May 14, 2019 · theory cognitive science of learning

Are Books Badly Designed?

Andy Matuschak asks an intriguing question, mostly because the answer doesn't require any thinking: of course not! Books are the best!

But as Matuschak points out, if your goal is to promote learning, non-fictional books are mostly under-designed. In other words, they rely on a model of cognition that isn't true: "transmissionism", or, the notion that understanding can be directly "transmitted" from speaker to listener or from text to reader. In fact, when we learn something new, it is almost always because of something meta-cognitive: we attach the new idea to a web of ideas we already have, we relate the new idea to questions or gaps in our knowledge, we test ourselves ("Do I understand this? What am I missing here, and how can I figure out where my gaps are?") and we try the new idea in novel contexts. Our brains are not bank accounts.

Ditto for lectures. To understand lectures or books, we have to engage in a variety of metacognitive activities. Where Matuschak complains that books and lectures "aren't pulling their weight," I would simply say: their "failure" as well-designed learning tools underscores the real goal of education, which is to help students develop the metacognitive skills and habits that will allow them to access materials and master them, in whatever form those materials appear.