Hacker News new | past | comments | ask | show | jobs | submit login
Introduction to the Zettelkasten Method (zettelkasten.de)
195 points by pcr910303 12 months ago | hide | past | favorite | 73 comments

This is your friendly reminder to avoid the productivity hacking tarpit. It feels productive to think about productivity but oftentimes you're confusing movement for progress. I'll give two specific pieces of advice for people who are thinking about switching from methodology / app A to methodology / app B.

1. You might find something that B does that is just so difficult to accomplish in A, so you want to switch. But before you do, you should spend a week living with the difficult path to do that thing in A. You will likely find that learning the muscle memory to do that thing in A was actually the difficult part, and once you've overcome that hurdle it isn't difficult any more.

2. You might switch to B and feel that "now, everything is so organized!" However, it's very likely that you could have "switched" from A to A and you'd feel the same way. What I mean by this is that by switching, you've forced yourself to revisit all of your notes. If you forced yourself to revisit your notes without switching, you'd see the same effect.

I wouldn't consider this to be a productivity tool in the usual sense (i.e., like Todoist or GTD). Instead, it's about facilitating deeper thought and making new connections between ideas. For a professional content creator, which includes most academics like Luhmann, the entirety of your career is your collection of notes. You can store them on paper, in the computer, in your head, or however you want, but those pieces of knowledge are all you have. Dismissing these tools as merely an arbitrary reorganization of a static collection of documents is like dismissing Javascript because we have html. Good luck creating Google Docs with only html.

Folks conflate TODO lists with a personal "library" because the first note they usually create is a TODO list, and then a set of sub-TODO-lists per project.

Often, what we spend time on most, are the things we need to get done, and our true reference "library" is just Google.

I would argue that a pure personal "library" is actually just the archive of past projects. In that context, searchability is more important that good organizing.

I have that sort of hybrid approach. My "main" doc is a TODO Journal where I list my todos for that day (often copied from the prior day). The journal maintains a record of what I've done every day since I started two years ago - each new day is set at the top of the doc. Many of the line items are themselves links to other documents (all in markdown), which are themselves a mix of project notes and TODO lists.

I use Typora to edit, and WinSCP to encrypt on a remote drive that's backed up each night.

Andy Matuschak makes a similar point. If I understand correctly, his perspective is that the notetaking is the work.


This, this is the post to read.

It is indeed a tarpit that traps people who mean well and legitimately want to Get Things Done. The knowledge work/"second brain" community will have you convinced that everyone who you run into online is a PhD candidate. The irony is that a lot of people intend to "simplify" their workflow but end up reaching for more things to pile onto it. The breaking point for me was the first time I loaded Obsidian. It is a nice program but it incentifies Doing Too Much, at least for my taste. Granted, what influences a person to even care about how they take notes varies. The chaps over at Zettelkasten.de actually advocate for a very simple approach to note taking that can work anywhere. Personally, I've found that it's best for me to do the bulk of my work in a Plain Text environment -- plain like the text box that I am using to type this comment plain.


As much as pragmatics would point in this direction, saying "let's be realistic here .. etc.", perhaps nurturing that part of us that dreams in alignment with Vannevar Bush's dreaming up the "memex" or Douglas Engelbart's vision of "augmenting the human intellect" is still valuable if the urge is strong enough. Maybe the seeds of the next information revolution are lurking in such dreams .. especially given the progress we've made in NLP.

I felt the same way about Obsidian. It seemed really helpful on the surface but after a few days of using it, I realized I was spending more time recording everything and making sure to use the correct backlinks than I was actually learning anything. Just bookmarking pages gets you the majority of the benefit of meticulously organizing your notes, while requiring next to zero effort.

IMO, the advantage of apps like Obsidian is that you don't have to organize everything meticulously.

1. You just create a new note, 2. give it some tags in square brackets "[[Tag]]" 3. and if it direcrly related to a note you recall making, make that connection

That's it. With time the amounf notes grows, and the amount of interesting connections grows too.

It is also important to note, that the process should note be automated. If you want to learn something, you need to make the time to write the note yourself, as opposed to simply copy/pasting.

Bookmarking a page doesn't mean you absorb its contents, and I don't think a few days is adequate time to perfect these techniques.

I'm the same as you, but a month or so in and still learning. Personally it's a price i'm willing to take for what will be a lifelong mission

A lifes work of bookmarks isn't very helpful in the grand scheme of things

>What I mean by this is that by switching, you've forced yourself to revisit all of your notes. If you forced yourself to revisit your notes without switching, you'd see the same effect.

At this point, I've learned that a good system of notes requires frequent visits and pruning. Otherwise, it turns into an overgrown, sprawling dumping ground of tags and random notes/links.

To me, Zettelkasten, and tools like Roam, help reinforce that behavior by providing a reward in connections. My brain can outsource the connection-making, context generating aspect. I just have to be prescriptive about setting time aside each day to tend the "garden". Not a ton of time such that I get burned out. Just a little each day. As a knowledge worker, it seems like that is an invaluable exercise to build a habit around.

Hey, I came across an earlier comment you made on a post and wanted to ask you some questions with regards to leveling up skills, would you be ok with me emailing you?

Sure thing.

>Niklas Luhmann ... published 50 books and over 600 articles. ... There are also over 150 unfinished manuscripts left in his estate. At least one of them is a text of 1000 pages.

A person at this level of productivity will have an overwhelming amount of notes, and therefore an advanced system of organization is necessary.

I would guess that most people reading this are not anywhere close to that level, and likely do not need an advanced note-taking system. They would be better off spending time on their work, rather than spending time perfecting their system. This is especially true now that notes are often digital, and can be searched in their entirety in an instant.

With anything productivity related there's an intense honeymoon period where it seems like you've found the magical productivity elixir. In reality, you're just high on placebo. It's far too easy to chase that high from app to app, method to method.

> It's far too easy to chase that high from app to app, method to method.

Likewise for languages and frameworks.

Your comment would be useful if it didn't so completely missed the point. Its rightful context would be productivity. But here it is sitting atop a thread of comments for an article about an approach to note taking, inducing people who it could greatly help (note takers) to dismiss it as "yet another productivity hack".

I think it depends what your goal is. The 'How to take smart notes' book is specifically about taking research notes for academia or at the very least non-fiction writing.

Zettelkasten Method can make a lot of sense in contexts like that. If you are writing a weekly blog, keep building notes up and polishing parts of them for the blog. Then your note taking process is in fact your blog writing process, but a bit more procedurally free form.

It's way more nuanced than that. Avoid the treadmill or cargo culting productivity, but don't avoid reckoning with your habits and learning the skills you need to stay organized. That's a real thing and best done as early as possible. In my experience, I was interested in productivity stuff as early as a teenager but nothing clicked until my late 20s and then I finally found things that worked for me, and I'm glad I did.

If you end up doing this stuff as an end in itself, as in it becomes the task of your day, well then you're in trouble.

On the other hand, I've found real value in actually adopting /a/ "productivity system". For a year now I've been following the GTD approach, and it has made my work more structured, and more targeted -- instead of floundering in my own scattered thoughts and notes.

But I do agree with you that constantly switching to find the new perfect system is very unproductive. Partly because of the pareto principle, and partly because you actually get better at using your current system overtime.

You need to start where you are, and realize that just doing more things isn't going to fix any holes in your life. If you have a ton of things to do and you are having trouble keeping track of them, then you might have use for a tool.

Separate your work and your life and your ambitions. I'm lucky enough that I can track my work stuff in my head. If it ever gets to where it's too hectic, I escalate to a written task list.

My home life is similar. All of my life context is kept physically. I can see that the dishes aren't done. I can see when I need to take out the trash.

My ambitions, on the other hand, well, I've spent man-years of effort on hacking away at solutions to manifest them faster. I don't consider the time I've spent writing scripts and stuff wasted because I love doing this stuff, it's what I naturally start doing once everything else is sorted. I don't take a 'ship it' strategy to them because to me the point is enjoying the journey, not to get to the end as fast as I can.

I don't have to build a colocated kubernetes cluster in order to host a blog, but I want to. It appeals to my sensibilities to not pay hundreds of dollars for compute when I have a perfectly good desktop machine with lots of its own compute.

I've spent more time on keyboard macros than I'll ever save. But the feeling of power I get from having my world accessible with just a few keystrokes is great. And learning about how evdev and xbindkeys and the like work is interesting as well.

The hole isn't filled when I get to the end. The end is when the hole stops being filled. This is what organization will do to you if you're not careful.

The reason we continue to look for improvement in productivity is the same reason I use Git instead of SVN or zsh/bash instead of GUI. Learning curve is steep but I feel so much more productive. The reward of learning something new can be massive. I've spent tons of hours on learning Vim and tweaking its config file, and it has rewarded me with an editor that I feel comfortable using for a lifetime. I've also wanted to try Emacs because of its potential benefits.

Zettel* is not for everyone, just as GTD or Pomodoro or Agile/Scrumm or Kanban. But when it works for those certain people, it is like magic, and the productivity level just skyrockets. When GTD was popular a decade ago, it really revolutionized how I worked, and its biggest benefit was that it offloaded brain so I can really focus on what is essential, not having to constantly remember what I needed to do next.

It's the same reason I use markdown, git, vim/emacs/vscode, etc.

I found it useful to put my yak shaving nice to have tasks in an org document. Ironic, for sure, but something about writing it down scratched the itch and allowed me to move on.

Productivity dont do shit for procrastination

My mother had a zettelkasten slide-cabinet. My father produced over 60,000 slides for her, over the course of their academic careers, by copying illustrations in art and architecture books (and by street photography on their many trips abroad)

She could walk up to the slide cabinet, and ask "give me a theme" and we'd say "animals" and she'd say "give me a period" and we'd say "modern" and she'd select slides about the history of representative art about animals, in modern times.

When she'd put the slides back, she could re-enter the cabinet and select slides about Michaelangelo and followers, or about the role of the pediment in architectural design, and select from the same slides, but for a different context.

Amazing stuff. Totally indexed by herself, operated for a lifetime of academic teaching.

I think she "invented" hypermedia in the 1960s basically.

This is stunning. I've never heard of another example of an offline zettelkasten before other than the original. I would love to peruse it!

But if someone asked for a theme or period that wasn't pre-coded, it wouldn't work.

Its true. You could do a talk on primitive art influence in modernist sculpture and find African tribal art, and cave painting from france so implicitly could do a talk on African tribal art, but without reference to south America or Asia or Inuit art, it was plainly incapable of doing a wide ranging talk on folk art in pre modern times.

You could do a talk on images of x in medium y, but only indirectly do a talk on x itself and no context not captured in images of x in medium y. Thats a fundamental limit.

If however there was no index to x in y, it relied on personal knowledge x in y existed. So there was a lot of path casting inside her head, to recall states to determine if the effort of an exhaustive search was justified.

Isn't that true of any pre index scheme?

thanks for sharing

Do you know what happened to it? That sounds amazing, I would love to see a video showing it off.

It was donated to the edinburgh college of art in the 1990s when she died. I suspect by now, its toast: The slide film process might be good for 50+ years if archival storage is used, but the university library probably had this out as a teaching resource before the web was big, and afterward, the digitization simply invited massive IPR complaints. Even the labour cost of separating out her own photos from the copies would have been massive.

I know a lot of the mounted slides had 'newtons rings' which are signs of the medium touching the glass carrier, which is not net beneficial to the slide ageing well.

I had a look in the catalog, It is not listed as available when her written work is, so I think its gone.

The slides hung in sheets from suspensors, about 5 rows of eight columns per sheet. You would stand at the cabinet (it was a couple of 4 or 5 draw filing cabinet) with a normal projector carousel, and then select slides from the sheets. The sheets had keys like "french baroque architecture" or "corbusier" or "golden mean" or whatever to her was a primary collation key, and then she would use her own head to eyeball the slides and select the stuff by 'secondary key' given what she knew. I think the hyperthreading here is mostly notional the primary keying, was the sufficient vehicle to find anything given "I know I have a slide of this, it was when i was researching post modern industrial design in iron" and then she'd strike out into the stack...

From 40x40x5 I get to more like 10,000 to 20,000 slides. I think you might get as many as 100 sheets to a drawer but even noting exaggeration it was a huge collection, wide ranging.

Regardless, it would be fascinating if you ever found the time to write more about your mother's system, how she used it, how you experienced it, etc!

Maybe she was inspired by Vannevar Bush's Memex concept?

Maybe she was. We used to talk sometimes about the value of a library catalogue as a research tool, and she'd worked in publishing so the concept of an index as a key value store was pretty apparent. The memex talk is maybe a stretch?

Neuron [1] is an exciting new Zettelkasten project.

In short, it’s a CLI tool that transforms a directory of markdown files with some special syntax into a static site.

A couple of highlights: It’s got support for tags, links, backlinks; it can generate trees of links to notes matching a given tag; the static site has search functionality; there are editor plugins for vim and Emacs.

The CLI has the ability to output your entire Zettelkasten graph as JSON, which you can then parse and use to write scripts. I’ve already used this feature to write a few tools for myself that greatly improve my note taking workflow.

Behind the scenes, Neuron uses Pandoc to convert the markdown into HTML. There is talk of a feature that would allow the user to write Pandoc filters which would essentially be programs that transform the Neuron AST prior to it being rendered with Pandoc. Such a feature would open up tons of opportunities for customization, e.g. you could write rules to parse custom syntax in your markdown files and generate inline HTML. I’ve already written a tool that does something similar [2] without the use of a Pandoc filter by operating directly on the generated HTML files, but a Pandoc filter would greatly improve the process.

All of your notes are simply markdown files stored on your local machine, so you’re free to do whatever you’d like with them: store them in git, write scripts to generate them, whatever.

The author is simultaneously developing Cerveau [3] which is a hosted web interface that allows you to edit your notes remotely via a web browser. It transparently syncs your edits to a GitHub repository. Unfortunately, Cerveau is currently closed source, but I believe the author has expressed interest in opening it up if he’s able to get enough GitHub sponsors to sustain its development.

If it matters to you, Neuron is written in Haskell and the project makes heavy use of Nix.

1: https://github.com/srid/neuron

2: https://github.com/srid/neuron/issues/228#issuecomment-67029...

3: https://www.cerveau.app

Sounds like that would work pretty well with this mobile markdown-based notes app I'm using: https://gitjournal.io/

For having been around for so long in its most basic form, it's pretty exciting to see an entire renascence of knowledge tools and methodologies.

Every few years I get bitten by the bug and I go on a hunt of what's out there. Most recently, some colleagues were very excited about Roam Research [0], so I took another look, and I was very pleasantly surprised to see that, among many great tools, the good old TittlyWiki [1] had evolved with the times, and now you have "distributions" tailored to apply the Zettelkasten methodologies.

I settled on Drift [2], and over a weekend, I moved hundreds of notes I had spread in several formats over to a small set of Drift files. I even moved my entire personal web site (mostly audio and foodie geeky stuff) into it, with close to 80 tiddlers by now [3].

Porting was phase 1. The fun part where the methodology becomes really powerful, is breaking down the monoliths, just like you would do with app modernization into microservices. Taking longer blog posts and decomposing them into discrete, reusable components that can hold meaning under different contexts. The more I looked at them, the more I could find I could break them down into more discrete ideas. And I'm sure I'm not done.

Several weeks into it, I've become more organized than I had been in years, and I realize it helps me structure my thinking, and how to connect ideas.

Regardless of what tool you choose (and there are plenty) it's a great moment to look at what this methodology can do for you.

[0] https://roamresearch.com/

[1] https://tiddlywiki.com/

[2] https://akhater.github.io/drift/

[3] https://ramirosalas.com

For any readers looking for a slick, local-first app, I've really enjoyed using Obsidian: https://obsidian.md/

I like the app, but cannot commit to a closed source solution.

I do appreciate the local-data first approach, but that is table stakes, especially for something as personal and sensitive as a knowledge base.

The "free for personal use" is a rug that can be yanked anytime without warning. After having a number of closed source but otherwise free pieces of software do this, I do not want to climb the learning curve, nor do I suggest anyone else do so.

I've recently played with an open source alternative called foam, which builds a similar experience to Obsidian but inside a VSCode workspace. It's a new project, but under active development, and is already pretty close to the base Obsidian experience of notes + backlinks + graphing https://github.com/foambubble/foam/

I can highly recommend Obsidian. I avoided Roam due to it being yet another cloud based platform (unless that has changed since it first came out?) My knowledge base is far, far too valuable to have it go up in smoke if the company ever collapses. I'm very much loving Obsidian. It just gets out of my way and lets me write.

Beside the knowledge-base aspects of obsidian, it is also an awesome markdown editor, for two reasons:

a) it blends source-view and render-view very well, I almost never use the actual render-view. Almost as good as Mark Text or Typora, and I wish VSCode would move in that direction either.

b) its BSP-style layout lets you make best use of your screen estate

c) (more of a gimmick, but I love it): When you want to change the theme, it randomizes the order of themes in the list. For any new vault I just pick the one that is on top, and have a different style for different projects.

I tried obsidian but when it starts it doesn’t open a window. I posted to the forum and got a shrug. I try every once in a while thinking an update might fix it, but I obviously don’t know if I could use it even if it started working.

IIRC only the AppImage works properly on Linux. Snap and Flatpak give trouble for many users.

Been using obsidian for my digital garden, found it to be very good.

I recently moved to Joplin from Evernote. I think I use Joplin in a Zettelkasten way (though no doubt it would upset a purist) nonetheless I find Joplin a good fit and it is interesting that I independently arrived at a similar (though less complete!) solution. Now I have to think about how much of that completeness I go back and apply to my existing x thousand notes... :)

Drift looks very nice. TiddlyWiki is interesting. But other than the name, one thing that I question is why it gives visitors the impression that they can edit posts. I know what the traditional concept of a Wiki is to facilitate collaboration, but is there a way to disable this for private instances?

I see why it can be confusing. The funny thing is that you can edit posts, but you can't save them on the server. In fact, you can save a local copy of a site built with TW, but it becomes your own copy. You can go to the settings and hide the save button. Not sure if you can totally disable it thou.

What I like about this model vs. others is that it's entirely self-contained in a single file. As long as you have a browser that supports Javascript, you can work with it. And as much as I like modern tools like Obsidian, TW pretty much guarantees me that if I save 1,000 tiddlers on a file (up to 100K are known to be possible), 20-30 years from now, as long as HTML and Javascript are still capable of being ran somewhere, my entire knowledge base will be there.

Heck, other than some Unicode issues I have, my entire site is happily in archive.org as a single, snappy file right now...

So there does seem to be a recent Hype-train around improved note taking and Zettelkasten in particular.

I would 100% echo the timesink/tarpit comments insofar as unless you have a specific learning or knowledge management goal devoting time to this can be a fun procrastination game.

Many of the sources rate the book "How to Take Smart Notes" by Sönke Ahrens as a starting point (ISBN10 1542866502 , ISBN13 9781542866507)

There are a plethora of youtube channels and Productivity websites to be found so YMMV, I liked:

Bryan Jenks' - https://www.youtube.com/playlist?list=PL5fd4SsfvECy0zzf8Cyo2...

Justin (Effective Remote work )- https://www.youtube.com/channel/UCkzyo69rqBoBJUyQ9jo53Bw/vid...

There is a podcast on "How to take smart notes", if you prefer to listen https://bookworm.fm/68/

Wow, thanks for the mention of my channel (Effective Remote Work). Glad you're liking the videos!

Hacked up a simple CLI tool to create and organize notes using Markdown files and publish that as a static website. The idea is very much inspired from Neuron - but a few differences:

- In zettel we have a concept of Connections and not uplinks/backlinks. This lets us have circular links across notes.

- For visualisation we use Graph structure instead a of a Tree. For reasons mentioned above.

zettel is a simple CLI written in Go, so very easy to just grab the binary and get started.

The project is in a nascent stage, feature requests or any relevant ideas are more than welcome!

Our GitHub Repo: https://github.com/hackstream/zettel

Documentation: https://zettel.hackstream.dev/

(Yes the docs are generated by zettel as well)

I like the idea.

I know about Baader-Meinhoff, but did something happen recently to lead to some sort of resurgence of zettelkasten interest?

I feel like I learned about the term a few months ago on HN, and then it's been cropping up quite frequently since. It's a pretty distinctive word, I feel as if I would have noticed and investigated if it had come up so frequently in the past.

Sounds cool, though I've seen this on hacker news several times already.

But this kind of regimental organisation would never work for me. I don't have the patience.

For me organising means putting all my stuff on the desktop, which then goes in a folder "Old Desktop" when my desktop becomes so full that I can't add more files to it :'). It's become a nested chain of old desktops over the years. Yet I never seen to have issues finding even old stuff back. In fact it's surprisingly effortless, somehow. When I navigate it, it feels like walking through an old neighbourhood I use to live in but haven't visited in a while. Same with my notes mess in OneNote. For some reason it makes sense to my brain.

I've been trying different organisation methods in the past but they never stuck. It was a lot of work and didn't really make retrieval easier because it was never that hard for me to begin with. So I'm taking that as a sign this lack of organisation is good enough for me :)

This again :) my five minute solution is to just have a “master learning log” which is just a glorified single google doc outline..

I have notes.org (emacs org mode) which contain basically all of my notes ever written.

Sure the org mode collapse thing is fine but the killer feature is being able to grep all of my notes in milliseconds.

org-rifle may be better for searching org notes, because it searches for multiple words occuring in individual notes which is more efficient than simple line-base grepping:


Please see also org-ql, which mostly obsoletes org-rifle and is much faster as well.

How do you use that day to day? I also use Org, but it has slowed down quite a bit for me. Perhaps one of my files is too big, and I also use the agenda, which takes at least a few seconds to regenerate.

If I see this on the front page of HN once more... I'll... I'll...

... click <hide>.

A Zettelkasten tool built on VS Code: Foam


For an academic zettelkasten that involves references and citations, are there any tools with good support for bibtex (or DOI references in general)?

Zettlr is a Zettelkasten tool that has great bibtex and Zotero integration [1].

[1] https://docs.zettlr.com/en/academic/citations/

Emacs, I'd imagine. (the same statement goes for Emacs and most anything, though) :)

You can also use org-roam-bibtex: https://github.com/org-roam/org-roam-bibtex


I'm working on a Zettelkasten app myself at the moment. Backlinks and tags are working.

Rather than using solely a time based ID like "202006110955", I'm adding the date to the file name. Links then show the file name instead of the id (e.g 2020-10-29 lessons learned.html → [[lessons learned]]).

Also working on ways to surface information from the Zettelkasten like a spreadsheet view and generated one-page summaries.

Quick demo here: https://youtu.be/hf2s8fZsojY

If, as the top post wisely indicates, you are in search of results and real impact you should carefully consider which app really helps you get things done.

Lots of hype and lots of features in an app is a recipe for lots of wasted time.

For making a zettelkasten with the minimum of hassle and friction check out https://scrapbox.io

I tried something with elixir, markdown, graphviz and feh.


Currently I add my thoughts manually. I'm still trying to find out if it has a significant advantage over classic note taking methods.

can we agree to stop upvoting these posts please? This has been topping HN for like each month for the last year.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact