The Case Against a Digital Folgezettel

A couple of weeks ago, I wrote about the case for a digital Folgezettel; after some time, I’ve spent around a whole week of deep work just researching about it. That resulted in another attempt to solve the Folgezettel debate.

One argued intensely against it (it’s controversial, after all), but rudely so. And so I talked to Christian, one of the founders of Zettelkasten.de, about the topic. Let me tell you, talking to him was totally insightful! I haven’t found others talk about ideas the same way since I met my close friends.

Anyway, the talk made me work on the development of a Folgezettel technique that works even better than Niklas Luhmann’s — the Note Sequence 2.0 — to solve the two problems inherent to it.

It looked promising…until I added hundreds of notes more.

To give some more context, I believed Folgezettel records the sequence by which ideas unfold, and that Structure Notes only organize the invisibly unfolding ideas created through interlinking outside the sequence. That was true…until I found through experience that without structure notes, these sequences would lose their implicit meaning in time. As of writing this sentence in my drafts folder, I’m still ambivalent on its redundancy to Structure Zettel, so let me elaborate to clarify my thinking.

I’ll start with what I’ve developed: the Folgezettel 2.0. (F2) when you try to link to a note in F2, the ID itself “activates” the previous notes. This is because of the notational features. In the F2 technique:

  • The addition of another character means a continuation of a thought. 0001-a continues 0001, and concurrently so between numbers and letters. (0001-a, 0001-a1, 0001-a1a, etc.) That made branching multiary rather than binary. (as compared to Luhmann’s Folgezettel)
  • The notation contains information about the conspicuous thought trails. In other words, the ID has an identity attached to it. The ID works like a Family Name, because you can say that 0005-a2f5h4 came all the way from 0005.

In contrast, I’ve started to realize Structure Notes “activate” ideas better than note sequences, albeit that only happens at the structure note level rather than at the note level. To put it in a concrete way, when you link a new zettel to an old one inside a structure note, you’ll be able to see other potentially relevant links in that structure zettel — not to mention they’re already presented in a logical sequence. (Which F2 will lack when you read syntopically)

A linking workflow using Structure Notes is like entering a bar, seeing one friend at a table, and then meeting his friends you’ll eventually get connect with; you’re not going to be friends with all of them, but potentially, yeah. It’s still consistent to how I’ve designed F2.

The way I see it, linking to a note in F2 is more like getting married, rather than meeting a friend at a bar table; when you become connected to someone in a family, the whole lineage becomes connected to you, too. (and you have no other choice, man) Here’s what it looks like graphically:

So far, so good. So what’s the problem? The problem is that in the first place, I’ve created this technique to lessen the dependence on structure notes in “activating” potential relations. The common argument against this is the topic of “meaning”.

You see, I’ve always thought of a Zettelkasten as a miniature Internet. (except without dogs that have social media accounts) Because of that, I looked up a couple of articles by Tim Berners-Lee, where he discusses about links and all. Here’s what I found out, and what @sfast has been saying all along.

Links themselves don’t carry meaning unless you give them. In F2, note sequences don’t have “meaning” by themselves, but the way they’re sequenced allows you, the creator of the sequence, to give meaning to them — in other words, that’s called implicit meaning.

On the other hand, when you link to another note and describe how they’re connected, you give explicit meaning. That’s exactly what happens in the Structure Zettel or when you link directly using a wiki-style link and explain how a note is connected to each other. (Not to mention you can arrange their sequence to follow a logical flow) To quote Berners-Lee:

So the existence of the link itself does not carry meaning. Of course the contents of the linking document can carry meaning, and often does. So, if one writes “See Fred’s web pages (link) which are way cool” that is clearly some kind of endorsement. If one writes “We go into this in more detail on our sales brochure (link)” there is an implication of common authorship. If one writes “Fred’s message (link) was written out of malice and is a downright lie” one is denigrating (possibly libellously) the linked document. So the content of hypertext documents carry meaning often about the linked document, and one should be responsible about this. In fact, clarifying the relative status of the linked document is often helpful to the reader.

While F2 depends on implicit meaning at first, the trouble is that implicit meaning gets lost over time and you’ll eventually have to create a structure note nevertheless.

Also, the inevitable fact is that the direct links from/to one note accumulate over time; they would eventually outnumber the note sequence branching, and so it would still depend on structure Zettel. (or hubs, as Luhmann called it)

Lastly, structure notes allow a Zettel to have multiple roles (what Sascha calls taking part in multiple hierarchies) and order them in a more logical, argumentative sequence. In contrast, F2 allows a note to be part of only one sequence — just like its predecessor, Luhmann’s Folgezettel. Meaning, there could only be one incoming link (also called in-degree in network theory) and multiple outcoming links.

I’m always saying that the ZK should work like a damn brain, but while the lacking network properties can be resolved by conjoining Folgezettel and the Structure Zettel, the structure zettel still becomes a requirement because of the aforementioned issues.

Therefore, the original purpose of the Folgezettel is defeated. Thus, I believe now that Folgezettel is a redundant technique.

If Folgezettel were my friend, you’d say that he’s experiencing existential crisis right now — preparing for his revenge against me and my clan… (Sorry, man.)

In summary:

  1. Folgezettel does spreading activation, but Structure Notes do it better.
  2. Without Structure Notes, it’s difficult to develop ideas in Folgezettel in a logical order. With Structure Notes, you can arrange them indefinitely.
  3. Folgezettel relies on implicit meaning that gets lost over time. If you don’t make it explicit (read: Structure Notes) you’ll have to keep the implicit meaning of the sequences inside your head. It defeats the purpose of the Zettelkasten itself.
  4. Incoming links from Folgezettel is limited to one. That means the links toward a note will easily outnumber that of the Folgezettel. (That’s perhaps what Sascha means about “single hierarchy”) So, how do you maintain them? Structure Notes.

Ironically, I’ve learned how to use structure notes properly by experimenting on the Folgezettel technique. Perhaps it’s because of the need to solve the gaps in the Folgezettel, like what happened to Luhmann.

I leave it to Sascha and Christian to end this war — it’s something the whole community needs. The most usual entry point to the Zettelkasten is Ahrens’ book, which discusses Folgezettel, so I think that creates a lot of confusion for people.

The other one is Roam Research. Frankly, I don’t really think how most people use Roam resembles a Zettelkasten at all, but if they can improve their productivity using interlinked notes, then more power to them.

Anyway, my efforts now lean toward streamlining the workflow, leaving some questions:

  • Do we keep a list of structure notes inside an overview note?
  • I know nothing can be ensured, so how does one approach maintaining the links made outside of a structure zettel? (“maintaining” kinda defeats the purpose of ZK, maybe “organizing” is a better word)
  • What does Ü1, Ü2, and Ü3 mean? Isn’t the qq1, qq2, qq3 trick enough?

P.S. @sfast@ctietze Thanks for your guidance on this topic, it made my experiments quicker. And I have around 300-400 real notes, I think. I deliberately didn’t abandon it right away to discover what happens and hopefully gain the same understanding of Folgezettel and Structure Notes. Whether it worked or not is a win, anyway.

You can share this post, if you want: