Why are different kinds of learning so differently compressible?
If I can work through a textbook in 20 1-hour sittings, I usually get similar results from 10 2-hour or 5 4-hour sessions. But piano isn’t that way at all: a 20x1hr piece simply can’t be learned in 5 sessions IME.
Conversation
One explanation might be that when learning piano pieces, successive sessions rely heavily on previous sessions having been consolidated, whereas many “book-learning” topics are somewhat more breadth-shaped.
Another might be that some tasks drain attention faster.
1
31
Other examples of learning which don’t seem to compress very well:
– learning how to draw
– learning how to design user interfaces
– learning how to write
Ones which seem to compress well:
– learning how to cook
– learning a new programming language
– learning a spoken language
Replying to
“Fundamentals of Piano Practice” makes the case that "mental play" is crucial for learning pieces. That does resonate; I catch myself silently/motionlessly “playing” difficult passages in bed. tals-of-piano-practice.readthedocs.io/chapter1/ch1_t
8
40
Replying to
Probably because:
Drawing, designing, writing, playing an instrument etc require you to use both System 1 (visual and motor skills) and System 2 (conceptual thinking) so you need to manage cognitive load.
REM sleep “compresses” the learnings for us and we are freed up again.
2
1
5
Can’t remember where I got this from but it’s informed by reading Why We Sleep, Superforecasting, and Peak.
The idea is sleep helps you “convert” system 2 learnings to something system 1 can use. They did studies on piano players and dreaming I think.
1
3
Replying to
Really curious because your entire list of "[doesn't] seem to compress well" aren't muscle memory memory tasks (at least, UIs and writing aren't), but instead those are things that require feedback and iteration based on discovery… something that takes time and (mental) space.
1
1
2
Right! Some amount of distance required for those, perhaps.
2
1
Show replies
Replying to
Research on retrieval practice and spaced learning might suggest otherwise. I wonder what happens when you try the spoken language two weeks later after the compressed session vs. spread out sessions, especially with self testing.
1
2
Right, it’s a great point: obviously massed learning produces less reliably recall than spaced learning! And yet language learning still feels compressible! When I spend more time on a language I really do make much more progress. Maybe it’s just *relatively* compressible.
1
2
Show replies
Replying to
Programming language or spoken language basically need you to follow a new set of rules in doing something you already have intuition / "motor skills" for (not sure how cooking fits here...), whereas piano, or drawing, or writing well, require deeper intuition?
1
1
Case in point: People who learn programming the first time struggle a lot more until they have an intuition for breaking down problems, approaching them computationally, etc.; vs. going from a few years of SomeLang to OtherLang
1
2
Replying to
Is learning a new spoken language compressible? I think it depends on how close is your native language to the new one. Also how much proficiency do you want? There are some counter intuitive elements of new languages that takes long time to get it(or possible never).