Lawrence Evalyn has been beeminding for almost 7 years and we’re very proud to have him guest posting for us. If this whets your appetite, check out his professional blog where he’s written about beeminding his comprehensive exams. (Graphs and charts and spreadsheets oh my!) And if you like spreadsheets and graphs about books you might also like his fortnightly newsletter (which he, of course, also beeminds).
Full disclosure: as a Ph.D. student in English, it’s basically my professional obligation to convince everyone around me to read more books, so you might say I have a bias in recommending them here. But I genuinely think this is a broadly-applicable goal-structuring methodology.
I was reflecting on some of my favourite, most successful Beeminder goals over the years, and I realised they all had something in common. I revived my rusty Latin for an exam, by making my way through a textbook of exercises. I studied for the GRE, by making my way through a textbook of exercises. I learned how to code in Python, by making my way through a textbook of exercises. Have you picked up on the common thread?
Here’s a hint, the common thread is not the exam: nobody was testing me on Python. Okay, I’ll tell you: it’s the textbook, with exercises.
Why I think it works
Unlike a lot of resources, a textbook written for classroom use will go into a level of depth sufficient for a person to actually learn the material. Teaching people how to really use the information they contain is what textbooks were designed for! This is why I’m using an Arabic textbook instead of Arabic Duolingo: the textbook systematically introduces, and then tests on, fundamental concepts. It’s not sexy — but it’s useful.
Moreover, whoever made the textbook has already done the hard part of figuring out how a person should learn whatever it is that you’re trying to learn. (This is where “a very thorough online class” could easily substitute for “a textbook”.) Instead of reinventing the wheel, you can follow a process made by somebody who knows what they’re talking about.
The downside of a textbook, though, is that it’s pretty monolithic. It’s intimidating. Who just sits down and reads a textbook? The trick of any big, overwhelming task, of course, is to break it down into bites. Taken one chapter at a time, or one exercise at a time, or even one page at a time, you never have to Go Read A Textbook. Since you know from the start how long the textbook is, as long as you stay on Beeminder’s yellow brick road you can even know when you will achieve the much-desired status of Having Read The Textbook.
Why Beeminder is so crucial
“Who just sits down and reads a textbook? The trick of any big, overwhelming task, of course, is to break it down into bites.”
The challenge with self-teaching is that learning stuff is usually way less fun than knowing stuff. When you’re learning something, you’re likely to feel stupid and frustrated, because you’re stretching your capabilities. A lot of online “learning” resources, in my experience, really exist to provide people with the much more pleasant feeling of knowing things. Duolingo is a ton of fun because it gives its rewards immediately: cute graphics, happy sounds, numbers that go steadily up — it just feels good to do even the most trivial Duolingo exercise. [1] Reading a textbook can be stimulating and satisfying, but its immediate pleasures are limited. The real payoff of a textbook is at the end.
A classic akrasia problem, right? It’ll feel way better overall to have done the textbook, but you’d always rather do the textbook… later.
Beeminder makes the long-term into the immediate. Beeminder’s job is to give you a good reason to do one more boring page of that textbook, tonight.
But what about weaseling?
Self-paced textbook-reading is the quintessential “you can cheat really easily, if you never want to accomplish your own dreams” kind of goal. Who is really checking that the exercise you did is “complete,” after all?
Well, the future chapters of the textbook are. This is the biggest challenge I’ve faced with my textbook goals. I’ll rush through a section without fully understanding it, and then I can’t make heads nor tails of the next exercise. I suggest writing something in the fine print of your goal about what you will do when this happens.
What I try to do, in these scenarios, is to have the good grace to accept the world as it is, rather than as I would wish it to be. And then I delete my previous datapoint, accept that I had not actually completed the exercise on that day, pay my derail charge, and re-do the last chapter. Maybe you want to come up with another way to handle this problem — if the Quantified Self aspect is less important to you, you could just give yourself points for the same chapter multiple times, and increase your total desired chapters accordingly. Or you could just work a lot harder before your next datapoint, and do your catching-up invisibly. But do plan ahead, and come up with something that will let you keep focused on how much you realistically have completed.
Picking a textbook
It’s worth doing some research and asking around before you get started, since you’ll be living with this book for a while. You can find promising textbooks by googling what you want to learn, plus the word “syllabus,” to see what textbooks get used in college classes. (As an advanced strategy, you could politely contact a professor in the field at a nearby university and ask what textbook they like best.)
Questions to ask of your candidate books:
- Does it have homework? It must have exercises that will make you practice what you learn. The closer these assignments are to the real use of the skill you’re learning, the better. (Typing out code in an online system is great — typing out code that you run on your own computer is even better.)
- Is it suited to my current skill level? The first chapter should be fully comprehensible to you (so you can actually do the exercises), but not something you could have taught someone else how to do (so you’re learning from it).
An easy textbook can be worse than one that’s too hard. In my experience, a lot of MOOCs are too easy — but many universities still have “online courses” where they post videos and assignments from the actual classes they’ve taught, and these are challenging enough to provide good learning experiences. (MIT has a lot, and Harvard and Duke — any university you like, you can probably google “university online courses” and find something.)
“Knowledge fit” is the hardest thing to find in a textbook when you’re outside of the formal system of a school, so word-of-mouth and patience are important here. I felt like the heavens had parted and blessings had rained down upon me the day someone recommended All The Arabic You Never Learned The First Time Around, which perfectly reflected my exact skill level with Arabic.
The world is your oyster!
Textbooks are great for really mastering the grammar of a language. That goes for programming languages, too. Languages and coding are probably the most common things people want to learn, but you can get creative — what about following an online course in statistics? Or history? Why not cook your way through a cookbook?
Anywhere in the world that you can find experts, you can find people who learned what they know. Learning things is challenging in the short term, but if you break it into bites and work your way through it, you can reap the rewards of being an expert, too.
Appendix: Complete list of textbooks I’ve read (at least in part) with Beeminder
(Links go to Beeminder goals)
- The Cambridge History of Eighteenth Century Political Thought
- All the Arabic You Never Learned the First Time Around
- The Norton Anthology of English Literature, Volume D: The Romantic Period
- The Longman Anthology of British Literature, Volume 2a: The Romantics and Their Contemporaries
- Learn Python the Hard Way
- Writing Your Journal Article in 12 Weeks: A Guide to Academic Publishing Success
- Breakthrough Rapid Reading
- Cracking the GRE Literature in English Subject Test
- Wheelock’s Latin
Footnotes
[1] This isn’t an anti-Duolingo piece — they actually do a pretty solid job of implementing Rosetta-Stone-style language-from-context learning exercises, with spaced repetition to boot — but Duolingo has limits. I spent a year doing Duolingo French, and then I attended 30% of a college French class, and it was the latter that equipped me to actually read in French, because the latter had very unpleasant exams.
Lawrence is a PhD Candidate studying eighteenth century British literature. He has 42 active Beeminder goals, 7 of which are supposed to make him write his dissertation. Arguably his coolest active goal is for reading an 18th century novel in French extremely slowly (your “coolness” mileage may vary). The oldest book he owns was printed in 1791; it is a collection of morally improving stories for children.