A Tale of Complexity – Structural Layers in Note Taking
A Tale of Complexity – Structural Layers in Note Taking
A Zettelkasten is neither a neatly structured filing system for notes easy to access nor a turmoil deep sea generating ideas out of the ununderstandable chaos. There are three layers in my archive which emerged from the years of working with the Zettelkasten Method. I didn’t plan them in advance. It rather was a organic process.
Howdy, Stranger!
Comments
Looks like the link to the screenshot is broken.
Fixed, thanks for pointing that out!
Author at Zettelkasten.de • https://christiantietze.de/
Great explanation of how a good learning workflow pays long-term dividends and how we often limit ourselves (say, to project-centered solutions) because we haven't got a good system in place. Love this post!
Quick question: I notice the screenshot of the note in the "Structure Notes" section has only double-hash tags. This would indicate that it's actually a Main Structure Note. Is that correct? Do structure notes and main structure notes look different?
No. They don't look different.
I came up with a spacial model but Christian doesn't think it is a good illustration.
Posts like this are much more about understanding the metaphysics of a Zettelkasten. I am tinkering with the models and heuristics because often I am suprised by issues people come up with. For example: Can a Zettelkasten can get cluttered? The answer is no. It scales organically with the content and the variability of topics. If you organize your archive properly. (Perhaps, the reason is that I orientate myself towards the work of Frederic Vester?)
I think I am suprised because my internal model of a archive (or Zettelkasten) seems to be very different from man others which is a reason for things like the big "Folgezettel"-war with Daniel Lüdecke. That is the reason I try to develop a theoretical corpus of texts like that.
Practically, the whole issue is very easily solved. I use double-hash tags when I want to get to some point quickly. The "##Morgenroutine" is just for getting there quickly with a few strokes. The "##Ü3" indicates an outline. Main structure notes emerge organically.
I am a Zettler
I'm using zim-wiki for my notes, and it does not have the separation between reading and producing. I'm very happy with it. Great article, and great site.
Thanks.
How many notes do you have? When I used it the zim-wiki it slowed remarkably after too many notes.
I am a Zettler
Thanks for this this is very helpful. Decided to just dive in to creating a zettelkasten and really getting a lot out of it, but was struggling with how to scale up as I add more notes. Was generally familiar with the idea of a "table of contents" or index note, but particularly eager to explore idea of "main structure notes" and double hashes to allow quick and easy way to filter to just see the main structure notes.
Hi
I am still starting out with my Zettelkasten so I have a lot of questions. I'm trying to wrap my head around this topic of structural notes. It would help if you would elaborate on how you actually create these structural notes. Particularly with something like ##Morgenroutine. Do you make one note with ##Morgenroutine that combines things that you consider to fall broadly under "morgenroutine"? In doing so, I imagine you would introduce some structure to different subsets of notes that fall under "morgenroutine". Or does the structural note somehow emerge on its own by adding ##morgenroutine to every note pertaining to the structural hashtag? I'm sorry if this is a stupid question. In my mind it is the former, but the double hashtag throws me off. Couldn't you just use a different title structure for structural notes as you have also done in the example "Ü3 morgenroutine". Searching for "Ü3" would give you all of your structural notes, and choosing "Ü3 morgenroutine" would direct you to your table of contents with all the notes on the topic, so why add ##Ü3 and ##morgenroutine? Maybe it's a way to keep the structural note somehow updating automatically, which is the part I'm missing from this post. Does the structural note have an automatic way of updating itself or do you revisit them regularly to update the table of contents?
All the best
P
This in particular is an outline a book on creating a morning routine. So, created it once and then linked to new Zettel if they were relevant. When I created this Zettel, I searched my Zettelkasten for relevant Zettel and sorted them in.
No. Only the Structure Zettel that deals with "##tag" gets the double-hash. The double hash allows me to distinguish between normal Zettel and the Zettel.
The double-hash in general is marker for meta-tags. The difference between "#morgenroutine" and "##morgenroutine" is that the former is asigned to any Zettel that is about the "morgenroutine" and the later is only asigned to the one Zettel that rules them all.
There is not automatisation method.
All is done by oneself. This is important because one of the main ingredients of a good system is a skilled system user. If a system automates structuring one couldn't use the system because one would be too incompetent to work with the skilled system.
This one a big reason why many systems fail: They don't assure that the system user is learning with the system. Many systems are designed to compensate for weaknesses of the user. They try to automate things, allow for sloppy work (just dumping files in, clipping, incomplete processing, postponing regularly etc.).
Take Luhmann for example: He was highly skilled even without his Zettelkasten because his Zettelkasten was a training tool. I encountered this effect in my own work. I know many topics inside out because I worked with my Zettelkasten and did many things myself. Don't model the Zettelkasten as a thing that does things for your but more like a training tool that mirrors your own knowledge work (for the better or worse).
I am a Zettler
Thanks! That was very helpful, as always.
You're welcome.
I am a Zettler
Hi,
Just to be 100% clear, the top level structure notes are also outlines and will have similar naming convention as the structure notes. i.e. They will have the ##Ü3 tag and ##top-level-topic tag.
No, the other way arround. Outlines are also top level Structure Zettel but there are some Structure Zettel that aren't outlines. Top-Level Structure Zettel will have a "##Ü1" and "##tag". Outlines will have a "##Ü3" and "##tag" and/or "##projecttag".
I am a Zettler
I understand that you are suggesting to use top level structure notes & structure notes to represent topic and subtopics. ##Ü1 will group all the top level structure notes while ##Ü3 will group all the structure notes.
I am still not clear about the meta-tags representing the topics/project.
Since you are suggesting to use metatags for topics and projects along with the ##Ü1 and ##Ü3 tags, I assume that this means that we can have multiple structure notes representing a same topic/project and apply ##topic or ##projecttag to those notes
But your example also suggests that we will only have one top level note representing a topic that will represent the tip of the iceberg. Then why do we need the metatag for the topic itself. Why cant I just use the note name as the topic name? If I want to search for a structure note, I can always search for
##Ü1 topicnameor##Ü3 topicname@Sascha
How organic was this formulation of 3 layers in your zettelkasten? As you approached 1500 notes, you recall, "I wasn't very concerned ... Then structure notes emerged."
As the notes became more difficult to navigate, were you actively looking for "a grip"? Then you came up with structure notes, so you began a directed process of "mapping" your ZK? If that's the case then did any alternatives suggest themself to you / do you think structure notes were the only way to do it?
Or did you naturally start constructing structure notes as you did your normal zettelkasten stuff, and you realised that this was giving a (useful) shape to your ZK?
Basically: is the fabrication of structure notes an active process by the human, or one that occurs "organically" (as the title suggests) in the zettelkasten? How should I view it?
This was an instructive blog for me as I have only ~50 notes and I hadn't heard a description of how a zettelkasten evolves before. I think that such an outline/timeline of a ZK is useful for beginners like me. Before this, I had been trying to work out how to integrate structure notes, and it's good to know that I can incorporate them gradually, when I feel that my ZK will benefit.
Final question: do you think that you will reach a point where you need another layer of complexity? Obviously they'll be minor adjustments to the system, but do you think that this current framework will be enough to handle a lifetime of notes?
I'll be interested to see your thoughts, and I look forward to nurturing my zettelkasten through the early stages of its life.
Thanks,
H
I'm not Sascha, but I believe I can answer that question. The answer would be no. The top-level is the top of the Zettelkasten. It's the tip of the iceberg. You could always keep their number small to not need further structure, though.
Both:
My advice: Start with structure notes right from the beginning. The direct links from content note to content note will not be harmed if you have the structural layer on top.
No. And to be honest, the more longer I use my ZK the less the difference of the layers become. If I'd start over I would just have one tag for all structure notes.
I am a Zettler
I can't entirely agree with this advice for a novice striving to learn about the ZKM. In the beginning, I had so much to learn about onboarding ideas in my ZK. It used to be talked about quite a bit in the forums, here, here, here, here, and here, that one of the essential factors in the ZKM was that you should resist the desire to impose top-down structure and let structure naturally emerge from a mass of notes. For me, the critical mass of notes was about 500, and it took me over six months of concentrated work to get there before my first structure note. Now I'm no longer a novice. My workflow has evolved, and I still don't start from a structure note. I sometimes forget I am talking with novices who have the same questions I had when I started. That was so long ago that I want to be careful about letting my Curse of Knowledge show.
@Sascha, do you remember how you understood structure notes when you had 25 notes? When was that? Just let us know which geologic era it was? Was that in the Triassic period or the Carboniferous period?
Will Simpson
My zettelkasten is for my ideas, not the ideas of others. I don’t want to waste my time tinkering with my ZK; I’d rather dive into the work itself. My peak cognition is behind me. One day soon, I will read my last book, write my last note, eat my last meal, and kiss my sweetie for the last time.
kestrelcreek.com
Let me just say @Will , I am certain I am far behind you in expertise in this method and have benefitted much from all your contributions. And I'm looking forward to @Sascha response to your inquiry.
But, for the sake of discussion I would like to push back a bit on your push back and pose a question: Is beginning with a structure note really equal to imposing a top-down structure?
In the context of what has been advocated (throughout the blog posts and here on the forums), I've understood "top-down structure" to mean a predefined hierarchical system of categorization (but perhaps I misunderstood). But if my understanding is correct, then beginning with a structure note from the start would still allow for organic growth, wouldn't it?
For instance: Lets say I want to begin building a zetelkasten, so my first note is about the Zettelkasten Method, so I make a structure note, "Zettelkasten Method." I then write two notes linking from that structure note, perhaps one about the "Principle of Atomicity" and the other about the "Principle of Connectivity". The next day I might be studying about reasoning so I create another structure note on "Reasoning," and create a new note linking from it about "Reasoning as Recognition of Patterns" and another on "Reasoning as Communication". From these six individual notes I see a connection between the note on "Reasoning as Recognition of Patterns" and the note on "Principle of Connectivity" and so I create my seventh note explaining their connection with the example of morphological patterns in language and then realize I need a new structure note on "Morphology Inflection and Case." I then refactor that note and hand it on my new structure note.
Of course, we could continue on, but I think it illustrates that beginning with a structure note is not equal to imposing a top-down structure. It still allows for organic growth.
I think part of the problem is the term, "structure," because structure is the arrangement of individual parts (notes for our purposes). But I think there are at least two ways to build a structure note: 1) organically arranged from the start, or 2) organically assembled from existing notes (e.g. through tags). I also think the term "critical mass" is a bit subjective because it seems to imply a point of recognized patterns that would benefit from (additional) structure. But what if the structure was begun from the beginning? Or to put it in a gardening analogy, what is the difference between a garden box started with one flower and added to over time, versus a garden box with flowers that have been transplanted from the surrounding garden? But of course, I still use both methods for creating structure notes (and the experience of finding/creating ones born from that "critical mass" is exciting!).
Personally, I wish that I would have started this way. It wasn't until somewhere around note 1,500 when I first read about starting from a structure note (I just hit 2,000). I think beginning from a structure note at the outset would have allowed me to better think through how each note I wrote connected to others because the cognitive load would have been manageable (i.e., beginning with two notes and their relation). And what's more, I think "orphaned notes" as we have called them (or "orphaned thoughts"), are really in their essence more akin to the "the collector's fallacy" because they are not connected to any other pice of knowledge (which violates the principle of connection).
What do you think? What have I missed? What am I misunderstanding? What have I not considered?
@Darryl, it is excellent to engage with you on this topic. I, too, wait for the sleeping giant gnome @Sascha to awake and unleash his secret army of Malinois puppies.
Maybe I am a bit out of line in my thinking. But starting with note number ONE as a structure note is counter to the ZKM as I understand it. At least, that is my understanding, and my thinking is fallible.
I wouldn't be so sure. You may be a lap ahead, and it only looks like we're even.
And I'm not so sure that the novice should concern herself with the intricacies of developing a structure note. You and I are not novices. We have experience building our ZK. How we act now is far different from when we started.
I was triggered by "My advice: Start with structure notes right from the beginning." If I had taken this advice, I'd be stumbling out of the gate and wouldn't be where I am. @Darryl, back at note three, did you start with a structure note? I know you said you wish you would have, but that is like a concert pianist saying she wishes she'd played Brahms No. 2 in B-flat major on her first day of practice.
You describe the top-down structure as I understand it.
You ask an interesting and contentious question. I'd say no. Imposing a top-down structure precludes and short-circuits organic growth.
I take organic in the context to mean that what structure is warranted is determined not by my wishful thinking but, in some sense, emerges out of the ZK chaos.
There is a lot to unpack here. Your ZK practice sounds as though it is running like a well oil machine. Again, I'm thinking about the beginner and how all this "organic structure emerging from notes" confuses a novice when the important thing is to capture novel ideas. Structure will come organically later.
It is plain that much of my thinking is muddled.
Nothing.
Organic is the opposite of top-down in this context.
I might be entirely off the mark.
Will Simpson
My zettelkasten is for my ideas, not the ideas of others. I don’t want to waste my time tinkering with my ZK; I’d rather dive into the work itself. My peak cognition is behind me. One day soon, I will read my last book, write my last note, eat my last meal, and kiss my sweetie for the last time.
kestrelcreek.com
Okay, I now want to see an army of Malinois puppies and die from cuteness. The image is soooo epic !
I would like to pinpoint you both forget something important : the purpose of notes, and the topic. I think we need an example here !
My first baby steps into Zettelkasten was about making my own painting. I already master something from the subject, and know the purpose of my first notes : to do not forget my recipes and the "why" behind the "how".
So I begin with a structure note, here an excerpt (sorry for the poor translation, I do my best here with technical words) :
How can we make the "organic" concept understandable to beginners ?
Organic : because I can change the topics and the organization inside it. Instead of classifying my binders with "natural" and "artificial" criterias, I can also classify it by the need of a polar (water) or apolare (Turpentine) solvants, or historical birth of their uses (Antiquity, Renaissance, Modern era, Prehistoric). The structure notes does not change the notes themselves or my Zettelkasten in a whole, but the structure note will change as my needs change and my Zettelkasten grows.
From my point of view and my (relative) understanding of making a Zettelkasten, structures notes should help you to have an overview of a subject, and help you to understand a topic. They are organic, because they change with time, needs and understanding.
Yeap, I think so too. To begin with a structure notes may give a direction to work, and gives ideas to dig deeper in a topic, to articulate knowledge, to obtain insghts and memory of what we put inside the Zettelkasten.
I am sure, and you're too kind. Your genuine humility is one of the qualities that make you such a great part of this community.
This is a good point, our practices are definitely different from when we first started. But as I'm sure that you can attest, much of the practice was through trial and error. Perhaps structure notes are only considered as intricate because of the route taken to get there (or I have an understanding of structure notes that is too simplistic).
My view is that the experience we share with others is beneficial in helping them to avoid the trial and error process. Similar to a strength and conditioning coach teaching good form from the beginning, rather than just letting the trainee figure things out on their own (although there will be a personal learning curve for application).
Haha, this was a good analogy. But I think that perhaps our view of structure notes come into play again at this point. Given that you stated, "the critical mass of notes was about 500, and it took me over six months of concentrated work to get there before my first structure note," I can understand why you would compare structure notes to Brahms No. 2 in B-flat major. You invested a lot of time and effort to produce your first structure note. And the examples I've seen on the forums testify to that. However, I think of structure notes as a place for structure, regardless of how complex or simple they may appear. In my thinking, which may be incorrect, a structure note does not have to have the complexity of Brahms No. 2 in B-flat major (although it may organically become so).
It seems like we're on the same page here. But just to be sure, I think the salient bit is, "predefined hierarchical system." I don't think a single structure note with two notes from the beginning would qualify as such. At this beginning stage, the structure note would be more like a structural workspace, than a structure note as we know them. But, if it is grown as interest allows, the growth would still be organic, and eventually the list of connected notes would need to be refactored through subheadings, rearranging, commentary, etc., and would be just like structure notes as we know them.
I suppose the difference is that I don't see structure notes as imposing a top-down structure, nor precluding organic growth (see my explanation above).
I think I can understand why you say this, again, perhaps related to your experience and the circumstances of developing your first structure note. However, it was only chaos because there was no previous order. The structure note brought order from the chaos. That would then seem to imply that beginning with a structure note will help to avoid (or at least mitigate) the chaos of unconnectedness, wouldn't it?
There is always room for improvement, and I still have many orphaned notes.
I agree with this. In the beginning it is indeed important to capture novel ideas.
But with all that said, I do concede that my comments are theoretical and your position regarding emerging structure notes is indeed valid, even from my own personal experience.
Yes, @Loni I agree with this. I tried to explain this aspect in my long-winded post (above, before I saw your post), but I think you said it more succinctly, especially the final sentence. Thanks for joining in!
I have a ZK, after 2 years, of about 500 zettels. I have rarely used structure notes. They have their uses, but they are not the first thing I think of when I'm entering one or several zettels. I only use them if it makes sense.
It doesn't matter whether you call the structure "top down" or not, structure notes do exactly what their name suggests - they impose structure, and that is sort of contrary to the whole idea of connecting zettels organically.
So, I'm with @Will .
I am so tempted to craft argumentation here… So tempted.
Let's get this.
You oppose structure and organic.
But organic has its own structure.
This is a neuron ^1 :
Cross section of a typical root - xylem and phloem ^2 :
DNA structure ^3 :
Life is about transmitting information. DNA itself may be a way to convey "programming" code between cells (and virus… are trully virus, hacking our cells to duplicate their own viral code, biology is beautiful).
So… If structure is good enough for life itself, it would be good for me as well. No lively system seems to exist without structuring functions or structured informations.
But you add the word "impose" in your sentence.
I would like to stay in the biological field : it seems like you see structure notes as an exosqueleton. From this perspective, yes, strucutre notes would impose something, and the only way to change the squeleton or to grow would be to moult. If they are squeletons, the only way to change it would be… painfull, at best.
But what if we see structure notes like a way to classify some lively inputs through criterias ? I put "orange" with "strawberry" because they are both fruits, but I can also put orange with tomatoes because they both grow in the South, or "strawberry" with "sausage" because they are both aliment with the "s" as first letter.
From the pool of notes I displayed on my previous post about making paint, I could make at least and without thinking four different classifications, and they would all be valid structure notes. They would evolve because of my need, me as a user, playing the role of the environment pressure. They would grow in the right conditions, and even giving birth to smaller part of self. They might even die, poor little things if I forget to feed them.
I admit that you don't feel the need of structure note, I don't have any problems with that. But they are definitively organic.
Sources :
Haha - good argument. I suppose I should have said an hierarchical structure - super-folder, folder, sub-folder, etc. I'm certainly not against structure, in a general sense, I'd just like it to reveal itself rather than me determining what it should be at the very start. Serves me right for being lazy
How to make a Loni happy, part I.
I had fun writting it, I am glad it talks to you
This could be seen as a squeletton, indeed. From what I understand, writting a Zettelkasten with links should prevent us for asking "where should I put it ?". Folders structure doesn't seem to accomplish that philosophy.
From my opinion, it will rarely stay "as it". As content evolves, the structure notes evolve as well. It is just like outlining a not yet written book : I make a first outline, the novel ideas evolve, the outline evolves as well. But the first outlining draft may gives a direction for the content and help to clean up and to organise the first thoughts, giving order to chaos. Structures notes can disappear and change with time without disturbing the whole Zettelkasten, just like my outlining drafts can disappear without changing my novel content.
In the end, those are method and sensibility matters, about how you appreciate to work, how you want to set up your workflow. If the way you work is fine to you, it is fine to me as well !
Yes, I agree with that statement. But if I understand some of the other posts in the forum, that is not the approach many others take. As you say further down though - a person should do what works for them. I think it's tied to why they are creating their ZK in the first place - for some purposes, imposing a lot of structure at the beginning might be the right thing to do. But it's so easy to head off in the "wrong" direction in one's ZK and then it takes a lot of time to recover, once realization dawns.
That's an interesting way to think of it. That perspective on using structure notes hadn't occurred to me, but then I don't use them much. Normally I create a structure note if I notice I'm accumulating a lot of zettels on a particular topic and I want to provide a more organized entry point (into the ZK) for that topic. Your description could certainly apply to my approach - I could reach a point where I want to split a structure note, or remold it, or morph it into one or several different structure notes.
Yes, for sure. I like that people share their ideas here; I learn a lot from them. On occasion, I will try something new that has worked for another person - if it works for me, great; if not, I back up and try something different.
I think the process is a lot like parenting. I have 5 kids (all adults now); each one is a unique person that requires different communication approaches and different ways of encouragement. Working with my ZK is something like working with an 8 year old - they are very enthusiastic, but I'm constantly trying to figure out how to interact with them, as they are growing and maturing so rapidly.
The people about whom I worry are the newcomers to ZK - they sometimes take forum advice much too literally. They are trying to learn how to get going on a ZK and sometimes treat it like they must follow a recipe. It's hard to grasp the idea that there are many ways to grow a ZK - that concept can be very frustrating to them.
I agree. If we focus on beginners however (as I was some monthes ago), using folders could be a trap as separate content in hierachical folders is a reflex. But they can be pertinent for some uses, of course !
Damned it, yes !
I experienced it first hand. 
It is frustrating, but it develops experiences and vigilance. really think that we can't fully learn and understand something until we live it by ourselves, with our own subjectivity.
However, mistakes and fixes take part of learning process. Begginers should'nt be afraid of failing. There is no fail, only lessons !
You use the parenting metaphor for the ZD itself, and I find it accurate. It can apply to beginners as well : as learning needsometimes first-hands experiences, they will make mistakes. There is something I tell my 5 years old son almost every days : "I know you are about to fall here… Wupsie. Told you, son. Come here. Take a magic healing kiss and then, we'll fix your mess." (and "go to the bathroom" too, but it's not the point) The Big Anciants can do that kind of things (where "magic healing kiss" is a metaphor, it would be weird for them to kiss their screens.)
We can't prevent chaos to emerge at some point. But you can offer guidance and advices ( I will stick with strange demonstrations as I am not a so much experimented user).
And there we have a nice and lively ecosystem ! \o/
From my perspective, being lazy and trying things that worked are the key to a clever process. What would be the point of reinventing the wheel or making something complicated if a more efficient solution exists ?