Background

My research area is machine learning. My first zettel was created on 02/23/2020 and most of them focus on my research area or research in general. This mostly consists of extra the key ideas in short 4-10 page article. Throughout my life I have read many books, but I remember little about them. For instance, I say books like “The Double Helix” or “The Art of Learning” made a big impact on me, but outside of a few anecdotes, I remember nothing. More than that, I have no notes to go back and refer to. My plan this year was to change this and systematically process any book I read.

The first book happened to be “Team of Rivals”. My first zettel was on 02/06/2021. At the time I had approximately 430 zettels.

The Process

I know little about Lincoln or the Civil War outside of what I remember from high school. Everything was new and everything seemed notable. As I read the book, I kept a notebook of fleeting notes. These were just page numbers with brief descriptions of what I found interesting. Later I went through the notebook and turned each fleeting note into a zettel.

I eventually became overwhelmed by the amount of information I was processing. Since this was just a hobby, something I did on top of my research, job, and family, the whole process took a protracted amount of time. Nearly six months passed between the time I read the first page and I processed the final note. From the time I entered the first note (02/06/2021), I added 252 notes to my Zettelkasten, 138 from this book.

Mistakes

No Plan

If reading is searching, I had no idea what I was searching for when I started. In hindsight I was interested in things like, examples of leadership, causes of the Civil War, history of racism, and great stories. If I had started with those things in mind, reading and processing would have been more productive.

Started Processing Too Early

Halfway through the book, I started processing fleeting notes. I think it was a mistake to start this early, especially given my lack of knowledge of the subject. I probably skipped over information that was related to later parts of the text. I likely missed relationships between information from the first half and the second half. Splitting my time between reading and processing meant I never had the whole text in mind before taking notes.

No Triage of Fleeting Notes

During processing, I went through my fleeting notes sequentially. For each note, I immediately started transforming it into a zettel. I never questioned whether it was worth translating. Towards the end I realized a fleeting note should just be treated as something potentially of interest. There is no obligation to keep it.

No Organization of Fleeting Notes

Since I processed the notes sequentially, I never took the step to look at them as a whole. I missed the opportunity to group notes and develop initial connections. It also made it difficult to synthesize information by relating several notes from different parts of the text into a new insight.

Focus on Exact References

I was meticulous about exact page numbers for quotes, facts, and ideas presented in the book. While it could be helpful to go back and fact check in the future, the need seems unlikely. This focus also made it difficult to extract information that was more nebulous. Some thoughts are based on impressions from a chapter or a book as a whole. Since I couldn’t point to specific page numbers, I was hesitant to create zettels for them.

My Tags are a Mess

I tagged every single note with ‘history’. Doubt that tag is going to be of much use. I chose not to tag notes specifically about Lincoln with a tag ‘lincoln’ because I thought I could just search for his name. Every single note contains a citation to the book with Lincoln in the title, so that doesn’t work.

Future

I am nearly finished reading a biography of George Washington and ready to start processing the fleeting notes. I’m hoping that I have learned something from this first attempt and the process will be smoother and more productive.

Sascha’s Comment: Very nice reflections. One thing leaped to my eye. @boxcariii had one insight in multiple incarnations: You have to know something to learn another thing. You need to know what you can find in text in general in ordcer to find specific things. You need to know the book as a whole to judge which parts are significant. But the issue with the tag. I think we all share this trauma to some extend :)

Christian’s Comment: Oh man, the “my tags are a mess” part resonated with me. I still have notes from almost a decade ago, before Sascha brought up the useful distinction of topic-vs-object tags in our discussions. I guess we all must suffer from experiences like this at leas once :) Thanks @boxcariii for sharing the story and the ‘lessons learned’.