Clarifications on Structure Notes

Mauro

Next month it'll be my zettelkasten first anniversary. Thanks to this forum as well, this has been a very prolific year. Since I started my zettelkasten I have produced more than a thousand zettels, but now I'm struggling with their organisation.

My current workflow consists of:
1. Read a book while highlighting interesting sections,
2. Process the interesting sections into zettels
3. Link the zettels in my structure note.
While this is surely better than no structure at all, I often find myself scrolling through an endless list of notes rather than an organised outline.

To solve my problem I wanted to slightly change my workflow:
1. Read a book and highlight interesting sections,
2. Create an outline in my structure note,
3. Fill that outline with the zettels I make.
Do you think this could be a better approach to properly organise my zettels?

Another concern of mine is about the "granularity" of a structure note. Eg. I have a structure note about philosophy that contains lots of zettels, and I want to subdivide it. Should I subdivide it for book, author or topic? I tend to have a topic-first approach, but maybe a book-first or an author-first approach could provide a simpler structure.

andang76

If I understand correctly, in your first method you first created the zettels, then linked them inside the structure note (bottom up approach), now you would first like to design the structure of the structure note, add over time the zettels you plan to make (top down approach).

If you feel problems with bottom up, you can surely try if using top down you obtain a better result. It's fine.

Both methods have benefits and limitations.
Top down produces more order than bottom up, but forces you more to follow the starting order you take. In your case, it is the same order of your book, so you'll tend to organize your zettel representation in the structure note with the same order of your book. This is not "good" or "bad", helps in something and limit in other things.

You can also mix both approaches, having an intermediate process. You can combine them in various ways.
Starting first levels in top down, then filling the others in bottom up.
Or alternate them, evaluating what is the most suitable for the content you are processing.

andang76

Regarding your second question, there would be a lot to say.
I develop in a writing how I manage your question in my case.
I don't like to write "the right way" (it doesn't exists), I prefer to explore the problem.
It's a little messy, I'm sorry, I've written and updated many times.
The short answer to the question is at the end

If it's too unclear, I can try to explain better and with greater focus the parts you interest you most. It's not a problem, If I can explain to you better I can learn better from my own explanations.

A broad topic that for you is Philosophy, for me it could be software development.

Even in this case there are two possible models, and I can follow one of them or combine them.

I don't have only the Development structure note with all the development notes are linked in that note.
In five years I would have a note with 10.000 links and an unmanageable mess.
In this case I think Development topic divided in smaller areas, and for every area I create a structure note. If I feel that these areas are still too big, I divide them too in even smaller areas.
Development topic become the structure note that contains links for its child areas, and maybe the second levels too.

So, what can be thinked as a very big thematic structure note, is splitted in a network of smaller structure notes.

I've written "network", even if it seems that I've described a "tree" (a hierarchy)
I don't have to build a hierarchy of structure notes. If I build it, I'll have the same problems of folder systems. For children of development I've used the term "area", not the term "subtopic".

Every note can belong to many structure notes, so you can create an infinite number of structure notes starting from a set of notes, and link notes in every structure you feel it is useful.

So, should I subdivide it for book, author or topic? I can have them all, if I want
I can create the structure note for Existentialism, and structure note of Kierkegaard into the Kierkegaard note, and structure note that collects notes originated from Critique of Pure Reason into the note of this book.
I can then connect Kierkegaard note into a philosophers note, if I want, and book notes of Kieregaard in his note.
Existentialism, Kierkegaard and Critique of Pure Reason are all structure notes, and linked together form a network, not a tree.

Dividing systematically by topic, "then" by author, "then" by book, instead, you'll develop the same issues of folder classification systems.

It's impossibile, of course, create a complete structure of this type. It's a lot of work. It's impratical and even useless insert every zettel generated reading Critique of Pure Reason in Kierkegaard, then Existentialism and in all the other possibile structure notes.

This is the theoretical possibility, within this possibility you have to choose a very partial coverage, reducing a lot the number of places where you actually put the links. You need do choose the most effective positions for every zettel or group of related zettels. You create links where you feel it's useful for your goals having links near each other in front of your eyes.
In my development network of zettels and structures, for example, if I see the final result, I feel less useful having zettel combined by language, or by books. What I use more often in my work are "functionality", so the most important criteria for making structures is composing zettels together that help me to implement or manage a feature. Those zettel can come from different books and describing different languages and tools.
This doesn't mean, anyway, that my zettels are directly composed into functions when they are created. Sometimes do, more often their first place is in a book structure note.
Studying phylosophy maybe requires a different model, so structuring by function could not be the right way for you.

Another aspect to consider, some of these relationships, besides, can be created using not links, but other constructs like tags and backlinks.

So, a possibile suggest, here, trying to answer "how can I choose my structure notes?" Just reason on what structures do you really need for your goals, then create them.

===

So, your area of knowledge is divided in a network of structure notes, according to the relationships you can create between the concepts that every structure note represents.

As already said, I have two ways to build this network.
Thinking to structures before my notes exist, or making notes first and only after creating the structures you feel need.

Designing the structure and force ourself to fill them in a systematic way is a big work, and this is the drawback of wanting a systematic organization of our notes according to the relationships between concepts. More complete we want it, more work do we have to do.

In the second approach, instead, you simply start to organize your notes as it emerges in your flow of thought or study, you don't start designing the network I've described up front, but you first take the notes and then build the structures that you feel need to organize them.
This is what I do (not always, sometimes I adopt the other, the design of the structure before the note).
I make an example, so it's more clear, maybe.
I read Critique of Pure Reason, and I have all the zettels of that book. The first structure note I create is not "Hesistentialism", but "Critique of Pure Reason" book note. In that note I start how to give a first organitation of my zettels.
During the organization of zettels in that note, some groups of notes into that note will emerge, distinct each others. I can give to some of those groups a name. That name can become a new structure note that contains links to those notes, and forms a note on itself. Maybe it could be "transcendental logic".
During time even notes that have the same level of transcendetal logic form one or more clusters, and I can repeat the process creating higher level notes. Until arriving, one day, to create clusters in the philosophy note.
When I read a second book that talks about transcendental logic, I create the second book note, but I don't create another transcendetal logic note. I will integrate the one I already have, that has already emerged as structure note of transcendental logic.

Even in this case there are benefits and drawbacks similar to those described in the first post.
In the designed-first approach (I decide to organize by topic, then by subtopic, and I add a second structuring by author) I maintain order, I'm systematic, but it's rigid and it's a lot of effort.
In the emergent approach, I obtain less order but it is a more agile and flexible method and, important thing, created structures follow my mental flows and my needs.

Even in this case, you can combine both, taking top down sessions and bottom up sessions

I've already read that you already tried having the zettel first and the structure note in a second moment, but maybe the issue is not in this aspect of your workflow, but in the fact that you have zettels, a very high-level structure note and nothing in the middle. That something in the middle could be intermediate structure notes built down-to-up. It could be hard doing a big jump in one step.

Your idea of having a book first structure note can lead in this direction, so for me is fine.
But I think it is not enough having only book structure notes.

• Every book structure note become like a silo
• there is the risk to develop the content of the book not considering what you already have written in the past

Book Structure notes are good for start, but It is useful have the same notes combined in other ways, over the time, for example when you have to process more books about the same thing.
It is very important being aware on how to manage this kind of workflow when you will have different sources about the same thing. you ideally should be able to combine all the zettel in one single network.
So, even in a book first structure note approach is a very good thing having, as the final result, notes that have their content "detached", abstracted, decontextualized from the book, even that they are created from the book.

Will

@Mauro said:
To solve my problem I wanted to slightly change my workflow:
1. Read a book and highlight interesting sections,
2. Create an outline in my structure note,
3. Fill that outline with the zettels I make.
Do you think this could be a better approach to properly organise my zettels?

Another concern of mine is about the "granularity" of a structure note. Eg. I have a structure note about philosophy that contains lots of zettels, and I want to subdivide it. Should I subdivide it for book, author or topic? I tend to have a topic-first approach, but maybe a book-first or an author-first approach could provide a simpler structure.

This is a great approach. Treat large structure notes as opportunities for refactoring. When an outline section of the structure note feels too large, break it off into a "sub-structure note" leaving behind a link.

The granularity is dictated by how you choose to refactor the structure note. The choices vary with the particulars of the moment.
When placing a note and its idea in a structure outline, think for a moment about why you are going to put it there. Is it because of the proximity to other books by other authors or because of other books by the same author? Or because it clarifies or argues a point. Or does it form a timeline of ideas?

The activity of reformating structure notes is a pleasure added to a growing zettelkasten. I have 31 structure notes and 69 sub-structure notes.

Here is a sample of some of my sub-structure note themes. This will give the idea that you can use a variety of themes to break up structure notes into manageable outlines.

S-ZK Didactics 202008220529
S-Biases 202002010738
S-Stoic Philosophy 202010110616
S-Python 202110091918
S-Eufriction 202401120930
S-Parkinson's Disease 202401050607
S-Becoming Blind 202401050603
S-Economics of Health Sciences 202401050559
S-Birding Garden 202105241530
S-Ivan Illich - Philosophy of Technology 202307310745
S-Philosophy in the Age of AI 202306130835
S-Outlining 202305190954
S-List Comprehensions 202305041905
S-Abundance Of Music Composers 202304141633
S-Keyboard Maestro 202102052228
S-Haiku Center 202007011434
S-The Archive 202006240540
S-Philosophy Of Travel 202208082023
S-Raymond Carver's Poetry 202212291740
S-Books Read With A Theme Of Solitude 202011291123
S-Poetry of Zettelkasting 202204251652
S-Notable Note Takers 202102240918
S-Guy Tal 202102262224

ZettelDistraction

I treat structure notes, category notes, and outlines identically, since I cannot see any substantive difference between them. Outlines are useful for top-down approaches, which I have not been terribly good at, in practice.

andang76

@ZettelDistraction said:
I treat structure notes, category notes, and outlines identically, since I cannot see any substantive difference between them. Outlines are useful for top-down approaches, which I have not been terribly good at, in practice.

Me too.
There are a dozen of names for what I consider the same thing.
Composition of notes according to some form of relationship.
The only relevant difference is if you build them top-down or bottom up.

Will

@ZettelDistraction said:
I treat structure notes, category notes, and outlines identically, since I cannot see any substantive difference between them. Outlines are useful for top-down approaches, which I have not been terribly good at, in practice.

Anything I've created top-down has languished.

Treat structure notes dynamically. Use a bottom-up approach, adding over time and refactoring as you go. When wrestling with a new idea/note, think about what it can be associated with. Bring it to a structure note, placing it surrounded by like ideas. Make a reference as to why you put it in a certain place on a certain structure note. In this way, the structure/outline grows until it bursts and gives birth to a new, more compact sub-hub of a thought stream. The process starts over again.