, and they link happily back and forth with my knowledge.
Additionally: Level 3 marks a direction and the distinction between level 2 and level 3 is marked by the difference between a natural and somewhat formal (or artificial?) language. @Andy nailed the connection to Rudolph Carnap's idea of explication.
Contrary to Carnap as I understood his intentions, I don't think that concepts should be fully explicated. The process of explication is an epistemological device, but the ordinary language is still the foundation for explications. For example: When I talk about knowledge I don't narrow the meaning of the word down to a set of building blocks for the same reason why I don't narrow down the idea of my dog to flesh, fur, and misbehaviour.
If you fully worked on an idea, you mastered it on all 4 levels, including level 1.
I am a Zettler
Can we have this a bit less abstract and a bit more concrete?
I'm trying to understand the relationship between "Zettelkasten Method", "Level 3" and "Knowledge Building Blocks".
You designed a model. I'm trying to understand the model you designed.
The levels and their labels are your personal design decision. You could have chosen 3 or 5 or 10 levels, but you chose 4. You could have chosen the label "Zettelkasten Method" for the entire model, but you chose to assign the label to level 3. Why?
The six knowledge building blocks are also your personal design decision. You seem to be happy with your design. You seem to have anecdotal evidence that it works for your students.
A wise man once said:
He had observed that:
So he made a clear statement:
How is this compatible with your current view?
If the answer to the question "What is this atom, then?" is not "knowledge building blocks", what is it then? I'm confused.
And going back to the iceberg, could you explain why you chose the label "Zettelkasten Method" for level 3? The Guide also states more generally, that "The Zettelkasten Method is about creating links."
The problem is on the logical level. Logic is by definition a bit abstract.
No, this is not how it works. It happened to be 4 levels.
I had to chose some label.
No, they are not a personal design decision. I don't know that this is a wording issue, but "personal design decision" seems to imply that it is about taste or preference.
The quote is too short. There are a couple of sentences that are missing to the full point I am making.
As stated above. "x is y" is perfectly compatible with "z is y". Just because I present a solution to the problem of adhering to a model for atomicity doesn't mean that this is the only solution.
If the logical issue is resolved, the confusion should be resolved, too.
I am a Zettler
Didn't help. I neither understand what you mean by "logical issue" nor how it might be relevant for my question. Would you mind giving a hint? Maybe replace the variables x, y and z with concrete values like "Zettelkasten Method", "level 3", "knowledge building blocks", "a very specifc theory of knowledge"?
I don't know how "it" works for you. That's why I'm asking about your design process. :-)
Yes, it is "about taste or preference". What else? :-)
If it's about something more, I'd like to learn about that something more.
You said a few months ago:
You also mentioned in a footnote1 that you are developing "a pattern language for knowledge". So I'm curious about your development process. How do you decide where to draw the boundaries between levels or blocks?
You seem to have been quite sure about the six blocks for a while. In your 2015 book2 you simply state on page 55 (machine translation):3
The claim comes with no explanation, no justification, no source. You state it like a fact, that there are 6 things. Not more, not less. Exactly six.
Your later writing is consistent with that claim. In August 2025 you said:
This doesn't sound like "a model for atomicity", but the model for atomicity.
I'm still confused. :-) What am I missing? What have I overlooked?
Footnote 3 of The Complete Guide to Atomic Note-Taking ↩︎
Sascha Fast: Die Zettelkastenmethode. Kontrolliere dein Wissen. 2015 https://www.goodreads.com/de/book/show/29981373-die-zettelkastenmethode ↩︎
"Es gibt 6 Dinge, die am Ende deines Lesens herauskommen können und gleichzeitig für deinen Zettelkasten relevant sind." ↩︎
To be fair, in august 2025, Sasha also posted this:
https://zettelkasten.de/posts/principle-of-atomicity-difference-between-principle-and-implementation/#everybody-talks-about-atomicity-nobody-aint-talking-about-the-atoms
I read that as saying that if you have your own system of building blocks, go ahead and use them. I don't disagree, I just think there can be other considerations.
An analogy is that the building blocks of everyday matter could be said to be the 92 naturally occurring atoms. How then should I write a card about the benzene ring, which is a specific chemical composed of hydrogen and carbon atoms in equal numbers? Its distinguishing feature is that these atoms have been assembled to form a stable closed ring.
Obviously the card cannot be about hydrogen atoms or carbon atoms. All right, we can say that molecules are the building blocks. But there are potentially an infinite number of them. Maybe we could say that the card is about the concept of a benzene ring, since "concept" is, I think one of Sasha's building blocks. But I don't think that fits either (depending on the purpose of the note, of course).
One could say that any card is about a concept if it isn't clearly about one of the other five building blocks. But that would be unhelpful!
I read an affirmation of the claim that "there is an inventory of knowledge building blocks".
I read an affirmation of the claim "of which there are six" plus some room for error: "there may be a seventh, or one item isn't an actual building block".
I read that "the knowledge building blocks are not arbitrarily cherry-picked constructions".
Where do you read that this particular inventory is just one of several possible inventories?
I generalized from "there may be a seventh, or one item isn't an actual building block", combined with "But that there is an inventory of knowledge building blocks is my claim".
I read that as saying "use building blocks, whatever ones that you recognize".
As I suppose you realize by now, I disagree about what should be considered as a building block to put on a card, and I also do not agree that the cognitive instrument we call a Zettelkasten has the primary or only purpose of "constructing knowledge" in a building-block manner. That is to say that the latter is valid and important but also limiting.
And it's fine that we don't all see it the same way. A ZK is at the very least an intensely personal construct.
@tomp said:
As I understand it, if you are using a set of knowledge building blocks (types) like Sascha's, an atomic concept note about a benzene ring would only define what a benzene ring is, like a dictionary definition. If your note "is about" a benzene ring but does not simply define what it is, then it wouldn't conform to the concept type and would either conform to another type or would not be atomic. So, it would be false to say "that any card is about a concept if it isn't clearly about one of the other five building blocks".
That a benzene ring is a chemical compound doesn't seem to be relevant; chemical elements are not closely analogous to knowledge elements (knowledge building blocks).
I don't think Sascha claims that is the primary or only purpose of a Zettelkasten. Here's the definition of a Zettelkasten in his introduction:
And in the iceberg model, it's very clear that the "knowledge building blocks" model is a thinking tool that is distinct from the Zettelkasten.
In another discussion, @harr drew attention to a passage from Mario Bunge's book The Myth of Simplicity that serves well as a basic explanation of the value of atomicity if one is using a Zettelkasten for scientific knowledge or something like it:
@tomp: Personally I disagree with many aspects of Sascha's method. :-)
But in order to have a meaningful debate about our differences we need to understand our differences.
At this point I'm only trying to understand Sascha's method as it is.
I thought I had understood the relationship between atomicity and the six blocks, but apparently I was wrong.
Molecules are the natural unit of chemistry. Benzene rings play an important role in organic chemistry. Neutron, protons, and atomic nucleii are fundamental building blocks of nuclear physics. Words can be seen as natural building blocks of written language. Syllogisms, which seem to be one of Sasha's building blocks, are particular kinds of patterns. Why not take a more generalized pattern as a building block?
My point is of course that what is considered to be a building block varies with the field of interest and the way it is going to be used. Even within a single Zettelkasten there will be many fields of interest and many ways that building blocks are going to be used.
Suppose I use my ZK to help me write a short story. I will probably use it to help develop and save the structure and story arc, to keep character notes, to remember descriptions or other memorable text bits I want to include, maps of imagined topography, and so on. Not one of those things corresponds to any knowledge building block so far as I can see. The story as a whole is not going to be an assembled work of knowledge (I would rather not get into the meaning of the word "knowledge" since not even philosophers agree).
Coherence, Cohesion, Consistency, for example.
You are missing that you jump from "x is y" to "only x is y". You are reading something into what I am writing that isn't there.
Here is another set of building blocks, specifically for using the Zettelkasten for story writing: https://zettelkasten.de/posts/zettelkasten-fiction-writing-part-2-elements-of-story/
I am a Zettler
I didn't mean to imply helplessness. I only meant to show how there isn't one set of building blocks to be used for every purpose (or by all people).
I know. I just took the chance to throw in another set of building blocks to confirm that there are more sets of building blocks.
I am a Zettler
I wish The Guide as "living document" would be updated to reflect that possibility more clearly.
It should explicitly say that the initial inventory of six blocks is not "universal" in the sense that it can and should be used to model any kind of knowledge.
@tomp said that he didn't want to get into the meaning of knowledge, but that may be a very helpful issue to address briefly.
In a previous discussion, a book about knowledge graphs was criticized for defining them as "a graph of data intended to accumulate and convey knowledge of the world". The counterexamples of fiction and mathematics were mentioned as not being "knowledge of the world". I suggested that philosopher Ernst von Glasersfeld's definition of knowledge could be broad enough to encompass subjects such as fiction and mathematics: "knowledge refers to conceptual structures that epistemic agents, given the range of present experience within their tradition of thought and language, consider viable."
Using von Glasersfeld's definition of knowledge, the name "knowledge building block" need not mean that the building blocks are used to assemble knowledge (although this is very often what a ZK is used for: even writing a story could be considered assembling knowledge in von Glasersfeld's sense); it can mean instead that the building blocks are conceptual structures that are viable for a given purpose. If you're doing science, you need conceptual structures that are viable for doing science; if you're writing fiction, you need conceptual structures that are viable for writing fiction.
In a personal knowledge base, what counts as viable is the business of the person creating the personal knowledge base, but if the purpose is one that other "epistemic agents" have shared in the past, it may help to learn from and adapt their knowledge—the conceptual structures that they found to be viable.
The deeper levels of the iceberg model are about having and using more, and more systematic, viable conceptual structures, i.e. knowledge building blocks, for whatever your purposes are.
I dedicated my morning run to thinking about this. I decided against it. These are my reasons:
This would contradict my opinion. I think that the building blocks are universal. I just don't think that the building blocks are the only model that can be used. However, if you'd offer a different set (for knowledge), I'd likely argue against it.
Again, mistaking "x is y" for "only x is y" is the underlying fallacy here. You are collecting evidence for me, saying "x is y" ("knowledge building blocks model knowledge"), and you quote the evidence.
I am a Zettler
Thanks!
I'd like to add context for readers who don't know the context. Sascha and I are referring to this:
I have difficulties to recognize such a "clear line". For example I thought that the inventory marked such a clear line. It doesn't help that relevant information is scattered all over the forum.
That's the kind of basic information I'd expect to read in a "complete" guide. Am I expecting too much of "The Complete Guide to Atomic Note-Taking"? Why should readers have to search the forum for such basic information?
Not sure if I understand your opinion correctly.
Are you saying that your inventory of knowledge building blocks is universal in the sense, that it can and should be used to model any kind of knowledge?
Again, I have no idea what you mean by x and y.
I think there are two ideas in play here and they tend to get blurred together:
Note that the phrase "building block" is a metaphor. We are talking about a supposedly rationally-based idea yet we are discussing it using embodied metaphors. This is common in language and even thinking. Nothing wrong here: one can't really get away from doing so. But we shouldn't forget it.
Don't forget building blocks in terms of technological components that make up the system. A paper-based system is composed of index cards. A digital hypertext system is composed of Markdown files, database records, …
I'd also like to add that the initial justification for those knowledge building blocks is reading. Sascha's post Reading for the Zettelkasten is Searching describes them as "a typology of items to serve me as an epistemiologic amplifier". He claims that they are "a small ontology of item types that are hidden in the various texts we read." He recommends: "If you read a text for your Zettelkasten you should search for items. This will deepen your knowledge work with the text and accelerate the reading process."
I'd like to add "blocks help with reading" as a third item to your list of ideas that tend to get blurred together. :-) And maybe "blocks can be found in texts" as a fourth?
There are blocks and then there are blocks. When a dry laid stone wall is built, the mason tries to find rocks of the right size and shape to fit in and lock to others. But sometimes a rock needs to be reshaped to fit a particular place, and that changes the choice and shaping of the next blocks to be placed.
After the discussions of the last few days, I am coming to view my Zettelkasten as more of a model than anything else. Like most good models it needs pruning and revisions over time.
In an earlier comment (probably on another thread), I quoted Elaine Svenonius on subject languages, and I quote her again here:
I view the ZK's collection of outline headings, or equivalently keywords, structure note titles, link types, etc. to constitute an informal, personal subject language. And even though finding may not be the main purpose, there is no point in storing a card in a ZK unless you can find it again later, and preferably find related cards too.
"...the subject language becomes an analog of knowledge itself". To my ears that says "model". In my terms, the structure of a ZK models the practitioner's understanding of the contents. One's understanding is dynamic, and tacit to a large degree, and that will be reflected in the ZK itself.
This is not "basic information". It is a correction to a fallacy that stems from a specific bad thinking habit that I don't assume the readers has.
I am a Zettler
How else would you call an accurate description of the relationship between atomicity, knowledge building blocks and inventory?
What "specific bad thinking habit" do you have in mind?
You are not requesting an accurate description. You are requesting accommodation to a specific bad thinking habit that I don't assume the readers has.
As stated above. When you read "only x is y" when "x is y" is written.
I am a Zettler
I gave you feedback that at least one reader (me) doesn't understand your writing as you intended.
I made a suggestion.
You don't find it useful.
I'm fine with that.
We're stuck in a loop. :-)
I noticed an error in the previous post. The last quote should read: