# Is there a benefit to Luhmann IDs vs. Date/Time IDs?

• @cobblepot said:
I agree, in that case it's easy/obvious to decide on a branch. But what about when I start taking notes on a source and these notes relates to many existing notes already in my ZK? How to decide where to put it? That's the issue I have trouble with.

pseudoevagrius is/has been attempting to solve this exact problem and this only person I have noticed in the forum actually dealing with it.

@pseudoevagrius said this here:
I’ve used folgezetteln in a way that I track in the header of each note. Described here: https://forum.zettelkasten.de/discussion/comment/3648#Comment_3648

I call them "folgechains" because it is not a strict use of folgezettel and one zettel can be a part of multiple folge chains.

And has a large Zettelkasten which gives me confidence in the perspective @pseudoevagrius can add to the discussion.

And said here:

My zettelkasten is currently around 12,000 interconnected notes. So that's the kind of size I'm dealing with for reference.

@pseudoevagrius my apologies if you feel I am throwing you under the fogezettel bus. I have great respect for your insights and comments.

@cobblepot, thank you for driving the bus to valuable destinations. Time well spent.

• @cobblepot said:
I just am not seeing how people are branching off notes in ways other than "this note is on a topic that is closest in topic to existing note X". So I am getting the sense that the branches are actually just based on one's momentary intuition or gut instinct about note relationships, which can include a ton of things including existing mental associations, and maybe the benefit is just that Luhmann IDs reflect those existing associations.

This. Folgezettel are a mean to rob meaning from hierarchy (so it does not matter where you put Zettel). Direkt links need to be intentional and the connection needs to be made explicit. Structure Zettel need hierarchical order to make sense.

The upcoming blogpost (edited, so you can read it in English and not sfastlish) could be summarised like this: Folgezettel remove meaning from hierarchy. Therefore, you can not extract anything from Folgezettel later on. The other techniques add meaning to connection and therefore enrich you Zettelkasten instead of just deliver a UID.

I am a Zettler

• edited April 2020

@cobblepot said:
"Don't you see @cobblepot ?! It's obvious! You make the direct, obvious links, rather than the irrelevant, pointless ones!" But...what makes these links seem direct and obvious to you? Is it possible they would not be direct and obvious to someone else?

Haha, that was kinda funny.

Anyway, I don't intend my system to be perfect. I just wanna work with it quickly. If I didn't see any obvious connection at first, I can always create a new sequence. That's how I make it simple. My most fundamental assumption is that my Zettelkasten replicates how I think. The existing connections, then, are merely external scaffolds to let me do the said thinking with 100% engagement. If I'm unable to recognize how a new note connects to anything in my collection at first, then by any means, I'll just create a new sequence using a new number. I can connect the two sequences in the future anyway once I realize they're connected in some way.

But for that to work, my rule is that new note sequences don't have initial connections with each other. That means I don't treat 6 as a "sequence" to 5, similar to how you wouldn't treat 202004011230 and 202004011231 as having a direct connection.

That said, I use Luhmann ID's only for easy sorting and UID's for easy linking. I don't even see a need to have a debate, rather than a discussion, about it.

I give up on this discussion, though. I prefer to learn by doing. Whether I make a mistake or not, I'd still learn a lesson.

Also, I recommend you try out each method using a mock ZK in case you haven't already—that's how I proved my thought experiments and initial assumptions about Folgezettel were wrong.

Post edited by alkhan on
• @Peter said:
I gave my input, explaining in detail, from different points of view, why I believe Folgezettels are not useful. Everybody has a system that works for their particular use case. The real newbies attach a lot of importance to Folgezettel, as I did when I 1st started - I remember having a discussion with Sascha about it. Some years and a few Zettels down the road and I have had to accept he was right. I also made the point about experience in my comments above.

For y'all, this is the old Disqus-based comment thread for context: https://zettelkasten.de/posts/luhmann-folgezettel-truth/#comment-2337566932

Author at Zettelkasten.de • https://christiantietze.de/

• @ctietze I finally read those Disqus-based comments. It was déjà vu all over again. Thank you for pointing us in that direction. I am not sure we have uncovered much new about this discussion. I think we have been finding and saying much of the same as from 5 years ago.

@cobblepot has a great skill of keeping things on topic and summarizing and distilling what's been said. Hopefully, he will review those disqus-based comments bring this together for us and keep the bus heading down the road.

• @cobblepot said:
I agree, in that case it's easy/obvious to decide on a branch. But what about when I start taking notes on a source and these notes relates to many existing notes already in my ZK? How to decide where to put it? That's the issue I have trouble with.

I think the challenge of finding other places to make insertions involves searching through using tags. links, the Ominar to find those places. You can add a direct link using a UID. But you don't have whatever benefit the foglezettel IDs provide.

Have we been thinking that a note can have only one folgezettel ID, one UID? We don't think that a note has to have just one tag. Why not multiple Folgezettel ID's?

A note title and its presentation in the note list could include multiple UIDs or foglezettel ID's. I am not advocating this because it may make the note list visually more difficult to discern. That might work for some but it may be better to stick with a UID and a foglezettel ID and the Title plug.

What if a note was assigned multiple foglezettel ids and given their proper place in the flogezettel chains? Then within the note, as metadata assigned to multiple Folgezettel ids.

Does this break the internet, create a disturbance in the force?
My first thought is that it doesn't. Maybe this was what @pseudoevagrius was saying and suggesting.

Within the note, if a list of folgezettel ids were listed that the note was inserted - added to:

Folgechains:
{301,1,a,3 } - Included as part of the Julius Ceasar tactics in war blah blah
{4083,1,a,1,j,22 } - Additional strategy for waging war against superior forces.
{3142,3 } - Spartan's also killed messengers when asked to accept submission terms.
{2733,1,h,43,a,40 } - Always seems to relate back to Sun Tzu

Searching any to these chains in the Ominbar would surface the note in the note list and other notes in the chain. Maybe one of these Folgezettel ID's is the note title but different than the one searched on. It would be noticeable maybe as a visual cue?

Is it worth it? It doesn't reduce the work to find the other places to add the note to the chains.
Maybe it provides a way to represent them when you do find where they should be added.

Lots of maybes. Does assigning a note multiple folgezettel ids break anything or help?

I am not sure where I stand on these folgezettel ids. I like trying to solve problems, thinking about things differently.

• @Peter said:

@cobblepot said:
NO! This is giving up! People (at least me) are talking because they are still trying to figure out how to make this system work best! It's easy to say "hey, whatever works for you" but it doesn't improve anything. If you don't want to engage in this masochistic nightmare with me, then fine, but don't rain on my obsession parade!

No, @cobblepot, this NOT giving up. I gave my input, explaining in detail, from different points of view, why I believe Folgezettels are not useful. Everybody has a system that works for their particular use case. The real newbies attach a lot of importance to Folgezettel, as I did when I 1st started - I remember having a discussion with Sascha about it. Some years and a few Zettels down the road and I have had to accept he was right. I also made the point about experience in my comments above.
Last but not least, you may have heard of the expression horses for courses: that applies to the Zk. There isn't 1 unique way managing one's notes and using one's Zk, esp. with the digital versions.
So, no, I don't rain on your obsession parade, I simply don't think it s useful talking about a subject ad infinitum when everything has been said. But, then again, YMMV, and it's your thread.

Hi @Peter, sorry if I misunderstood you. But I was not making a claim about your general position on Folgezettels or anything else you have said elsewhere in the forum, including months or years ago before I was even aware of the ZK system. And I am not defending Luhmann IDs (at least not yet), I'm trying to figure out how to solve my own problem with placing new notes in the ZK, and wondering whether Luhmann IDs are part of the solution.

When I said it was giving up, I was talking very specifically about your claim that "it is probably best to leave it at that" - i.e., to stop the debate with the notion that everyone works best with different systems, or in the English idiom, "different strokes for different folks" (horses for courses? what on Earth is that craziness? ). It is not best, at least for me, to leave it at that, because my system is not working well for me right now and am trying to figure out how to make it work. If your system is working well for you, then great! But I don't want you to try to persuade others that there is no point in talking about this further and that no intellectual progress can be made past the claim that everyone has their own opinion. That will leave me in my current situation, which is not working for me.

You have said a few times that people who advocate Luhmann IDs are mostly beginners that only support that system because of some baseless devotion to "the master". But that's not the case with @argonsnorts or @pseudoevagrius - they both used date/time IDs for years, accumulating thousands of notes, before switching to Luhmann IDs and found the latter more useful. And they gave specific reasons why the think the lumen ID system is more useful, none of which were based on any appeal to Luhmann's authority.

I have described in many of my previous posts the specific type of problems I'm trying to solve – namely, how to link new notes in the system where I have a number of different writing projects and complex, interrelated topics, in a way that will allow me to write about specific questions productively. If your arguments against Luhmann IDs generally can help me solve that problem, I would absolutely love for you to explain it to me – no sarcasm.

• @sfast said:

@cobblepot said:
So I am getting the sense that the branches are actually just based on one's momentary intuition or gut instinct about note relationships, which can include a ton of things including existing mental associations, and maybe the benefit is just that Luhmann IDs reflect those existing associations.

This. Folgezettel are a mean to rob meaning from hierarchy (so it does not matter where you put Zettel).

@sfast, can you explain why you interpreted my description as equivalent to the claim that assigning a Luhmann ID "robs meaning from hierarchy"? One's existing mental associations are not meaningless, they are highly meaningful, especially to the individual doing the thinking interaction with the ZK. To give just one example, if you made an association at one time, you are likely (although not guaranteed) to search for a note later on based on that association. Also, we often make mental associations between ideas for reasons that we cannot yet articulate, but in the process of recognizing our inability to articulate the reason behind the association, come to discover previously unrecognized similarities between things.

• @cobblepot said:
Hi @Peter, sorry if I misunderstood you. But I was not making a claim about your general position on Folgezettels or anything else you have said elsewhere in the forum, including months or years ago before I was even aware of the ZK system. And I am not defending Luhmann IDs (at least not yet), I'm trying to figure out how to solve my own problem with placing new notes in the ZK, and wondering whether Luhmann IDs are part of the solution.

Thanks, and I will try to be less sensitive

@cobblepot said:
When I said it was giving up, I was talking very specifically about your claim that "it is probably best to leave it at that" - i.e., to stop the debate with the notion that everyone works best with different systems, or in the English idiom, "different strokes for different folks" (horses for courses? what on Earth is that craziness? ). It is not best, at least for me, to leave it at that, because my system is not working well for me right now and am trying to figure out how to make it work. If your system is working well for you, then great! But I don't want you to try to persuade others that there is no point in talking about this further and that no intellectual progress can be made past the claim that everyone has their own opinion. That will leave me in my current situation, which is not working for me.

No, my suggestion to leave it at that was stated in my discussion with @argonsnorts
and was meant for that particular discussion, NOT for this thread that was your initiative in the 1st place.

And I definitely do NOT try to persuade others not to use Folgezettel, why should I? If you look through my comments you'll see that I challenged Sascha about his statement that he will try to "drag people out of this Folgezettel madness". I even said to him that everyone should use whatever system suits them best.

@cobblepot said:
You have said a few times that people who advocate Luhmann IDs are mostly beginners that only support that system because of some baseless devotion to "the master". But that's not the case with @argonsnorts or @pseudoevagrius - they both used date/time IDs for years, accumulating thousands of notes, before switching to Luhmann IDs and found the latter more useful. And they gave specific reasons why the think the lumen ID system is more useful, none of which were based on any appeal to Luhmann's authority.

That may be true, and if Folgezettels works for them, good for them and I will not argue against it. I personally have yet to see an argument that convinces me to revert to Folgezettels.

@cobblepot said:
I have described in many of my previous posts the specific type of problems I'm trying to solve – namely, how to link new notes in the system where I have a number of different writing projects and complex, interrelated topics, in a way that will allow me to write about specific questions productively. If your arguments against Luhmann IDs generally can help me solve that problem, I would absolutely love for you to explain it to me – no sarcasm.

In this thread I have explained all the reasons why I do not think Folgezettel are useful, I do not have more reasons. You have not been persuaded by them, that's OK. From what I gather so far you have not been persuaded by the reasons to use Folgezettel either. Perhaps you are using or trying to use the wrong system, I cannot judge that, but neither argument is swaying you in one particular direction.

Thanks for a stimulating, pleasant discussion. I am sorry for you that you are in a bind, but you'll undoubtedly work it out.

• I wonder if the practice of "inserting an association right where you are", one of the features attributed to Folgezettel, gets wiped under the rug when using date/time Ids. What I have in mind is at least not mentioned here, so let me recap what was probably said elsewhere already:

Now imagine you have an idea right now as you read Zettel XYZ somewhere in the middle of the file. With the notion of "each Zettel is a file" out of the way, you don't have to hit the "New Note" button or whatever to create a blank slate to write. You just insert a new ID right where you are, and write the note. Through this magical act of creation, you push the Zettel that were adjacent further apart and intersperse a new thought. Stop typing, and you're done. There's no ceremony to create a new Zettel, and there's no cleaning up you have to do once you stop writing. The new Zettel just appeared where you typed, in the very long text file full of other Zettel.

So this is another implementation or realization of the "create a Zettel" step in one's work.

And you can do this variant it even if your usual realization consists of creating separate files.

Have a thought but the editor shows a totally unrelated note? Just create a new ID and write away. Once you're finished, execute the imaginary Extract Zettel refactoring: cut the new Zettel content out, paste it into a new file, and leave a link from the original place to the new place.

This is about as close to @argonsnorts organic growth of loose association as it gets. (See his post about "heart → shaped like weird potato → origin of the heart icon")

In other words, do not create Zettel in a vacuum. Always leave a trail. I do a similar thing in my Epstein videos when I create "forwards links" that lead nowhere first, then create the target note second. Just like a wiki encourages us to do.

Once a Zettel is finished, look through likely related structure notes and other topics to find who else might be interested in having a link to the shiny new note.

If it turns out that the place where this note originates is a totally useless association, you can always sever it later.

Author at Zettelkasten.de • https://christiantietze.de/

• @cobblepot said:

@sfast said:

@cobblepot said:
So I am getting the sense that the branches are actually just based on one's momentary intuition or gut instinct about note relationships, which can include a ton of things including existing mental associations, and maybe the benefit is just that Luhmann IDs reflect those existing associations.

This. Folgezettel are a mean to rob meaning from hierarchy (so it does not matter where you put Zettel).

@sfast, can you explain why you interpreted my description as equivalent to the claim that assigning a Luhmann ID "robs meaning from hierarchy"? One's existing mental associations are not meaningless, they are highly meaningful, especially to the individual doing the thinking interaction with the ZK. To give just one example, if you made an association at one time, you are likely (although not guaranteed) to search for a note later on based on that association. Also, we often make mental associations between ideas for reasons that we cannot yet articulate, but in the process of recognizing our inability to articulate the reason behind the association, come to discover previously unrecognized similarities between things.

I am not interpreted as equivalent but expanded on the quote.

The momentary associations could be meaningful if they made explicit. But they are not by just placing a number in the title. And further: It is difficult to make associations explicit and even more difficult to make those associations meaningful because they are inherently not. This is not how associations work. The more creative they are the less meaningful they are from a very early threshold. Meaning is maximised when you find the right "distance" of association.

To add on to the problem: By placing a Folgezettel you are making momentary gut instinct a context. The context will fade as the moment goes by the flow of time and will take its meaning with it.

I am a Zettler

• @sfast said:
And further: It is difficult to make associations explicit and even more difficult to make those associations meaningful because they are inherently not. This is not how associations work. The more creative they are the less meaningful they are from a very early threshold. Meaning is maximised when you find the right "distance" of association.

This section totally lost me. Can you explain what the words explicit, meaningful, and creative mean to you? On my understanding of "explicit," it is easy to make associations explicit: I might link a note about chocolate ice cream with a note about a particular song because I was eating chocolate ice cream when I listened to the song. I can easily write down the reason for that association next to the link. That association has meaning because it represents a concrete situation I was once in. I have no idea what it means to say that associations can be more or less creative. By "creative" do you mean surprising, or unexpected? What do you man that meaning is "maximized"?

• edited April 2020

Some Zettlers here are describing practices that focus, and even depend, on clearly understanding and explicitly articulating the nature of the relationship between two notes prior to (or in the moment of) establishing that relationship. @sfast you seem to be saying that without that articulation, the relationship is effectively meaningless—or, to use another word, valueless. But the establishment of a relationship between two notes (by whatever means) is not a random, meaningless, valueless act. It is something done based on the understanding (perception, intuition) that there exists a valuable, meaningful relationship between those two notes. So the existence of a valuable meaningful relationship between two notes precedes (and is therefore not dependent on) the articulation of the nature, meaning, value of that relationship.

Luhmann numbers are simply a method of encoding/establishing the existence of a relationship prior to the clear articulation of the precise nature of that relationship. In other words, Luhmann numbers are a non-semantic mode of indicating the existence of a relationship.

In contrast, Structure Notes are a semantic mode of establishing/encoding the existence of relationships between notes. They require one to understand and be able to articulate the nature of a relationship prior to the establishment (encoding) of that relationship.

Yes, you can use direct links in notes to relate one note to another, but as I have described elsewhere, this relationship is encoded only in the note itself. When using Luhmann numbers, the relationship gets encoded (non-semantically) in the title and therefore in the index. When using Structure Notes, one has to work to understand and be able to articulate the nature of the relationship between the notes prior to its inclusion in a structure (web of relations) that is indicated somewhere outside of itself.

In some cases, Structure Notes work very well, either because one already understands the nature of the relationship between two (or more) notes, or because the nature of the relationship between two notes is clear. The latter is the case when notes are related hierarchically—when note 2 is an example of what is described in note 1.

But using Luhmann numbers allows one to build a structure (web of relations) before one understands (or can clearly articulate) the exact nature of the relationship between the notes—when the nature of the relationship between notes (or between the phenomena you are studying) is highly complex, yet to be defined, described.

• ^^^^^ THIS is a different way saying of what I am saying in this thread (not sure how to link to particular comments and too long to quote): https://forum.zettelkasten.de/discussion/928/hierarchical-branched-note-taking-and-the-archive-app-is-topography-important#latest.

My work is figuring out how to articulate that relationship between two notes or ideas, but to do this I need the context of a ton of notes. Thus, I need a way of articulating possible relationships before I can articulate explicitly what those relationships are or how they work. The structure note front loads this cognitive task or skips it by operating on assumptions or on conclusions from other work/notes as opposed to the ones internal to the structure note. Of course, you can let the internal structure of the notes change as you process them further and add new notes, but this can get untenable quick depending on how many notes you are dealing with.

To referentially lead my way out of this labyrinth, I think it is worth unpacking what Luhmann did differently in his second zettelkasten.

Of course, don't do something just because Luhmann did it. Who cares about Luhmann? All I'm saying, is we too quickly assume that the only reason he did what he did was because of physical limitations. I think this is too simple of an assumption.

Further I want to highlight @ctietze comment about how what we are used to interface with - what kind of machine we imagine the zettelkasten to be - really shapes how we talk about it here. I think there are probably points of impasse in our imaginations: is your zettelkasten a "file based system," a "single file system," a set of drawers, a stack of index cards, a forum? In some ways, this is where we are talking past each other.

So much of the zettelkasten is the doing not the thinking about doing. That is not to undermine the discussions here. They are important! This is where I come for inspiration if I get stuck. But it's hard to articulate the particular practical aspects of
a) what one needs to do
b) what one wants to do
c) what one imagines themselves doing.

And that is saying more than: hey whatever floats your boat.
It's also hard to exemplify this because communication with the zettelkasten is not the sum of the notes themselves but how we find them, put them back, and what we do with them.

Also @sfast please keep don't turn down the sfastlish!

• @pseudoevagrius right-clicking the time link at the top of the comment and copy the link will give you the direct link to a comment.

https://forum.zettelkasten.de/discussion/comment/5025/#Comment_5025

Here is the second:

https://forum.zettelkasten.de/discussion/comment/5026/#Comment_5026

right-clicking the time link at the top of the comment and copy the link will give you the direct link to a comment.

Thanks for that, Mike, I was wondering how people were linking to specific comments.

I think of folgezettel as a method I do on top of the other basic practices of zettelkasten taught here.

I follow two key meta principles:
1. I don't delete (even if I change an Id preserve all old id's in the metadata)
2. when I adopt a new method I don't go back and convert all my notes.

Some methods work. Some don't. But if you continue to make links, and if you don't delete then nothing is lost.

Further, I've need my zettelkasten to serve me in different ways in the past so different approaches were necessary. All of these structure notes, folgechains, and folgezettel are still embedded in interrelated links, but they represent different trails of thinking or wandering through the zettelkasten at different years.

With the Luhmann numbers I find I worry less about finding the one exact right note or keeping track of the right note. Sometimes when I clicked on a note I would worry about the notes that I'm not seeing. What related notes did I forget to link here?
I had this anxiety because often enough I would wander to a pocket of notes that for whatever reason (forgetfulness or development in my own thinking) I had not linked to notes that were relevant to me.

When I make a Luhmann number it connects one note to another note. Easy. That's the focus. Just link to one note.
But it also produces a lot of implicit relationships to other notes. Whether these relationships will be relevant or not is up to me to work out later, but what the Luhmann number provides is added context to the note in a systematic way so that it is never lost.

This was the anxiety that grew in me as my zettelkasten grew to a larger size.

That voice in my head saying: what if you aren't putting this in the right place? What if you're not executing this system perfectly with each note? What if you're forgetting something? How would you find it?

Whether true or not, that voice was seriously raining on my parade.

• @cobblepot said:

@sfast said:
And further: It is difficult to make associations explicit and even more difficult to make those associations meaningful because they are inherently not. This is not how associations work. The more creative they are the less meaningful they are from a very early threshold. Meaning is maximised when you find the right "distance" of association.

This section totally lost me. Can you explain what the words explicit, meaningful, and creative mean to you?

• Implicit associations are present when the nature of association is not made explicit. Anything can be associated to everything. Therefore, the mere assertion that something is associated to something other does not inform us because it is true for anything. Explicit associations present you with the nature of the association. Example for implicit association: My right leg has something to do with Donald Trump. Example for explicit association: My right leg has a birth mark that looks like Donald Trump.
• Meaningful means that an association is far enough that it is not downright obvious (I am identical to me. So a Zettel about me that links to itself with this statement would not be meaningful). And it is close enough that it does not slip into randomness. (If you were ever stoned you know what I mean). Anything can be associated with everything. Meaningful associations are useful.
• Creative as in creation: Something new. The more distance between the association the the more new and more creative.

What do you man that meaning is "maximized"?

See above. I can be maximally creative when I create new patterns on a canvas. But if I reduce the element of creation, by re-using known elements, I can draw a picture that has way more meaning.

@argonsnorts said:
Some Zettlers here are describing practices that focus, and even depend, on clearly understanding and explicitly articulating the nature of the relationship between two notes prior to (or in the moment of) establishing that relationship. @sfast you seem to be saying that without that articulation, the relationship is effectively meaningless—or, to use another word, valueless. But the establishment of a relationship between two notes (by whatever means) is not a random, meaningless, valueless act. It is something done based on the understanding (perception, intuition) that there exists a valuable, meaningful relationship between those two notes. So the existence of a valuable meaningful relationship between two notes precedes (and is therefore not dependent on) the articulation of the nature, meaning, value of that relationship.

No. The premise is not true. You can establish a relationship from anything to everything. That makes it effectively meaningless. There is some merit to the assumption that you didn't act just randomly when you placed a Zettel as a Folgezettel (which is what you are doing. You are dealing with the assumed intentions of your past self).

This is the very reason I make sure that every link I place has an explicit context of why I place this link. After a while you have no access to any of your momentary intuitions. And you shouldn't. It is the job of your Zettelkasten and your past self to make those momentary intuitions explicit.

Luhmann numbers are simply a method of encoding/establishing the existence of a relationship prior to the clear articulation of the precise nature of that relationship. In other words, Luhmann numbers are a non-semantic mode of indicating the existence of a relationship.

Yes. Semantics comes from sēmantikós which means "significant". Semantics is the linguistic and philosophical study of meaning in language, programming languages, formal logics, and semiotics. (Wikipedia)

Or: Luhmann-IDs are a method of encoding meaningless (not semantically accessable) relationships. And they are by design! The relationship must be as meaningless as possible to make it as unimportant as possible where you put a Zettel.

In contrast, Structure Notes are a semantic mode of establishing/encoding the existence of relationships between notes. They require one to understand and be able to articulate the nature of a relationship prior to the establishment (encoding) of that relationship.

Not necessarily. You could postpone that step and just place a list under the heading: "Don't know how or if, but perhaps those Zettel could have something to do with this."

Yes, you can use direct links in notes to relate one note to another, but as I have described elsewhere, this relationship is encoded only in the note itself. When using Luhmann numbers, the relationship gets encoded (non-semantically) in the title and therefore in the index.

This is indeed a difference. But why would I want to know the position of a Zettel in a meaningless hierarchy? The Folgezettel is designed to erase hierarchy which you could do by making it invisible.

@pseudoevagrius said:
^^^^^ THIS is a different way saying of what I am saying in this thread (not sure how to link to particular comments and too long to quote): https://forum.zettelkasten.de/discussion/928/hierarchical-branched-note-taking-and-the-archive-app-is-topography-important#latest.

My work is figuring out how to articulate that relationship between two notes or ideas, but to do this I need the context of a ton of notes. Thus, I need a way of articulating possible relationships before I can articulate explicitly what those relationships are or how they work. The structure note front loads this cognitive task or skips it by operating on assumptions or on conclusions from other work/notes as opposed to the ones internal to the structure note.

Structure Zettel are a tool for the exact thing you are describing: To figure things out.

Of course, you can let the internal structure of the notes change as you process them further and add new notes, but this can get untenable quick depending on how many notes you are dealing with.

No. The scope of the Zettelkasten does not influence the usefulness. One of the main points is the self-scaling nature of the Zettelkasten and the Structure Zettel are a tool to allow for this self-scaling.

Of course, don't do something just because Luhmann did it. Who cares about Luhmann? All I'm saying, is we too quickly assume that the only reason he did what he did was because of physical limitations. I think this is too simple of an assumption.

It is not an assumption. He explained it himself. Not that it is because of the physical nature of his Zettelkasten but what he aimed to do with his decisions. He said: It does not matter where I put a Zettel as long as I can refer to it. The time-stamp ID is just the final consequence of this.

But I think this is what the Folgezettel seem to deliver very good:

With the Luhmann numbers I find I worry less about finding the one exact right note or keeping track of the right note. Sometimes when I clicked on a note I would worry about the notes that I'm not seeing. What related notes did I forget to link here?
I had this anxiety because often enough I would wander to a pocket of notes that for whatever reason (forgetfulness or development in my own thinking) I had not linked to notes that were relevant to me.

When I make a Luhmann number it connects one note to another note. Easy. That's the focus. Just link to one note.
But it also produces a lot of implicit relationships to other notes. Whether these relationships will be relevant or not is up to me to work out later, but what the Luhmann number provides is added context to the note in a systematic way so that it is never lost.

This was the anxiety that grew in me as my zettelkasten grew to a larger size.

That voice in my head saying: what if you aren't putting this in the right place? What if you're not executing this system perfectly with each note? What if you're forgetting something? How would you find it?

Whether true or not, that voice was seriously raining on my parade.

My suspicion is that this is a main driver for the Folgezettel-Technique. There is a common theme in this domain on note-taking, Zettelkasten etc. It is FOMO, feeling overwhelmed and similar things. And quite a lot of advice is based on the emotional load due to the uncertainty. Therefore, a need for order and control arises and the Folgezettel-Technique satisfies this need.

I think most of us remember when Evernote came to the market and many people were excited about it: Finally, you could snip and store and don't worry about losing any information. It was no wonder that back then Christian's article on the Collector's Fallacy got traction. It pointed to a problem that is always present in the field: You need to trust that your current self is not a sabotaging past-self to you future-self. And the collector's fallacy is one of the failings of this relationship.

What matters in the future is the meaningful enrichment of the Zettelkasten. Christian was more mindless than I was and has loads of waste in his Zettelkasten. Meaningsless links, uncomprehensible Zettel, unprocessed material and more. I was way more careful and still have wasted quite some time with such low-effort techniques like the Folgezettel (I am talking from experience). That is a re-occuring theme.

Luhmann was no exeption by the way. His style is exeptionally bad in many texts. I think it is due is practice of being very idiosyncratic. Don't get me wrong: He was quite a genious and listen to him is really fun and interesting. But his writing is as bad as my English (which says something). He could be compared to Kant who is a really, REALLY bad writer (and moaned a lot about it and was jealous of Hume's writing ability.

This is always the danger when you apply techniques that allow you to do something later. This afternoon I am FINALLY (sorry guys for the delay) doing the Zettelkasten Coaching (A LOT more difficult in English for me). I some of my points will be cleared up.

I am a Zettler

• @sfast the relationships of objects in a series is not trivial. A sign does not have have meaning outside of a series. It derives meaning from its similarity/difference to other signs (Eco would call them segments) in that series. [see Ducrot, Qu'est-ce aue le structuralisme?; or Umberto Eco, A Theory of Semitoics; or Ferdinand de Saussure, Cours de linguistique générale; or Guattari, Anti-oedipus Papers; or Levi-Strauss, The Raw and the Cooked; or Augustine, De doctrina Christiana]. Theoretically, I'm not sure how this practice is meaningless, and practically (my own experience) I can definitely say it is not meaningless.

In fact, naming a relation, commenting on its value, only adds to this continuum of signification, which is how structure notes can get so confusing. This is also how structure notes can be useful too.

• I acknowledge @sfast Sascha’s comments above. But kicking the folgezettel can around some more I feel has merit. In support of @pseudoevagrius comments above, I offer the following.

Outlines, nonlinear networks, hierarchal networks, any network, folgezettels, folgechains, Luhmann's numbering scheme all have something in common, nodes with node ids.

When these nodes are decentralized and residing who knows where willy nilly in our zettel boxes, they wait to be gathered together into some order or view. Reaching into the zettel box and grabbing and placing these nodes onto a virtual desktop into our view and arranging them in some sequence for thinking interaction is where the money hides. Returning to the view in the future is an essential aspect of the Zettelkasten Method.

Structured notes are centralized views. To return to the view, you find and open the structured note. There is everything just as you left it in the past, all the nodes gathered together. All the thinking and context you created is there to revisit and to pick up where you left off.

Decentralized views, virtual views, just in time views, can exist as well. We have these now with UID links, tags, save searches, save search links. These are all networks with nodes as well, but they lack sequencing, order, and provide only weak context (options may vary about this weakness). Sometimes although a view is fun, cool, exciting, they can also not be beneficial.

Folgezettels and folgechains are or can be decentralized nodes of outlines, nonlinear networks, hierarchal networks, or any network. For them to be useful, we have to gather them together with their sequencing into a helpful view. What @argonsnorts and @pseudoevagrius have done is shown us a way to do this using KB Maestro macro and a Hazel macro. At this time, The Archive doesn't provide an automated way to centralize these nodes into a meaningful virtual dynamic view. Including the Folgezettel node id in the title, then using the Omnibar to search the folgechains and walk up and down the chain is as close as we come as of today. But that lacks the sequencing of the nodes to completely useful.

@argonsnorts and @pseudoevagrius found a workaround for this using KB Maestro and Hazel.
They move the usefulness of this approach beyond seeing the Folgezettel node id's in the note list when included in the title/filename of the zettel.

Useful how? @argonsnorts found relief from having to take a top-down approach of building structured notes. He can whittle out these decentralize outlines, networks, as he interacts with his zettel box. It is his lightweight bottom-up approach to having structured notes without having to have structured notes.

@pseudoevagrius takes it a step further. Using the folgechains and Hazel, @pseudoevagrius can extract first drafts for writing projects with zettels ordered and sequenced based on how he assigned the mode ids in the folgechain. Instead of using a strict Luhmann numbering scheme, @pseudoevagrius assigned nicknames, textual abbreviations, for what the folgechain was about as the first node of the chain—combining a tagging and sequencing into the folgechain notional scheme.

Combining @argonsnorts and @pseudoevagrius approaches could be a means of creating a virtual structure notes, visualizing (working) outlines, thought networks, lightweight and dynamic highlighting marginal inside our zettel boxes.

Also, zettels can be apart of an outline or network, multiple outlines, or networks by merely assigning a zettel to different node ids that are part of different outlines, networks. There are no limitations here.

These node id's, folgezettel ID's resenting some else as well. They are cues, markers, representation of thought work you applied, or accomplished while using and working with your zettels. They are audit trails tagging for thinking work you have accomplished or that you have in progress.

I think we have collectively thought through most of the aspects and design of using Folgechains. Also, we have two working prototypes following @argonsnortss and @pseudoevagrius approaches. There may be more work to do to help lift folgezettels and folgechains from the bottom shelf of the toolbox to top with structured notes.

First, I think the mode id scheme would benefit from a more generally and widely accept convention. Not sure how that will look. We have Luhmann's schema as a guide. I like @argonsorts, and @pseudoevagrius, the nickname textual tagging addition to the scheme has additional potential. More exploration in this area would be beneficial.

Second, realizing these virtual views of the decentralized nodes within The Archive, I think, would be the Holy Grail. If and until then, more experimentation with KB Maestro and Hazel and potential other scripting. Maybe even some resourceful young or not so young developers might build an app we can put in the external editor open list.

Things I would like to see more discussion and experimentation about folgezettels are -

• Node-id scheme convention becomes generally accepted and easily understood
• An easy way to quickly open and see a useful view or views of from any node id in a Folgezettel

If we had these things, then I feel it would greatly help with @cobblepot struggles. He could think in terms of these virtual outlines-networks and using these when-they-become-available conventions and tools and find flow.

• I've said it before and will say it again: I think @MikeBraddock understands me better than I do.

He also is introducing to me a really useful concept for this conversation: "virtual outline."

Correct this definition if I'm off. This is a kind of just-in-time outline summoned from a search. Compare virtual outlines to traditional outlines (for drafts) and structure notes.

• edited April 2020

@pseudoevagrius the force is with us!

Yep we are getting on the same page and the page is getting clearer.

I think when implementing these ideas becomes as easy as assign a UID, slap double brackets around it, and let the link loose end drag, many will steal these ideas and make them their own.

My sense we don't have a a complete toolset yet.

Right now there is friction in how to use this which is slowing acceptance.

• I would like to add some additional thoughts to the conversation.

While the deep-thinking, yay-saying, nay-saying, opinions, insights, suggestions, complaints, comments keep rolling along (the more, the better), I would like to request input from another type of resource that exists here in the forum, developers.

My feeling is that these concepts and ideas have surfaced enough and have enough clarity to begin thinking about them in terms of software solutions. I ask developers to start thinking about how these ideas get implemented in the context of the following constraints:

1. Stay with our current plain text format.
2. Stay within the principles of the Zettelkasten Method.
3. Easily invokable from The Archive either via existing tools that complement and extend The Archive’s functionality or
4. The Holy Grail from directly as part of The Archive’s current framework, functionality, or possibly via a future framework or functionality.

I know coming up in the Road Map is scripting. Maybe as part of that.

Or can someone figure out ways to do this within The Archive now, with the existing toolset? It maybe is possible using the right conventions and possibly search-links. Thinking about it like this helped me to imagine how this might work.

As I said, we have two prototypes as starting points. Can we enhance those to make them easier to flow and easy to use and beneficial for various use cases?

Great things happen in the forums here - it could work!

• I know I'm late to the party here, but I've been following this thread with great interest and I think there's a benefit to the Luhmann-IDs that has been hinted at here (especially by @pseudoevagrius) but has not yet been made explicit. As a disclaimer: I've only ever used date-IDs and at this point I don't think I'm going to switch or add Luhmann-IDs, but I also think I made a lot of errors over the years that I may have avoided if I had started with Luhmann-IDs.

This is a key difference: with Luhmann-IDs, finding a note to link to is a prerequisite of adding a new note to the Zettelkasten (or, if there is no relevant note, determining that there is no relevant note to link to is a prerequisite). With date-IDs, that's not the case.

Asking "what note should this new note follow?" is like asking "what train of thought in my Zettelkasten would this new note contribute to?" With date-IDs, you're not forced to ask yourself that question.

Of course, you can implement a rule that says you ought to look through your archive for relevant notes to link to before adding a new note. And, indeed, Christian has stated this rule. But, it's easy to fall into bad habits, and I think the Luhmann-IDs force you to review your zettels in a more thorough way than tags or even structure notes.

For instance, my current Zettelkasten has notes from the last two years on Heidegger, Merleau-Ponty, Hegel, Plato, Wittgenstein, Brandom, and Aristotle. However, most of these notes exist in their own silos, not connecting with each other. Or, if they do, these connections are tentative or forced. That's because whenever I started reading a new philosopher, I'd think, "well, I don't have anything on Aristotle yet, so I should just start a new note..." I would resist reviewing what was in my Zettelkasten and just dump in new stuff. So, much of my Zettelkasten sits in unconnected fragments and just stagnates. I felt in the back of my mind that this was a problem, that I wasn't reviewing and integrating enough, but I couldn't see a straightforward solution to it at the time. The solution just seemed to be to review the zettels more often. But why would I randomly review zettels when I have more urgent research projects to be working on?

Luhmann-IDs force you to review your zettels each time you add a new note. And not just skimming a list of notes with a relevant tag, or skimming the titles of notes in a structure note, but actually reading your zettels and following their links searching for the most appropriate place to put the new note. And if you review the relevant zettels prior to writing the note, you'll write a note that integrates much better with what is already in the Zettelkasten, and you'll take notes and read with an eye to growing the trains of thought already in your Zettelkasten. This seems to me how we ought to be using the ZK.

Of course, I said I'm not going to switch to Luhmann-IDs. There are still downsides I think. Luhmann-IDs can't unambiguously specify child/parent relations: Luhmann had to use additional footnote-style numbers and letters in red to disambiguate whether, for instance, 1/1b was a child of 1/1a or 1/1. And there might be organizational issues at scale, I'm not sure. But most importantly I think the kind of thinking I described above can be achieved with date-IDs. Date-IDs just don't force you to think that way like Luhmann-IDs do.

My process now, using date-IDs, is to keep a Luhmann-style index of topics with links to particular notes. After I take literature notes on a text, I open my index and ask myself if there are any topics that my literature notes can expand on. If there are, I follow the link. Typically the note linked to from the index will not be relevant enough to my literature notes to expand on with a new note. So I'll follow a few links (reviewing my zettels!) until I find a note that I think is sufficiently related to my literature notes. I'll then write my new note as an expansion or comment on the existing note and add any other links I think are relevant. This ensures a degree of integration in my notes that I wasn't achieving previously. It's also a lot more fun (and reduces the worry and anxiety @pseudoevagrius was talking about!)

Do the advocates of Luhmann-IDs see a greater benefit than what I've described here? If so I'd like to know!

• @Taylor said:
My process now, using date-IDs, is to keep a Luhmann-style index of topics with links to particular notes. After I take literature notes on a text, I open my index and ask myself if there are any topics that my literature notes can expand on. If there are, I follow the link.

This sounds very sensible to me, and like a habit that is worth picking up! In modern terms, Foglezettel nudge you to find another note to branch off of, while date-time (or just sequential number-based) Ids require an extra spark of volition, or habit, to check through the existing material. (I won't spoil details, but the Course and Book got you covered in that direction )

@MikeBraddock : Regarding an implementation of virtual outlines -- that should indeed be scriptable. Even though The Archive does not yet support defining triggers or manual invocation of scripts, the actual script should work just as well on plain text.

As for outlines, I have started years ago with the draft compiler script. That expected a list of notes as input. With the Epstein video series, I added the network traversal & visualization script to my personal toolkit. One variant of the script debug-prints an indented list of related notes. "Relatedness" is defined by forward links. That's essentially a snapshot of the web of notes that can be a virtual outline.

Andy Matuschak's backlink finder supplies the other direction. I don't know if that's useful for a virtual outline, but it's useful for completeness of network visualization.

The problem with the network traversal for e.g. the Epstein notes: every link is treated equally.

That's fine for some (most?) cases, at least early on. But virtual, implicit outlines will eventually give way to purposeful, manual outlines, to distinguish one way to read the web of notes from any other way. Automation can be tweaked to ignore some keywords, for example, but much like full-text search is a very loose tie, and links are very strong ties, automatted graph creation is less meaningful than manual outline creation. You can, but will not always like, an automatically generated draft from all your notes starting at point A, and going X levels deep.

For example, you can take my BA draft notes from Evernote/GitHub. I have an outline in there (aka structure note) to bring a narrative order to the notes. An automation tool can create any order, starting anywhere and traversing the network from there, and may even come up with the exact same outline I have (sans commentary/annotations). But then we end up in a discussion about value and computer generation, aka "Would you think the exact Mona Lisa painting is worth anything if it was painted by a monkey?", or a variant of "If a computer generated enough colorful pixel images, one will eventually look just like the Mona Lisa; is that worth anything?" (The usual answer is: no, because the computer lacks intention, and we generally assume the monkey does, too.) So automation's use breaks down, eventually.

So what can we do with virtual outlines instead? It can become an intermediary, the input to a visualization tool, that shows a network of connections. I like visualizations for their ability to bring a lot of things to the table at once through their two-dimensional nature and the use of distance and proximity -- which a sequential text outline cannot offer.

Author at Zettelkasten.de • https://christiantietze.de/

• I think @Taylor is right in his assessment, that using Luhman-IDs in the title (and forcing their usage) forces a good habit that you should have. The question really comes down to "Is forcing that habit worth the tradeoffs for putting that hierarchy in the title statically"? For those links, the main Luhman links, are but one way to put the notes together, and we all admit that linking the notes in other ways besides this should be a goal as well. Since we already have to put more links to other areas/notes to get more "use" from the ideas, why not make the original Luhman-ID-inspired hierarchy those same kinds of links?

If the problem is one of "ease of presentation", that could be solved by a tool, or by an (admittedly longer) manual practice. Having that one hierarchy always present, while handy, can also hinder you somewhat with how you connect notes (which itself could be fixed with better habits).

I think in the end, we all agree to the core ideas: notes should be linked or not left dangling, and being able to mix and match different sets of notes for different purposes is highly desirable if not almost required for "deeper thinking". After that, the rest of this comes down to questions of implementations and tradeoffs between them. It's an area that could use more examples/discussion for sure, as long as we are aware that's what we are doing.

• @ctietze obviously you have spent lots of time in this subject. Thank you for the feedback.

I know you are busy to spend additional time on this. The upcoming class might indeed clear all this up. Where and when can I sign up?

I need to think through what you provided and ponder. Thanks for the feedback.

@Taylor I see additional benefit, not necessarily greater benefit. I see the two approaches as complementary.

• edited April 2020

@ctietze : I am addressing this to you, but with the others in mind

Regarding your indented list of related notes and the idea of a virtual outline, one of the features of markdown is 6 different levels of headers, indicated by the number of hashes.
So, for example, # Text is a header 1, ## Text is a header 2, and so on.

I don't use The Archive but downloaded a version for trial. From what I can see it is possible to use the hashed titles; TA does not seem to distinguish between them. On the Welcome note, for example, there are the different hash headers but they all look the same, which made me wonder what their function is. Depending on the type of markdown TA uses, there could be more formatting possibilities that can aid in removing some of the barriers/hurdles discussed here.

For example I know it is possible to write a script such that the app makes a distinction between the headers and form an outline on the basis of the level of headers. (Note: we're talking about markdown.)

In addition, it is possible to get the headers numbered automatically, with any subsequent changes reflected automatically too.

This is what is would look like if TA had an Edit mode, as well as a Read mode that renders the raw syntaxed editable text into pleasantly readable text:

in Edit mode:

# Text header
### A bit more
## Again

## Yep, some more


1. Text header
1.1.1. A bit more
1.2 Again

2.1 Yep, some more


Furthermore, it is even possible to have the indented, numbered outline shown in the list in the left pane.

It seems to me that would contribute to making note inter-relationships clearer. For someone like @Taylor, who is not using Luhmann-IDs, this could be helpful; maybe for struggling @cobblepot too; I am not sure about the others already using Luhmann-IDs.

Some markdown background info:
Markdown cheatsheet
Extended syntax

These don't show the nicely numbered indented headings because that is part of the Read/Rendering mode, i.e. not part of the raw markdown syntax.

Post edited by Peter on
• edited April 2020

@pseudoevagrius They don't have meaning in sequences but in semantic webs.

Besides all of my points: If it measurably works it works. If you'd tell me that your Zettelkasten needs pink color and it boosts your productivity than I'd pull out my pinkest pink. In the end, this is what matters. This is not what I am arguing against.

Ok, small point I need to make:

In fact, naming a relation, commenting on its value, only adds to this continuum of signification, which is how structure notes can get so confusing. This is also how structure notes can be useful too.

If you do this along with using the Folgezettel-Technique I would have taken a different road of criticism. Then, in my opinion, Folgezettel is a subpar technique only. I am arguing against that using Folgezettel to establish a link first and then later or much later uncovering its nature as a valuable practice. I am not arguing against its usefulness in isolation but that its opportunity costs are not justified. You could invest the same time in knowledge work that results in higher meaning accumulation (sorry @cobblepot, I am using the concept as a vague everyday term..).

@MikeBraddock Great diplomat! It is a pitty that I don't have you in my discussions in general to balance my sometimes stubburn disagreeableness..

I take a little break from this topic and let some points sink. I need to switch my brain to German for a couple of days..

Thanks for the great discussion! @all
As long as all of you use both my helper's syndrome will be managable.

As a closing point: It seems that I did quite a bad job to present what Structure Zettel (why is everbody now reverting back to "structure note" which I prefered back then? You are all against me.. conspiring to bend my will and twist my mind until both break.. and my soul will fall into the dark pit named hell.. NO! I will survive!). Many statements on Structure Zettel are inaccurate and its my fault: I wasn't careful and comprehensive enough presenting them.

Perhaps, as a teaser: They are a swiss knife to capture patterns and patterns of patterns. Argumentations instead only arguments, complex models comprised of submodels, capturing abstractions and their relation to the concrete, dealing with emerging strength of evidence etc.

And they provide the workflow that the gestalt of the Zettelkasten will resemble the living to make it more durable:

(Picture C) The book is: https://www.amazon.de/Neuland-Denkens-Frederic-Vester/dp/3423330015/

I am a Zettler

• @sfast you ARE the eight hundred pound gorilla in the room. We, well most of us , love you and your sfastlish. Keep growling.