From digital to paper-based
Warning: long post ahead.
Since a number of months, and certainly for about a year ahead I am transferring my Zettelkasten from digital to paper. It will also start using Luhmann-like IDs.
The below text is partly from my own task management and thinking of how to do this. I wanted to share it in case anyone has comparable plans.
My starting situation is:
- I have been working on my Zettelkasten in The Archive since 2018,
- I have a total of about 2500 notes, of which about 750 reference / reading notes (I also do reference management inside my Zettelkasten)
- I used the conventional Date-time IDs
- I did a major overhaul of my Zettelkasten in 2022 after about 1500 notes, bringing much more structure and discipline
- I have another Zettelkasten for fiction writing that will remain completely digital for the moment
First let’s get into the why of the transfer.
TL;DR I am happier devoting less screen time; the increased friction (need to make my notes more concise) is a major benefit for me.
The are good arguments for digital (The Archive of course!) and paper. Though the note IDs and the medium are in theory independent decisions, I discuss them together here.
Paper with Luhmann IDs
- The organization of notes is more hierarchical, especially in the beginning. Of course lateral connections are added, but hierarchy is the primary organizing mechanism. For me, this fits well (better) with the way my mind works.
- It is more flexible (drawings, scribblings). A digital file has a linear order, paper is non-linear. But the difference is not that big: for me The Archive also has all the features I need. That said, I like taking notes on paper more than digitally.
- It solves the analog-to-digital conversion step (paper notes to be transcribed into digital notes), but it creates a digital-to-analog conversion step (e.g., excepts or quotes have to be transcribed).
- It creates more friction, it is more work, and it forces more focus, conciseness and attention. For example you cannot just excerpt 20% of a source you read, because you’d be writing forever. For me, this is a good exercise in discipline. For others, it may be a limitation. Of course, if you are very disciplined, you can replicate such focus and conciseness easily in digital form.
- It invites browsing through your notes, certainly in combination with the Luhmann-like IDs, because all of a sudden there is a logic in the way the notes follow up on each other. There have been many discussions about this on the forum, following the great Folgezettel debate of ancient times.
- It is available offline and forces offline working. For me, there are pro’s and con’s here. The biggest pro is that because I spend my life in front of screens it feels good, healthy and pleasant to be off-screen for my note taking. Then again, I cannot easily copy-paste notes into a digital outline and it is much more difficult to share notes with other people.
- It is, for me, more fun and relaxing.
- I personally do not experience the assumed psychological advantage of writing paper notes over typing digital notes. Personally, I can think as well when I write with a pen as when I type on keyboard.
Digital (The Archive) with Date-time IDs
- It is less of a hierarchy and more of a heterarchy from the very beginning. Because of the Date-time IDs, no note takes precedence over another just by its ID. Notes are always (if you make links of course) both hierarchically and laterally related. This creates a big “sea” or “ball-bath” of notes, which is a more modern interpretation of a Zettelkasten that takes advantage of computer storage.
- Digital readability is superior (my handwriting is bad).
- Editability is a draw: digital is very easy to rename, adjust or add to, while with paper there’s always a limit, like paper size, or the number of times you can scratch through a word and then change it back, etc. Then again, paper is better for fast, messy scribbling.
- Digital portability is superior: I can access The Archive on any computer I work on. Bringing along my paper ZK is at this moment just possible, but will become impossible once I’m finished transferring.
- Digital backup is far superior. Especially with The Achieve and text files, my entire Zettelkasten is just a few MB of data. It automatically links with the sources in my /media folder, etc.
- Digital is faster, much faster. For example: copy-pasting a reference from Google Scholar, or adding a quote of even moderate length is fast and frictionless. This is a big advantage when you need it, although it can lead to lack of discipline (guilty M’lud).
- Sunk cost. In my case I’m dealing with about 2500 digital notes, many of them not strictly (ahum) adhering to the atomic principle (see also with lack of discipline). This will translate into anywhere around 4000-5000 paper notes.
Sources that I found helpful in taking my decision to transfer:
- Luhmann’s 1981 “Kommunikation mit Zettelkasten”, and the translation by @Sascha , which I contributed to (as an aside: if you want to really, deeply understand a text: try translating it!).
- The great talk by Johannes Schmidt “Der Zettelkasten Las Zweitgedächtnis Niklas Luhmanns (you can find it on YouTube)
- Scott Scheper’s book Antinet Zettelkasten, which is basically a long, passionate advocacy of paper-based Zettelkasten.
- Many, many posts here on the forum, but specifically https://forum.zettelkasten.de/discussion/978/is-there-a-benefit-to-luhmann-ids-vs-date-time-ids/p1
Steps I took till now:
- Duplicated the alphabetical index, initially without the links (new numbering so links not available yet). Also duplicated the people index. Integrated other index types (location, organization, etc.) with the alphabetical index.
- Decided on the main categories and the logic of the note numbering system. This is paradoxically “not that important” on the one hand, because any numbering system will do, and “very important” on the other hand, because you cannot change a numbering system afterwards (or only by starting anew). So once you decide you are stuck with it, and of course it needs to accommodate your own preferences. I spent a few months thinking about this, eventually settling on seven main departments and note numbering like 1A1/2B2/… . I am aware of the advantages and disadvantages of this system. Again the great Folgezettel debate and all its offspring come into play. A very interesting contribution was made by @ctietze . in the tread mentioned above, showing that a drawback of the Luhmann-like numbering is that you can only branch off twice. @scottscheper . in his book showed a number of alternatives to branching off, some of which look quite messy (negative numbers, anyone?), but then again: numbering and placement does not matter that much. What matters is that you can find it back!
- Based on step 2, made the skeleton of the Zettelkasten organization: main departments, sub-departments, and sub-sub-departments as far as I had them in the digital Zettelkasten (these were my “super-hubs” or main access points in the digital system). This was a top-down process (see hierarchy). Added cross-references while going along.
- Duplicated the existing structurenotes (and their sub-notes, sub-sub-notes, etc. as far as I had them). Added cross-references while going along. Luhmann had structure notes, but the big credit here goes to @Sascha . In Luhmann’s Zettelkasten structurnotes are but one of the ways of connecting notes. In a digital Zettelkasten they are a primary way of organizing notes, as “paths through the heterarchy”. They happen to be super-useful on paper as well to create the necessary structure and overview. Luckily I had but few structurenotes (< 20, though each one has numerous subs and sub-subs, sometimes sub-sub-subs), mostly related to existing long-running projects I’m working on. I use to start with hubs, i.e., unstructured collections, moving only very slowly to structures, in a process that can take years of reading, researching, and understanding a topic.
This is where I am now. About 800-900 notes in the system now, plus about 125 in the index, and fewer than 25 literature notes for now. To date most notes are “structure” and “skeleton”.
I made a digital duplicate of the skeleton in Apple Notes (my GTD app) so that if I take a digital note I can easily determine its approximate placement. The text-folding feature of Apple Notes was very helpful in what is essentially a long outline.
This process took about 9 months, which is long, but included a lot of thinking time (structure and numbering system) and a lot of hesitation (getting over the sunk cost of 2500 digital notes and assessing the mountain of work ahead).
My main learnings until now:
- If you want to make your Zettelkasten effective, you need to “live in it”, that is use it on a daily basis. This makes the pleasure of using it (more pleasure = more use) a big factor in its effectiveness.
- You cannot easily transfer a Zettelkasten from one medium to another. If you do so, you are effectively building a new one. That is, unless you would make an exact carbon copy, not changing anything in number or structure. Do not underestimate the amount of work.
- A paper-based ZK has many more notes than a digital Zettelkasten. In digital it's so very easy to sin against the principle of atomicity. But even without that sin, for example a hub with 20 links (one digital note, not sinning against atomicity) will become three A6 paper notes.
- The usefulness of structurenotes emphasized again.
Next steps are:
- Update the index based on steps 3 and 4 (doing now).
- Duplicate the existing hub notes (about 300-350) and literature lists. In my system a hub (unstructured collection) normally has a literature list attached to it with the most important references on the topic. While making hubs, update the index with links to the hub topics. Also create cross-references. While making the hubs, convert some hubs into structurenotes. This last action is more Zettalkasten maintenance than migration. I expect this to take a few months.
- Duplicate main notes & literature notes. This can be done in parallel. Once step 2 is finished, I can gradually duplicate the old main notes (> 1200) and the literature notes (about 750). I decided not to heave over literature (reading) notes completely. Literature notes will only contain the reference plus brief information or summary, much like Luhmann’s reading notes (see also Scott Scheper). Longer literature summaries will remain digital (yes, The Archive!), and are to be gradually processed if that had not been done.
Hope this helpful to anyone who considers migrating to paper. I’ll give an update in a few months on how it’s going.
(Edited by @ctietze 2025-02-16 17:24 to fix list markup)
Post edited by ctietze on
Howdy, Stranger!
Comments
Oof that a monumental undertaking! Best of luck!
Can you elaborate what you did before and after this turning point?
Author at Zettelkasten.de • https://christiantietze.de/
@erikh Yikes! You must have a lot of time left
Me - not so much.
@ctietze sure, I was creating many notes, OK in themselves, but not systematically linking them. I have broad interest in many topics, so creating new notes is not my constraint so to speak.
At the moment of my first post on this forum I noticed that I was often creating duplicates of notes: the same or almost the same thought, in almost the same words. Déja-vu. After about 1000-1200 notes, despite extensive tagging, it was not possible to find the original note back, and therefore faster to recreate it. Comments by @Sascha and @GeoEng51 pointed to better structure as a way of preventing it. Interestingly, @Loni and @Will did not consider it a problem.
I thought about it for some time, and then decided that the more structured approach fits better with my habits and ways of working. So, I started creating more hubs, super-hubs (hubs of hubs) and literature lists with every hub, next to structurenotes. I put myself the rule that every new note would have to be connected to a hub, a structurenote, of (in case of a reference) to a literature list.
This required going through all notes till that moment to connect them. The result is a slightly overdone structure, but also that I can find anything back in about 2 steps. I kept on tagging but that became more or less pro forma. With paper now tagging is obviously out of the window.
@GeoEng51 haha, yes it must sound like that. It's not so much the amount of time left, but that I'd like to spend at least some of it away from the screen.
Both an interesting read and an interesting prospect for your experience report!
I am lurking around.
I am a Zettler