[Link] Holistic learning ebook

This ebook is kind of dopey, but it’s one of the few re­sources I’ve seen where some­one who’s rea­son­ably good at learn­ing stuff tries to dis­sect and com­mu­ni­cate the men­tal mechanisms they use for learn­ing:

http://​​www.scot­thy­oung.com/​​blog/​​Pro­grams/​​Holis­ticLearn­ingEBook.pdf

Here’s a quick sum­mary.

  • You can learn things faster and bet­ter by im­prov­ing the strate­gies you use for learn­ing stuff.

  • “Holis­tic” learn­ing is op­posed to “rote” learn­ing. Holis­tic learn­ers make lots of con­nec­tions be­tween differ­ent things they learn, and be­tween things they learn and things that are per­son­ally rele­vant to them. An ex­am­ple might be this di­a­gram of var­i­ous con­cepts in elec­tro­stat­ics, which I no longer know how to in­ter­pret. Another ex­am­ple might be me re­mem­ber­ing about that di­a­gram when read­ing the book.

  • Holis­tic learn­ers un­der­stand con­cepts in many differ­ent ways in or­der to re­ally “get” them. They fo­cus on build­ing men­tal mod­els in­stead of mem­o­riz­ing facts or pro­ce­dures.

  • If you un­der­stand a body of knowl­edge well enough, and for­get a spe­cific thing, you should be able to re­con­struct your un­der­stand­ing of it based on re­lated things you un­der­stand.

  • The book refers to a “model” as some­thing spe­cific you un­der­stand par­tic­u­larly well that you can ex­plain other things in terms of. For ex­am­ple, your “model” of a sub­space (in lin­ear alge­bra) might be a plane cut­ting through 3d space. Not all sub­spaces are planes, but think­ing of a plane could be a way to quickly preload a bunch of rele­vant con­cepts in to your head.

  • To learn holis­ti­cally:

    • “Viscer­al­ize” con­cepts by sum­ma­riz­ing them with a spe­cific image, sound, feel­ing, and/​or tex­ture. Ex­am­ple: when learn­ing pro­gram­ming, think of an ar­ray as a bunch of col­ored cubes sus­pended along a cord.

    • Use metaphors to un­der­stand things bet­ter. Whenen­ever you learn some­thing new, try to figure out what it re­minds you of. If it’s some­thing from a to­tally un­re­lated do­main, that’s great.

    • Ex­plore your un­der­stand­ing net­work, ideally by solv­ing prob­lems, in or­der to fix glitches in your un­der­stand­ing and re­fresh it.

  • Holis­tic learn­ing works great for some sub­jects, like sci­ence and math, but it’s not as good for oth­ers, like his­tory and law. It also helps less with con­crete skills, like play­ing golf.

The au­thor sells var­i­ous in­for­ma­tion & coach­ing prod­ucts in this vein, but as far as I can tell the ebook I linked to is the only free one: http://​​www.scot­thy­oung.com/​​lm­slvid­course/​​2.html. (If any­one pays for any of these, they should sum­ma­rize them (to un­der­stand them bet­ter) and post the sum­maries to LW ;].) I’m definitely in­ter­ested in hear­ing about other re­sources peo­ple know of on the me­chan­ics of learn­ing.


Some­one once told me that if you’re a grad stu­dent study­ing un­der a No­bel lau­re­ate, you’re much more likely to later win the No­bel your­self. (I just searched the in­ter­net for ev­i­dence re­gard­ing this claim and couldn’t find any, so I’m now less con­fi­dent in it.) This claim sug­gests that do­ing good re­search is learn­able.

The per­son who told me this thought these re­search skills couldn’t be de­scribed with words, and could only be trans­mit­ted through ac­tual re­search part­ner­ships. I think it’s more likely that they can be de­scribed with words, but no No­bel lau­re­ate has both­ered to sit down and write a book called “How I Do Re­search”. (Please leave a com­ment if you know of a book like this!)

Even if your fluid in­tel­li­gence is static and difficult to im­prove, that doesn’t pre­vent you from im­prov­ing the men­tal al­gorithms and habits you use to ac­com­plish tasks.