# Unlearning vs Reinforcing Behaviors

edited November 2021

For most people (myself included), creating and maintaining a zettelkasten involves a significant amount of unlearning with regards to how we think captured information should be stored and categorized (going from so-called "top-down" to "bottom-up," etc).

Because this is a process, and few things are ever complete or "perfect," I'm curious to know in what ways you've found yourself catering to or reinforcing old beliefs about information mgmt when creating and maintaining your zettelkasten. Places where you've "bent the rules," or just felt it was OK to keep some old praxis in place.

I'd also be curious to know what you feel you've successfully unlearned, that is, what about your ZK practice you feel comes out of having successfully unlearned "the old ways" (so to speak).

• @taurusnoises

I wonder what's the purpose of this discussion beyond sharing experiences. If it means that it will be useful to others, I'd like to share.

I've only "bent the rules" once, I believe. In No, Luhmann was not about Folgezettel, Sascha Fast talks about following the principles instead of the techniques. That's not very practical for me, so I developed techniques anyways. However, I made sure that they were as close as possible to the principles and didn't carry any drawbacks.

For example, to apply the Principle of Connectivity, one practice or "technique" I use is this: Make the Zettels about the connections. You see, Zettels should be about ideas, the term idea is an umbrella for types of ideas, and these types of ideas have their own structure. So, the place for connections is in the structure of the idea. If this is not clear enough, let me know and I'll add examples.

Then, for enforcing rules, I got two for that. One is thinking with my Zettelkasten. Using my Zettelkasten to think about problems, ideas, and so on has been helpful. The other is applying the three layers of evidence model. I thought it was garbage at first. The problem turned out to be using a strict sense of note types. Now that I understand the model better, I follow it religiously.

I hope that this is useful in some way!

• @taurusnoises

Good questions for discussion - thanks!

In regard to old habits, I think I moved into the new paradigm smoothly - in fact, so much so that I was quite resistant to the idea of using structure notes. However, I've learned they do have a place in my ZK but I use them sparingly.

So - no problem leaving the old hierarchical organization behind.

I have few "rules" about my ZK. I do keep as fairly sacrosanct the atomization of zettels (it took some practice to learn how to do that) and the importance of connecting zettels. Other than that, the organization and form of my ZK is pretty flexible.

• @Annabella_Zelsky hopefully the purpose will reveal itself in the answers and discussion that happens in the comments (if any). Bottom-up emergence as we say. Thanks for the response.

@GeoEng51 That's funny to hear about the structure notes. I'm still ver "Huh?" about them. But I see how other people really appreciate and get use out of them, so to each their own!

It took about a month of constant reading and multiple attempts to get mine started. I had an incredibly hard time conceptualizing it. Like, "what even *is it??" was the constant question in my head. Actually lost some sleep over it 😂. But, then when it clicked, it clicked hard, and it was as if I had known about it for lifetimes. Really strange actually.

For anyone else who's interested in sharing, I'm really curious about what sort of difficulties people find in the unlearning process. Especially if there were specific aspects that just took a while to click for you.

• edited November 2021

I'm not sure that I had to unlearn anything, because my note-taking was too disorganized to begin with.

## Unlearned

Except possibly unlearning my fascination with Folgezettel IDs. Mathematically Luhmann's Folgezettel IDs define a distinguished spanning tree within each connected component of a ZK that uses them. The interpretation of the spanning tree is that it describes the connections to neighboring Zettels at the time a new Zettel was added to the ZK. I find this feature useful.

However, a spanning tree can be obtained with timestamp IDs as well. Since there seems to be more discussion of principles than examples useful to a novice, I provide templates at my github site, which hasn't generated much interest. (On this score I am patient, and will settle for posthumous recognition.) The templates do show how to get a spanning tree with the same property as Luhmann's Folgezettel ID system, but with timestamp IDs. I keep track of what I call CONTEXT links. A backbone-like structure can also be obtained with structure notes, but there I am concerned with arthritis. Just kidding.

Like @GeoEng51 I have tended to avoid structure notes, except for an index that is becoming increasingly difficult to maintain. However, structure notes serve the function of outlines (and tables with links, etc), which are useful and which should not be ruled out. If I have an outline in mind for a series of notes, I would relent.

Structure notes can be arbitrarily nested, so they amount to the general Folgezettel, which do the same thing "externally" (through an encoded ID) as structure notes do internally, with annotated links. Using the generalized Folgezettel ID requires maintaining a cumbersome ID encoding that has to be assigned to every note. So you might as well use the structure note, which is a more flexible technique, and not use the general Folgezettel IDs, if you had to make a choice.

A funny thing happens when I reflect on something that @sfast has thought about years ago: I find myself in agreement with him. Only it takes me an inordinate amount of time to come to the same conclusion.

## Reinforced

• The use of the Zotero reference manager
• Checklists, after Atul Gawande
• Interest in mathematics, reawakened after a period of inanition
• Likewise for LaTeX, Mathematica and experimental mathematics
• A habit of purchasing expensive books
• Since it's often less expensive to buy foreign language books, a renewed interest in foreign languages (really the persistent nagging sense that I am failing to meet the world on its own terms, pun intended)
• The use of GitHub
• The conviction that much that is meaningful in life is inseparable from pain, and that pain avoidance—the path of "no harm"—leads to meaninglessness. "Meaningful" for me is "capable of restoring an affective attachment to life."
Post edited by ZettelDistraction on

Erdős #2. ZK software components. “If you’re thinking without writing, you only think you’re thinking.” -- Leslie Lamport.

• Like @ZettelDistraction, learning how to work in my zettelkasting garden has meant that I don't know if I underwent or am undergoing any classical "unlearning." I feel like I'm "learning" new ways of connecting ideas and thoughts. If I "unlearn" anything, is it unlearning my laziness. Complex and challenging work is where the rewards are.

Unlike @GeoEng51 and @ZettelDistraction, I tend to embrace structure notes. When the muse cuddles up next to me and I notice several zettel ideas based on a topic, I'll capture them in a structure note. I started one today.

G-Metacognition 202111260910


My workflow is such that I will hang almost every zettel on a structure note. This assures that structure notes grow. Lots of thinking and refactoring happen. The more a structure note grows through refactoring, the more its ideas are repeatedly imprinted in my mind. Surprising new and novel connections appear with this structure note activity, some structure notes have gone quiet through lack of attention, but they are there when my interest is rekindled. They are a genealogy of my thinking.

Will Simpson
“Read Poetry, Listen to Good Music, and Get Exercise”
kestrelcreek.com

• edited November 2021

For most people (myself included), creating and maintaining a zettelkasten involves a significant amount of unlearning with regards to how we think captured information should be stored and categorized (going from so-called "top-down" to "bottom-up," etc).

From my coaching experience, the unlearning part is minimal. Within the context of this quote

I'm curious to know in what ways you've found yourself catering to or reinforcing old beliefs about information mgmt when creating and maintaining your zettelkasten. Places where you've "bent the rules," or just felt it was OK to keep some old praxis in place.

the bottle neck for 95% of the people does not change. It is the ability and practice to actually process knowledge instead of capturing it on a surface level. It is very, very rare that I see a fully developed thought in a note.

My personal challenge is the same that I see why people have difficulties: Sometimes, I don't fully develop the thought, postpone work that should be done in the present. So, my "unlearning" success is mostly that I take each note very serious and focus on the development of each thought with effort and concentration.

Now, I don't just take note but have clear intent why I take this note.

However, I'm bending the rule a little bit when it comes to capturing spontanously developed thoughts (with my correspondance, I often don't write emails, I rant..) and just dump the whole mail in a note instead of honing it and extracting the knowledge cleanly.

I am a Zettler

• @sfast That's interesting that you find unlearning minimal, because in my experience unlearning is the #1 issue. This is especially so when it comes to unlearning decades of top-down approaches to storage and categorization which goes unchallenged throughout most people's lives (at least in the west). We must be engaging with very different communities of people!

• @taurusnoises You might like to investigate this: https://en.wikipedia.org/wiki/Einstellung_effect. There is also the associated notion of functional fixedness https://en.wikipedia.org/wiki/Functional_fixedness. And I'm sure most people here know of the Kuhnian paradigm shift, to which one might add the concept of the Revolution in Military Affairs (RMA) which also involves a radical alteration in ways of thinking about a subject.

My PhD was on the difficulties that British generals encountered in adapting their thinking and methods to new conditions during the First World War. Broadly speaking, people do what they have been trained to do. One of the reasons for training is that it obviates the need to think about what you are going to do in a situation. This is fine until you encounter a situation which is new and for which you have no training. At that point, you have to analyse the situation to work out what to do. During the First World War this was highly problematic for the generals on the Western Front for a whole host of reasons which I won't try to summarise. (It took me about 85,000 words to do the issue some justice in my thesis.) Suffice to say, adaptation is usually difficult when the situation is complex, or time is lacking. But there have been psychological studies that show that prior knowledge is a barrier to adaptation. It often seems to come up when people try new software, because they take with them assumptions that it will behave like software they already use. Such assumptions cause a fair amount of angst and difficulty.

• Thanks, @MartinBB. I know a little about these effects, but very cursory. So, I'll defs give them a closer look.

I think it'd prove interesting to get a sense of what about the Zettelkasten proved most confusing or difficult to grasp for people when first starting out. Was there anything in particular that you struggled with in the beginning?

• edited November 2021

Amendment--concerning Zettelkasten, I had nothing to unlearn, aside from suffering from each of the growing list of cognitive biases, none of which I'm able to identify or correct...

Post edited by ZettelDistraction on

Erdős #2. ZK software components. “If you’re thinking without writing, you only think you’re thinking.” -- Leslie Lamport.

• edited November 2021

@sfast said:
It is very, very rare that I see a fully developed thought in a note.

Curious, what properties do you look for to determine if a thought is fully developed?

I imagine it is very contextual and depends what end you are processing knowledge for. If you are just fleshing out notes, maybe past a certain point you are just developing the thought for the sake of developing the thought, not to an end. I find myself doing this sometimes at least.

Zettelkasten is love. Zettelkasten is life.

• @taurusnoises said:
@sfast That's interesting that you find unlearning minimal, because in my experience unlearning is the #1 issue. This is especially so when it comes to unlearning decades of top-down approaches to storage and categorization which goes unchallenged throughout most people's lives (at least in the west). We must be engaging with very different communities of people!

As a coach, I don't concentrate on what people concern the most but on what is the bottle neck to be expanded in their personal user case.

The switch from top-down to bottom-up is such a small part of the Zettelkasten Method and knowledge work in general that I don't want to dedicate any significant amount of time or energy to it. It does not help people to the degree necessary to up their level.

@JoshA said:

@sfast said:
It is very, very rare that I see a fully developed thought in a note.

Curious, what properties do you look for to determine if a thought is fully developed?

It depends on what piece of knowledge is thought about. An argument is different than a plot line different than a model different than a checklist.

I imagine it is very contextual and depends what end you are processing knowledge for. If you are just fleshing out notes, maybe past a certain point you are just developing the thought for the sake of developing the thought, not to an end. I find myself doing this sometimes at least.

Each piece of knowledge has intrinsic properties that govern its boundaries. Take an argument for example. You have internal aspects:

1. Premises
2. Conclusion
3. Logial form

Then you have external aspects:

1. Additional premises (potential, missing etc.)
2. Alternative conclusions by different inferences fromt he same premises
3. Related arguments

But it is not necessary to go that far: It is even suprisingly rar for me to see processed knowledge opposed to collected information.

I am a Zettler

• @taurusnoises said:
Was there anything in particular that you struggled with in the beginning?

It took me a while to work out that Luhmann's field of study and his interests had moulded his method -- perhaps even demanded a certain method -- and that other fields of study would not be so well suited to the method, or would require adaptations of the method. I think the idea of having short, atomic notes which have individual uniques IDs, and are linked together in a network, is valuable. But I have worked in many fields, and they all seem to have differing imperatives and needs (some of them demand that you collect a lot of direct quotes, for example). My own story is that I started out with Italian language and the history of art, moved on to teaching English language and literature, wrote a history book, took up psychology, then moved into counselling and psychotherapy. (As you can tell, I am indecisive.) I am now retired.

If I take the example of history, I found myself collecting extensive quotes from various sources, and not knowing how to handle them. I also found myself with a LOT of maps. After about five years I had an illumination (the simple things are often hidden from us) and put all the quotes in chronological order. When I had done that, certain themes, patterns and ideas became much clearer to me, and the narrative I wanted to write emerged fairly simply from the material. Similarly, when I was doing my PhD in psychology, I was working with a large amount of source material (journal entries) on which I was doing a textual analysis. This involves reading and re-reading the same material over and over again in order to try and perceive what is being expressed between the lines or below the surface. You collect and use a lot of quotes that exemplify the themes you find in the corpus of text.

I'm not sure how much a Zettelkasten might have helped me with those particular enterprises (it might indeed have been useful in the second case), but anyway, that is all in the past. However, when I first came across the idea of the Zettelkasten, I went through a process of putting a lot of original quotes into one, then later taking them all out again to leave only personal notes, and now I am in the process of building something that is more like a personal wiki while maintaining a separate Zettelkasten with my own notes. I was a heavy user of DEVONthink for over a decade, and I have recently gone back to it after a hiatus (the release of version 3 of the software was a bit of a nightmare for me) and it could be that the Zettelkasten gets absorbed into the "personal wiki" in DEVONthink by a kind of gravitational force. We shall see. Adaptation can be difficult, but is often necessary. I often think in terms of a "system" consisting of person+method+work-to-be-done, and every one of those "systems" is unique -- a specialised adaptation. Vaguely Darwinian, I know!

• Thanks @MartinBB. It's really interesting to hear about how for some people different subject matter and interests can feel incongruous with ZK. I often forget this can be the case. Also, I love how much ground you've covered!

• edited November 2021

@sfast said:
...
But it is not necessary to go that far: It is even surprisingly rare for me to see processed knowledge [as] opposed to collected information.

Would 20211121160510 A technical example count?

Incidentally, I have used the ZK to do some mathematics over the Thanksgiving weekend. At first the computations were going well, but there were two difficulties that I needed to overcome, which I did this evening, the day before Cyber Monday. I'm spending my time verifying calculations—a suspenseful process. The earlier calculations are still useful enough to retain and link to. One day I will say what they are.

Erdős #2. ZK software components. “If you’re thinking without writing, you only think you’re thinking.” -- Leslie Lamport.

1. The title is incomplete and does not represent the content.
2. If this is a clarification of something there is no explicitly stated relationship to what is clarified
3. The sentence "This is a trivial example, but according to a member of the National Academy of the Sciences, it's difficult to write a page of mathematics without errors." is based on too many conditions to be understood on its own.
4. The tags are not specific to the piece of knowledge at hand.

I cannot be comprehensive since I cannot understand the math.

I am a Zettler

• edited November 2021

A tough critic! Thank you for that. About number 2, the note observes (for me) that the remark after the semicolon in the original text is misleading and false as stated. My restatement following is a complete statement. Perhaps it could be clearer. The thoughts are telegraphed about the National Academy of Sciences member (a mathematician speaking to another mathematician in the dining common--I overheard). I know what it means, and I have reason to believe my future self will know--but I could omit it. Point taken about the title and the tags.

There really is a need for a checklist.

Post edited by ZettelDistraction on

Erdős #2. ZK software components. “If you’re thinking without writing, you only think you’re thinking.” -- Leslie Lamport.

• As for personal un-learning, relying on the process required to (also) ditch what I felt was the security of strict categories. Hard to say in hindsight and after so many years, but something carried me over an initial phase of doubt. Could have been a belief in the power of Luhmann plus desperation Still my early years of note-taking produced far less useful pieces than what came after. Again, hard to pin-point an exact turning point, but in general the last X years are more orderly than the first X years.

This could be related to what @sfast posted above, collecting too much (and wanting to categorize as a consequence) as opposed to connecting individual "things". (I don't want to postulate that this was the main hindrance, because there's no way to be certain. From today's perspective, looking back, it sounds plausible though.)

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

• @taurusnoises a lot of people don't like the comparison but I feel an apt analogy is describing the zettelkasten as a second brain. And what we are essentially doing is helping our second brain learn different topics than communicating with it to further develop ideas and communicate them to others.

In this context you need to unlearn the "bad study habits" and adopt better ones. For example people will reread textbooks and highlight as a method of studying, which gives them an illusion of fluency with the knowledge. The zettelkasten equivalent would be to copy-paste quotes into the zettelkasten and add a few keyword links. You see this echoed by @sfast saying

"the bottle neck for 95% of the people does not change. It is the ability and practice to actually process knowledge instead of capturing it on a surface level. It is very, very rare that I see a fully developed thought in a note."

Instead, you want to do what learning scientists call "elaboration", where you write the concepts in your own words. This applies both for the zettelkasten and our own learning. It is a way of checking yourself and making sure you actually understand what you are reading.

Another example would be creating shallow vs. deep connections. Shallow ones basically involve looking at the keywords in a body of text and creating links. Those are of minimal value, especially if they are obvious connections. Instead you want to create what I would call deep connections, the non-obvious ones that take more work to find or how a concept connects to the bigger picture.

The unifying theme here is that people tend to do what is most cognitively easy instead of the difficult but fruitful tasks. Often the reason people take the easier route is that it feels better and like you are making more progress. It is cool to see the big network of notes, even if they don't form a meaningful whole.

It took me a while to work out that Luhmann's field of study and his interests had molded his method -- perhaps even demanded a certain method -- and that other fields of study would not be so well suited to the method, or would require adaptations of the method.

@MartinBB that is the biggest hold up when thinking about the zettelkasten because it seems so perfectly suited towards what he was trying to do. It makes me wary of recommending the zettelkasten method to new people.

When I had done that, certain themes, patterns and ideas became much clearer to me, and the narrative I wanted to write emerged fairly simply from the material.

This involves reading and re-reading the same material over and over again in order to try and perceive what is being expressed between the lines or below the surface.

That is what I think is one of the harder parts that has been on the back of my mind a lot. You can spend a lot of time building up knowledge/information in your zettelkasten, the hard part is finding the conceptual connections between all the different concepts.

I am in the process of building something that is more like a personal wiki while maintaining a separate Zettelkasten with my own notes

This is essentially where I am at but I don't differentiate between the wiki and zettelkasten. I just have them combined into one. Instead of turning every idea into an atomic note, I keep wiki like pages on topics, and only create an atomic note for a subtopic if I find myself referencing it in multiple places.

• @Nick I was interested to see that you changed my British "moulded" to the American "molded". An eye for detail? A declaration of independence?

I am in the middle of a comprehensive reorganisation of a lot of material that I have gathered over the years. I'm not sure what the final outcome will be, but the wiki side has taken over for the moment.

Cheers!

• @MartinBB

Haha! I am tempted to do the same thing when I see something written in American English and want to switch it to Canadian English (which sometimes has British spellings and sometimes has spellings inspired by our French half and sometimes spellings from other cultures/sources). It's a difficult choice to make or perhaps a difficult habit to break. In either case, I have a really hard time not "correcting" things (from my perspective only, of course).

• edited November 2021

@Nick said:
@taurusnoises a lot of people don't like the comparison but I feel an apt analogy is describing the zettelkasten as a second brain. And what we are essentially doing is helping our second brain learn different topics than communicating with it to further develop ideas and communicate them to others.

When I read the phrase "second brain," I am reminded of the now debunked hypothesis of the second brain of the dinosaurs. The location of that second auxiliary brain inspired a cartoon character of mine, whose development took years of heroic struggle and sacrifice.

Some of these aren't bad, such as Crohns Devil or the hands of the Sentient Trashcan, the Excommunication of Glutanus, the Immiseration and a sketch The Trash of Ages Resonates Within. In digital charcoal: Platonic Solid Removal in the Underworld and My Private Hell. Some "lunch scenes:" newspaper and a sleeping artist. And speaking of attempts to cultivate a second brain, here is a prescient remark of John von Neumann on the development of the stored program computer during the Manhattan Project

Post edited by ZettelDistraction on

Erdős #2. ZK software components. “If you’re thinking without writing, you only think you’re thinking.” -- Leslie Lamport.

• I wonder about the distinctions we can and should make between a second brain and a second or secondary memory.
(Manfred Kuehn used the term "secondary memory" in his translation "Communicating with Slip Boxes". Luhmann used the term "Zweitgedächtnis" here.)

I still see massive potential in engineering practices and structures that transform ZKs and ZKoids into actual thinking aids. Take any random list of brain teasers from the internet. How can we design ZK elements that help us to solve these puzzles? How can we make the transition from less relevant brain teasers to more relevant problems?

The article on Leibniz's manuscripts (already mentioned here) describes how Leibniz made 38 documented attempts to solve a problem from number theory. How can we plan and carry out stubborn work like this meaningfully in a ZK framework?

• Sorry. Couldn't resist.

• @MartinBB Haha! Good one. Really miss the Far Side.

• Little insertion: The purpose of the memory is not information storage. It is to work together with other functions of the mind to extract out the proper way to act in the world.

I am a Zettler