I am a literary omnivore, at least regarding good books, but I suffered from the chronic illness of bookworms for a long time: gluttony. My default pastime is reading (and sports), but for a long while, I did not ask myself:
Why and how do I read? And what do I want to get out of it?
Focusing only on the amount surely honed some reading skills, but there is more beyond the joy of reading. You can read for joy and simultaneously learn something. At least you think you will remember something until two days go by and you forget everything.
There needs to be a better way. Here is how my thinking about reading evolved, with a few cul-de-sacs and wrong turns along the way.
Fire and forget
Fire and forget is a simple algorithm in control theory, which is remarkably good: if you want to go from A to B, you check whether you are going in towards B; if not, you turn towards B, then repeat until you are there. The name plays to the fact that it only looks at the current situation.
How does this relate to reading?
When I threw myself into papers and textbooks, without even organizing, horribile dictu, highlighting, the material, I was doing fire and forget control. It was efficient (not a lot of work required) but not efficacious (looking up information was a gamble).
To add insult to injury, if you speed up in a frenzy to catch up with the literature in exploding fields such as machine learning, your comprehension will equal to nothing.
The perils of information hoarding
The fire-and-forget scheme shines with its low overhead but fails when looking up information: it scales with the number of papers; in the worst case, you need to go over each to find what you want (you are at least saving the papers to specific folders on your computer, aren’t you?).
At the beginning of my PhD, I started highlighting important passages in my pdf reader. I associated different meanings, such as assumption/experiments to different colours—If you wonder, I do make highlights in fiction and poetry as well. Mostly to capture quotes and elaborate pieces of writing. That just brings me joy.
Highlighting and organizing made reading slower but accelerated information lookup: if I knew that I needed information on a robotics algorithm, I just needed to open that paper, and a quick glance gave me the details. This is where it gets nuanced though:
if you spend more time on extracting information, you end up with reading less papers and having more data.
Let Sherlock Holmes say the obvious for me:
“I consider that a man’s brain originally is like a little empty attic, and you have to stock it with such furniture as you choose. A fool takes in all the lumber of every sort that he comes across so that the knowledge that might be useful to him gets crowded out, or at best is jumbled up with a lot of other things so that he has difficulty in laying his hands.”
Thus, we need to be strategic. This is the reason why I chose the below Thoreau quote for my reading list emails:
Read the best books first, otherwise you will find you do not have time.
This keeps the quantity in check and enables you to organize the information well.
So you need to read less, but better material.1
I leave it to your best judgment for now how you triage reading material. For now, back to the highlights.
Now, you have a bunch of highlights. So what?
You can believe (as I did) that you will be able to recall where to find what. Given that the human brain can hallucinate memories (reminiscent of ChatGPT, isn't it?), I am not so sure. What turns information into your favourite precious metal is the network of tidbits connecting them, providing a scaffolding you can traverse at ease and potentially spot your next good idea. Indeed, there is research showing that networks foster innovation.
If you do not connect the dots, why are you reading in the first place?
Into the gold mine
Highlights give you the information, but since highlighting can become second nature, it will not prompt your System 2, your conscious processing part of your brain (more on System 1 vs 2 can be found in Thinking, Fast and Slow)
You need to think about what you read. And re-read. And finds its place in the puzzle of your information network.
This is much harder than simply reading and takes a lot of time. But it will hopefully develop an organic fabric of knowledge which you can rely on. It will be your intellectual safety net, ensuring that you will (almost always) find what you once read. Even more importantly, thinking about what you read helps you spot inconsistencies, understand the reasoning, and plant the seeds of your own ideas.
Down to the weeds
The perils of technical advice
Before spending countless hours on adapting my system, ask yourself whether you really do need a complex solution? Start with the simplest one and upgrade if you find something is missing.
Managing books
For (non-)fiction, my choice is Calibre (free)
I usually read on my Kindle, and Calibre can export my highlights
to process them I use this script (it outputs markdown files for Obsidian; see below)
Managing academic papers (Zotero)
I use Zotero (free; additional storage can be purchased) for storing academic papers and textbooks
Zotero has a built-in pdf reader; I would suggest you use that because
you can make screenshots as "visual notes"
the highlights can be exported (more on this below)
useful plugins for Zotero:
Zotfile (to store your files in the cloud; thus, avoiding paying for Zotero storage)
Night for Zotero: why are you not using dark mode?
Better BibTex: for better Latex citations
Note-taking (Obsidian)
For my notes, I use Obsidian (free)
it uses the markdown format, so you can read your notes with any text editor
I grab my Zotero highlights with the Zotero plugin (works best with Zotero's built-in viewer)
I can comment and link to another papers more easily (than in Zotero)
I can refer to my meeting notes, etc, seamlessly integrating all the information I have
Well, the reason why September's reading list email got delayed by weeks is exactly because I did not follow my own advice.