Why Books Don't Work: Constructivism over Transmissionism

Although there are many good points in this post, it is strange and unnecessary to speak of books or lectures “not working” because transmission doesn’t really exist. I, too, am a constructivist. But books and lectures work just fine, thank you— because they help me construct.

This is why I read slowly, and I read few books, yet own thousands of them (about 5000). Because a book is not full of knowledge bullets that are shot into my head. Books are full of clues that I use in my sensemaking process. I own books in anticipation of needing to binge on them when I encounter certain problems.

What I wrote about in my book is designed to help other intellectuals put the clues together to construct their own education. That strategy depends on agency. Reading a book using your agency means you are in control, not the author.

Anyway just because some dumb theory about how books work is wrong doesn’t mean books are broken. Search the term “swashbooking” on youtube for a different attitude.

One is note-taking - if you know your topic well, you'll be able to speed read many books in your field because you're familiar with the underlying patterns, and making handwritten notes about what's different (or little code snippets embodying edge cases, or similar) will help enormously. The great thing about notes is that they're just for you so you can be as opinionated as you like about what's important and omit context. Even if your notes turn out to be wrong or superficial, they'll still function well for information retrieval when you go back to them (possibly with updates).

The other, which I've found very helpful with topics I don't know well where I'm trying to get a grip on the fundamentals, is to visualize future me teaching the material to someone else. As I think I've grasped a concept, I think about delivering a lecture with the confidence of my new understanding. This opens up space for other parts of my brain to ask questions or challenge assumptions, which I then try to address with the knowledge I've accrued thus far. If I can't, then I didn't really get it and I need to go through the material again.

A small but delicious bonus of this technique is when you do grasp the concept well and then encounter your imaginary question and best-effort answer on a subsequent page, or even in a subsequent book.

The key to learning from a book is to interact with it kinesthetically and through metacognition. Skimming, outlining, writing notes in the margins, questioning, are all part of analytical reading. Furthermore, there's syntopical reading which tries to fit things into a larger tree and seeing how things interrelate through comparative reading. It's all really hard work, but the goal is deep comprehension (the kind that enables synthesis of new ideas).

Memory is not the goal, though having a good memory helps when one is diving into a new book or journal paper. A memory aid like Anki can help one get up to speed quickly when there are lots of new vocabulary and concepts. [1] But memory alone will not lead to comprehension -- one needs to converse with the text as if one were picking the brain of someone who is our better. (if the author is not our better, best to just read the book cursorily, extract a few points and discard -- no need to do a deep reading)

That said, having Anki as a memory aid can lower the activation needed to read the literature say of a new unfamiliar field.

Another way to lower the activation to reading the literature is the "land and expand strategy" [2] where you read stories about thing you're trying to read rather than starting with the thing itself.

[1] http://augmentingcognition.com/ltm.html

[2] https://commoncog.com/the-land-and-expand-strategy-for-readi...

It is not an easy feat, but some author can write a book that target their audience’s previous knowledge so well, that it is a joy to read. I believe many have such a science book as an example.

What make books such a timeless tool of learning is that it is “self-pacing”. If a given idea is hard to grasp, I can take my time on it, while easier parts can be quickly skimmed. This is not possible on a lecture or even with video. While the latter can be slowed down/sped up, you don’t know what you miss by skipping even just a few seconds ahead. You loose this spatial orientation that is a given for texts.

EDIT: Mentioning Halmos’s How To Write Mathematics book is probably a good fit here.

You may forget most of a book in a very literal sense but that doesn't mean that connections you've made while reading haven't left some impressions on how you 'think'.

that's particularly obvious if instead of reading you take something like making music. You often 'construct' or build on things you've learned in ways that aren't conscious choices at all. What you learned doesn't just manifest by playing scales but maybe by some subtle stylistic choice or technique you can't recall picking up.

So I think it's just a false dichotomy. A lot of what you construct, in particular subconsciously is based on transmission, and transmission always requires reconstruction anyway, because we're not databases, and every act of transmission is interpretation.

1. Read the assigned material before class. 2. Attend the lecture and any recitations; listen and take notes. 3. Do the homework.

Note that the material is read, heard, written, and used. It worked so well (for me) that I basically didn't have to study for finals.

A lecture or a book by itself is useless, but when used in combination with other forms of learning, can be powerful.

I agree with the comment that constructivism and transmission is a false dichotomy. We use both all the time, together.

I read books and really do remember stuff, so clearly the issue is that the author is an idiot, no?

I'll go further, what kind of idiot reads books to not remember them? The whole point of reading a book is to hopefully enrich your life in some way--even if only for entertainment. How can that happen if you don't remember anything from it?

As for lectures, the problem is that a "lecture" works best when the lecturer and audience are completely prepared. I have only been to two lectures in my entire life where that was the case. Once as audience and once as lecturer--both times were glorious.

My comprehension - and grades - went up when I just stopped taking notes in lecture and listened.

I read books and remember a lot of what is in them.

Are you reading it to find an answer to a problem? Then you'll probably skim, search, scan, and then closely read the part that (may) have your answer.

Are you reading to be exposed to something new, in a structured way? Then maybe it's not important to "learn" it, just to become aware. (btw, awareness is valuable, which is why talks and lectures are still important: to raise awareness and maybe provoke questions to explore later.)

Are you reading to be able to apply the information? And in what way? Just a bunch of facts? Then memorization techniques (Anki, et al) are useful.

But for learning to build expertise? No, a book is not enough. Memorization isn't enough. Even "integrating it with your relevant prior knowledge" (important!) isn't enough. You must do the thing you want to learn and _get feedback_. Generating summaries, explaining it to someone else, etc., are all great techniques, but if nobody is telling you whether what you've summarized or explained is right or wrong, how do you know if you've learned it correctly?

While there's something to be said about conceptual/abstract hierarchies merging (the prior learning effect is strong and can be exploited using curiosity and order control for what you're studying) I do think that with an appropriate theory, there is a universal conceptual schema that can be acquired. Many subjects are like learning a foreign language.

What I'm getting at is the difference between trundling through any particular piece of content and responding/learning based on a philosophical Worldview vs. the handy-dandy hierarchical list of the chapter headings and subtopics already laid out. It is an interesting question. Perhaps it depends on the type of content, and the end goal for learning the info.