Analysis of the possible origin, role, and use of folgezettel in Luhmann's method

davecan

Great article from Sean Lawson.

https://www.seanlawson.net/2021/03/the-role-of-folgezettel-in-the-zettelkasten-method/

TLDR:

• Theory: Folgezettel may be a result of Luhmann's exposure to the idea of note slips from other great thinkers combined with his legal training (which in the US emphasizes case briefs (literature notes) and outlines (permanent notes) to build a synthesis understanding)
• Folgezettel are not useless but are not necessary for everyone in all cases, and can be implemented as outlines if needed

My own hunch is that Luhmann had at least some classical humanities education that also emphasizes decomposition and synthesis, but his legal training likely at least reinforced the ideas if it didn't introduce them. Plus his readings on the topics he was interested in would invariably show lines of thought being traced and it is unlikely he would not have noticed that as well.

I do agree with the second major point, that folgezettel can be implemented as outlines and do not require Luhmann-style IDs. When I experimented with those IDs early on I found I was often using them in place of creating inter-note links, relying on the note sequences to establish relationships, which is not really any different than using folders.

IMO outline notes are the way to go for most people. There is an additional approach, however, if the system supports transclusion of sections of a note. A PhD researcher on reddit previously described how he would write down the main ideas in his literature notes with headings like ## authorNameYYYY Idea name here with bullets/quotes/ etc under the heading. In his permanent notes he would have a # Scholarship heading and then transclude in the relevant headings from the various literature notes relevant to that synthesized idea. This is an innovative approach to folgezettel that may be useful for those who require that capability, though it forces dependence on software that supports it. (as opposed to simple links between markdown files, which is more tool-agnostic)

ctietze

Theory: Folgezettel may be a result of Luhmann's exposure to the idea of note slips from other great thinkers combined with his legal training (which in the US emphasizes case briefs (literature notes) and outlines (permanent notes) to build a synthesis understanding)

I panicked at this at first, but Sean actually points out only what he knows about the U.S. system and explicitly states he doesn't know how the German system works. Because Luhmann sure wasn't trained in U.S. law. -- Just wanted to leave this here before the urban myth spreads that Luhmann used Folgezettel because the U.S. way to create outlines in law school influenced him so much

GeoEng51

@davecan or @ctietze or @Will

I've seen several people use the term "transclusion" on this site. I've looked it up in dictionaries and in Wikipedia, but for some reason the explanations I've seen just don't click in my brain. Can someone please explain what transclusion is, in a simple way, and why someone might use it in a ZK?

Will

@davecan said:
Folgezettel are not useless but are not necessary for everyone in all cases, and can be implemented as outlines if needed
... folgezettel can be implemented as outlines and do not require Luhmann-style IDs.
... outline notes are the way to go for most people.

Brainstorm.

My zettelkasten is growing, evolving, a messy outline of ideas. A structure note is also a growing, evolving, messy outline of a subset of those ideas. When creating a structure note, I put ideas one below or above another and group them in the white space. I am implicitly "folgezettelizing".

Sometimes.

I'm lazy, and some structure notes follow the order presented by the author of a book. Some are purely chronological. I feel the best about I have created my own "Idea Index" and have "folgezettelized" the structure.

I'm thinking about this process in a more formal way. We'll see where this goes.

prometheanhindsight

@GeoEng51 said:
@davecan or @ctietze or @Will

I've seen several people use the term "transclusion" on this site. I've looked it up in dictionaries and in Wikipedia, but for some reason the explanations I've seen just don't click in my brain. Can someone please explain what transclusion is, in a simple way, and why someone might use it in a ZK?

Transclusion is essentially where part of a note (or a whole note) is copied into another note, but is still tied to the original note that it comes from. It isn't just a link to the original note, it still is part of that original note but is displayed (and sometimes interact-able) from within the second note.

Roam, for example, does this with references. When you reference a block ID, the contents of that block appear as underlined text. Modifying the original block's contents will simultaneously modify that text in every place that it is referenced. It is transcluded in multiple places. The neat trick that you can do with this is to have a note appear in multiple contexts simultaneously. For example, you can take reading notes while you read, then directly transclude parts of those notes into permanent notes. Or, you can transclude parts of permanent notes into other permanent notes. These transcluded notes act as links between notes, and if you refine the wording of the original note the refinement will be echoed across all of your notes.

I think Obsidian does something similar and allows you to embed parts of notes into other notes. It's a bit rougher, though, since you can't view the embedded text while in edit mode. You can only view its contents when in preview mode.

GeoEng51

@prometheanhindsight

OK; I think I get it. Thank you for the explanation. It sounds a bit wild, like you could go down a rabbit hole very quickly. For example, what if one note transcluded a portion of another note that already contained a transclusion of a third note?

Sascha

I disagree with the article on the following points:

1. Outlining as a tool is necessary and cannot be avoided when you deal with more than just a single entity of knowledge and with a field of knowledge instead if you do it in written form. This comes from the issue of grouping. Cases and instances are grouped by their type. Type and instance can be used in both ways. You can frame it in different ways: Type-Token Analysis. Deduction vs Induction. Categorisation. Generalisation vs Specification. etc.
   • Example: A template for a training regime is a type and the application and individualisation of it means generating instances from the type. Analysing the different training systems and plans results in type generating (a training plan can have more than one type: It can be both a specific body part split and a HIT-Training (opposed to volume training))
2. If focusses just on the specifics of a field. I could use the same evidence to make a case for the universality of general principles of knowledge work (which I do regularly). The general issue stems from the conceptual problems of similarity.
   • How similar am I to a banana? The intuitive answer might be not very much (depending on how you judge me as a person). But if I come from a general perspective I might argue that we are both entities of organic matter which makes us part of a tiny fracture of the whole mass of the universe, therefore making us very similar.
   • How similar are men and women? Very similar on the one side. We are all humans. But we differ genetically as much from each other as we as humans differ from each other from chimps. Of course, one make the case that we are indeed very similar to chimps. But the same evidence could be used to make the case for huge differences. (idealogical and emotional immaturity might cloud the ability of people to explore this topic)
   • Similarity depends on value judgements that are very difficult to make.
3. "However, folgezettel-style outlines are also not required for everyone. Webb, Mills, and many other scholars over the last century or more who have relied on the use of notecards for taking "atomic notes" did quite well for themselves without necessarily putting those notes into ever-growing outlines (or, for what it’s worth, making copious links between individual notes as Luhmann also did)." This might be correct but distracts from the actual issue. The interesting question is the usability of Luhmann's approach. Many scholars are using shit-systems and do quite fine. But how many would be way better at their job if the use a good system. Webb, Mills and the others could have been more productive if they used Luhmann's approach or even less. But this is the question to be asked.
4. The article does not support the claim that the principle of atomicity is the most important. (I don't think he is far of. But there is neither evidence nor arguments to support this claim.)

These are my 2c.

prometheanhindsight

@GeoEng51 said:
@prometheanhindsight

OK; I think I get it. Thank you for the explanation. It sounds a bit wild, like you could go down a rabbit hole very quickly. For example, what if one note transcluded a portion of another note that already contained a transclusion of a third note?

I think it mainly works for Roam because each "note" is a bullet point of text, so transcluding blocks tends to only bring small chunks of text along. In your hypothetical, the third note would show the text of the second note plus the transcluded part of the first note. In practical application, though, I don't think it is common for all of that text to end up in a single block. If a single block starts getting too large and complex, then I would think about how I could split it up. I might, for example, move the first note's transcluded text into a nested block in the second note. Then I can easily transclude the second note's text into the third note with or without the first note's text.

And if the structure starts to get too difficult to parse, then I would probably just opt for a link instead.

zk_1000

I don't understand the role of Luhmanns categories. When using note sequences the preceding note only follows the context of its parent note. There is no reason to assume that this note is, simultaneously, related to the context of the category it is belonging to. It is possible yet could also result in any context whatsoever.

Without them, Luhmann could have simply continued this notes in a single sequence. Instead, he created another kind of organization on the root. He didn't start with a complete set of categories from the beginning. How many did he have from the start? At which point does he create a new one? What kind of problems does he solve with this? Is he trying to reduce the amount of connections between different trees? Probably not - some of his root nodes have very generic titles. Then, maybe the opposite?

Does Luhmann ID follow the pure concept of note sequences or is it a combined technique?

Most of his categories look like topic tags, to me. How does he know he can find a reference in politics but not in political party?

davecan

@GeoEng51 To answer your nesting question, Obsidian supports transclusion of pages, sections (everything under a single heading), and blocks (single paragraph or bullet) up to 10 "layers" or "levels" deep. So in your example it would keep transcluding until it hit the 10th layer and then stop.

Like @prometheanhindsight it's not quite the same as Roam's transclusion because Roam doesn't have the concept of a "document" per se, the app just has you interact with essentially an endless stream of potentially infinitely-nested bullets. You can make some bullets headings, tag them, etc and the bullets under those become the "notes" but they are still all individual bullets and collections of bullets. But its a web app only so you lose control of your notes. Obsidian is like The Archive in that it works on local Markdown files, which means it can't be as "slick" as Roam but you own all the data at all times.