Zettelkasten Forum


What is the boundary between your ZK and other notes?

What goes in your ZK and what goes elsewhere?

I've been thinking about the different kinds of notes I take.
For my academic work, I'm maintaining a ZK in ST3 that has mostly extracts from publications as well as some topical overview notes I make related to academic science.

But there are many other kinds of notes I make that are knowledge work, but not directly related to the main theme of my academic work. Such notes include:

  • Ideas (scientific, medical, inventions, etc.)
  • Code snippets
  • Logs/records of events and meetings (personal or job)
  • Planning notes (where I do some planning of some project)
  • Brainstorming notes
  • Random thoughts (thinking in writing)
  • Lists of useful resources for something (say, learning Python)
  • Records of miscellaneous things (addresses, directions, etc.)
  • Reflections (journaling)

Right now, all these other kinds of notes go into nvALT, which is my dumping bucket for everything because it's just so fast.

My desire to keep them separate is so that the signal/noise ratio in my academic ZK stays high. With that said, there is stuff in my nvALT repository that could conceivably interact with the think I keep in my ZK. Like code snippets or ideas or brainstorming.

What do you think are best practices? Strict boundaries? Or should a ZK be just an externalization of thinking, no matter how formal or germane to a particular topic?

Comments

  • In my understanding, a Zettelkasten is intended to be a both long-lived and highly ordered. These qualities help determine what to add to it.

    I intend to keep my Zettelkasten for life, so before adding something to my Zettelkasten, I always ask if it's possible that I will be interested in this thing twenty years from now. If it's not reasonably possible (as in a set of meeting notes that I took), I don't add it.

    My Zettelkasten is a web of authoritative knowledge that I have acquired, and when I add something to my Zettelkasten it should be clear to me how to link it into that web. If I don't know how to link it in, I don't add it yet.

    This means my Zettelkasten contains no meeting notes, to-do lists, brainstorming stuff, resource lists, or any of that. All that goes into my monthly org-mode files, which are much messier and chaotic than my Zettelkasten.

    The main role of my Zettelkasten is in helping me to understand a large area of knowledge more effectively. So I wouldn't add just any useful code snippet to my Zettelkasten. But if I were making a study of a particular style of programming, or different programmatic approaches to solving a problem, then code snippets might be appropriate to add.

    Just how I use it, of course.

  • Thanks @galen. This is very useful. This is how I've been thinking about my ZK, which is why I wanted to clarify. All those other kinds of notes - resource lists, ideas, free thoughts, etc. They are also very important and useful. But they would clutter the ZK, which I use to digest and respond to external knowledge. It seems like the way many people use a ZK is to to acquire and respond to other people's thoughts and ideas.

    So I'll keep using nvALT or ST3 (with a different directory) to hold all these other bits. With that said, many of the principles that hold for using/making a plain-text ZK also are valuable in these other disorganized notes. Like the unique IDs and the inter-textual links. @rene 's ST3 ZK package is awesome, so I'm working with my nvALT notes in ZK as if it were a ZK, even though it's not.

  • I use the same ID scheme to assign unique numbers to invoices I send to clients, receipts for stuff that I buy and have in my bookkeeping list, etc. -- it's not "taking away" the ID from my archive (since I cannot do two things at the same time anyway) and makes cross-referencing files between systems super easy thanks to macOS Sportlight indexing the whole file system :+1:

    I cannot seem to find where I wrote about my non-Zettel notes in the past. I use the Zettelkasten as a reference system (in the GTD sense), so my task lists live outside of it. So do ephemeral project files, like article drafts, book manuscripts, book cover images and their various iterations, related PDF files, bug and crash reports, etc. -- things I one day would want to delete when the project is finished.

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

  • I distinguish between knowledge work and everything else.

    Notes on writing projects and things like that are not knowledge work but part of project or self management. That goes into my org-files. Adresses and things I use for something else, too.

    Then, there are articles and longer texts I write. They are excluded, too. Shorter aphorism-like text are in my archive. This is for practical reasons. Texts that don't adhere to the principle of atomicity are excluded.

    Therefore, I have three containers:

    1. Zettelkasten (The Archive)
    2. Self and projekt management (org)
    3. Longer Texts (iA Writer at this moment)

    I am a Zettler

  • @Sascha so is the distinction you're making for ZK inclusion based off definition alone or are you also making a value analysis of the information before including? For example, under the circumstance in which you're developing research around a topic you don't think will be very important in the future -- do you make that determination or as long as it's "knowledge work" it goes into the system? For people like me in the typical desk job, we often conduct research for work to output a small report or writeups -- let's say research on a small company to send a summary to my boss -- and it's hard to imagine being useful in the future. Although this company research is technically knowledge work -- does it deserve some filtering? If so do you have any tangible criteria or is it totally subjective?

    I see potential problems with this because things can get grey and again you'll fall into the issue you all have written about before that thinking too much about it will result in no note at all....

    You also can't predict where the connections might take you right?

    But anything you can add about boundaries for "knowledge work" outside of just the definition of "knowledge work" would be helpful.
    Because I interpret this to mean technically any research on any topic, as long at its atomic, etc, could go in.

  • I cannot seem to find where I wrote about my non-Zettel notes in the past.

    @ctietze if you were able to find your writing on this it would be a great addition to the topic!

  • @galen said:

    My Zettelkasten is a web of authoritative knowledge that I have acquired, and when I add something to my Zettelkasten it should be clear to me how to link it into that web. If I don't know how to link it in, I don't add it yet.

    I've been trying to set some clear conceptual boundaries in my mind between different kinds of input buffers and miscellaneous storage categories such as:

    • "general inboxes" — e.g. my e-mail inbox, which I do want to at least sort of periodically be processing toward emptiness (but also don't want to devote too much time to)
    • "input streams" — e.g. Twitter, RSS feeds — no expectation of processing "everything" even to the extent of deleting stuff I don't want. I can dip into these if and when I want to but never have any "obligation" to. A few items get pulled out and placed into specific inboxes.
    • "slushpiles" — basically secondary inboxes where I keep things I want to clear out of a general inbox, but don't yet even want to invest the time to decide whether the items are actually worth keeping or not. These are automatically deleted by a script if they still haven't been processed after a certain length of time has passed. Functionally they're similar to input streams except that what's in them originally came from a general inbox.
    • "specific inboxes" that belong to apps or to other areas where I at least know what an item "is" but haven't yet processed it any further, such as my OmniFocus and DEVONthink inboxes.
    • "junk drawers" — containers for things I haven't classified yet (and may never). Not relevant to the ZK since it eschews categories, but relevant for other systems that do have them. Unlike slushpile items, these are things I have decided are worth keeping, but don't yet (or possibly ever) want to take the time to fully classify.

    So I'm particularly interested in where you keep notes/information that you're not ready to add to your ZK "yet" but may want to someday. For me, at least for now, it seems to make sense to have an input buffer in DEVONthink, which I otherwise use only for my ZK. (So technically it's not "part" of my ZK proper, but it is part of my ZK system.) Where do you keep this material if you keep your ZK in a pure-text environment like The Archive?

  • @zwhaley You seem to have found the topic on Org mode already, which maybe is the one I had in mind, only I remembered it differently.

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

  • @zwhaley said:
    @Sascha so is the distinction you're making for ZK inclusion based off definition alone or are you also making a value analysis of the information before including? For example, under the circumstance in which you're developing research around a topic you don't think will be very important in the future -- do you make that determination or as long as it's "knowledge work" it goes into the system?

    One of the advantages of the Zettelkasten Method is that you don't have to decide wether something is important in the future or not. :smile:

    I am doing knowledge work in my The Archive. The results just stay there.

    For people like me in the typical desk job, we often conduct research for work to output a small report or writeups -- let's say research on a small company to send a summary to my boss -- and it's hard to imagine being useful in the future. Although this company research is technically knowledge work -- does it deserve some filtering? If so do you have any tangible criteria or is it totally subjective?

    The whole point is to acknowledge that you don't know if and what something is important in the future.

    I have the feeling that you are afraid that the archive can get cluttered. If you follow the Method accordingly that won't happen. I have notes about religion, my family, future directions of the archive, fiction book ideas, philosophy and much more -- happily co-exististing in my archive with nothing but enriching each other the the occational connection.

    I am a Zettler

  • @Sascha, yes that is the fear for me. You have written before about not worrying, just start entering. This is me holding onto the last thread of "OCD" before letting go :-)

  • I write a blog post on that topic.

    At least with my ("the") Zettelkasten Method there is no problem at all. It scales and focuses on the fly. In one second I can work on an abstract problem of ethics and the next I am switching to a area that focusses on my future home because I have a sudden inspiration. :smile:

    I am a Zettler

  • edited January 2018

    @Eurobubba said:

    So I'm particularly interested in where you keep notes/information that you're not ready to add to your ZK "yet" but may want to someday. For me, at least for now, it seems to make sense to have an input buffer in DEVONthink, which I otherwise use only for my ZK. (So technically it's not "part" of my ZK proper, but it is part of my ZK system.) Where do you keep this material if you keep your ZK in a pure-text environment like The Archive?

    I keep an org-mode file for each month. I would just keep a single file forever, but org bogs down with very large files, so I keep one per month. Pretty much everything I work on gets dumped in there. Something new comes in, I tack it on at the end of the current month's file, as its own top-level item. Most stuff never makes it into the ZK, but stays tagged and searchable in the org archive.

    It's slightly frustrating because org is almost a perfect ZK environment on its own. But org was really designed for relatively few relatively large files rather than many small files. And worse than that, the search capability, while far, far more sophisticated than The Archive, disrupts my flow by forever realigning windows, so somehow I can't get "in the zone" with org the way I can with The Archive.

  • edited January 2018

    I think it's interesting that org-file management does have Zettel-like qualities like @galen pointed out. Is there a reason why project management isn't structured like a ZK in the same way research / knowledge work is?

    I realized when consolidating my projects and deliverables that the process is very hierarchical -- tasks lead to completion of projects, multiple projects lead to completion of larger projects, until finally an objective is reached. @ctietze this is similar visually to the mind map you shared with us before. Mind mapping is a great way to visualize your workflow of larger projects, and shows just how hierarchical things are.

    Also, there's a lot of linking involved with projects, literally or figuratively. Projects reference the knowledge work and research from the ZK as inputs. As mentioned above, they link vertically when directly contributing to larger or smaller projects, and also laterally when contributing indirectly to another project. There can also be indexes for projects. For example, a project can link back to a specific customer or topic index, showing all relevant projects related to that customer / topic and providing another entry point to dive into any project like you would a research zettel. Projects may even be placed on a completion timeline index to show work done over time, or easy entry to dive into the Structure Project Note from the WIP list of current projects.

    You can also give projects a project tag to bring in meeting notes, research zettel, and other ideas into one project "workspace".

    Does this not seem like a ZK to you guys? @ctietze, @Sascha I know you have a very separate distinction between knowledge work and org work / task management. I'm not disagreeing with that distinction, but I'm wondering for the same reasons a ZK structure is helpful in your knowledge work would a similar structure not be helpful in task/project management, just accessed in a difference folder? Especially for more complex or longer term projects that are broken down into small sub-projects.

    Why go to a straight linear folder, sub-folder hierarchy the second you take on a project and forgo the advantages of the network / inter-connectivity capabilities you had with the ZK knowledge work?

  • edited January 2018

    @zwhaley said:
    Why go to a straight linear folder, sub-folder hierarchy the second you take on a project and forgo the advantages of the network / inter-connectivity capabilities you had with the ZK knowledge work?

    I'd be interested to hear what others have to say. For me it's simply the fact that entering something properly into my ZK takes time and energy to consider how it fits into the rest of the web of knowledge, and this is usually not worth the investment for project-management stuff.

    I find that my project management has to stay super lean or else I fall into productivity masturbation and end up just wasting time coming up with "better" systems for handling tasks that aren't complex enough to warrant them. This means I end up doing most project management with a combination of paper and email. I don't use org-mode for project management, since I can't manage to keep org-mode organized. I use it as a digital notebook and idea dumping ground that I can probably find stuff in later, as well as for composing document drafts. Trying to rely on org-mode for more than that ends up being a huge time sink, where I spend all my energy customizing emacs rather than just getting work done.

    The Archive is relatively fiddle-proof, which is one of the reasons I find working with it productive, even if it doesn't have all the features I want.

    Also, just a guess, but projects tend to have a natural hierarchical structure, and a ZK, while able to contain hierarchical information, in my experience lends itself structurally to a general graph of information. This is perfect for knowledge work, where you are trying to track and discover connections and abstractions amongst a bunch of disparate sources, and where a predefined hierarchy could obstruct your understanding. It's less good for project-management, where a predefined hierarchy is often extremely useful in setting priorities, etc.

  • Since I was tagged by @zwhaley, I want to jump in and say that I agree with @galen's points, even though I hadn't thought about them this way before :)

    Rehashing what was already said:

    • While projects may be interconnected horizontally, to pick up a term from above, tasks are supposed to be short-lived, so there's not much benefit in creating a network of task references. The overhead wouldn't be worth it.
    • Tasks lists may serve the same objective as continuing to write on your text in fragments, that is: in your Zettelkasten note archive. There is a one-way vertical connection from tasks to notes. But I have found that I, the user, suffice to realize the connection and make it work: no need to (the term is misleading since it's all digital, but I cannot find a better expression:) "physically" tie tasks to notes. At least not besides the casual Zettel reference by ID.
    • Also, as @galen points out, task management produces very hierarchical and linear project plans, while the notes in your Zettelkasten archive are always free to be recombined and may form circular graphs instead of the tree structures of projects which are necessary to be tree-like and uni-directional to accomplish anything.

    I wonder if people like Twyla Tharp (The Creative Habit, 2003) would disagree about the necessity of linear project plans. She's not a writer, she's a dancer and choreographer. I remember reading in her book, which I seem to have misplaced, she collects notes and physical objects in shoe boxes as incubators of projects, eventually digesting the mess inside the shoe box and making a plan from it. I remember the result to be rather profane and linear, too.

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

  • @zwhaley said:
    Does this not seem like a ZK to you guys? @ctietze, @Sascha I know you have a very separate distinction between knowledge work and org work / task management. I'm not disagreeing with that distinction, but I'm wondering for the same reasons a ZK structure is helpful in your knowledge work would a similar structure not be helpful in task/project management, just accessed in a difference folder? Especially for more complex or longer term projects that are broken down into small sub-projects.

    There are underlying principles the Zettelkasten Method is build on. These principles are founded in universal traits of thinking, epistemology, science theory etc. But ultimately the principles of thinking are the most fundamental I think.

    Take for example the principle of hierarchy. You will find that in all areas of knowledge and information work. You write in chapters and subchapters. You think in structures embedded in structures. In the Zettelkasten Method you get hierarchies of structures. In project management you have projects, subprojects and tasks.

    That's why I think that the software should do one thing: Make you good at manipulating text.

    There a some differences between them and therefore some optimisation for think-writing and project work could be done. We optimize The Archive for the former. But it is totally possible to do you project management with The Archive or even do both of the tasks in one archive with The Archive. It isn't an optimal solution but you will be still effective.

    I think your obversvation is spot on. :smile:

    Why go to a straight linear folder, sub-folder hierarchy the second you take on a project and forgo the advantages of the network / inter-connectivity capabilities you had with the ZK knowledge work?

    I don't. I have eight active org-files. Nothing more.

    @galen said:
    I'd be interested to hear what others have to say. For me it's simply the fact that entering something properly into my ZK takes time and energy to consider how it fits into the rest of the web of knowledge, and this is usually not worth the investment for project-management stuff.

    True for me, too.

    @galen said:
    I find that my project management has to stay super lean or else I fall into productivity masturbation and end up just wasting time coming up with "better" systems for handling tasks that aren't complex enough to warrant them. This means I end up doing most project management with a combination of paper and email. I don't use org-mode for project management, since I can't manage to keep org-mode organized.

    That's interesting. What kind of mess do you produce? One of the main benefits of org for me is the ability to keep them tidy... and I had a hard time of keeping my project management in order.

    @galen said:
    The Archive is relatively fiddle-proof, which is one of the reasons I find working with it productive, even if it doesn't have all the features I want.

    Did you write Christian your feature requests?

    @galen said:
    Also, just a guess, but projects tend to have a natural hierarchical structure, and a ZK, while able to contain hierarchical information, in my experience lends itself structurally to a general graph of information. This is perfect for knowledge work, where you are trying to track and discover connections and abstractions amongst a bunch of disparate sources, and where a predefined hierarchy could obstruct your understanding. It's less good for project-management, where a predefined hierarchy is often extremely useful in setting priorities, etc.

    I agree. Project management is very hierarchical while knowledge work is way more lateral.

    @ctietze said:
    I wonder if people like Twyla Tharp (The Creative Habit, 2003) would disagree about the necessity of linear project plans. She's not a writer, she's a dancer and choreographer. I remember reading in her book, which I seem to have misplaced, she collects notes and physical objects in shoe boxes as incubators of projects, eventually digesting the mess inside the shoe box and making a plan from it. I remember the result to be rather profane and linear, too.

    I think there is good possibility for that. Arts is very iterative. I compared it once during lunch with pugging a clay figure rather than distinct iterative steps.

    I am a Zettler

  • @Sascha and @galen thank you both for your incredibly insightful responses. After hearing your opinions I agree with what you're saying. It's very interesting for me to explore the realms of this method, mostly really just understanding the way that we think in ways I've never thought of before, and testing ways this can be enhanced. But I agree project management should probably be dealt with differently.

    @Sascha you said your project management used to be messy and now it's orderly. Can you talk about what changed? It might offer some helpful insights for those of us who still feel messy :smile: . Also when you say org-files, are you referring to the org-mode program or just plain text files and folders? More details on your PM workflow in general would be very interesting.

  • What did I change? I cannot tell. It clicked when I switched to org-mode and since them I am being a pain in the ass to Christian because I want him to make this change, too. (Which is totally unnecessary because his PM is quite fine, but nothing is really fine until you switch to org/emacs)

    A big component is ongoing practice. I still invest ca. 30 minutes (not more!!!) per week to try new things. If I mess something up I do it a couple of times over to ingrain my habits.

    I have a blogpost on my workflow with writing in the pipel. Only a matter of time.. :smile:

    I am a Zettler

  • edited January 2018

    @Sascha said:
    I have a blogpost on my workflow with writing in the pipel. Only a matter of time.. :smile:

    Can you direct us to that blog post?

  • "In the pipe" means: Is roughly drafted, awaits editing, polishing and Christians critic. :smile:

    I am a Zettler

  • @Sascha said:
    "In the pipe" means: Is roughly drafted, awaits editing, polishing and Christians critic. :smile:

    I do hope the sequence does not get out of order ... ! :)

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

  • @Sascha said:

    That's interesting. What kind of mess do you produce? One of the main benefits of org for me is the ability to keep them tidy... and I had a hard time of keeping my project management in order.

    Well, org is an open-ended tool, and I use org as my general scratch pad for a lot of different projects as well as keeping code snippets and various other assorted bits of useful or interesting information I come across. So to use it effectively for project management, I find that I have to come up with pretty stringent conventions to follow, otherwise there's not a clear distinction between the project itself and the project-management section of the org file. I know that one of the ideas of org is to be able to keep your todos inside of the projects as you're working on them, and then rely on agenda views to find all your tasks, but this just ends up a mess for me.

    If org had general agenda views that narrowed progressively as you searched, that updated automatically as files changed, and that stayed in the same freaking place in the editor, I would surely use it for more. But the visual context switch required when emacs decides to replace or split the current window is too disruptive. Yes I know about shackle but now we enter into "forever customizing emacs" territory.

    @Sascha said:
    Did you write Christian your feature requests?

    I've written him about a few. I'm waiting on a lot of them to see if they are really features I want, or only features that I think I want, if you see the difference.

  • @galen Ah, ok. I think I am able to handle this problem because I stick to the "one app one function rule". Emacs is my project manager only.

    I thought a bit on this issue the last couple of days. I think there is a good argument out there to do all of the knowledge and information related work in the Zettelkasten/Archive which then cleans up the other areas like writing or coding.

    I think that I will find something when I understand properly how a Zettelkasten scales to different sizes of work. I can write anything from short articles to books in the archive (better: rough-drafting). But I need to clarify that before I can come up with a good argument.

    I am a Zettler

  • edited February 2018

    @Sascha said:
    I thought a bit on this issue the last couple of days. I think there is a good argument out there to do all of the knowledge and information related work in the Zettelkasten/Archive which then cleans up the other areas like writing or coding.

    Can you clarify what you mean by this? My initial interpretation is that you're saying there could be no limitations on the quality of knowledge / information notes that go into the ZK and you could potentially just "dump it all". This would include brainstorming, planning, loose thoughts, etc.

    But how does this clean up other areas like writing or coding?

    P.S. This reminds me of Dan Sheffler's use of engagement notes vs. memory notes, which is an opposing argument to my interpretation of Sasha's comment.

  • Still, haven't not thought it through, properly.

    The boundaries between a Zettelkasten, your writing, self-org and more are somewhat flexible. Some boundaries make sense. For example: Tasks in self-org and not in the archive. But still more thinking is necessary on my part.

    I am a Zettler

Sign In or Register to comment.