1. 56
  1. 29

    TL;DR: Organize notes with tags and links instead of folders. Notes should be small & digestible “ideas” instead of lengthy analysis so they can be better linked.

    1. 20

      I would say there is a more important, not explicitly mentioned overall concept here. You train yourself to, rather than gradually forgetting about the past, keep revisiting your accumulated knowledge, everytime from a different angle, because you actively try to fit in new ideas into an existing body of previous ideas. You award yourself for doing so (by neatly tugging away a card in a drawer), creating an incentive to keep doing it.

      This way you’ll be much more inclined to regularly overthink what you know and keep previous ideas in a semi-active working state. I suspect that the Zettelkasten maps very well to a mode of functioning that the brain is surprisingly good at (and thus feels good and thus readily induces a state of flow): maintaining relatively quick access to a lot of information by linking ideas together in the form of a web.

      1. 5

        That’s a good summary.

        I’d also emphasize the organically evolving heterarchy (which the Zettelkasten facilitates) as opposed to a pre-defined hierarchy (that is the norm for outliners like Workflow, Dynalist, etc.).

        1. 1

          For anybody coming to this thread late, we are developing an official community zettelkasten here: https://www.zettel.page/

        2. 1

          So basically what I’ve been doing for the past 15 years with a private wiki. Didn’t know it had a name!

          (Although in my case, due to other issues, my extensive notes and organization only bring me up to functionally productive.)

        3. 13

          I read it but there is something I am still missing.

          So say I have 10,000 notes. I write a new one. How on earth could I link it with others that I’ve forgotten about? Surely I cannot re-read all of 10,000 to see which ones I should link together. I do not get this part of it.

          1. 11

            Incidently, I just stumbled upon this Zettelkasten-like note: Notes should surprise you:

            If reading and writing notes doesn’t lead to surprises, what’s the point?

            If we just wanted to remember things, we have spaced repetition for that. If we just wanted to understand a particular idea thoroughly in some local context, we wouldn’t bother maintaining a system of notes over time.

            This is why we have dense networks of links (Evergreen notes should be densely linked): so that searches help us see unexpected connections.

            This is why we take Evergreen notes should be concept-oriented: so that when writing about an idea that seems new, we stumble onto what we’ve already written about it (perhaps unexpectedly).

            The linked note inside starts with this:

            It’s best to factor Evergreen notes by concept (rather than by author, book, event, project, topic, etc). This way, you discover connections across books and domains as you update and link to the note over time (Evergreen notes should be densely linked).

            So one aspect of the answer suggested here: By writing notes about general concepts, you provoke revisiting them. Through heavy linking you now create connections between related ideas. Through further maintenance like overview notes transitive connections become more direct connections and thus closer. (As a self-refential example: This reminds me of union-find and how it merges disjoint sets).

            1. 5

              How does one find “related notes”?

              I believe the answer lies here:

              Luhmann described his Zettelkasten in different ways. Sometimes he called it a conversation partner and sometimes he described it as a second memory, cybernetic system, a ruminant, or septic tank.

              You actually just keep a big chunk of the notes in your memory. You don’t memorize their literal content, but you know of their existence, so you know approximately where to look. I would say that the Zettelkasten was a tool to enhance his own memory rather than a tool to do the remembering for him.

              The secret to keeping the existence of the largest part of 90.000 notes in memory is probably revisiting them regularly, by following links and browsing at random, both of which the Zettelkasten invites you to do. Luhmann’s notes were all centered around his single general focus area: philosophy and social science, so he was thinking about the full body of work all the time.

              I think the brain is better at this sort of thing than we generally acknowledge. I actually believe that this may be an argument for using an analogue Zettelkasten. You come up with an idea and your brain hints that you have thought about something related before, you just don’t know when and what exactly. You might vaguely remember a particularity of a card though, maybe a tear, a stain or even the smell. So you go browsing, and soon enough you run into a connected idea, which leads you to the subweb of existing ideas that are relevant to the new idea. This rummaging around has the additional benefit of refreshing your memory of existing notes.

              The efficiency on such a large scale is probably for a large part determined by your ability to leverage the hints your unconsciousness gives you, which I think is why Luhmann describes his Zettelkasten in varying, not very rational or strict ways, in an effort to capture this unconscious process.

              1. 4

                My interpretation is that you would have meta-notes that serve as an index for notes about a given topic. (I do that and it’s very useful) So you might not remember the existing 10,000 notes, but surely you can relate your new note to some topics you already touched, and then you can go look at the index for those, and follow the links that look relevant.

                Then, I guess one assumption is that, while exploring the existing web of notes to connect it to your new note, one should dedicate a bit of time and make some new connections if they feel relevant, i.e. do some general “maintenance” of the system.

                1. 2

                  If you completely forgot about something, then you have no trigger to even look for something to link to. However, if you have a vague memory of something, a system of indices can be helpful to refresh your memories.

                2. 3

                  It’s about destruction and creation. You deconstruct something to use it’s parts for creating something new.

                  1. 2

                    This is exactly the question none of my reading has answered. I haven’t read “Taking Notes the Smart Way”, but I’ve otherwise read a lot, and nothing seems to address that question.

                    I have my own ideas, but I’d love to see how other people solve this.

                    1. 7

                      Yeah I’m starting to think that “do X the smart way” is mostly BS for most values of X.

                      The more I go on the more it seems to me that the system is not important as long as you’re consistent with its use (whether it’s note taking or organising or anything else).

                      Also, the most productive people I’ve seen… They don’t use fancy things that make them “super productive”. They just sit down and do the work, and they got good and fast at it by means of practice.

                      The more I go on with life the stronger I feel the smell of BS.

                      1. 4

                        I gave it some thought and this is what I have:

                        Linking problem is not solved. If it was solved there would not be any need for adding links in the first place. Why add links at all if you can easily find other notes related to the current one? So the fact that links exist leads me to the conclusion that linking is the actual hard part of this method.

                        Second - the whole approach is a hyperlink data base structure before data bases became a thing. This system can be implemented in a single table having 3 columns: 1) ID, 2) Note, 3) References. We have software now so maintaining a data-base in a furniture is probably obsolete.

                        But most importantly, like the article said, Zettelkasten  becomes better the more notes are added to it. OK. So the logical conclusion is to invite other people to add their notes to it. Then it grows faster. Everyone is adding their ideas, and creating links between ideas. To me it seems such a state is the ultimate goal of this system. But we have this now, it’s called the internet.

                        So I think whatever made that one german scholar so productive probably wasn’t the system itself.

                        1. 1

                          You say that as if the internet isn’t a hugely transformative, productivity increasing tool.

                          1. 1

                            I would go as far as to say that for many people it’s the opposite. It can be used as productivity increasing tool and an amazing one at that. But many people use it for other things completely.

                        2. 1

                          I think the theory is that if you’re taking a note, you’ve probably already got a bunch of notes about the thing you’re researching handy, so you link to those. For my part that doesn’t match how I generally take notes (I often find myself making note of something I stumble on, and writing summaries/self-tutorials of stuff I’m studying to ‘learn it’ more thoroughly).

                          That said, the Zettelkasten core idea seems to me to be ‘smaller notes, more often’, and also to follow the old adage: “Graphs are a set of Edges that incidentally have some Vertices attached.” Links and granularity are the key takeaways, and there are good ways to do that that aren’t exactly Zettelkasten.

                          1. 1

                            I’m currently trying TiddlyWiki because its UI encourages smaller notes.

                            With a physical Zettelkasten you implicit see neighbor slips as you search a linked one. This is lost in a digital version where searching is delegated to computer and effectively instant. It would be easy to track all kind of implicit connections (backlinks, created before and after) but how to present that in a helpful way?

                        3. 2

                          I don’t think the system is supposed to answer that question. Connecting ideas is the job of the human, not the system. However, if you do connect some ideas, you don’t have to remember that connection. Your web of thinking is externalized.

                          I’m sure there are strategies. Randomly showing any two cards seems as good as any strategy. This would only increase the probability (marginally) that two ideas get connected. It would still be up to you to figure out how they are connected.

                          1. 4

                            But to create the connection you have to know that the connection is possible, so you have to be aware of the other cards to which you can connect this one. That’s the essence of what a zettelkasten is supposed to be, and what no one is explaining.

                            Here, I’ve read something, I’ve written a short, pithy note. Where do I file it? To what other cards is it connected? How do I find them?

                            I have some ideas, but it’s supposed to be a solved problem via this method, and I’m seeing zero discussion of it in any of my reading.

                            There’s the puzzle.

                            1. 4

                              I think of zettelkasten a bit differently. In database terminology, I see it as normal form applied to ideas. Ideas are often tightly coupled to their original context. This is why it’s natural to apply a hierarchical structure to notes. You just place the note under a folder which represents the original context. In zettelkasten you make the idea atomic and if you want to give it context then you have to reify that context with a link (a foreign key). Zettelkasten isn’t your schema, zettelkasten is relational algebra. You’re free to come up with any schema you want.

                              Where do I file it?

                              In the same folder with everything else. It’s flat. The structure is provided by the links.

                              How do I find them?

                              I use org-roam in emacs. A digital system helps with search. The original zettelkasten had physical organization by topic and you could traverse the links from there. I’ve only been doing this for a few weeks so maybe search might not scale.

                              Btw, I think your questions are great ones. All I’m saying is I don’t think the system has an answer.

                              1. 2

                                Thanks for the reply … I pick up on this:

                                How do I find them?

                                I use org-roam in emacs. A digital system helps with search.

                                I agree that search solves/avoids a lot of the problems that a purely analog system would/did have.

                                But the whole point of a zettelkasten is that it helps with the search, either to avoid it, or to guide it, or to augment it. Luhmann described “having a conversation” with his zettelkasten.

                                So I’m here, I have a new “card” … I can search the existing ZK for cards that have the same words, but that suffers the problem of combinatorial explosion, and it’s not using the zettelkasten in any kind of clever way.

                                I think there are answers to be found, and perhaps Luhmann had some, and even his enthusiasts don’t really “get it” beyond having a huge box of index cards with some sort of indexing system. It feels from my reading that Luhmann had more than that. Otherwise it’s just a wiki with search.

                                As you may be able to tell, I’ve thought about this a lot.

                                1. 1

                                  I think the idea is to never make a note without linking to it from some context of an existing note. So you might have a note on when is the best time to do exercise. That note should be referenced from an index card on exercise, or maybe a note on daily routines.

                                  I think there is some hope it’ll be like a wiki, but maybe something more like tvtropes or c2 than Wikipedia where all the ideas are enmeshed are easy to move between.

                              2. 3

                                You connect nodes you know of. Since nodes are connected to nodes, this lets you wander the node links. No node contains every link, because that’s not their job. Their job is to link to whatever you still remember so you can later link it to what you’ve long forgotten.

                                1. 2

                                  I have some ideas, but it’s supposed to be a solved problem via this method, and I’m seeing zero discussion of it in any of my reading.

                                  How to Take Smart Notes is worth a read, IMO, though it is still kind of nebulous on some points.

                                  There’s an “Everything You Need to Do” section. It says:

                                  Now add your new permanent notes to the slip-box by:

                                  a) Filing each one behind one or more related notes (with a program, you can put one note “behind” multiple notes; if you use pen and paper like Luhmann, you have to decide where it fits best and add manual links to the other notes). Look to which note the new one directly relates or, if it does not relate directly to any other note yet, just file it behind the last one.

                                  b) Adding links to related notes

                                  c) Making sure you will be able to find this note later by either linking to it from your index or by making a link to it on a note that you use as an entry point to a discussion or topic and is itself linked to the index.

                                  Presumably, in order to find things to link if you don’t have them ready to hand, you use the existing index.

                                  1. 1

                                    The magic seems to lie here:

                                    a) Filing each one behind one or more related notes … b) Adding links to related notes

                                    This is the question no material or article seems to be answering: How does one find “related notes”? It refers to “the index”, but that’s rarely referred to elsewhere and seems utterly mysterious. How are things indexed?

                                    I should write up my musings as best I can to further reveal my confusion and incomprehension. I’ll try to do that.

                                    1. 1

                                      So, full disclosure, I’ve taken notes on cards before, but at the time I gave them either meaningful names or datestamps rather than Luhmann-style sequential identifiers or what-have-you, so I just kept them in a card box sorted by alpha / date. The collection never grew substantial enough for me to think about other forms of indexing. The cards I’ve held onto are in a single cardboard file box with little alphabet dividers I bought at an office supply store.

                                      That said, indexes for paper information storage are a pretty well-established technology.

                                      If I were going to take a crack at it for a paper Zettelkasten, I would:

                                      • Set up a large card box with alphabet dividers.
                                      • When adding a note to my permanent collection:
                                        • Search the index for keywords pertaining to the new note
                                        • If no card exists in the index for the keyword (name, phrase, concept, etc.) I want to be able to track down again:
                                          • Print the keyword on top of a fresh card
                                        • Write down the ID of the related note on any relevant cards (either existing ones or the ones I’ve just created)
                                        • File the cards with keywords alphabetically in my index

                                      Here’s an abbreviated example from a random page of the index in the first reference volume I could find on my office bookshelf, The Chicago Manual of Style:

                                      ornaments for text break, 1.56 
                                      orphans (lines), 2.113, p.899
                                      o.s. (old series), 14.132
                                          basic principles, 6.121
                                          parts of a book, 1.4
                                          publishing process, 2.2, fig. 2.1, fig. 2.2
                                          punctuation and format, 6.94, 6.126
                                          See also: lists
                                      Oxford comma, 6.18-21. See also commas.

                                      This is a ~900 page reference work with fairly dense text, and about a hundred pages of index, so that ought to give you some very rough idea what ratio is useful for a working reference system, and how far an alphabetized index can scale. The references are mostly to section numbers or figures rather than pages, which seems like a pretty useful parallel to how things could work with numbered cards.

                                      The other thing you might want to research is library card catalog approaches. The paper card catalog systems I was taught at length in elementary school have all pretty well been obliterated by electronic databases by now, but there was once a range of well-developed techniques there for indexing into very large collections by author and subject matter.

                                      There’s nothing to stop anyone from translating these techniques directly to software, though there are probably more automated ways to get most of the same benefits in any given system. (i.e., tagging systems, automatic keyword indexing, and good old grep.)

                              3. 1

                                I’m thinking of trying this out and my idea was to try some sort of transitive closure view and “show me five random notes” thingymajiggy.

                                1. 1

                                  How on earth could I link it with others that I’ve forgotten about? Surely I cannot re-read all of 10,000 to see which ones I should link together. I do not get this part of it.

                                  Practically speaking:

                                  • in a paper system, the answer is probably an index and a sorted reference collection
                                  • in an electronic system, the answer is some combination of search and tagging

                                  Relatedly, the cards (or other unit of note-taking) aren’t intended to be an append-only log. You’re supposed to interact with the system and refine the web of connections as you go, so it may not matter if something is initially orphaned.

                                2. 3

                                  We “practice” the Zettelkasten technique using Neuron. Our process-to-date: folks create one fact (zettel) about a tool in our tool chain. Example: sample pull request. Over time we organize and reorganize our zettels into different trees. When we have a large collection, we combine them into documentation and move them to our Sphinx platform (we are a Python shop). Similarly we organize facts about our content curation process (we create life-science applications). Our developers are not biochemists, so we relate the handling of specific biochemical issues to our toolchain. Neuron <linking> is the service that allows us to connect specific domain information with specific technical processes/methods. The common “technology” here is a markdown document. We do some light editing of Neuron documents before combining them in Sphinx (which uses reStructureText)

                                  1. 2

                                    I would love to read it, but I have to pay to do so…

                                    1. 3

                                      Jesus Christ this article is complete trash.

                                      Two thirds of the article is pointless propaganda, only after a good 60% the author actually starts explaining how this zettelkasten system works.

                                      What’s the point? Why are you even trying to convince me? Are trying to sell me a piece of furniture later?

                                      Regarding the system itself: I am tempted to say that a good wiki software like confluence would do the same, but the real advantage of the furniture is that it’s likely going to keep working in 20 years. I wouldn’t bet the same in confluence (or MediaWiki or whatever).

                                      1. 4

                                        Mediawiki is 18 years old. Confluence is 16. Considering the Lindy effect, there is a good chance they will still work in 20 years.

                                        1. 4

                                          Well, I wouldn’t call Confluence “good wiki software” any more than I would call Visual Basic a robust development environment, but I take your point. My database of personal and professional notes for the past 15 years has been a private instance of Dokuwiki and the more I read about Zettelkasten, it just sounds like a curated personal wiki like I have, once you get past all the gushing.

                                          1. 2

                                            Editing in confluence is light years ahead of pretty much everything else.

                                            1. 1

                                              Does dokuwiki have backlinks and tags?

                                              1. 2

                                                Backlinks yes, tags no. Maybe with a plug-in.

                                                1. 1


                                                  So perhaps one could have a page called “Tag:Something” that only holds a description of what could have been the tag “#Something”. All pages that are related to “Something” should have a link to “Tag:Something” on their taglist, and the backlinks on page “Tag:Something” will show relevant pages.

                                                  It’s a bit like Wikepedia’s “Category:Something”, isn’t it.

                                            2. 3

                                              The space around the concept “Zettelkasten” has all the features of an emerging marketing space. There are already multiple software solutions. Just wait for the custom-made physical slip-boxes, the note cards in different colors, the books, the pay-for videos, and the webinars.

                                              1. 7

                                                As I wrote here a couple weeks ago:

                                                The Zettelkasten thing sure has been hitting the zeitgeist hard these last few months - right around when I started poking at those ideas myself after kind of edging around them for a decade or two. It’s interesting to feel a burgeoning nerd methodology cult wash over and through the system of my own thinking. I was a lot less self-aware the last few times this really happened to me (the first big wiki wave back in the era of thousand-line Perl CGI wiki software comes to mind), and I never got drawn into GTD or Agile on any deeply felt personal level, so it’s almost like a new experience.

                                                That said, I think it’s also been quietly bubbling along in the background of the note-taking nerd memespace for many years now. I think I first ran across the word “Zettelkasten” on Taking Note, a blog I’ve probably been following since 2008 or so, but index card approaches that are clear relatives to it in one way or another have been popping up now and then for most of my adult life, I think. It just seems to have reached a critical mass lately. Or, as you say, become an emerging marketing space. Establishing itself as a working methodology-cult ecosystem with an in-group vocabulary, defined rituals, canonical texts & standard arguments, and mystique about True Process. You can see it happening in realtime over at the Zettelkasten Forum, which is run by the authors of The Archive.

                                                …and which is an interesting forum to skim now and then. I don’t want to be disparaging, this is just how these sorts of cultural phenomena seem to unfold. I’m trying to stay self-aware about all this while I spend a fair amount of time building up my own system of notes.

                                                (I did some ranting about notes about notes / writing about writing and so forth last night, inspired partly by this thread and others like it.)

                                                1. 3

                                                  The memory of your comment inspired mine.

                                                  There will always be a market for selling tools that magically replace hard work and time with a “process”. I’m not really judging. My work/life doesn’t require anything like Zettelkasten, but I’m sure it would interest my dad, who has been buying old handheld computers just to keep using their database software.

                                              2. 4

                                                It’s weird too, that it sells the idea, then starts explaining how it works, then it goes back to selling it again for a few more paragraphs! And only after that second set of propaganda it finishes the explanation.

                                                1. 3

                                                  It’s not well written, but all the tools linked in the article are free (and most of them not harvesting your data).

                                                  1. 2

                                                    What’s the point? Why are you even trying to convince me?

                                                    I think that it is targeted towards a particular audience: “The main component of The Writing Cooperative is our publication, which is one of Medium’s largest. […] Everything we publish falls within our mission statement: Helping each other write better.” But yeah, the tempo was a bit choppy and it reminded me of one of those “weird thing” articles. Then with big promises it dumps a board game on the reader without explaining the rules.

                                                    It seems like there is something promising in Luhmann’s system, but I don’t want to risk getting a hand-me-down cargo culted version of it.