Zettelkasten Forum


Folgezettel is More than Mechanism

Here's a longer piece critiquing the mechanistic view of folgezettel.

"Folgezettel is More than Mechanism"

This article intends to:

  1. Show that the current state of the FZ debate is one mired in a mechanistic reading of folgezettel
  2. Add to the discussion the often overlooked experiential aspects of employing folgezettel
  3. Show how the practice of folgezettel positively influences the note making process
  4. Diversify and broaden the discussion in an effort to show that zettelkasten techniques are more compliments than competitors

From the article:

"A mechanistic view is one that privileges the mechanics of a process over the experiential. It ignores the emotional. It tries to establish cohesion often through reductive arguments, determining value (more or less of it) by how parts function together as a system divorced from the subjective experience of the user. As it stands, the folgezettel debate is mired in this mechanistic systems thinking...."

«1

Comments

  • @taurusnoises

    Thanks for this link. I've only had a skim through the article so far, but it looks like it will widen out the discussion considerably, (and I will read it through properly).

    I suspect whether or not you employ folgezettel is a little like buying and operating a car. Everyone is going to find a model that suits them. They will be quite happy with it and tell their friends all about it (and how good it is). Then, after a while, they may find they like another car their friend drives even better, or at least can appreciate it's value rather than disdain her choice :smile: There are lots of excellent cars on the road - some we haven't even seen yet.

  • "There are lots of excellent cars on the road - some we haven't even seen yet."

    Agreed! There's room for all the styles.

  • edited March 2022

    Added to the Folgezettel-Garden.

    I will address your points in a separate article. :)

    Just in short:

    The article I have been referencing is just one of a handful that have been authored by non-folgezettel practitioners who take a top-down, mechanistic view of folgezettel.

    There is a chance that I used the Folgezettel-Technique longer that you yourself. ;) (A decade ago)

    To quote myself:

    I am yet to witness an example of Folgezettel being able to provide or even to establish an even remotely as rich context.

    I didn't see any example or demonstration in your article.

    I am a Zettler

  • edited March 2022

    @taurusnoises Thanks for adding the psychological dimension to this. As you point out, it's possible to have both Folgezettel and structure notes (or notes with other kinds of context). Any argument that supports more "context" (the shifting uses of this term figures fallacies of equivocation) in some way, either by adding text to links or within structure notes is a stronger argument for Folgezettel including that context than with Folgezettel without it or with notes linked by timestamp IDs lacking that context.

    The fallacy of equivocation on the word context amounts to sneaking in one meaning of context, which includes the surrounding or adjacent notes visible in a directory or file viewing pane, for example, and ignoring another meaning of context, which is the local context provided by the Folgezettel ID itself, in the mind of someone routinely familiar with them enough to visualize the tree--so thank you again for emphasizing the psychological dimension of this. These various meanings are complementary. Really this should be the subject of psychological research.

    Since so much of Folgezettel debate is an attempt to argue from first principles and from home-grown ideas of "knowledge" that assume that Wissenschaft ist Wissenschaft (no it isn't--subject methodology varies by subject matter) that I have lost patience with the argumentation and leave it to others to pursue. What little interest I have left is to show how various kinds of Folgezettel and related IDs can be implemented in the system I happen to be using.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • @ZettelDistraction

    "As you point out, it's possible to have both Folgezettel and structure notes (or notes with other kinds of context). "

    This is basically the tldr of the entire piece.

    "I have lost patience with the argumentation and leave it to others to pursue. What little interest I have left is to show how various kinds of Folgezettel and related IDs can be implemented in the system I happen to be using."

    Same. At this point, I feel I've got it on record that reductionist takes that attempt to dismiss others' practices are lame. And, have hopefully provided a diversified vision that makes it possible to include more. Now, I can talk about other things, and let the rest deal with it.

  • edited March 2022

    @ZettelDistraction said:
    @taurusnoises Thanks for adding the psychological dimension to this. As you point out, it's possible to have both Folgezettel and structure notes (or notes with other kinds of context).

    Nobody disputed that that you could. The question is if you should.

    Any argument that supports more "context" (the shifting uses of this term figures fallacies of equivocation) in some way, either by adding text to links or within structure notes is a stronger argument for Folgezettel including that context than with Folgezettel without it or with notes linked by timestamp IDs lacking that context.

    How it the use of the term shifting?

    The fallacy of equivocation on the word context amounts to sneaking in one meaning of context, which includes the surrounding or adjacent notes visible in a directory or file viewing pane, for example, and ignoring another meaning of context, which is the local context provided by the Folgezettel ID itself, in the mind of someone routinely familiar with them enough to visualize the tree--so thank you again for emphasizing the psychological dimension of this. These various meanings are complementary. Really this should be the subject of psychological research.

    I am not ignoring this meaning. I reject the assumption that Folgezettel is the best tool to provide the user with cues to create an internal map.

    I am a Zettler

  • @taurusnoises As people of my profession usually do, I'd like to point out a technical problem :) The footnote links are app://obsidian.md/... links that are essentially dead, but footnotes are still available at the bottom. Their link doesn't work either, though. Just in case you're not aware.

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

  • edited March 2022

    @ctietze Aha! I forgot to change those. Great catch!

  • @taurusnoises bravo! Thanks for sharing and writing this. I hold that more openness to the idea of making FZ working in a digital Zettelkasten is a good thing. I’ll be reading your article closely later.

    Scott P. Scheper
    Website | Twitter | Reddit | YouTube

  • I haven't seen the use of the term “mechanistic view” in the discussion of Folgezettel, and I think it’s a useful distinction to make when we discuss the use of the technique.

    From the article:

    the idea that folgezettel "doesn't establish meaningful connections," is an unfortunate misrepresentation of what takes place when a thoughtful user directly engages with the alphanumeric system of identification. The above understanding mistakenly concludes that the meaninglessness of the letters and numbers themselves, and the places where notes fall based on these letters and numbers, eradicates any meaningful connections that can be garnered from engaging with a list of notes whose top-level connections are represented visually.

    This is something I’ve been investigating more lately as I dive into more Folgezettel discourse: the idea of “meaningful” connection seems to be a pain point in our communication to each other on why we do or do not advocate for Folgezettel. When we talk about “meaning,” some people use it to say that there is a lack of explicit semantic connection being made (I believe that this is the argument @Sascha makes, though I could be reading into things too much). Others use it to say that the connections literally have no meaning, which is reductive, as you say.

    Also from the article:

    Folgezettel establishes a "paper trail" of meaning, showing how an idea was interpreted in its infancy, establishing a historical record…

    I do like that Luhmannesque Folgezettel (i.e. Folgezettel that encode connection using IDs) establish a progression of understanding. I'd gander this understanding can then be followed and excavated over time in a manner akin to sedimentary rock. The idea of a “paper trail” particularly stands out to me since we often conceptualize our ZK as a communication partner – the paper trail documents our past understandings and, I’d assume, can help us combat the curse of knowledge.

    Finally:

    Folgezettel allows the note maker to see top-level connections between all their notes without having to pull out a single one, be it a zettel or structure note.

    Just a point of clarification, especially since you mention it a few times throughout the article: what do you mean by “top-level connections?” I think I would be better able to tease out some of the meaning in your argument if I knew, more specifically, what you were articulating here.

  • @iamaustinha said:

    Folgezettel establishes a "paper trail" of meaning, showing how an idea was interpreted in its infancy, establishing a historical record…

    I do like that Luhmannesque Folgezettel (i.e. Folgezettel that encode connection using IDs) establish a progression of understanding.

    I'd like to err on the side of caution when it comes to Luhmann exegesis and inferring from samples of his notes -- If by 'progression' you mean "what came first in the timeline", alright; but there's not necessarily any inherent order of thoughts, or progress as in depth-of-knowledge.

    Counter-point to illustrate

    Luhmann didn't start the ZK as a toddler, so he came fully loaded with ideas when he started his "ZK II", and had to start somewhere. And then add details later to fill in the gaps, even though that doesn't have to imply he understood more by writing things down he already knew.

    The fault is in the Reddits, I believe

    On the web, people sometimes share examples they pull out of their hats to illustrate a point. The limited nature of thought experiments or hypothetical examples like that doesn't always replicate a chaotic result from the real world, but is biased towards an order we as example-writers impose to make them easier to come up with, to understand, etc.

    Examples on e.g. Reddit sometimes (maybe even "often") look too orderly to be true, and thus impress upon the reader that there's a "progression of understanding" (if I may borrow the term) similar to ordering that is sometimes top-down/general-to-specific, sometimes bottom-up/from pieces to abstraction, but seldom, well, organic, where one may jump to conclusions, revise them, connect semi-abstract ideas, then add more ground-level examples that refute old implications, and so on.

    TL;DR

    Ok, rant over -- Bottom line is: "progression of understanding" as a progression in time, okay, but not in an orderly fashion when it comes to the content. I do see of course that both time and depth of understanding correlate, and thus the paper trail can somewhat map both at once. But I'd say only one by necessity (time) and the other incidentally (depth of understanding).

    If that's not what you mean by 'progress', please nevermind 🙃

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

  • edited March 2022

    Finished the article now! Thanks for sharing, @taurusnoises!

    Want to point to a very minor detail in the OG article:

    And yet, although Luhmann preferred to flip through notes without having to pull out each one, Luhmann saw the importance of the quick scan as divorced from the slip box's function, stating that "These are mere externalities, which concern only how easy the card index can be used. They do not concern its functionality." Respectfully, this is where Luhmann and I disagree, for I would argue that scanability is directly related to functionality.

    I believe Luhmann may have been underselling the importance of the quick, scanable, zettelkasten. Ease of use has a direct effect on functionality, whether we define functionality as the ability to make new notes or as the ability to make connections between notes.

    I see where you're coming from and understand your point @taurusnoises but also see a potential to mislead the reader:

    But Luhmann was rather pedantic, one might say, and when he writes 'functionality', he really means it. In a way that works in his theory. So more like 'function' as in: the shared function of a large rock, a chair, and a chaiselonge to sit on.

    (If anyone is academically inclined: check out Talcott Parsons, in whose framework Luhmann thought and wrote in his early career; and related to Parsons, there are notions of 'structural functionalism' with very specific connotations that are only borderline relevant.)

    So the function is not to be pleasant to use, thanks to a nice UX -- try to remove all feelings from your interpretation and reduce it to almost mathematical abstraction, and then you're on the level Luhmann argues :) You don't really disagree on the same level of analysis, it's rather you open another can of worms with same-sounding words, maybe. Instead of merely underselling, I'd say Luhmann is rather ignoring that for the sake of his article.

    Relevant? -- Only if Luhmann exegesis is one's hobby, or maybe to train one's thinking in abstract terms and stay in that realm for longer.

    To immerse oneself in Luhmann's own systems theory is like stepping into a fairy tale, only instead of fairies, all you have is relations of relations, and jokes that aren't funny.


    Edit: forgot this; later in the article:

    Folgezettel has the added benefit of keeping the note maker in touch with the note long enough to situate the note among its peers. [...] How does a timestamp improve on this? The short answer: It doesn't.

    That is almost an example of (correctly!) pointing out functional equivalence 👍 Although it seems you want to establish a FZ-is-better-than--relation in this case, which there's no need to as far as I see.

    The statement "How does a timestamp improve on this? The short answer: It doesn't." should cut both ways, because both Folgezettel and, say, overview notes can have "the added benefit of keeping the note maker in touch with the note long enough to situate the note among its peers."

    Unless you automatically, blindly, and mindlessly append new notes to some list. That doesn't make for a competetive alternative to FZ though.

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

  • edited March 2022

    @iamaustinha

    "Just a point of clarification, especially since you mention it a few times throughout the article: what do you mean by “top-level connections?” I think I would be better able to tease out some of the meaning in your argument if I knew, more specifically, what you were articulating here.

    This is me being a bit lazy, interchanging "top-level" and "initial connections." By "top-level" I'm referring to the FZ referred to in the title that connect the note to those in the immediate vicinity, as opposed to those connections made inside the note itself. Bc notes are atomic, they can link up across topical threads (I know you know this already). So, the FZ connection is the initial connection that was made, shown outside the zettelkasten, or "at the top of the note." Other connections, either expanding on this topical thread or jumping to other threads, are revealed inside the note. You can see an example here, where FZ is at the top of the note, thread vicinity in middle, and "jumps" and/or supporting/conflicting notes toward the bottom:

    Post edited by Sascha on
  • edited March 2022

    @ctietze

    "The statement "How does a timestamp improve on this? The short answer: It doesn't." should cut both ways, because both Folgezettel and, say, overview notes can have "the added benefit of keeping the note maker in touch with the note long enough to situate the note among its peers."

    Thanks for taking the time to read the piece.

    I'd love to hear a way in which automated timestamps force the note maker to stick with their note longer during import. I can't think of any off the top of my head. Defs interested in what you got in that regard.

    Although, rereading what you wrote, you start with timestamps and end with overview notes. So, now I'm not sure what you're saying! 🤣 I could definitely see how an overview note, or the creation of structure notes (et al) would naturally extend the time you found yourself "in the note." But, not sure with timestamps. Illuminate I. 👦🏻

  • @ctietze said:

    @iamaustinha said:

    Folgezettel establishes a "paper trail" of meaning, showing how an idea was interpreted in its infancy, establishing a historical record…

    I do like that Luhmannesque Folgezettel (i.e. Folgezettel that encode connection using IDs) establish a progression of understanding.

    …there's not necessarily any inherent order of thoughts, or progress as in depth-of-knowledge.
    […]
    Bottom line is: "progression of understanding" as a progression in time, okay, but not in an orderly fashion when it comes to the content. I do see of course that both time and depth of understanding correlate, and thus the paper trail can somewhat map both at once. But I'd say only one by necessity (time) and the other incidentally (depth of understanding).

    Completely agree, and thank you for the point of clarification!

  • edited March 2022

    @Sascha said:
    I reject the assumption that Folgezettel is the best tool to provide the user with cues to create an internal map.

    Fortunately, I didn't say it was the best--besides, no one has bothered to define what "the best" means. I will restate my point as an algorithm in which context does not decrease.

    Step 1: do whatever it is you do to furnish a note with context, using timestamps links.
    Step 2: replace the timestamp links with (generalized) FZ, where the (generalized) FZ indicate the spanning tree formed by the original placement of notes in the ZK. All the inter-connections are the same, all the text other than the links is identical. This can be done in Zettlr , following the procedure here.

    Conclusion:
    Unless you have an interesting or non-standard notion of context, the notes now have all of the context your way of doing things has from step 1, with the additional context provided by the FZ.

    Instead of Step 2, you can do a variation:

    Step 2a: replace the timestamp links with combined FZ/timestamp links. This can be done in Zettlr , following the procedure here.

    Now the notes have timestamps and spanning tree information from the FZ, with at least the same context as before and, it's fair to say, the additional context that the FZ provides.

    That's it. Now I must go communicate with my ZK about a homology theory...

    † by (generalized) FZ I refer to the generalized Folgezettel of Folgezettel Formalized in case its necessary or desirable to use IDs that are coordinates of outlines with any degree of nesting, or any number of alternative parallel outlines. All such IDs are definable in Zettlr by changing the defaults. The IDs don't have to indicate the original ordering of notes--they could be the generalized outline numbers of Folgezettel Formalized. In any case they will provide a spanning tree.

    P.S. Meanwhile, elsewhere on the Internet, a mathematician provides a formal grammar for his ID scheme, shows how it's used, and cites Folgezttel Formalized. So there's someone who understood what I was doing and who used a special case (which is nevertheless pretty general). Oh ye of little faith!

    P.P.S. Of course Gödel invented sequence numbers and encodings in formal arithmetic. This is old hat.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • @taurusnoises said:
    @ctietze

    "The statement "How does a timestamp improve on this? The short answer: It doesn't." should cut both ways, because both Folgezettel and, say, overview notes can have "the added benefit of keeping the note maker in touch with the note long enough to situate the note among its peers."

    I'd love to hear a way in which automated timestamps force the note maker to stick with their note longer during import. I can't think of any off the top of my head. Defs interested in what you got in that regard.

    Although, rereading what you wrote, you start with timestamps and end with overview notes. So, now I'm not sure what you're saying! 🤣 I could definitely see how an overview note, or the creation of structure notes (et al) would naturally extend the time you found yourself "in the note." But, not sure with timestamps. Illuminate I. 👦🏻

    Argh, sorry for the confusion -- I edited too hastily and didn't tie things together.

    One fair comparison of actions (as in: the things you do manually) would be:

    non-FZ FZ
    generate timestamp pick a random (!) Folgezettel ID not used yet
    save file save file
    leave desk leave desk
    good luck finding the note later good luck finding the note later

    That'd focus on the mere acts of, let's call it: filing.

    But you're comparing the mindless generation of timestamps to the left in that table with the careful picking of a meaningful location in the existing note sequences aka FZ.

    non-FZ FZ
    (be mindless? :)) spend time finding a good-enough place
    generate timestamp use Folgezettel ID for that place
    save file save file
    leave desk leave desk
    good luck finding the note later find note thanks to the sequence

    Another fair comparison that produces functionally equivalent results (to some extent) would be:

    non-FZ FZ
    spend time finding a good-enough place spend time finding a good-enough place
    generate timestamp + link use Folgezettel ID for that place
    save file save file
    leave desk leave desk
    find note thanks to links find note thanks to the sequence

    The order to the left is not fixed. You can create a note first and link later, or create link to nowhere first and fill int he gap later. But that doesn't matter much.

    Point is -- no matter what one does, one has to be careful. Otherwise results are shitty. You point out in the article that the FZ workflow produces a kind of good stress, so the user doesn't slap a note just anywhere. Which would defeat the purpose of any ZK, of course. Totally agreed there. I do find you pick on a falsely limited understanding of the alternative to FZ, which is the mindless timestamp generation, and to which a similarly mindless, automated free-FZ-ID-finder would be the best comparison, but that'd be an almost un-thinkable option.

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

  • @ctietze

    "But you're comparing the mindless generation of timestamps to the left in that table with the careful picking of a meaningful location in the existing note sequences aka FZ."

    Yeah, cuz that's A. what was referenced in the source article, and B. that's (as far as I know) the dominant use-case scenarios.

    Question: do you organize the notes in your stack according to theme/thread/topic? Or do you let the timestamps organize the notes at that level?

  • edited March 2022

    @taurusnoises said:
    Yeah, cuz that's A. what was referenced in the source article, and B. that's (as far as I know) the dominant use-case scenarios.

    A To which quote are you refering to?
    B No, it isn't.

    Question: do you organize the notes in your stack according to theme/thread/topic? Or do you let the timestamps organize the notes at that level?

    The question is answered in the very article you referenced.

    I am a Zettler

  • edited March 2022

    @taurusnoises We can't advance the discussion in a useful way unless we take care of making the trails transparent. Please supply quotes so we can follow what your interpretation is based on:

    that's A. what was referenced in the source article,

    Which part gave that impression?

    If by 'the source article' you mean Understanding Hierarchy by Translating Folgezettel and Structure Zettel that you mention in your article, I cannot find any advice that says "create IDs and you're good". On the contrary, the summary section says:

    • Folgezettel and Structure Zettel use the form of hierarchy.
    • Folgezettel uses one single hierarchy to eliminate any meaning from hierarchy. Structure Zettel does the same by introducing infinite possible hierarchies.
    • Folgezettel do provide an address via ID and therefore hypertextuality through linking. Structure Zettel don’t, and we rely on time-based IDs.

    (Source: Sascha, in 2020: Understanding Hierarchy by Translating Folgezettel and Structure Zettel)

    The 3rd item mentions Structure Zettel and ID, because you need both: the structure for, well, the structure; and the ID to address the note.

    B. that's (as far as I know) the dominant use-case scenarios.

    I wouldn't know, so I wonder where you picked that up?

    Or, if you want to be strict: That's an empirical argument for which you would need to provide examples to give it any weight.

    If that sounds wild: I do collect observations like that, too, to support my claims (e.g. from code samples in the realm of programming); it can only be sample-based, because nobody is able to scrape the whole internet for an exhaustive list, of course. But it's a start.

    Question: do you organize the notes in your stack according to theme/thread/topic? Or do you let the timestamps organize the notes at that level?

    Of course I spend time to connect notes manually. Notes with an ID but no connections (aka "orphaned notes") are waaaaay less useful. Can be retrieved via full text search, yes, but that's it, no magic of the ZK happening. That's defeating the purpose somewhat. The time part of a time-based ID is only helpful for one use case: to see things at a glance that happened around the same time. I don't want to over-emphasize the inspiration from the Noguchi filing method because I do not access my ~7.5k notes by remembering the date of anything.

    Do I organize them in the structure of a "stack", maybe! Often it's more a tree of things, an outline. Or a pro/contra comparison. Or a table. There's examples aplenty on the forums and on the blog where you see hierarchical lists/outlines with links.

    I'm confused you ask, because this somewhat implies you haven't read my posts?!?!?! 🥲

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

  • @ctietze >"I'm confused you ask, because this somewhat implies you haven't read my posts?!?!?! 🥲"

    Ha, no. I have! I think we're misunderstanding one another.

    When I say "referenced in the source article," I'm referring to the reference to timestamps and automation mentioned in "Understanding Hierarchy..." .

    Re. The list of possible scenarios you referred to re timestamps and fz.... I'm not talking about what a person does after automation. There's a million things a person can do. They could resurrect Luhammn and have him take over!

    I'm talking about the impulse created (or lack thereof) with automation vs having to physically devise fz codes. It's straight forward. I'm not talking about what one does after (as you are). It's about friction vs automation. If you honestly think that being forced to slow down and not being forced to slow down produce the same impulses, and it's just a matter of shuffling the scenarios around, we may be further apart in this matter than is bridge-able.

    When I say, "dominant use case" I'm simply referring to the fact that when people us timestamps, they are usually automated. They aren't pausing to think about the day, time, seconds. Apps automate this process.

    If you have any other questions, happy to jump on a Zoom call and answer. I can't unpack every detail that gets lost in translation through texting.

  • edited March 2022

    @ZettelDistraction said:

    @Sascha said:
    I reject the assumption that Folgezettel is the best tool to provide the user with cues to create an internal map.

    Fortunately, I didn't say it was the best--besides, no one has bothered to define what "the best" means. I will restate my point as an algorithm in which context does not decrease.

    Step 1: do whatever it is you do to furnish a note with context, using timestamps links.
    Step 2: replace the timestamp links with (generalized) FZ, where the (generalized) FZ indicate the spanning tree formed by the original placement of notes in the ZK. All the inter-connections are the same, all the text other than the links is identical. This can be done in Zettlr , following the procedure here.

    Conclusion:
    Unless you have an interesting or non-standard notion of context, the notes now have all of the context your way of doing things has from step 1, with the additional context provided by the FZ.

    Instead of Step 2, you can do a variation:

    Step 2a: replace the timestamp links with combined FZ/timestamp links. This can be done in Zettlr , following the procedure here.

    Now the notes have timestamps and spanning tree information from the FZ, with at least the same context as before and, it's fair to say, the additional context that the FZ provides.

    It is a given that you could use a combination and I don't claim that the context decreases. My claim, however, is that FZ do not provide additional meaningful context to the Zettler. I quote myself:

    I am yet to witness an example of Folgezettel being able to provide or to establish an even remotely as rich context. (slightly edited)

    The Folgezettel technique is designed in a way that the relations established to other notes are not meaningful: Being a parent, a sibling or a child could mean anything and has nothing to do with the content. It can't because if you you tie meaning to the hierarchical position you run into several problems.

    One of the problems is that the whole Zettelkasten hierarchy just has two vectors of meaning: Vertical and horizontical. If you do it right and being inconsistent with the attribution of what each position means you'll run into problems that show their face within (depending on your productivity) a couple of years of keeping on top.(1) If you are consistent, you are limited to two vectors of meaning and have the same problem as categories.

    You will forget and this forgeting is dependent on how your brain works and not how it would be optimal for your work with you Zettelkasten. That is the reason why the use cases for FZ is limited (look up Dan Sheffler for one).

    This is one of the reasons why I run every feature through a infinite scalability test: Everything what I do needs to work if my Zettelkasten had infinite notes. (Apart from technical issues like finity of matter).

    (1): The issue with such problems is its year long delay. There was this phenomenon called Evernote. The problems of the behavioral incentives and poor practices showed up after years of work. The result was high usability in the beginning and either decreasing usability over time, or (even worse) a moment of crisis.

    I am a Zettler

  • edited March 2022

    @Sascha said:

    It is a given that you could use a combination and I don't claim that the context decreases. My claim, however, is that FZ do not provide additional meaningful context to the Zettler. I quote myself:

    I am yet to witness an example of Folgezettel being able to provide or to establish an even remotely as rich context. (slightly edited)

    Remotely as rich a context as what? This is a rhetorical question.

    The Folgezettel technique is designed in a way that the relations established to other notes are not meaningful: Being a parent, a sibling or a child could mean anything and has nothing to do with the content. It can't because if you you tie meaning to the hierarchical position you run into several problems.

    This is absolutely false. Reread Luhmann's Communicating with Slip Boxes, any of Schmidt's papers or even my own notes. The use of Folgezettel comes with a discipline. From Niklas Luhmann‘s Card Index: Thinking Tool, Communication Partner, Publication Machine section 12.2.2 System of Numbering:

    Illustration of the Method of Card Integration and Numbering
    1/1 Card with notes
    1/1a Card containing notes referring to a concept/idea from card 1/1
    1/1b Continuation of notes from card 1/1a
    1/1b1 Card containing notes referring to a concept/idea from card 1/1b
    1/1b2 Continuation of notes from card 1/1b1
    1/2 Continuation of notes from card 1/1

    There's a method in assigning numbers, and it concerns the relation of one note to the next. It's not the only context one can provide. In fact it's implicit in the notation that 1/1a refers to a concept from 1/1, that 1/1b continues a note from 1/1a and so on. That this is obviously more context than only "horizontal and vertical context" is glaringly evident.

    One combines this with the system of references (links) in section 12.2.3. I won't repeat this here. But I will include a table to illustrate the sophistical argumentation.

    Timestamp IDs Folgezettel IDs Generalized Folgezettel IDs
    Links with explanatory text and structure notes 1-way branching links. Augmented link context. 2-way branching links. Augmented link context including contination or comment from ID. n-way branching links. Augmented link context including continuation or comment from ID.
    Links without explanatory text, no structure notes 1-way branching links. Link context (time) from ID only 2-way branching links. Link context (continuation or comment) from ID only n-way branching links. Link context (continuation or comment) from ID only

    Instead of comparing the top lines, you are comparing the upper left with the lower right and lower middle. Not a fair comparison, but now I see a pattern here.

    One of the problems is that the whole Zettelkasten hierarchy just has two vectors of meaning: Vertical and horizontical. If you do it right and being inconsistent with the attribution of what each position means you'll run into problems that show their face within (depending on your productivity) a couple of years of keeping on top.(1) If you are consistent, you are limited to two vectors of meaning and have the same problem as categories.

    This ignores the practice that goes with it--see above. As for horizontal and vertical being the only vectors of meaning, this is utterly false--see above. How did you become appointed the arbiter of admissible interpretations of Folgezettel? Is their a Council of Arbiters of Folgezettel Interpretation? I'll submit my application for consideration.

    You will forget and this forgeting is dependent on how your brain works and not how it would be optimal for your work with you Zettelkasten. That is the reason why the use cases for FZ is limited (look up Dan Sheffler for one).

    That's what Zetteln are for. You write the technique into one of them in case you forget. On the contrary you will forget your technique for adding context, which amounts to adding some explanatory text to links, possibly in a supplementary structure note.

    If you can blithely assert that a user of FZ--even a user doing exactly what you do while adding FZ instead of timestamps-- will forget the technique that Luhmann did over a longer timespan than you allow, and that Luhmann, Schmidt and others describe, then I can assert with equal justification, which is to say none at all, that you will forget just as easily. You write as if you've already forgotten, it's just that I don't believe you.

    This is one of the reasons why I run every feature through a infinite scalability test: Everything what I do needs to work if my Zettelkasten had infinite notes.

    Then you can't count. Eventually the IDs will become arbitrarily long--longer than the non-ID text of your Zettels.

    (Apart from technical issues like finity of matter).

    Then you are imposing a bound and it's not an infinite scalability test. Assuming a large bound is a sensible thing to do, but this is already provided by 12-digit timestamp IDs, which allows for one trillion notes. A 14-digit timestamp allows for 100-trillion notes. Likewise for Folgezettel. That effectively removes ID-length from the discussion of how to organize a ZK.

    "Infinite scalability" is a cool-sounding slogan that doesn't mean what it seems to suggest.

    (1): The issue with such problems is its year long delay. There was this phenomenon called Evernote. The problems of the behavioral incentives and poor practices showed up after years of work. The result was high usability in the beginning and either decreasing usability over time, or (even worse) a moment of crisis.

    OK that's an important empirical finding, at long last! (It could be made more precise, but empirical finding is such a rarity in this discussion that I'll take it.)

    In a sentence: a ZK with your technique of explanatory text around links (let's forget about FZ for the moment), together with structure notes (with the same type of explanatory text) is a sound and workable technique for growing a (digital) Zettlekasten, without the drawbacks that afflict systems like Evernote, etc.

    This is the kind of thing you want to emphasize, instead of arguing minuscule points ad nauseum, not to mention logically invalid points to ensure Zettelkasten Methodological supremacy at any cost.

    Friends don't let friends do this to themselves.

    You can apply the same technique you suggest with FZ instead of timestamps, with no loss of context (and some additional context). I don't have the time or leisure to pursue this further.

    In the end, I think @GeoEng51 is right: it's a matter of taste.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • edited March 2022

    I wish to apologize to @Sascha and to everyone else for taking an unnecessarily adversarial tone in my previous post. I'm going to answer Sascha's comment below with an example: my own ZK. I think Sascha's comment deserves a constructive example.

    @Sascha said:

    I am yet to witness an example of Folgezettel being able to provide or to establish an even remotely as rich context. (slightly edited)

    My answer is in four parts, beginning with a partial listing of my ZK illustrating the combined Folgezettel+timestamp ID I worked out. To make sense of the numbering, I'll include three Zettels.

    1. a comment on my Index with a rough set of categories, with three methods of handling the limitations of categories;
    2. a description of the IDs that I use; and,
    3. a note on how these IDs allow for n-way branching, so they are Generalized Folgezettel (GZ as opposed to FZ). These IDs are easier to work with in practice than the general IDs defined in Folgezettel Formalized.

    Partial File listing

    0.0.22.0306.1900 Comment on the ZK index

    CONTEXT [[0000.0000.0000]] Index

    #index #folgezettel #zkmethod

    Presupposition: an ID definition (timestamps, Lumann-Folgezettel, Generalized Folgezettel) and a procedure for generating and assigning these IDs to Zetteln have been given in advance. See [[0.1.22.0305.1829]] ID format.

    The index [[0000.0000.0000]] is a structure note. It contains a rough classification and a few entry points into the Zettelkasten. The index is developed with the expectation that its rough classification will eventually break down. A breakdown means that the classification will fail to properly classify new notes after a certain point, even if the intention is to add new notes to the Zettelkasten only if they are related to pre-existing notes. At this point, one either gives up; or one uses structure notes to mitigate the limitations of pre-defined categories; or else one makes a virtue of the necessity to recover.

    There are three ways of handling this breakdown.

    1. Give up. This happens with systems such as Evernote, with a delay of months to a year. (Sascha Fast pointed this out.)

    2. Attempt to avoid a breakdown altogether.  Use structure notes with annotated links, and do not use Folgezettel. This is the approach of Sacha Fast, Christian Tietze, Will Simpson etc. This method distributes localized indexes of related Zettels around the Zettelkasten, which postpones but does not completely eliminate the possibility that the localized indexes will fail to accommodate new Zettels, even if existing Zettels (and structure notes) are revised. Will Simpson speaks of refactoring Zettels. In that case, new structure notes are added to the system, or existing structure notes and Zettels undergo refactoring, or both.

    3. Internal ramification. Attempt to recover by attaching a new Zettel not classified by the categories of the index to some related Zettel, either as a child or sibling of the related Zettel. This new Zettel becomes the entry point of a new category. For example, the new Zettel may be a structure note with annotated cross-references (links).  The recovered Zettelkasten improves over an indexed collection of notes with fixed categories. Over time, this process repeats. Starting over with new Zettels takes priority over revision of an existing Zettel (unless the revision is minor and affects only that Zettel). An exception is to update the Index with a new entry point.

    The third method uses Luhmann-Folgezettel (FZ) or Generalized Folgezettel (GZ).

    • FZ. Attempt to find a related note with the Luhmann-Folgezettel ID ID, and either add the new note as a successor to ID if it is a continuation or else add the new note as a sibling of IDif the new note is a comment on ID. If there is no obviously related Zettel, then find a roughly related Zettel (see Internal Ramification above) or begin a new category number and FZID. Add the new Zettel ID as an entry point for the new category to the Index. Add structure notes with annotated links as needed. (Niklas Luhmann, Daniel Ludecke, Johannes F. K. Schmidt)

    • GZ. Use Generalized Folgezettel following the procedure of FZ, prioritizing internal ramification over the addition of new categories. The Generalized Folgezettel are combined FZ with timestamps. This allows for multiple branching by reusing the FZ portion of the ID and appending a later (or different) timestamp! See [[0.1.22.0305.1829]] ID format and [[0.1.22.0306.1951]] Generalized Folgezettel and multiple branching IDs.

    0.1.22.0305.1829 ID format

    CONTEXT [[0000.0000.0000]] Index

    #folgezettel #zettlr #generalizedfolgezettel

    The ID format is a dot-delimited Category Number, Folgezettel ID and TIMESTAMP.

    CategoryNumber.FZID.TIMESTAMP

    In this note, the ID is 0.1.22.0305.1829.

    Note: think of the initial segment of the above ID 0.1 as 0/1 as in Niklas Luhmann's notation. This portion of the ID is the FZ (Folgezettel, by an abuse of language). The dots of the ID divide the ID into places. There are four dots and five places in the ID 0.1.22.0305.1829. The FZ portion of the ID occupies the first and second places of the ID. The remaining places of the ID constitute the TIMESTAMP, in this case 22.0305.1829. The entire ID is an example of Generalized Folgezettel (GZ).

    How does it work?

    In Zettlr,

    • ID generation matches (%Y.%M%D.%h%s).
    • The ID regex is (\d[\w\.]{12,}\d).

    See [[0.1a.22.0222.1123]] Backward-compatible Folgezettel IDs in Zettlr

    CategoryNumber

    This is a rough category specification. In my case

    1. Index, Zettelkasten-related, administrivia
    2. Philosophy and Religion
    3. Computing
    4. Mathematics
    5. Design, Drawing and Art
    6. Linguistics and Foreign Language
    7. Literature
    8. Medical
    9. Behavioral Science, Strategic Interaction, Economics, Law, Politics
    10. Hard Science, Physics, Chemistry, non-strategic interaction

    In this note, the CategoryNumber is 0.

    FZID (Folgezettel ID)

    This is borrowed from J. Schmidt, with the slash / rewritten as a period ., where the first number is a CategoryNumber, and where the FZID starts after the initial period.

    1.1 Card with notes  
    1.1a Card containing notes referring to a concept/idea from card 1.1  
    1.1b Continuation of notes from card 1.1a  
    1.1b1 Card containing notes referring to a concept/idea from card 1.1b  
    1.1b2 Continuation of notes from card 1.1b1  
    1.2 Continuation of notes from card 1.1
    

    In this note, the FZID is 1.

    TIMESTAMP

    This the portion after the CategoryNumber and the FZID. A dotted timestamp, depending on the length of the CategoryNumber + FZID. In this case, the TIMESTAMP is 22.0305.1829.

    See [[0.1.22.0306.1951]] Generalized Folgezettel and multiple branching IDs

    0.1.22.0306.1951 Generalized Folgezettel and multiple branching IDs

    CONTEXT [[0.1.22.0305.1829]] ID format

    #multiplebranching #sibling

    The ID format given in [[0.1.22.0305.1829]] is an example of Generalized Folgezettel. Luhmann Folgezettel (FZ) allow for $(2)$-way branching, either one sibling (across) or one child (down). A Generalized Folgezettel allows for $(n)$-way branching, where there are $(n)$ siblings (across) or child (down). Multiple ($(n)$-way) branching is possible by fixing the CategoryNumber and FZID of an ID and assigning a later (or a different) TIMESTAMP. In this way, revisions and additional comments on aspects of earlier Zettels can be accommodated. In fact, entire sub-hierarchies can be translated to a later TIMESTAMP with this method.

    Notice that this ID 0.1.22.0306.1951 has the same CategoryNumber and FZID as its older sibling [[0.1.22.0305.1829]] ID format. Both IDs are Generalized Folgezettel, and this Zettel is a comment on 0.1.22.0305.1829.

    With Generalized Folgezettel (GZ), the FZID portion of the ID can be thought of as a subsection number, and the TIMESTAMP as a revision of Category.FZID or as a comment on Category.FZID.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • @ZettelDistraction Awesome posts. I will need some time to process them. But the answer is coming.

    I am a Zettler

  • edited March 2022

    @Sascha said:
    @ZettelDistraction Awesome posts. I will need some time to process them. But the answer is coming.

    I'll need to update these, since I soon realized that I needed to separate the Folgezettel portion of the ID from the timestamp. I settled on CategoryNumber.FZID.0.TIMESTAMP. The IDs are longer, but this is a workable system (for me). I clarified the argument the github wiki, but I will add the updated posts here. Let me know if I understand the procedure that you, @ctietze and @Will follow.

    The route I'm taking is to start with some top-level categories with the expectation that they will break down. At that point, you'll end up linking a note somewhere in the middle.

    There is an example in a paper of Johannes F. K. Schmidt illustrating that important notions for Niklas Luhmann ended up in the Zettelkasten somewhere other than the top level.

    Consider, for instance, the cards on autopoiesis, which is one of the key concepts in Luhmann’s more recent social theory. Most of these cards in the second collection are filed under what seems to be a subordinate position in this system bearing the number 21/3d26g1i. The same is true for his notes on his major project of a theory of society, which are located in the same subsection under the number 21/3d27fB.
    -- Johannes F.K. Schmidt. Niklas Luhmann’s Card Index: Thinking Tool, Communication Partner, Publication Machine. In: Forgetting Machines: Knowledge Management Evolution in Early Modern Europe Type: Chapter Pages: 287-311 DOI: https://doi.org/10.1163/9789004325258_014

    I interpret this to mean that not only will the initial top-level categories break down, you want the top-level categories to break down. This will lead to unexpected branching sequences further away from the original notes. The system only gets going once the initial classification into categories collapses in on itself.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • edited March 2022

    @Sascha said:
    @ZettelDistraction Awesome posts. I will need some time to process them. But the answer is coming.

    Previously I revised some of the posts above on a wiki. The wiki included a sample index note with 10 "top-level" categories, each of which was assigned a link within the index to a "category note." Under the link to the category note (within the index) were a few links to representative notes, and each of those notes linked back to the category note. The index and category notes were assigned special identifiers to make them easy to locate. There were also some detailed remarks on workflow.

    Within Zettlr, the notes that link to the currently viewed note are visible in the "Related files" pane. This is how category notes work in Zettlr: the notes assigned a category are provided with a link back to the corresponding category note. By visiting a category note, the notes assigned to that category are visible in the "Related files" pane. The ten categories were chosen with the expectation that they would fail sooner or later to properly classify new notes. The workflow is to find a somewhat related place within the graph of notes, and to continue anyway. The new note can be back-linked from other notes as if it were a top-level category note and it can be linked from the index as another entry point into the Zettelkasten.

    However, after being reminded that "most people who write about note-taking don’t seem particularly accomplished in their own fields," I decided not to document the index and categories, and to cease writing about this subject. Potential exposure to the charge of not having "a serious context of use" is to be avoided. The deleted and cut wiki pages are now Pythagorean secrets.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • edited March 2022

    The Folgezettel technique is designed in a way that the relations established to other notes are not meaningful: Being a parent, a sibling or a child could mean anything and has nothing to do with the content. It can't because if you you tie meaning to the hierarchical position you run into several problems.

    >

    This is absolutely false. Reread Luhmann's Communicating with Slip Boxes, any of Schmidt's papers or even my own notes. The use of Folgezettel comes with a discipline. From Niklas Luhmann‘s Card Index: Thinking Tool, Communication Partner, Publication Machine section 12.2.2 System of Numbering:

    >

    Illustration of the Method of Card Integration and Numbering
    1/1 Card with notes
    1/1a Card containing notes referring to a concept/idea from card 1/1
    1/1b Continuation of notes from card 1/1a
    1/1b1 Card containing notes referring to a concept/idea from card 1/1b
    1/1b2 Continuation of notes from card 1/1b1
    1/2 Continuation of notes from card 1/1

    >

    There's a method in assigning numbers, and it concerns the relation of one note to the next. It's not the only context one can provide. In fact it's implicit in the notation that 1/1a refers to a concept from 1/1, that 1/1b continues a note from 1/1a and so on. That this is obviously more context than only "horizontal and vertical context" is glaringly evident.

    This is illustrating my point: There are only two directions:

    1. Card containing notes referring to a concept/idea from parent note
    2. Continuation of notes from (older) sibling note

    You can add the reverse: A note can be simultaneously be older and younger sibling as well as being parent and child.

    One combines this with the system of references (links) in section 12.2.3. I won't repeat this here. But I will include a table to illustrate the sophistical argumentation.

    >

    Timestamp IDs Folgezettel IDs Generalized Folgezettel IDs
    Links with explanatory text and structure notes 1-way branching links. Augmented link context. 2-way branching links. Augmented link context including contination or comment from ID. n-way branching links. Augmented link context including continuation or comment from ID.
    Links without explanatory text, no structure notes 1-way branching links. Link context (time) from ID only 2-way branching links. Link context (continuation or comment) from ID only n-way branching links. Link context (continuation or comment) from ID only

    >

    Instead of comparing the top lines, you are comparing the upper left with the lower right and lower middle. Not a fair comparison, but now I see a pattern here.

    I don't compare Folgezettel and Timestamps. :) I compare structure (context) provided by Folgezettel and structure provided by structure notes.

    You will forget and this forgeting is dependent on how your brain works and not how it would be optimal for your work with you Zettelkasten. That is the reason why the use cases for FZ is limited (look up Dan Sheffler for one).

    >

    That's what Zetteln are for. You write the technique into one of them in case you forget. On the contrary you will forget your technique for adding context, which amounts to adding some explanatory text to links, possibly in a supplementary structure note.

    I cannot what was never there. There is no technique to the context I provide. But the content of whatever I am doing is explicitly written down. This is the contrast: Context provided by Folgezettel provide that there is connection. Context provided by link context (the surrounding text) provide the how the connection is established.


    @ZettelDistraction said:
    I wish to apologize to @Sascha and to everyone else for taking an unnecessarily adversarial tone in my previous post.

    It's fine. :)

    I am thinking on how to respond to your second post. To me, you are developing a structural layer in which each note has a place. You seem to do it by combining techniques (FZ + Timestamp) that have little to no connection to the content of the note.

    Even both directions of FZ "Being a continuation to" or "referring to a concept of" are highly abstract.

    It is not to deny that FZ provide something and those "somethings" can be called context, connection etc. However, they provide a number that encodes on the first level categories and on the following levels an abstract relationship (sibling, child, parent).

    But this puts a layer between the actual content and what we see. With FZ you see a hierarchical structure or nested list in your file viewer (an analog Zettelkasten would even provide you with that unless you invest quite some labor). But this accomplishes not that much if you compare the alternative. I mean "alternative" as exclusive alternatives since attention is not dividable. You have only one focal point and one only. If I'd combine FZ with context enriched structure notes (see sample image bellow) I either can look at the FZ structure, the structure note or switch between them. But the structure note already provides everything the FZ provides and more.

    One aspect of context is phenomenological: It is what you see in the moment. Because of this, the link context should be next to the link.

    Another aspect of context is how your working memory and your Zettelkasten interact: By viewing a couple of notes you upload a strange mix of details and compressed items into your working memory. Because your brain is inherently pattern seeking it tries to make sense and creates all kinds of connections between those items. If the pattern is not feedbacked directly (you see the pattern in your Zettelkasten) you will have another new idea. Because the working memory is limited in size you want to have exclusively non-arbitrary items and relationships in it. So, by adding arbitrary or arbitrarily encode information to the working memory load you decrease the creative effect (the ZK feels less like a communication partner?) of your Zettelkasten.

    I don' want to exhaust this topic. But you see my point, hopefully.

    Post edited by Sascha on

    I am a Zettler

  • edited March 2022

    @Sascha said:

    This is illustrating my point: There are only two directions:

    The links are provided with adjacent text. I know that your position is that the adjacent text isn't present I add the text, but they are if you add the text. My own setup allows for any number of siblings, and other annotated links are not precluded.

    I don't compare Folgezettel and Timestamps. :) I compare structure (context) provided by Folgezettel and structure provided by structure notes.

    That was in the table, including text surrounding links and even combining folgezttel with structure notes. Why pretend otherwise? :smile: .

    I cannot what was never there. There is no technique to the context I provide. But the content of whatever I am doing is explicitly written down. This is the contrast: Context provided by Folgezettel provide that there is connection. Context provided by link context (the surrounding text) provide the how the connection is established.

    I have surrounding text also :smile: It provides the same function, but for some reason you need to keep denying this. There is no point in continuing this discussion.

    I am thinking on how to respond to your second post. To me, you are developing a structural layer in which each note has a place. You seem to do it by combining techniques (FZ + Timestamp) that have little to no connection to the content of the note.

    LOL. Whatever. The second post is out of date.

    Post edited by ZettelDistraction on

    GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Armchair theorists unite, you have nothing to lose but your meetings! --Phil Edwards

  • Ich stehe auf dem Schlauch.

    I understand that you are insisting that you are doing more than just using FZ. But I don't see your point in that insistence since I don't argue that those techniques are incompatible. I argue that FZ do nothing and should do nothing more than providing an ID instead.

    Basically, my points can be divided into two categories:

    1. Folgezettel do not provide what they claim.
    2. Structure Notes and link context actually provide what FZ claim to provide.

    But by the your insistence that it is no either or, from perspective, you are not adressing the points I am making.

    I am a Zettler

Sign In or Register to comment.