Specificity of "top level" categories
I'm following Bob Doto's advice and using Luhmann's numbering scheme, which in effect imposes a tree structure on top of a more free-form linking structure. My top-level titles are weird:
- 1.5: The brain is expensive.
- 3.3: Republic of Letters letters were often sent multi-hop
- 4.3: Russian fairy tales are constructed from linearly organized event types.
- 7.1: What does the abstraction abstract away?
- 14.1: People hate negativity.
- 15.1: Socialism aimed to change the environment in order to change mankind.
(All my top-level titles – as PDF)
This is reminiscent of nothing so much as Borges' top-level classification of animals, supposedly taken from the (imaginary) Celestial Emporium of Benevolent Knowledge:
- those belonging to the Emperor
- embalmed ones
- trained ones
- suckling pigs
- mermaids (or sirens)
- ...
- those that have just broken the vase
- those that from afar look like flies
I'm wondering if this is typical. I'm happy enough with it, to the point I consistently use that list when I start to decide where to put new notes. I have a keyword index from jargon words to notes, but I don't use it. (It may be relevant that I have only 284 main notes.)
(It may also be relevant that I started my Zettelkasten mostly by taking my unfinished drafts, extracting their claims, and loading them into the Zettelkasten. But when I take notes on other people's texts, I still tend to put highest whatever's first in my Inbox today. Maybe I should scan my Inbox and do the most abstract claims first?)
Howdy, Stranger!
Comments
I think they are fairly normal, especially in an early stage.
Some of my titles:
@Mauro Thank you.
I've been experimenting with Folgezettel (essentially the numbering scheme Bob Doto introduces in his book) for a while.
The biggest frustration with this method is organizing top-level categories. Although it claims to be non-hierarchical, it inevitably introduces an order that feels like a hierarchy, no matter what you call it.
One strength of Folgezettel, as presented in Bob's book, is that it naturally lets relationships emerge. As you create (child) notes, those at the same level of numbering depth effectively cluster around an emergent topic.
This works well, except for top-level categories. You increment the most significant digit when you encounter a topic unrelated to existing notes, starting a new tree. Later, as your Zettelkasten matures, you might realize a better position for that note based on connections that have developed since it was written. The temptation to refile the note arises, but actually doing so creates all sorts of problems.
This complexity makes me sympathize with the view that Folgezettel is just a distraction.
@marick At the beginning, it is weird that way. Thanks for the funny Borges reference, I never connected these
But it won't matter much in the long run as the heterarchy grows and the placement becomes less important. You need to start somewhere, so there's no avoiding weird 'top level' notes for a while, until 'top level' stops being associated with "things that were there first". (I don't use positional, hierarchical ID's for many reasons discussed on this forum.)
Author at Zettelkasten.de • https://christiantietze.de/
bob doto in his book absolutely doesn't suggest use of "semantic hierarchy" but a "train of idea sequence style. I took some time to understand this point, hierarchy isn't semanticaly relevant. Hub and structure note are sematically relevant in bob doto zettelkasten declination
Maurizio Boriani
GPG key: 0xCC0FBF8F
@marick Is making a different statement from Bob suggesting a hierarchy. To stealman your position a little bit: Bob repeatedly insist that Folgezettel mustn't be seen as some form of hierarchy.
@marick is making a factual statement:
So, while Bob is making an either a possibilistic statement ("doesn't necessary mean"), a normative statement ("mustn't be used for"), or (which would be the appropriate case) both to make his point, the other position is that Folgezettel does in fact impose a hierarchical structure.
You may read this article to explore the relationship of Folgezettel and hierarchy: https://zettelkasten.de/posts/understanding-hierarchy-translating-folgezettel/
I am a Zettler
When I was referring to tree structure, I meant no more than that:
A hierarchy, I think, implies something more: that there are rules determining where a new note should go – rules more rigid than "it seemed like an interesting linkage to make at the time."
I like using the
1.3a3bstyle as a failsafe measure. Some times I'll put notes in the notebox without adding enough links. Yes, over time, encountering that note again, I'll adjust the links on it. In the meantime, the tree structure means I'll (almost) always have implicit links to neighbors that stand a good-enough chance of being interestingly related to the note I'm looking at.And a habit of running through my top levels before I decide where to put the note helps keep me from lazily making the first link on the new note something obvious.
I've made too many mistakes in my life to depend on myself always doing my best. "Suspenders and belt," we say in the States.
@zettelsan couldn't you just copy/paste the existing note into the new, more appropriate spot, delete the contents of the note in the old spot and place a link to the new spot?
I believe Luhmann created top level categories at some point with his department overview. Seems like that is what is needed.
My intuition says that folgezettel is more so just a distraction more because it solved a problem with the limitations of physical systems. With digital systems, it is every easy to structure sets of ideas on the fly, so an elaborate numbering system isn't needed.
This is a good way to deal with the situation especially with analog zettelkastens. With digital zettelkastens, you have an option to change note ID and make global edits on all the links referencing the obsolete ID. I think that’s personal aesthetics.
While I can sympathize with the view that folgezettels are just a distraction, they do have semantic benefits in that folgezettels with the same prefix from a cluster of notes around an “emergent” topic that you didn’t explicitly attempt to create. You can achieve the same thing with well-maintained structure notes, but with folgezettels, prefix search is an alternative way and can help you with serendipitous discovery. And serendipity can result in interesting connections.
Once you realize that ideas in a Zettelkasten are meant to exist in a heterarchy, you will not worry so much about trees, hierarchies, topics and categories. The numbering is a convention to find the idea again. Numbering also allows local fitting-placing a new note behind the one it most relates to.
All this talk about the purpose of folgezettel seems tangential to @marick’s original request for advice but @Sascha’s article on structure notes does much to settle the questions about folgezettel and what's referenced in Bob Doto’s book. But the recourse to structure notes is appropriate and justifies why talk about folgezettel and order is being raised.
I’m not sure where the phrase “semantic hierarchy” originated from in this discussion but I get how folgezettel may be interpreted as one depending on how you make sense of the word “semantic” relative to how folgezettel is supposed to work.
The ID is supposed to signify something...just not in relation to actual order, but proximity. Sascha’s term “meaningless hierarchy” addresses this well.
I hope the passages that come before the quote above also support what I’m trying to say...
Perhaps some of us are misinterpreting what @marick meant when he said that Bob Doto referred to folgezettel as a tree structure “in effect”. And it is one. Just not for most of our purposes. Luhmann’s use of folgezettel can be represented as a tree structure but that wasn’t his intention. The purpose of Niklas Luhmann Archive is to show off just that, his archive—not his method. Doto refers to this case as the “tree structure fallacy”, but only from the perspective of notetakers and not the note archivists of Luhmann’s collection.
Folgezettel is not meant to portray an ordered procession of ideas. It’s just a useful way to approximate the location of a note and like @marick says it also affords you with the healthy friction of having to think about where a note goes and what it relates to as you compose it. It’s an approach analogous to how we think. A swing in the thinking environment for those interested in having one there.
However the sheer assembly of notes possessing these “semantic identifiers” portraying “meaningless hierarchy” and even hyperlinks in a note’s body do not satisfy the need to make sense out of all the atoms in your zettelkasten beyond what you may have arrived at within the contents of the individual note.
Amalgamating atomic notes is a necessary practice in the slipbox. So is building structure and sense out of them. This is where the following article becomes useful: https://zettelkasten.de/posts/three-layers-structure-zettelkasten/
Emphasis my own.
The only thing I would provoke Sascha about is the amount of pencil and rubber Luhmann would run through in devising his own structure notes. Could he have expanded on a method similar to John Locke's? Or was the excess pencil shaving and erasing spent on his second zettelkasten?
Anyway, @marick I’m not sure if the lists you shared all belong to a single document or not. But you might as well break them up into structure notes if that’s the case. Then yes, I’d find what you’ve put together so far to be “typical” and you may find yourself doing away with the keyword index altogether.
HTH
For that, I'd rather use a Zettel's approach: Using small slips that together would be what is the structure note.
I am a Zettler
This is correct, but you have to get relaxed about what a "hierarchy" means. It does not need to be rigid and pre-defined. Luhmann's numbering system is exactly the same as is used for numbering most technical articles and papers. In a paper, you can stick in subdivisions wherever you like and they don't have to conform to some strict classification. They can be about whatever they need to be. In practice, a subsection will have some relation to its parent section but that relation doesn't have to be well-defined and can differ between siblings of the same parent.
Luhmann had to do things this way because of the severe constraints of using a paper-based system. He couldn't renumber any cards in case some other cards pointed to them, so yes, the system became rigid. If he could have had an integrated, flexible outline that could grow and change with him, he could have decoupled card identifiers from their physical locations.
I have no doubt that he would have used such an outline if he could have had one.
Yes. For those with digital systems, I think that Folgezettel ID scheme violates separation of concern, introducing unnecessary complications. Note IDs, first and foremost, are for uniquely identifying/locating notes. Trying to assign more meaning to them simply isn't a good idea, if notes are considered to constitute a database system.
Folgezettel can be a neat way to organize notes if you must write and store notes physically. But digital Zettelkastenisticians have other, more convenient and effective ways to achieve similar goals. You can stick with Folgezettel to get its (traditional) benefits, but you'd also inherit its limitations.
Comparable yes, "exactly the same" no. A paper still needs to maintain some overall coherence and it has an end. Luhmann's system is open-ended.
Luhmann was facing the same constraints as everybody else. But he used a different technique than everybody else.
He could have done so. But he decided against it.
A paper that hasn't been finished is just as open-ended. You can add on new higher level sections and insert new subsections where ever you want. A paper is coherent so far as the author wants it to be and is able to make it that way.
He could not have done because there was no way a paper outline could have been flexible and changeable enough to work with for his purposes. And if he had somehow been able to change it enough, he wouldn't have been able to find his cards. It would have been as if one put the back-of-the-book index for a first edition into vastly expanded second edition. Nothing in the index would match up any more.
One can just think of Luhmann's Zettelkasten (or anyone's) as a very large book in progress.
You were talking about technical articles and papers. They have a limited scope. They have a certain structure. And even an unfinished paper is being written with the intention of being finished one day.
We could study how Luhmann outlined his manuscripts. However, I don't have enough information available to do so.
Depends on what you mean by "his purposes". Could you expand on that?
You don't need an index to find zettels. You don't need numbers to organize zettels. As far as I know, the vast majority of personal slip boxes in the last 250 years had neither numbers nor indexes.
John Locke didn't invent the index for a slip box, but for books, where the pages are in a fixed order.
I also think of Luhmann's Zettelkasten as a book. But I do think of it more like a commonplace book that can grow and branch indefinitely. The idea of "progress" doesn't make sense for me in that context. Progress toward what goal?
The idea here is that organizing your personal notes and writing an outline for a published work are different tasks.
I'm saying Luhmann's numbering system is just like that. I'm not saying anything about his intent or scope.
Like all of us, he had to find cards and find related cards. There are various ways to find "related" items, and point-to-point links are only one way and not necessarily the best way.
If you don't want to riffle through hundreds or thousands of cards you need an index or some other starting point. You want to find something you wrote two years ago on "Equality and Identity". How can you find it? In Luhmann's collection, much of that material had ID numbers 6.3xx. If you didn't remember that, you couldn't get started. You need an index. Maybe you can remember all your index headings but I can't.
Card catalogs have been in very active development over, say, the last century. How to find "works" and "documents" is the subject of Library Science and there is good information to be learned from it. A zettel can certainly be considered to be a document.
I used the phrase "in progress" with the common meaning of "being worked on".
Aside from the fact that there can be an overlap, one for sure needs to use some kind of organizing principles. There is no one right answer, but as I said, library science has something to offer.
Johannes Schmidt writes:
Have you heard about alphabetical or systematic ordering? They have been around for at least 250 years. (I'm currently reading a zettelkasten tutorial from 1773 that explicitly mentions such techniques.) :-)
Why did Luhmann not use such techniques, considering that so many of his contemporaries did?
I can't either. That's why I use digital tools. :-)
The first card catalogs for libraries have been developed in the late 18th century, for example in Vienna.
Digital technology obviously added new possibilities.
At some level yes. I treat zettels from Luhmann's zettelkasten as source documents like any other source.
No doubt! There's a long history that proves it, with Melville Dewey being a better known example.
But for some reason more recent developments like semantic technologies haven't been widely adopted in personal note-making. Maybe it's a lack of tools? I wouldn't know how to apply TM or RDF in real life. I use some very basic semantic linking in Obsidian, but that's about it.
Exactly. Digital tools make searches so much easier. You can run a search and hope that the card you want is linked to one of the cards you start with based on the search. But you might not be able to find it that way. In that case you (or at least I) want to find it because my target is somehow similar to ones that you can find. That's what subject languages (a library science concept) are for. Subject languages in connection with a ZK are another topic; we can discuss that another time, perhaps.
I call ways in which one can create or find relationships "semantic channels". Point-to-point links are one kind of channel. I have identified at least eight; cards themselves only seem to provide two so far as I can see. How we can get more of them is something else we can talk about another time.
I was thinking in terms of finding documents that one wants. BTW, there can be a lot of value in finding cards that one wasn't looking for, and I want my ZK to be useful for that too. This relates back to some of those semantic channels.
The library science folks have been working diligently on adapting time-honored techniques to the computer era, with mixed results. In some ways the old card catalogs were better, and the indexes written in special books apparently had some advantages over card catalogs.
The technologists seem to have gone off and running without digesting the library science part.
There is a difference between what the library science people are working on and how we use Zettelkastens. But there is an overlap, and I've been trying to make use of it.
Well, three and nine, if you count full-text search. Full-text search is a last resort for me and I hardly ever use it.