What are your views on Zettelkasten Note taking System

The quotation from Andy Matuschak vaguely reminds me of what Argyris and Schön said about single-loop and double-loop learning. It might be crudely said that single-loop learning is trying to answer the question “Are we doing it right?” whereas double-loop learning is asking “Are we doing the right thing?” I think Argyris and Schön compared this to a thermostat that shuts off at 24 degrees C, as against a thermostat that asks “Is 24 degrees C the right temperature at which to shut off?”

@satikusala, given your interests, you might find it interesting to look in particular at the 1982 article listed below, if you don’t already know it.

Cheers, Martin BB.

Argyris, C., & Schön, D. A. (1978). Organizational Learning: A Theory of Action Perspective. Addison-Wesley.
Argyris, C. (1982). The Executive Mind and Double-Loop Learning. Oganizational Dynamics, Autumn. https://doi.org/10.1016/0090-2616(82)90002-X
Argyris, C. (1999). On Organizational Learning (2nd ed.). Blackwell Business.

1 Like

Fantastic! Thanks.

@satikusala You’re welcome. I also managed to dredge from my unreliable memory the fact that Chapter 6 of the 1999 book was entitled “Teaching smart people how to learn”, which is more pertinent to the original post in this thread than it might appear. I also managed to find a pdf from the article that seems to have been the precursor of that chapter: https://www.safepilots.org/wp-content/uploads/Teaching-Smart-People-How-to-Learn.pdf. I think it would probably be worth your time.

Cheers, Martin.


You, sir, are AWESOME!!!

Not really, but I thank you, and I hope you find it useful! :wink:

1 Like

Thanks @satikusala for these interesting comments. They immediately made me reflect on my own note-taking practices, and the reasons for my initial attraction to Tinderbox.

I have to confess that I didn’t know of the Zettelkasten system, but after reading about it I realised that it encompasses much of what I already do, or have done in the past. What I take from your posting, though, and want to comment on are (1) the importance of fundamental principles, as distinct from particular systems and software packages and (2) the myth of a ‘one size fits all’ solution.

Given the second point one might ask if there’s any value in outlining one’s own approach, if it’s so individual that it’s unlikely to be applicable to anyone else’s situation. That said, if such an account contains even only some elements that others may be able to adapt to their own purposes, then the exercise is justified.

I’ll mention first one method used for my PhD, a study of the literature of the Beat Generation and its social contexts. My aim in this instance was to create databases of fictional and critical references to the concepts and topics I was discussing in my dissertation. Each fictional record, for example, contained the fields Theme, Details, Text and Page Number, Character(s), Location and Comments. At the time I used Bento for the database software, and when that became obsolete I transferred the information to Tap Forms. While reading a book I would insert a sticky note in the appropriate place (so as not to interrupt the reading unduly), then complete the Bento entry at a later stage. This method worked very well for its intended purpose. For example, if I needed to locate all the fictional references to ‘jazz’, I would enter it into the search box and all the relevant records would be selected.

Of course, for other purposes, a different approach is needed. For complex theoretical texts I’ve found that there is no substitute for detailed, sequential notes. I use Apple Pages for this. I would disagree that this method is ineffective, but it all depends on how the task is approached. Some level of analysis is required before anything is written down, that is, a block of text or section of a chapter needs to be read and thought about and the main point(s) clearly identified and summarised. In my experience this leads to a much more thorough understanding of the text than can be obtained simply by reading it. Another benefit is that after a lapse of time the notes can refresh one’s sense of the text without the need to reread the entire book. It’s an easy matter, too, if one wants to quote a particular passage or discuss a particular point to locate the precise reference.

Neither of the above systems copes adequately with a constant problem that is often mentioned: how to manage the countless pieces of information relating to a research project that one comes across in the course of reading. I’ve discovered I’m not alone in trying to remember an elusive detail that suddenly seems pertinent to what I’m writing, and spending hours trawling through websites, PDFs and my print library in a vain effort to track it down. It was this that led me to try Tinderbox.

In recent years my interests have revolved around early modern philosophy, and in particular the translation from Latin into English of various hitherto untranslated texts. The references to concepts, historical events, academic and theological institutions, and the works and views of many often obscure writing can be bewilderingly complex. Tinderbox provides a means of mapping all this material and its interconnections that is not only effective but enjoyable. By this I mean that, rather than just jotting a new piece of information down somewhere, one can always find a suitable Tinderbox note, or if none exists create one within an appropriate container to accommodate it. Being a sucker for nice visuals, I take great pleasure in organising colour schemes and attractive layouts in map view. And while this is not strictly necessary, it helps to keep one motivated in a way that other approaches may not. As a relative novice–I have to say I’m completely out of my depth with most of the technical discussion that occurs in the forum–I’m still using Tinderbox at a fairly basic level, but it’s certainly proved its value. If it’s of any interest, I’m happy to post an outline of the way I’ve set it up, and I’d certainly welcome suggestions as to where to take it next.

I’ve probably strayed a little far from the topic, so I’ll conclude by saying yes, I completely agree that the most important aspect of any method of note taking or organisation of research data is the quality of thought and analysis that goes into assembling and recording the material. No software is a substitute for this vital step, however much it may contribute to its subsequent usefulness. And there are certainly programs I wouldn’t want to be without, not because they are necessarily the best in any absolute sense, but because they complement effectively my particular way of working: apart from Tinderbox, these include Scrivener, Yep, Keep It, Bookends and Mellel.


Yes, there absolutely is. Think the scientific method: Hyptoehses, test/document, review. Lather, repeat, rinse. Even if you’re only doing this work for yourself, writing down the process is critical, else you’ll find yourself repeating work, reproducing mistakes, etc. And, as you point you, you may find a hidden gem that you can contribute. Again, this is what the whole knowledge-making and publishing process is about.

You are not. this is universal. As I note, you’ll want to develop a process that helps you re-discover your insights when you need them, not necessarily to retain them in your RAM memory.

Would love to see it.

…totally agree RE this. Tinderbox is a tool in the toolbox. I too use many other tools. Not sure I’ve you’ve seen this article–it is a year old but still relevant: An end-to-end Professional, Academic, & Personal 5Cs of Knowledge Management Workflow (Updated).

1 Like

I agree. Another reason, if nothing else, is to remind your future self either/both why you used that approach in this context or the steps needed to reproduce it. Unless one is fortunate enough to study/work in one homogenous niche or are very dogmatic about process, remembering the nuances of the structure of past note-taking is value easily lost in time.

1 Like

It seems to me that a significant motivating factor for Luhmann was to create a method that enabled him to trace theoretical ideas relating to his views on systems theory. He was a sociologist, but he was also a systems theorist, writing volumes on the topic including:

1993 - Law as a social system
1998 - The economy as a social system
1999 - Science as a social system
2000 - Art as a social system

Anyone of these topics would take me a least a few lifetimes to cover:-).

Please please, post your outline (with pictures :slight_smile: ) when you get the chance. I always find learning how others do research, how they wield the tools they use, and how they adapt over time is one of the more fascinating aspects of this forum. The technical depths of Tinderbox are interesting, but the real world uses of this or other techniques is really the point of participating here, IMO.


Thanks for your interest and encouragement. I’m about to reorganise my Tinderbox projects as I’ve reached a point where, rather than have everything related to the very broad field of Early Modern Philosophy in one document, I think it may work better to create separate Tinderbox files for particular projects within this field. I’ll see how this goes before documenting the whole process. One thing I’ve taken away from the knowledge management article you mentioned, @satikusala, is in addition to any external interest, the value for oneself of reflecting on and formally documenting the details of one’s approach.

Before you embark on this, I’d recommend you take a look a the last few weeks of meetups. You’ll see the positive results that others are having with the linking and indexing (terms, categories, themes, contacts, location, etc.) strategies that I’ve been developing. I have a feeling you’ll experience great efficiencies by centralizing and linking your knowledge.

1 Like

Thanks for this advice. It may turn out to be a case of restructuring the existing document, starting with a clearer sense of its overall purpose and parameters. When I began using Tinderbox mid-2021 I started with a few notes that progressively became multi-generational containers on a rather ad-hoc basis. I ended up with notes on the same author or topic in different places, and it took a lot of drilling up and down to get from one to the other. Re-visioning the field, and adopting either a flatter structure or more use of aliases (or both) may be a more productive approach.

Some time back I accidentally started off quite a discussion that you might find useful:


I remember that James Fallows made the observation that associations are better than hierarchies (or similar wording) and that idea has stuck with me.

1 Like

Don’t be afraid of re-structuring. Yes, it is work you don’t ‘have’ to do, but I find enlightened self-interest wins out. Tinderbox is very forgiving in this respect. Generally some re-structuring improves messy old document as I go back and, for instance, apply prototypes my old self would not have thought to use. If I need to export, e.g. for a report, I might review the doc’s structure (the doc’s outline) to make it easier to generate a better linear output of info.

I recommend @satikusala’s suggestion. Even if you don’t want to fully embrace the sort of modularity he shows, I’m sure some of the techniques can easily be borrowed into your own work. I think the pattern of auto-generating notes about specific subjects (terms, keywords, tags - whatever noun you use) is very powerful. Not least there is a single note in your doc explaining that thing. It might be to explain to others (e.g. as an endnote or appendix in a report) or simply a note to self to remind you what that thing is and/or why you are taking the trouble to note it.

Indeed, extracting terms to a canonical (single source reference) note seems very much in the spirit of Luhmann. It’s too easy to follow his ‘system’ rather than the ideas driving it. for instance, Luhmann didn’t have digital links.


Agreed. amplifying that I presume you mean organisational/decompositional hierarchies rather than where a note lives within your TBX.

I make the latter point as you generally don’t want a map (i.e. an outline view container) with 00s or 000s of notes. As long as they are linked—or have attribute values allowing for easy querying—where they live in the TBX stored structure (hint: it is the outline view) matters less. It is worth noting as some people are uncomfortable with things being stored where they can’t ‘see’ them. The problem with that it scales badly, even with a very big screen.

“Restructuring” notes and documents – aka incremental formalization – can be a very effective learning exercise. Perhaps the most important aspect of working with notes. As you work with prior notes, form new notes from the ideas or data in them, or summarize them, or create linkage between them, or even discard irrelevant ones, you are headed down the path of emergence and discovery. That is what makes note taking exciting, regardless of how you go about it.


Really good points there.

A point to consider here—with no ‘correct’ answer—is it may help to keep those ‘old’ notes if only for provenance reasons. In such a case you can still shunt the old notes off to a different part of the doc, e.g. not your current map/view. Also for such old notes you might want to ensure they aren’t running rules/edicts. Perhaps apply a new/different template so they also get excluded from queries you actually intend to run against newer notes.

1 Like

Just a quick update: After considering the various suggestions made (thanks @mwra and others for these) I’ve decided that, rather than structuring the document as notionally accommodating the whole of Early Modern Philosophy, I would make the projects on which I’m currently working the primary focus. I have two recent conference papers that I’m looking to expand into journal articles, as well as an ongoing translation project. The three projects now have their own dedicated containers which occupy a prominent position in the opening map view. I’ve just started moving all subsidiary notes (and child containers) into these top-level containers. For shared notes I’ll use aliases. A further intention is to expand use of prototypes and user attributes. The prospect of a rejuvenated document with enhanced functionality is a great motivation, and I’m really looking forward to seeing how it works out.


Indeed, I keep old notes because I often find they mean different things to me at different times of consultation because my thinking has invariably changed. Sometimes I see things in my notes I had not seen before, other times I am reminded of how I used to see things and that such a perspective once seemed natural.