Vision on implementing a universal Zettelkasten
I was trying to set up a long-term digital Zettelkasten but was immediately hit by problems around multimedia content, proprietary data formats, heterogeneous Markdown syntax, device and platform limitations, and Internet connectivity. I propose the following vision as a response to the challenge of setting up an all-encompassing Zettelkasten across devices and platforms.
- In the Information Age, a Zettel should be more than just notes. It should be the case that every piece of self-produced content that can link to other contents AND be linked to may be considered a Zettel, whether it’s mostly textual or not.
- Tagging a Zettel can be seen as linking the Zettel to the tag. Since a tag is a purely structural piece of ‘content’ that links to Zettels AND can be linked to, tags themselves can be seen as Zettels from the perspective under the previous point.
- Furthermore, a Zettelkasten doesn’t have to be contained in a single software, service, or directory. It’s somewhat analogous to the argument that the order of Zettels doesn’t matter as long as proper linking is done. I think that it’s not only the order, but also the environments to which the Zettels belong that don’t matter, as long as one Zettel can be properly and conveniently linked to another.
- It should be possible to build a ‘universal Zettelkasten’ where Zettels are linked with each other across platforms and devices. As long as frequently traversed links can be accessed without having to change the platform or device, it should be fine even if no single software or device could properly read the entire Zettelkasten at any one time.
- Thinking of tags and Zettels as two sides of the same coin lends itself to alleviate the challenge of setting up universal tags in a universal Zettelkasten. Although canonical tags usually aren’t shared between software services, it is possible to set up a reopsitory of mostly structural Zettels linkable from all services in question. Universal tagging can then be done via linking to and back-referencing from this repository of Zettels that function here as tags.
- Ideally, the most crucial pieces of Zettels should be stored offline-first on multiple synced devices, in case of low network availability under urgent situations.
- Explicit user-generated IDs needn't be strictly necessary as long as links remain unambiguous.
- The above arguments can perhaps extend to incorporate physical repositories of Zettels into the universal Zettelkasten.
Howdy, Stranger!
Comments
I don't know that I agree with this first point. A ZK is primarily a system for the storage and development of notes and thoughts, with the usual output being written composition. This isn't to say that it can't be stretched to incorporate other types of content, but it is ideally suited to handle written word.
I, personally, would not store all self-produced content in my ZK. There are other formats and structures that are more suited to this. If I want to reference some of my writing in my ZK, then I do so as a reference. I don't need the actual article stored in the ZK to write my thoughts about it or respond to it in my ZK.
I also would worry that having non-textual content zettel risks falling into the collector's fallacy. I can either download a video and link to that, or I can take notes while watching the video and turn those notes into textual zettel. Linking directly to the video risks me not taking the time to process the contents of the video fully. Similarly, as is more common in my line of work, there are often diagrams and figures in scientific manuscripts that feel like they could be useful to clip and embed in a zettel. Alternatively, I can create a textual zettel that discusses my interpretation of the figure, its consequences, and the conclusions that might be drawn from it. I then add the scientific paper as a reference, and I can always return to that paper later. Storing the figure itself in my ZK means that I don't feel obligated to process my thoughts on that figure, and risks me filling up my ZK with random figures with little time spent processing their contents in a meaningful way.
In some ways, I think that ZK is antithetical to the average person's experience of the Information Age. New information and content comes out like a fire hose, and it can feel like we have to drink from the fire hose constantly or risk falling behind. Building a ZK, though, is a slow process that is specialized towards building a depth of knowledge and thought. I think that building a ZK forces you to slow down and think more carefully about what information you want to engage with. You have to curate your attention and engagement.
A tag can only be considered a zettel if you have made a note on that tag. The tag itself is meaningless, except as a way of stumbling across zettel that have been tagged. The tag is a context in which you place a zettel, but a tag is not inherently a zettel. There are software, such as Obsidian or Roam, that can treat tags as their own individual notes. I think that relying too much on this feature would be a mistake. Sure, this software can list all of the contexts in which you used a tag, but that association is only meaningful if you put it down in concrete terms. In other words, that association is only meaningful if you make a zettel out of it.
I don't know that convenience is something to be worried about when it comes to linking. Convenient linking risks falling back on proprietary formatting, which negates the goal of future proofing a ZK. That is the benefit of the UID. It is not convenient to have to search out a UID to follow a link, but it does not require relying on in-built metadata that may not transfer between different software.
The point of saying that the order of your zettel doesn't matter is to try to remove artificial hierarchy. The order of your zettel shouldn't impose a hierarchy on your notes. By storing your "zettel" in other folders and files, you are imposing a hierarchy.
I do agree with you that it is possible and potentially useful to reference files that are stored outside of your ZK, but I treat those links as references. Otherwise, I once again risk not processing the contents of that non-zettel file properly. I might reference figures that I have stored in a powerpoint (making sure the powerpoint has its own UID), but if I don't go on to describe the contents and conclusions of those figures then I risk not remembering the point of that connection in the future. This is why zettel are primarily textual and self contained. By writing my zettel out such that a stranger can understand them, I ensure that my future self will be able to understand that zettel. If I just link to another file that isn't textual, and I don't write out what the purpose of that file's contents are, then I may not remember the reason for that link in the future. I might not remember why that figure was meaningful.
At this point, I think you've strayed away from the purpose of a ZK. A ZK is not just a method of linking. It seems to me that what you are describing here is just wanting an easy way to link between files. The process of building and working with a ZK is so much more than just linking between files.
I think that you have some good ideas here. I definitely encourage exploring ways to incorporate more of your thinking into your ZK, which seems to be what you're trying to do here. I think that you are over-generalizing what a ZK is, though, and in doing so you are losing some of the benefits that having a focused approach to ZK can bring.
This reminds me a little of Digital Minimalism by Cal Newport. Newport is mainly focused on technological services like social media, but I think his conclusion may be useful here. It may feel like there could be a benefit in expanding ZK to incorporate all of these other file types, but there is also a cost associated with attempting to do that. It's important to think about what specifically you're looking to get out of your ZK, and then to optimize the ZK to accomplish that. You may be better served adapting your process to fit a more limited ZK approach. It may be convenient to link directly to a non-textual file, but that convenience does not mean that is a meaningful or useful link in the long run. I would worry that a ZK built upon that type of linking would risk quickly losing meaning a year, or even a few months down the road.
That said, you might look into DevonThink as an application that could be helpful with the approach you've outlined here. It can index all types of files and formats, as well as suggest potential relevance of files to one another. It can read markdown, as well as markdown links. It also has its own tagging system and linking system, though relying on these risks having broken links in the future as with any software-dependent linking or metadata tagging. I use DevonThink to index and store all of my files, including my ZK.
ORCID: 0000-0003-2213-2533
@RobynThorn - what @prometheanhindsight said - it was well reasoned and covered a lot of your points.
My only extra comment is that the more pieces of software you use to connect various files, web sites, and other pieces of information, the greater the risk of something breaking and not working.
I've been using computers since my undergrad days (early 1970s). I've seen all sorts of software, file formats, storage media, etc. Nothing much has had longevity, except text files. I have some real "floppy" disks (about a foot in diameter) and some tape drives that no one can read now, and if they could, I'd have to write my own program to extract and interpret the data.
But why would you need or want a "universal Zettelkasten"? If you asked what that meant on this forum, you'd get as many answers as they are forum members.
This right here deserves to be highlighted and should come as a warning label with each new ZK, fresh out of the box. ZK is the craft equivalent to a wiki-style data dump, and that's what appeals to many people. Building a ZK requires commitment, an eye for the long game, and all those skills we typically associate with building furniture or creating art: precise thinking, grit, patience, but also an openness to the 'happy accident' and the unexpected moment.
As Cal Newport puts it in Digital Minimalism, we live in a society where "screens replace craft," and optionality has a bred in us a willingness to tolerate insubstantiality and what he calls "shallowness" (180). In other words, craft has a presence which most manifestations of digital culture lack. Newport is talking about the embodied physicality of craft in the non-digital sense, but I think his comments apply equally to the ZK method.
Which is why, I think, my ZK feels more 'real', has more of a presence to it, than my other digital projects.
Started ZK 4.2018. "The path is at your feet, see? Now carry on."
Welcome, @robyn_cotton!
It is easy to disagree with all of these points without reading them carefully.
It looks like you are sharing your implementation details on a Zettelkasten with us. It would be interesting to expand on the challenges you where facing to better understand your design decisions.
my first Zettel uid: 202008120915
@Phil I like your comment on the craft of building a ZK. It seems that Cal Newport keys in on that idea in number of his books (I've only read "So Good They Can't Ignore You"). But it's a solid concept. What are we looking for in our creations? Something solid or something ephemeral?
Thank you everyone for your thoughts and input!
@prometheanhindsight Thanks for writing such a detailed comment. I think it's well argued and I'll study it in more detail. I will reply a bit later, when I'm more prepared to do so. For now I'd like to say that I welcome and enjoy reading these critical comments.
@GeoEng51 Thanks for your perspective on time. I'm only an undergraduate student myself, so I could only imagine how much change there could have been; but now that I'm considering it, the concern that the system may break apart is actually very real. By being a user of mainly iOS devices, I found the majority of my data in the hands of a few vendors. With my data dependent on them I cannot decide but suffer from any bug or backward-incompatible change.
Thanks for pointing out that 'universal' can be seen from different angles. Regrettably I wasn't thinking about that while writing this post; I was mostly concerned with the very tangible situation in front of me, for which 'universal' has a mostly unambiguous meaning.
@Phil This is a completely new idea to me. I dare not say I fully understand it, but I'll take note of the advice.
@zk_1000 Thanks a lot for the welcome and interest in the implementation. My primary intention in making these controversial points is to experiment ways in which one may extend (generalise) the Zettelkasten principles to make digital task management work well with digital knowledge network, meanwhile overcoming the many technical challenges arising in this process. I will write in more detail in my reply to @prometheanhindsight.
@robyn_cotton I see a list of claims. Claims are properly challenged with just counter-claims. Perhaps, you create arguments and provide evidence.
I am a Zettler
Like any craft activity, something worth keeping, something that reflects our best impulses to create something of value in this impermanent world. Even if that value is something only we can derive from it.
Started ZK 4.2018. "The path is at your feet, see? Now carry on."