I've definitely done the "reading nonfiction and forgetting all of it" thing. But I'm also a big highlighter and note taker. I have a harder time actually doing stuff with my highlights and notes once I've finished the book. And then I feel kind of bad about it (and this is the "bashful confessional" that Matuschak is talking about).
The amount that I read oscillates, in terms of time spent and content. My idealized self would be producing output during the downswing in input, though I'm getting better about it.
Bashful confessional, just like he said.
I'm somewhat skeptical about books being "surprisingly bad at conveying knowledge"... We'll see what he has to say as I read through this.
Lectures not working is something I can agree on. Particularly when not everyone listening is on the same page as far as getting it goes.
The are definitely different cognitive models. Everyone is able to categorize themselves as visual/reading/listening/doing type Learner, but it does go deeper & these aren't mutually exclusive.
A long time ago I read something (I think in the copywriting field) that had different color groups assigned to learning type and personality traits. There were different ways to structure the marketing message. I wonder if that'd apply here?
In a way I guess lectures are like a sales pitch to put in the work to understand something?
There's a note about lecture speed being slower than reading speed. Something I've recently started trying is playing talks on YouTube at 2x with captions on and pausing as needed.
While I do agree that transmitting knowledge is more than just writing down things one hears or reads, it is a good starting point and better than nothing. But that's rather obvious, isn't it?
To understand something, you must actively engage with it.
So maybe when designing a course there should be a transfer exercise part way through? Something that is "same but different"?
It's kind of funny that he talks about some readers having the inner monologue of what things remind them of. I got that. It's actually part of why I take notes and highlight while I read. If I don't, I'm much more likely to do that thing where I've flipped/scrolled several pages and realize my mind has been wandering away on a different but usually somewhat related tangent.
He mentions that these skills don't come easy, and I'm trying to remember how I learned them... Pausing to think about it a bit, I'm thinking it was in Junior high when we had some sort of unit on study skills. That's where I first heard about Cornell notes, even though I don't use them.
The section about metacognition is interesting, and reminds me of some of the stuff I read in college. I'll have to go through my old notes, which I definitely have around.
So revisiting the skepticism I had as to whether or not books are bad at conveying knowledge, I guess it's the metacognition angle that I have going for me. He makes a great argument, though.
I also just remembered there being an entire page on metacognition in the Head First books.
Textbooks can be super rough, though there are definitely some more readable than others. (Speaking of unreadable books, I'm looking at you, unfinished copies of Infinite Jest and Goedel, Escher, Bach).
Also on another side note, fuck textbook companies and University bookstores, and professors who assign copies of books they wrote and don't provide to the library.
In most cases, I don't know that having machines grade questions is with the amount of work that would go into it.
for many students, courses offer a helpful accountability structure.
He probably didn't intend it this way, but this brought back memories of "we're in this together" pacts for the shittier required courses I had to take with my friends at school.
So we need a cognitive model, then.
A "Post-book" medium.
I think MDX will be in that mix (from an implementation standpoint, that is).
Spaced Repetition is an obvious choice.
I was saved many times by Anki. I'd have marathon study sessions listening to binaural beats or Steve Reich and do the sets of cards I built for art history, discrete math proofs, computer science... My greatest SRS achievement was going from a midterm low-C to a B+ in Mandarin. I developed some mild synesthesia from color coding the characters in my cards. When I looked at the tests, they turned colors and I knew I knew it. But again, now all I can do in Mandarin is politely ask if I may ask a person's honorable last name, what I should call them, tell them my family and first name, and that I'm an American college student.
Someday I'll write up the process I adapted from the Supermemo rules to formulating knowledge.
The Quantum Country page was really interesting when scrolling through it, though I didn't know about the SRS via email angle... Very interesting.
Andy Matuschak, “Why books don’t work”, https://andymatuschak.org/books, San Francisco (2019).