Introductions and some ramblings
I've been lurking for a few months, so I thought I should introduce myself. My name is Benjamin. I'm a chemistry graduate student that transitioning into a chemical engineering post-doctoral research position within the next few months.
Back in April, I was searching for a more future proof method of managing my notes. I had been using Evernote for years, but a glitch in their software shuffled around all of my file links, breaking them all. I had to spend days combing through notes and fixing the links or risk losing those links entirely. In searching for alternatives, I stumbled across Zettlr, which has references to Zettelkasten in its preferences menu. That's how I stumbled across this blog and forum.
I spent a weekend reading How to Take Smart Notes by Sönke Ahrens, and I was immediately sold on Zettelkasten as a method. I had always been discontent with my previous note taking, trying everything from annotating directly on the PDF to writing notes in a notebook to making a new text file for each paper I read. I often cycled between all of these approaches, generating a lot of scattered notes. My largest frustration was the feeling that I was losing things. I read a paper, I understood the paper and learned from it, but if I didn't put the paper's concepts directly into action then I would always forget many of the important details. Once I forgot the details, returning to my notes didn't help, and so I was forced to reread the paper (taking a new set of notes), restarting the whole process over again.
My first zettel is from the end of April. I'm just now starting to build up a critical mass of notes, as Ahrens describes it. I think the hardest part has been optimizing and smoothing out my work flow and organization to make working with my zettelkasten as frictionless as possible. I'm still making some tweaks to it, but I think the overall flow is settling into a comfortable rhythm.
I've tried Zettlr, Obsidian, and making notes directly in Devonthink, but I finally settled on using The Archive. One of the largest deciding factors for me was the ability to correctly sort notes that use Luhmann style IDs (1, 1a, 1a1, etc.). I've seen a lot of discussions in the forums recently about this, and I definitely come down on the side of folgezettel being useful to my workflow. I don't use them as a rigid hierarchy, as I've seen @sfast describe. Instead, I use them as a loose association. I put each note behind the note that it is most closely related to.
I think that the concept of folgezettel as a rigid hierarchy breaks down in practice, as it requires that you start with a hierarchal category label, and then work your way down through sub-categories until you reach the note that you want to make. In specialized research, this is an unwieldy approach to note taking. Often, the useful concept to my work is a sub-sub-sub category, and so this sub-sub-sub category note is created firs. The larger "categories" are added later, as they become relevant . So, 27 might be "Localized Surface Plasmon Resonance", 27a is about tuning the localized surface plasmon resonance, 27b is a description of "Surface Plasmon Resonances", and 27c is "Plasmons". This isn't an hierarchy, it's more of a locational association.
If I was to do this as a hierarchy, it would look like:
- 1a Surface Plasmon Resonances
- 1a1 Localized Surface Plasmon Resonances
- 1a1a Tuning LSPR through control of morphology
- 1a1 Localized Surface Plasmon Resonances
This requires too much forethought and planning to be quickly implemented in my notes. So, instead I number them only based on the note that I feel is most related.
- Localized Surface Plasmon Resonances
- 1a Tuning LSPR through control of morphology
- 1a1. Field enhancement around edges and corners
- 1b Surface Plasmon Resonances
- 1c Plasmons
Notice that I didn't tab out each new item to create a hierarchy. It isn't that 1a and 1b are on the same level, and 1a1 is on a lower level. Instead, these notes are related through their physical proximity. They will all still have concrete links, which is a stronger way of associating these notes together. They will still have tags so that I can relate them to the broader context of my notes. I think that Ahrens describes tags as the context in which you want to stumble across a note. This style of numbering is similar. The ID defines another sort of context that I can stumble across that note again without rigidly defining a hierarchy.
The benefit here, as I see it, is in creating different levels of association between ideas.
The strongest association is a direct link. You are defining that this note is related to that note. I think this is the most flexible type of association, as it has its own in-built form of diminishing association strength. Each note is x-many degrees separated from every other note by these links. I think one of the difficulties with this as your only level of association between notes is that you have to click through your links to find the other, lower associations.
A medium association would be something like a tag or a structure note. They define a broader context that relates two notes. Flexible in implementation, but once again you have to decide to search for a tag or open a structure note in order to recover that association.
The ID that you give your note is a low level association. It created a locational proximity. Now, when I go to a note, I can see the notes with similar ID numbers that might, very loosely, relate. In a way, using the date as your ID (i.e. 202008190906) is doing something similar. It tells you which notes you made in a given time period, and these notes may be related because you made them on the same day or in the same week. By using Luhmann style numbering, I am allowing myself to choose which notes to associate locationally, rather than letting my calendar decide for me.
To show this in action, I can use a few notes that I made recently. I'm interested in methods for using light to measure the temperature of a material, i.e. optical thermometry.
Last month, I read a paper about using an approach that relies on the physical expansion and contraction of a material as it gets hotter or colder. I label this note as 27, as I don't have any other notes that are related to this yet.
Last week, I was working on putting some of my publications into zettel format. One of these papers describes my own version of optical thermometry which uses a material's fluorescence to estimate its temperature. I number this note 27a. This note is related to 27 only in that they both describe optical methods of thermometry. The method outlined in 27 is based on the physical changes that occur to a material when its temperature changes. The method outlined in 27a is related to the way that heat is packaged and stored in the material, not on a physical change.
Yesterday, I read about a method for optical thermometry that uses the scattering of light off of the material. This becomes 27b. This method is closely related to 27a as both methods relate to the packaging of heat in the material, not physical changes in the material. As such, they both get links to each other. Neither 27a nor 27b has a strong enough relationship with 27 to get a link. All three are tagged with #optical-thermometry.
If they all got tagged, then why does their locational proximity matter?
27a1 discusses an experiment I did using the method in 27a. This note, 27a1, is not tagged with #optical-thermometry since the point of the experiment wasn't the thermometry technique. 27b1 is, similarly, an experiment I did using technique 27b. Once again, the point of this experiment wasn't #optical-thermometry, so it doesn't get that tag.
I can get from 27b1 to 27a1 by clicking through links (27b1 -> 27b -> 27a -> 27a1). Or, I can see their association because they have the same tag (#optical-cooling, the point of those experiments), or I can see their association in the structure note that outlines a paper that I'm working on writing that includes these experiments.
Getting from 27a1 to 27, though, requires a mixture of links and tags (27a1 -> 27a -> #optical-thermometry -> 27). The step that I take issue with is the movement from following links to searching for tags. This requires the thought, "what other optical thermometry techniques do I have notes on?"
Locational proximity, though, means that I can be thinking about my experiment 27a1, and immediately be reminded of the method outlined in 27 without clicking a single link. I don't have to search for other optical thermometry techniques in my notes, I am just reminded that I have other methods. That could give me a spark of inspiration to try the same experiment using 27.
If I used date IDs, these notes would be spread out over hundreds of other notes. The only way to know that they could be related is through hard links and tags, which ironically requires more rigid categorization than just having those notes close to each other in location. (Note that "more rigid" does not mean that links and tags are actually rigid categorization, just that they are more rigid than this loose locational association.) It's certainly possible that I would find my way from experiment 27a1 to optical thermometry technique 27. They do have an association that I can find if I think to search through my tags as well as my links. Personally, though, I like the spontaneous realizations that can come from seeing two notes in close proximity.
It looks like you're new here. If you want to get involved, click one of these buttons!