Way back in February, I wrote a post on how I capture my reading notes in Obsidian. I had been using Obsidian for a little over a month and was blown away by its features. It had just about everything I’d been looking for in a notes tool. I quickly hacked together some scripts and proceeded to begin gathering all of my reading notes in Obsidian, using a kinda-sorta Zettelkasten framework.
In the nine months since I wrote that post, a lot has changed. This was brought home to me when a reader asked me how I got notes from paper books and articles in Obsidian. My original post focused on Kindle notes. What about paper books? It just so happened that I recently finished re-reading One Man’s Meat by E. B. White, and I marked up a bunch of stuff along the way–in the paperback copy of the book. It occurred to me that over these nine months, I’ve started to stray from the automated model in favor of a more manual approach to my notes. This is a lesson I learned from the automation I originally set up: the automation was nice, and efficient, but it was also detached and remote. The notes felt distant; I frequently never even reviewed them. And even the UID titles my script generated was distant and meaningless. What started out as something cool and interesting, proved less useful than I thought.
That’s the way it goes when experimenting to find the best way something works for me. Here, then, is how I capture my reading notes in Obsidian today–call it de-automating my reading notes, or switching to manual override.
Reading Notes In Obsidian: A Case Study Using One Man’s Meat.
There are three things that I do when reading a book to get my notes into Obsidian. First, I read the book and mark it up along the way. Second, I created a source note for the book, that acts as a kind of map of content (MOC) for that book. Third, I review the notes I took, decide which are worth capturing as a separate linked not, which are worth noting in the source MOC, and which are safe to ignore. Call these three parts reading, prepping, and curating, and creating.
Everyone has their own way to read to learn. For me, if I’ve got a book in my lap, I also have a pen handy to underline passages or make marginal notes. To help find the places I markup after I’m finished, I keep spare sets of Post-It Flags handy. Each place I mark up, I stick a flag. When I’ve finished reading a book, it frequently looks something like this:
The colors don’t mean anything. I rotate through the 5 available colors in the pack. The placeholders they represent vary. Sometimes it is an underlined passage with a marginal note. Sometimes it is an underlined passage without a note. Sometimes I’ll put the letters “ph” in the margin beside the passage because I like the way it was phrased. If I pause to think about what I just read, I err toward marking it. I don’t go and create my note in Obsidian right then and there. Experimentation has taught me that if I wait until I finish, some of the passages I marked aren’t really worth capturing.
Prepping and curating
When I finish reading a book, I do a couple of things. I update Goodreads. I update my list of books I’ve read since 1996. And then, I go back to the book and review the flags I’ve left. I go through all of them, deciding in the overall scheme of what I just read, they are worth keeping. If I don’t think the note is worth keeping or recording, I’ll pull out the flag. Of course, my underlined passages and notes are still there, but the flag is there to tell me if I should collect that note in Obsidian. Whatever flags remain mark things I want to capture.
Next, I create a source note in Obsidian. A source note is a note that represents the book I am reading. In this case, the source note is my note that represents One Man’s Meat. I have a top level “Reading” folder in Obsidian with two sub-folders: Sources and Commonplace. Source notes go into my Sources folder. I have a template for Source notes to ensure there is consistency in their formatting and content. I discuss and other templates I use with Obsidian in Episode 8 of my Practically Paperless with Obsidian series. When I first created the source note for One Man’s Meat, it looked like this:
Once I’ve got my source note, and I’ve whittled the flags down to those that I want to keep, I begin adding my notes to Obsidian. I do this manually, reading from the highlighted passages, typing them into a note in Obsidian, and then adding any necessary meta-data. In general, I capture notes in 2 ways:
- Capturing a passage in its own separate note and then link to that note from the source note (MOC) using a transcluded link so that the full note will show up in preview mode. I use this method when I think the highlighted passage and my related notes on it might ultimately be linked to other things than the source note itself. For instance, here is a note from White’s essay “Removal” which I created as a separate note. I then link back to the note on the Source note. (I’ll give an example shortly.)
2. Sometimes, I just want to capture a thought that isn’t really a separate note itself. In this case, I’ll just include it as a bullet in the source note. For instance, in White’s essay “Movies”, I have the following bullet point in the source note:
When all of these notes are created, the source note serves as a master for everything I noted in the book. Since One Man’s Meat is a collection of essays, I list out all of the essays and then using transcluded note links to link to the notes I created for each essay; or, I just add bullet points beneath that essay. The images below show what my source note for One Man’s Meat looks like. It is a long note with the transcluded links so I’m just showing the first few screens of the note:
The notes that appear in boxes are actually separate note files in my “Commonplace” folder that are included in the source as transcluded links. You’ll also see that for some essays, I had no notes, but I included the essays there because (a) thought it would be useful, and (b) if I read the book again, I may add more notes. There are 55 essays in the book so you can imagine this source file going on and on like above with my notes.
Some lessons learned
Just because I changed the way I get my reading notes into Obsidian does not mean my original method didn’t work. It just didn’t work as well for me as I thought it would. My initial idea was that I’d save myself a ton of time by automating as much of the process as possible. But I found that I didn’t really absorb my notes when doing that. It was only when I was culling, curating, and typing in my notes manually that they began to resonate with me.
For that reason, I no longer automate my Kindle notes either. I use much the same process as I’ve outlined above, only instead of using the Post-It flags, I use Kindle bookmarks to mark those notes that I want to keep. This has worked much better for me.
I also found that I was defeating myself by automating the titles of the reading notes with UIDs instead of readable titles. I found that for search purposes, it was far better to have a title like “E. B. White on productivity” than something like “0afe6881-6060-4dd0-8a27-2e91ed323ebc” even if the important text was searchable within the note. For skimming search results, I’ve found that a useful title for a note is key.
One more challenge
Between paper, Ebooks, and audiobooks, the toughest challenge is gathering notes for audio books. But I’ll save that for a future post.
Did you enjoy this post?
If so, consider subscribing to the blog using the form below or clicking on the button below to follow the blog. And consider telling a friend about it. Already a reader or subscriber to the blog? Thanks for reading!
Thanks for the clarification and ideas. I used to fold a sheet of paper in half and keep it as a book mark and also jot down` notes and page numbers for printed books. Some of the notes would get pretty long to write out. I would then keep the page/s of notes in the book for reference.
I am currently reading a book from the library so I can’t mark in it. As I mentioned in previous comment – I am using Google keep with dictation / transcription to take notes on the book to save retyping info. I am planning on exporting to a note or notes to be used with Obsidian.
I was interested also to see how you have changed the titles from UID. Your current method seems to make it much easier to locate.
Dan, I still have the habit of using the first or last couple of blank pages that books have to jot page numbers and notes occasionally. For library books, I use a modified version of what I do for audio books. With a library book, I’ll use my Post-It flags to mark places, and then jot notes about passages of interest in my current Field Notes notebook with page number references. Of course, I remove the Post-It flags before I return the library books.
On the UIDs, that seemed like a great idea from in my developer perspective, but didn’t work well in practice for the exact reason you point out: it is much easier to locate a note titled “E. B. White on democracy” than some scrambled UID.
I think it was in Charles Duhigg’s book “Smarter Faster Better” that I came across a name for exactly what you’re experiencing: That making information a little bit “harder” to deal with, by removing automation and requiring some manual interaction with it, enhances your ability to draw valuable conclusions from it. It’s called disfluency.
In the book, the example is that of teachers at a school who ditched their fancy automated reporting dashboards in favor of manually transcribing their student’s homework and test scores and in doing so discovered a number of insights that enhanced their teaching.
Clemens, I love that there is a word for this. I’ve read one of Duhigg’s books after I saw him speak at a convention I attended. Now I think I want to read this one as well. Thanks for passing this along.
Hi Jamie. I found your work this evening after searching for some info on obsidian. Thankyou for sharing your thoughtful evaluation of your notes process after a long trial period. The ability to self-evaluate and adjust is a key skill and a quick look at the rest of your work here shows that you have been doing that for a long time! I look forward to browsing your work in future coffee breaks.
I wanted to share with you a book, in case you haven’t seen it, called “How to take smart notes” by Sonke Ahrens. Ahrens writes about the Zettelkasten method which you are already familiar with. His description is the best I have found to properly describe the real value of the process, which is more about the steps of reviewing notes, re-writing notes and placing them into context within your existing reading and original work. Much of the information about the Zettelkasten method focusses on tools (eg index cards or digital-“second brain”); just like any tool it is the process and technique of the user that matters the most. Those critical steps of reviewing and re-writing notes manually is what you have discovered as more effective with your own reflective evaluation of your process. Ahren’s book talks through all this (and more); it is targeted towards researchers (and also non-fiction writers), so it can seem “too much” for many people, but I am going to make a judgement and suggest that you will get a lot out of it, especially given where you are at in your note-taking journey (and use of obsidian).
Also on the topic of hand-written versus digital notes – there is substantial evidence to show there are measurable differences between the effectiveness of those methods, with hand-writing giving superior results (this is for the case of noting external information; producing original content is an entirely different process). It sounds like you have already observed that the hand-writing and manual processing of your notes is a more effective process for you – the science backs it up!
I think these points relate to the concept of “disfluency” mentioned by Clemens, although I have never heard that term used before.
kind regards, lmo.
Excellent essay. Thanks for sharing. I recently discovered this blog and it’s now one of my favorites.
Why take notes on a book like “One Man’s Meat”? What will you do with those notes once you have them?
Thank for the kind words on the blog, Mitch. To answer you question, there are 3 reasons why I take notes on a book like One Man’s Meat:
I also use these sticky notes to annotate! https://twitter.com/NotionwareHQ/status/1431639245878345729
Hand-scrawling notes in the margins or inside-covers stopped cutting it for me. But writing comprehension notes on my iPad (Notion) doesn’t either, since I read at night in bed and I don’t want to use iPad in bed.
To me, the notes represent salient points, possibly for retrieval. I also have taken book notes in Notion, but t’s really just a series of indented (toggled) lists, a page per book.
I’d like to move to something “proper”. I recently created a “tag” database and have been tagging some notes, but it’s not what the Obsidian/Zettelkasten types would do.
I’m currently reading How To Take Smart Notes, but I think it’s a pretty poor book that shows very little about the system and is geared specifically toward academic types. Indeed, such a full-on note-taking system may not be what I want to do anyway.
But I do want to ratchet things up a bit. In my mind, that is not only about a better software/capturing system, but it may also be about formalising the actual note-taking in the morning, the morning after reading, using those stickies as prompts.
I might return to this good article!
“I found that for search purposes, it was far better to have a title like “E. B. White on productivity” than something like “0afe6881-6060-4dd0-8a27-2e91ed323ebc” even if the important text was searchable within the note. For skimming search results, I’ve found that a useful title for a note is key.”
Yeah, with all due respect, I could not wrap my head around why you would make something that long and confusing your title 😂
NylaTheWolf: I was experimenting with Zettelkasten and looking for a way to auto-name the notes. In true Zettelkasten fashion, as I understand it, notes get titles like 1a, 1a1, 1b1a, etc. I wanted to see if something along those lines would work for me. Clearly it didn’t. 😀