# Mind Maps and The Archive / Zettelkasten

edited March 2021

I am working on an interesting project with about 5 other people. It covers a broad range of engineering and science specialties, and has a strong research component to it (although that is not its main purpose).

In thinking about how we were going to undertake the work, we created a mind map (using Miro) to visually show the relationships between the technical aspects of the problem, and the processes and event paths that lead from various causes to various effects of the phenomena/behaviour that we are studying (apologies for being so vague, but it would take a thesis to get into this any deeper). So this is a technical mind map, not an organizational one.

One of our tasks is to gather and integrate information about poorly understood factors and processes driving the studied behaviour. It occurred to me that our gathering and integrating efforts would be greatly assisted by bringing all of the information into a Zettelkasten, and that in our initial mind map we have already established some underlying connections between the bits of information (although not all, by a long stretch). We are not starting with an apparently unrelated set of ideas, but rather already know they are related, although we don’t necessarily appreciate just how. As we gather information, I see us expanding the “nodes” on the mind map into zettels (or several zettels) and using the connections between the nodes to guide initial connections between zettels. And as the Zettelkasten evolves, I think so too will the mind map and its connections.

I hope this isn’t too general to make sense to you. I am wondering if others have experience with a similar process, of using a mind map as a starting point for a zettelkasten, and if so, if you could share your experiences in doing so. I would like to be aware of pitfalls (and hopefully avoid them), and also understand ways in which this can be done effectively and efficiently. And I’d like to better understand how the two tools can interact with each other.

• When I read your description of the process, I immediately thought of Tinderbox. Not everyone would want to use such a complex program, but it will do things others will not. If you are not familiar with it, you might want to take a look at it just "for reference". There is a series of training videos for it: https://forum.eastgate.com/c/Videos-to-help-you-get-started/17. A user also posted an example Zettelkasten using Tinderbox -- https://forum.eastgate.com/t/a-tinderbox-zettelkasten-starter-file/3271. There are more videos listed here: https://forum.eastgate.com/t/dominique-renauld-video-tutorials/21 and here: https://youtube.com/channel/UCOdcySa5EQVRz_I6ZXFRc8A/videos.

For interest only -- Tinderbox is not for everyone. But it is often worth knowing what is out there.

• I had a bunch of notes from a book lying around and I just could not figure out how to turn them into good notes for my Zettelkasten. Around that time @Phil made this post on idea indices. I immediately took a paper and drew a mind map based upon the notes and it really helped me see the big picture of what kind of notes I would be writing.

So yeah, I'd say that mind maps are quite beneficial for seeing a high-level view of things.

• @MartinBB Thank-you for the suggestion. I have looked at and played a bit with Tinderbox, but I don't think I'm looking for another tool. I'm just trying to understand how to progress from a mind map to a Zettelkasten, and how the two forms of information and ideas can interact with one another.

@henrikenggaard Do you see a correspondence between elements in a mind map and elements in your ZK, or is that taking things too far? I'd like to figure out how to use a mind map to start populating a ZK, but then conversely, as the ZK evolves, how to revise and improve the mind map. I find mind maps particularly useful in communicating complex relationships to people in our engineering team and to clients, but I think the process of creating the ZK will result in a more comprehensive and integrated set of ideas (when all is said and done).

• @GeoEng51 Interesting thought with updating the mind map after the fact. I've only ever used mind maps for "exploration" of ideas.

Looking at my own "mind map" of my notes, I think that a big help in them was that they were very high-level and abstract. Some items on the map are connected to many notes and some to few. The issue is then that I can't really see this high-level view in my Zettelkasten. Perhaps through structure notes? Perhaps?

• edited March 2021

@henrikenggaard said:
I had a bunch of notes from a book lying around and I just could not figure out how to turn them into good notes for my Zettelkasten. Around that time @Phil made this post on idea indices. I immediately took a paper and drew a mind map based upon the notes and it really helped me see the big picture of what kind of notes I would be writing.

So yeah, I'd say that mind maps are quite beneficial for seeing a high-level view of things.

@Phil's post supercharged me. I'd not considered the mind mapping connection but your right on the mark that they help gather the big picture together with the parts showing the path.

A mind map is a 3D structure note. The difference is the level of detail. For a mind map to have a structure note's details, it would be huge and cumbersome. It would grow huge with the base note links and level-1, level-2, level-3 linked notes. It would be too easy to become lost in the forest and miss the individual trees. As an example, my Coding Structure Note has 27 links on the base note and 63 level-1 linked notes, all with their own set of links. I've not yet figured out to get a count of the level-2 and level-3 linked notes. I suspect that by the time I got to account for the level-3 links, my entire zettelkasten would be included! I've been building this huge 3D mind map call my zettelkasten for a while now.

Don't tell anyone. This is a super-secret between you and me. I just had the idea that it would be fun to see if I can work Keyboard Maestro magic or conjure some bash pixy dust and figure a way to count the 'degrees of separation' two notes have.

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

• @GeoEng51 said:
I am wondering if others have experience with a similar process, of using a mind map as a starting point for a zettelkasten, and if so, if you could share your experiences in doing so. I would like to be aware of pitfalls (and hopefully avoid them), and also understand ways in which this can be done effectively and efficiently. And I’d like to better understand how the two tools can interact with each other.

I should have said a bit more. I was struck by this, and the reason why I thought of Tinderbox is that you don't have to have an interaction between two tools. You have both "Zettelkasten" -- or something similar -- and mind map in the same program. In fact, it is more than that -- you can have many different mind maps (or visual representations of the data, if you want to think of it that way) available in the same program.

I've recently been helping someone to analyse some data, and my approach to this was to pull it all into Tinderbox and work inside that. I still have The Archive as my repository for notes, but I find Tinderbox is a better tool for analysis of material pertaining to a discrete project. One of its advantages is the facility for "incremental formalisation", a concept that I learned about on the Tinderbox forums. Stated a bit crudely, you can steadily add metadata to items as you go along, and build a more complex picture of relationships, links, hierarchies, and so forth.

I'm not suggesting you rush off and start working with Tinderbox, but from your description of the work you are doing it does seem exactly the sort of task that Tinderbox was designed for. It is not an easy option, though. It can be a bewildering program!

Some views of the same Tinderbox file -- the same dataset presented in different ways. You just switch from one tab to another.

• @henrikenggaard said:
@GeoEng51 Interesting thought with updating the mind map after the fact. I've only ever used mind maps for "exploration" of ideas.

Looking at my own "mind map" of my notes, I think that a big help in them was that they were very high-level and abstract. Some items on the map are connected to many notes and some to few. The issue is then that I can't really see this high-level view in my Zettelkasten. Perhaps through structure notes? Perhaps?

That is one way to look at mind maps - as providing a high-level view of a ZK, and the ZK as providing all the detail. However, that kind of relationship belittles the function of both tools. It might imply that the mind map "dumps" and shows what is initially (and somehow fully formed) in our brain, and all the ZK has to do is fill in the details.

I believe (without yet any solid proof) that the relationship between a mind map and a ZK can be much more complex and more integrated. For example (and not meaning to limit the discussion), in the process of creating a ZK we will hopefully discover connections, associations and perhaps whole lines of reasoning that were not initially apparent to us and thus not initially included in our mind map. In part, we are creating the ZK to multiply and leverage our thinking capabilities, generate new insights and challenge our brains to consider concepts in different ways. Hence, there is a need to go back and revise the mind map, to show what we have discovered in the process of creating the ZK.

I would like to think that the mind map and the ZK can be interactive tools, both of which allow us to progress with our consideration of a topic or set of ideas. And there could be many cycles to this interaction. Perhaps this cyclic interaction, working on our ZK, updating our mind map, working further and perhaps revising our ZK, updating our mind map, etc., will result in a superior product to just working in the ZK (with no thought of a mind map).

I've often wondered - once my ZK gets sufficiently complex, how will I ever explain what it is about to other people? Writing a set of papers or a book is of course one way. But I think that a well-illustrated mind map, that is integrated with the essential findings of the ZK, could also be a powerful tool to communicate those findings. And if someone could figure out a way to illustrate a truly 3D mind map, the communication could be even better.

Note that at no point do I suggest that the mind map contains anywhere near the level of connections or details that exist in our ZK. It is not meant to be a one-to-one mapping of the connections between zettels - I think that would be counter-productive. Rather, it is meant to communicate, at a more general level, the insights that we have gained from our ZK.

Does this make sense or am I just dreaming in technicolour?

• edited March 2021

@MartinBB Thank you for your detailed post. It is both causing me to stretch my brain and at the same time scaring me (do I really have to dig into Tinderbox to understand what you are talking about?). Could you read my response to @henrikenggaard (just posted) and see if that makes any sense to you? It looks like Tinderbox is providing more of a one-to-one mapping of all the zettels in a ZK, which I think is too complicated and perhaps counter-productive from the point of view of understanding (or communicating) the essential findings of a ZK. But I'm not arguing the point; I'd really just like to understand this myself. It may be that you are way ahead of the game and I will have to struggle to keep up, but I'm wondering if what I tried to express in the other post has any resonance with you.

• @Will The idea of a mind map as a simplified, 3D structure note is an interesting and certainly valid one. Could a mind map be more than that? Could it go beyond illustrating structure to providing insight?

• @GeoEng51 You deserve a fuller reply than I can write at this moment, but I will try to draft something better a bit later. However, I think one of the key concepts you have mentioned is "cyclic interaction". One of the standard texts in qualitative data analysis is Miles and Huberman's book on the subject -- https://books.google.co.uk/books?id=3CNrUbTu6CsC&source=gbs_similarbooks -- and it contains a graphic depiction of the methodology adopted by researchers in this field.

I hope this will show how similar this is to what you describe -- and quite a bit of my work in psychology (not to mention history) has this sort of pattern. In my doctoral thesis (finished over ten years ago) I wrote: "The collection of data together with three kinds of analytical activity (data reduction, data display, and conclusion drawing / verification) form an interactive, cyclical process, as described by Miles and Huberman (1994)".

So I think we are either on the same page, or adjacent ones.

• Sorry, I couldn't resist.

More on the background here.

• Thanks for the reference, @MartinBB! That seems to supports the knowledge cycle stuff we have on our blog:
https://zettelkasten.de/posts/knowledge-cycle-efficiently-organize-writing-projects/

I realize I never uploaded the diagrams anywhere that picture the cycle(s) mentioned, so here you go: knowledge cycle in blue, writing cycle in green.

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

• @GeoEng51 I believe (without yet any solid proof) that the relationship between a mind map and a ZK can be much more complex and more integrated. For example (and not meaning to limit the discussion), in the process of creating a ZK we will hopefully discover connections, associations and perhaps whole lines of reasoning that were not initially apparent to us and thus not initially included in our mind map.

This! Definitely agreed. I was speaking from a point of where the tools (I know of) are today, but I can follow your train of thought to what-could-be.

In TiddlyWiki I have added a plug-in for visualizing connections between notes. I don't use it all the time, but sometimes I decide to dive in and see if there are interesting connections to find. One of the things I like about it is that it exposes 2nd and 3rd connections. If the 1st order connections are those directly linked from a note, then the higher order are those which are 2, 3, etc. steps removed. Like the Kevin Bacon game.

The difficulty, as you get at, is an integration which is deep enough to allow for editing in the mind map space and note space.

• @thomasteepe

Interesting way of creating thought chains when you don't have them mapped out at all. I like this approach - is it of your own devising? I will have to test drive it. In the past, I've used mind maps mostly to flush out processes and event paths of which I already have a general idea - as a minimum knowing some starting conditions or events and some end outcome. It seems your approach would lend itself to generating a mind map when at first the process or chain that you are attempting to define is open-ended.

I can see how it makes the generation of mind maps more flexible and at the same time might integrate well with progressing your Zettelkasten.

I like the thoughts and questions that you have posed, as well. Thank you for spending the time on this!

• @GeoEng51 said:
In the past, I've used mind maps mostly to flush out processes and event paths of which I already have a general idea - as a minimum knowing some starting conditions or events and some end outcome.

A hurried response again as I am pressed for time -- I have never thought of mind maps this way. I first saw them on a BBC program back in the 1970s, presented by Tony Buzan, and called "Use Your Head". He gives a summary of the method in this video:

This is how I conceptualise a mind map, and how I have always used them.

• I have used The Brain for years and it has recently added better functionality for Zettelkasten in v.12. It's also useful for team use. But, expensive. https://www.thebrain.com/products/thebrain

• @allan said:
I have used The Brain for years and it has recently added better functionality for Zettelkasten in v.12. It's also useful for team use. But, expensive. https://www.thebrain.com/products/thebrain

WOW I felt in love with TheBrain since the first versions, but I've never entered full because of:

1) afraid of the export options: I always evaluate the exit plan.
I see TheBrain exports in Json, is it?
How flexible and "universal" is this Json?

2) poor support on mobile: since I use lot of iPhone and iPad, a good mobile App is mandatory for me

May I ask you what was your experience with TheBrain for knowledge management, and why you moved away?

• @allan said:
I have used The Brain for years and it has recently added better functionality for Zettelkasten in v.12. It's also useful for team use. But, expensive. https://www.thebrain.com/products/thebrain

We're trying out Miro, which also seems quite capable.

https://miro.com/

• Here are some miscellaneous ideas about mind mapping.

• Mind mapping software has seen a rapid evolution over the last 25 or so years - from a basic imitation of paper mind maps to the whopping enterprise editions of today. Paper mind mapping however hasn't seen anything like it, in spite of the optimistic claims about ushering in a new era of mental literacy. But, without irony: If we adapt our expectations about timescales, the Buzan brothers may be right after all.
• Imagine the evolution of a paper-based mind mapping method driven by the tools used for software development: a community of users, buyers or subscribers sends complaints and feature requests etc. to a team of developers - I think we would have something really useful after a couple of years. As far as I understand it, we have no viable business model that supports this - Andy Matuschak and Michael Nielsen deal with economic aspects of "transformative tools for thought" in their essay of that name. (Side note: I have the vague impression that the Bullet Journal concept has been developped by its community of users with some success.)
• I suspect there is a massive discrepancy between Buzan's canonical set of rules for mind maps and the vast majority of maps people use in their lives, and for very good reasons. Most maps shown in this forum arguably have a defiantly low degree of rule conformity.
• The 4x4 method shown above is the most recent release in my tinkering with problem solving methods. I am unaware of exactly the same combination of well known elements, so in that very limited sense it is of my own devising. (@GeoEng51)
• @thomasteepe said:

• I suspect there is a massive discrepancy between Buzan's canonical set of rules for mind maps and the vast majority of maps people use in their lives, and for very good reasons. Most maps shown in this forum arguably have a defiantly low degree of rule conformity.

I agree (and to respond to @MartinBB as well) - I started off using mind mapping software about 10 years ago and was immersed in the Buzan way of creating and using mind maps. That worked well for certain kinds of tasks, but later, as I gained experience and tried to apply mind mapping to other types of tasks, I found that I needed to change the Buzan rules considerably.

For example, I am currently working on a mind map that starts with different ways that gas can be generated in an artificial or "created" landscape (think large scale). It attempts to "map" all the processes whereby gas generation can cause failure in the final landform - typically this involves impacts on ground strength, groundwater flow, and surface settlement. Some of these processes I understand, some I can only guess at, and there are others of which I am ignorant.

There are a number of process paths that can be defined that lead from the initial conditions to the final results. They are not all unique but can branch, interact and/or re-combine. This is not a matter for a "central image" with radiating, simple spikes, but a case for starting with a number of factors (on one side of the mind map) leading to a number of outcomes (on the other side of the mind map), following non-unique and interacting paths. My intent isn't to get into the details here but just to indicate how far from the original Buzan concept we can get. Purposefully at low resolution, so that you just see an overview of the pattern, is one portion of the mind map (still under development):

This is a far cry from the early "Buzan"-type mind map and is the level of complexity to which I was referring in my earlier questions.

• The 4x4 method shown above is the most recent release in my tinkering with problem-solving methods. I am unaware of exactly the same combination of well-known elements, so in that very limited sense it is of my own devising. (@GeoEng51)

I see the matrix method that you describe as a way of asking questions, pursuing avenues of thought, and generating new ideas. If you desired, you could use this exercise to generate a more "traditional" mind map (whatever that term means) or you could use it to drive the organization and connections in your Zettelkasten. I like that it enables your thinking process.

This leads me back to one of my original questions, which is what are good strategies for using a mind map (of any type) to interact with our Zettelkasten? I believe one of Buzan's early ideas was good - try to keep each "arc" of the mind map simple. I don't limit myself to one word or one image per arc of the mind map, but I don't try to enter an atomic note on each arc either. Or is there even a one-to-one correspondence? Maybe, in one case, you generate one zettel from an entire arm (sequence of arcs) on your mind map and in another case, you generate 10 zettels from one arc.

Perhaps I am getting obsessive about this. I appreciate everyone's responses and comments on this post, which have already led me in a number of new directions. I think I just need to try out some ways of interacting between the two tools and having them inform one another in a cyclic fashion.

• Interesting! I find myself wondering at what point a mind map becomes a diagram, and vice versa. But I think I will leave exploration of that rabbit hole to others!

As an aside, my elder brother is a geophysicist who worked in exploration in the oil and gas industries, and is still involved in analysing data from gas fields. And my neighbour is a geoscientist who has been doing research into earthquakes. Small world?

What we seem to be discussing here is the difficult problem of needing or wanting to have different representations of the same data. I have run up against this countless times, and have never found a solution that satisfied me. On the one hand I find that visual representations like mind maps or diagrams are best for an overview or for seeing networks of relationships, but they don't usually allow for much detail or text (though iThoughts, for example, will allow notes to be attached to any item in a mind map). One reason why I have dabbled with Tinderbox is that is possible to combine visual and textual in the same program, but it is not an easy program to work with, and is mostly beyond me. Moreover, it doesn't take long before a mind map/diagram has too many branches and links for it to be easily intelligible (I think the classic Buzan mind map resists this problem because of the insistence on a central node, whereas diagrams without one are more likely to get chaotic).

At the moment I am experimenting with Hook to see if simply linking items is a useful strategy. I haven't done a lot with this yet, but I know that I can insert a link into an iThoughts map, or almost any other document, that will point to a single note in The Archive. I also know that I can copy an "item link" in iThoughts that will take me from a note in The Archive (for example) straight to the arm/branch in iThoughts. However, all such links are fragile (alter a filename and they break), and will usually only work on a Mac, and not on iOS.

I get the vague impression that there is a growing fashion for linking, as more and more programs seem to be offering the feature. If deep linking (not to a file, but to an element in a file) could be made robust I think it would solve some of the problems I've encountered. There are always flies in the ointment, but at the moment some sort of linking is about the only option I can see.

• @IvanFerrero said:
May I ask [...] why you moved away?

"I have used ... for years" indicates that the person is still using the thing. "I used ... for years" indicates that they no longer use it.

• @MartinBB said:
What we seem to be discussing here is the difficult problem of needing or wanting to have different representations of the same data. I have run up against this countless times, and have never found a solution that satisfied me. On the one hand I find that visual representations like mind maps or diagrams are best for an overview or for seeing networks of relationships, but they don't usually allow for much detail or text (though iThoughts, for example, will allow notes to be attached to any item in a mind map). One reason why I have dabbled with Tinderbox is that is possible to combine visual and textual in the same program, but it is not an easy program to work with, and is mostly beyond me. Moreover, it doesn't take long before a mind map/diagram has too many branches and links for it to be easily intelligible (I think the classic Buzan mind map resists this problem because of the insistence on a central node, whereas diagrams without one are more likely to get chaotic).

Agreed.

I use MindMap in the brainstorming step, then I export the image / outline and work on my Apple Notes.
I think too complex mindmaps generate confusion, since you lose clarity.
A plain text note, on the other hand, allows you to focus on the specific note.

• @MartinBB said:
"I have used ... for years" indicates that the person is still using the thing. "I used ... for years" indicates that they no longer use it.

Eheh thank you...again... :-D
I’m a self taught, and I’ve never understood (or “I never understood”?) the difference ;-)
The past tenses have always been hard to manage, since I see they are much different from Italian.

• @IvanFerrero said:
The past tenses have always been hard to manage, since I see they are much different from Italian.

The problem for the Italian speaker is that this English tense looks like the passato prossimo. But in English grammar it is called the present perfect. The name alerts us to a significant difference. More details here: https://ef.co.uk/english-resources/english-grammar/present-perfect/.

• @GeoEng51 said:
I see the matrix method that you describe as a way of asking questions, pursuing avenues of thought, and generating new ideas. If you desired, you could use this exercise to generate a more "traditional" mind map (whatever that term means) ...

I found my experiments with this frustrating, for two reasons:

• My mind maps for problem solving were highly asymmetrical - there was one dominant branch and the rest of the sheet was almost empty. Hacks like "hoisting" a subtopic trom the sheet's edge to the central topic didn't work for me.
• I found it awkward to reflect on a part of the map - a "reflection subbranch" with ideas about obstacles, gaps, criticism and open questions always felt ill placed. (And these items are crucial for my thought processes.)

... or you could use it to drive the organization and connections in your Zettelkasten.

In general, this worked very well for me. Nevertheless, the matrix method has for me the following benefits:

• smaller "thought units" in one cell, with a higher frequency of refocusing,
• a better synopsis of several thought units and
• a workflow with less interruptions - I can spend some time with one A4 sheet.

This leads me back to one of my original questions, which is what are good strategies for using a mind map (of any type) to interact with our Zettelkasten? I believe one of Buzan's early ideas was good - try to keep each "arc" of the mind map simple. I don't limit myself to one word or one image per arc of the mind map, but I don't try to enter an atomic note on each arc either. Or is there even a one-to-one correspondence? Maybe, in one case, you generate one zettel from an entire arm (sequence of arcs) on your mind map and in another case, you generate 10 zettels from one arc.

Unprecise as it is, the rule of thumb "one arc - one unit of thought" might be useful.
But how to set the boundaries for a "unit of thought"? Arguably not as a single key word, and arguably not too large - the smallish mind maps used in the matrix with a maximum of 2 branch levels are the pragmatic option I use now.

To use one last quote: