Zettelkasten Forum

Zettelkasten Structure: Luhman vs. Automated Numbering

I know that this has been discussed on the blog and the forums in the past, but I'm interested to see how people's thinking on this question has evolved.

For my notes, I create unique identifiers for each note based on a date string (e.g. 201802150955) . I also use tags. But as I've been reading more deeply into the Zettelkasten system, I realize that Luhman's numbering system was a way to create structure in his notes, i.e. creating lines of thought with branches and sub-branches. This is a built-in feature of Zkn3 but, honestly, I much prefer the plain-text approach pioneered on this blog and in The Archive app.

I'm curious if anyone uses the Luhman-type system for numbering notes or if everyone just uses date strings. If you're using date strings, how do you create structure? Through tags? Do you think there is some qualitative difference between a tag-based system and a one that develops through the linear creation of branches and sub-branches?


  • edited February 15

    I am not sure whether this is an evolution of thought :smile: but I just use date strings and try to keep it simple.

    Adding structure via identifiers is only relevant if you also navigate your Zettelkasten by ID, which I do not.

    I navigate based on search results (by tag, note-id for finding other notes referencing a specific one, citekeys, or just regular search terms) and by links within notes.

    When I want to make clear that one link is the start of a branch, I just remark that in the text and thus separate the branch structure from the standard link structure. No need for a unique tag just to create a branch. It requires going back to the parent note to insert the link, though. Usually, this happens automatically when I create the child note from within the parent note, e.g. via implicit link creation.

    # The note (1)
    tags = #topic 
    Text goes here. Interesting idea. 
    <!-- Folgezettel -->
    [[20180102030405]] Expanding on this great idea (1a)
    [[20180102030406]] Second element of the branch (1b)
    <!-- See also -->
    [[20180203040507]] Another great idea (2)
    [[20180203040508]] And another one (3)
    [[20180203040509]] Maybe a good idea, too (4)

    Note that I added the (1), (1a) references just for clarity. I also do not go to the length of providing a link pointing back to the parent from a child note, and I also don't place links from (1a) to (1b) nor would I update the latter to (1a1) when branching off from (1a), just to mimic what would come next in a Luhmann style Zettelkasten.

    So I suppose, the linearity gets lost to some extent this way, but so far this has never really been an issue for me.

  • Luhmann's system was developed to serve a paper-based Zettelkasten. It enabled some things that normal categorical systems didn't:

    1. Branching
    2. Connecting
    3. Hierarchy

    This are the principles. Luhmann's system is a technique that oriented itself towards those principles.

    With the digitalisation and simple IDs you can have it all with some added benefits (automatisation, date-related stuff)

  • @sfast what's the difference between branching and hierarchy? When you say hierarchy I'm assuming that you're referring to vertical relationships (i.e. argument and supporting idea) and when you say branching are you referring to horizontal relationships (inductive / deductive supporting ideas)?

    Also what's your personal method for achieving these two things with your archive?

  • Branching is lateral. Hierarchy is vertical. In tendendy at least. You got it right.

    My personal method:

    1. I branch
    2. I write structure notes


    It is not as complicated as you think.

Sign In or Register to comment.