Walking and Learning Update

I discovered a year ago that walking makes it pleasant to study boring stuff — as I put it then, boring + boring = pleasant. I am still a little amazed.

Like any scientific discovery, I suppose, I had to do serious engineering to make good use of it. In particular:

1. Make walking easier. I use a treadmill in my apartment, which eliminates travel time (to where I do it), eliminates distractions, provides climate control, and allows me to walk barefoot.

2. Steady stream of study materials. Now I am using an Anki deck of Chinese characters put together by someone else. This saves a lot of time. (Anki is an open-source version of SuperMemo, a flashcard program that tries to optimize repetition.)

3. Figure out how much new stuff to study each day. Without plenty of repetition, you are wasting your time — you will forget what you’ve learned. Most of a study session is repetition. This means it’s not obvious how much new material to introduce each day. I found that 10 new Chinese characters is about right.

4. Put laptop on treadmill. To use Anki while on my treadmill, I need to use my laptop on my treadmill. At the Beijing Wal-mart, I found a piece of Sunor metal shelving that works perfectly. I put the shelf (about 90 cm long) across the arms of the treadmill, put the laptop on the shelf.

5. Minimize complications. I first noticed the effect using Anki. But Anki had several features I disliked, so I switched to ordinary flashcards. But they were too complicated — hard to schedule appropriately (you need to slowly expand the time between tests), time-consuming to keep track of progress. I had to keep stopping to make marks on the cards. So I am back to using Anki. Anki lacks a graph of progress — a graph that shows amount of learning versus date. But it is better than flashcards.

Each improvement made things better. With all of them, I lose track of time. Study, study, study, walk, walk, walk. Then it’s over. Not just painless, pleasant — different than any pleasure I have felt before. It feels a little like a new energy source (I imagine it can be used to learn many things), a little like teleportation.

The science aspect of it also interests me. Learning is the core topic of experimental psychology. Thousands of experiments have been done about human learning, thousands more about animal learning. Experimental psychologists are good methodologists; the average experimental psychology experiment makes the average medical-school experiment look retarded. But the walking/learning effect (walking makes learning pleasant) is outside anything anyone has ever reported. Only Michel Cabanac (not an experimental psychologist) has studied how variation in pleasantness regulates action (e.g., eating). Experimental psychologists lack good ways to find new effects. By missing this effect, they are missing a bigger idea:  learning is regulated, just as a thousand other things inside our bodies are regulated.

16 Responses to “Walking and Learning Update”

  1. Joshua Conner Says:

    I’d love to see a picture of your laptop “desk” set up. Is the metro shelving (with your laptop on it) just resting on the arms of the treadmill?

  2. Seth Roberts Says:

    Joshua, that’s a good request, I will post a picture. Yes, the shelving just rests on the arms of the treadmill.

  3. M Says:

    Great post. What brand/model of treadmill are you using? Where did you buy it in Beijing?

    You might try the “Hanzi stats” plugin for anki. I don’t believe it has graphs – but it gives you more statistics about how many characters you have “seen”, which you can follow over time.

  4. Darrin Thompson Says:

    ooohhh snap! The arrrrr word!

  5. bjk Says:

    I’ve always found riding on a bus/train and reading is very pleasant (not so much cars or planes). I thought it was that it gave the illusion of progress while doing something sedentary. Or maybe it was the moving scenery, which provided a break from the reading.

  6. gwern Says:

    > Anki lacks a graph of progress — a graph that shows amount of learning versus date. But it is better than flashcards.

    I use Mnemosyne, but that would surprise me. Leaving aside the issue that ‘learning’ isn’t a clear metric (# of cards ranked 4? Average ranking? # of cards with intervals longer than a month? A week?), I had understood that Anki had tons of statistics. Even Mnemosyne can display basic statistics like # of cards ranked 4, or 5.

  7. Thomas Johnson Says:

    Anki definitely has progress graphs. I forget the exact menu to go to, but you can see how many reviews you’ve done, how many “mature” (i.e., hard-to-forget) cards you have, etc

  8. vic Says:

    Any idea how one can make writing grant proposals pleasant?

  9. Tom Says:

    Fascinating. I wonder if the synergy between walking and learning has been selected for?

    Literature is a recent development, and our brains were huge long before it. Perhaps this gift is in part to map the resources and dangers of large amounts of terrain while walking through it.

  10. Seth Roberts Says:

    Darrin, by the “arrrr word” you mean “regulate”? Yeah, I think there is some sort of bigger case to be made for regulation of learning — more evidence that points in the same direction — but I cannot figure out what it is.

    Thomas, you say “Anki definitely has progress graphs.” My version (the latest) has eight graphs. Four show something (y axis) versus day in the past (x axis). The four graphs are reps, review time, number added, and first seen.  None of those four is how much you have learned (which is the progress I care about). One of them shows reviews of mature cards. The total number of mature cards would be a measure of learning but they don’t show that as a function of day. The Pinyin Plugin includes a graph of progress, which is one sign that plain Anki lacks such a thing.

    M, my treadmill goes under two names. In America it is the AFG 5.0 treadmill; in China, it is the Elite T5000. (They are slightly different — the American version, with the 3.0 HP motor, is better.) I bought from a store in Beijing with the website http://www.93bill.com.

  11. Oli M Says:

    Damien Elmes, the developer of Anki, is pretty friendly and very responsive to user input about Anki. I’m sure he would be happy to hear the observations of an experimental psychologist on how to improve his learning tool, if you were to post on the forum. He seems to respond to just about every thread personally.

  12. Seth Roberts Says:

    Oli M, thanks for the suggestion.

    Tom, yeah, the evolutionary reason you give (“to map the resources and dangers of large amounts of terrain while walking through it”) is close to what I suggested earlier (“to push people to walk in new places (which provide something to learn) rather than old places (which don’t)”).

    vic, to make writing grant proposals pleasant I suggest you divide the task into tiny pieces and after finishing each tiny piece give yourself a reward.

  13. Evelyn Majidi Says:

    In the middle east it is common to see young people walking up and down streets or in public parks studying for final exams. They repeat the material to themselves in a sing-song manner until it has been memorized.

  14. Kevin Page Says:

    Aristotle, the “paripatetic” philosopher, allegedly taught in the Lyceum as he walked around… Presumably, his students followed him about… He may have been onto something… I would rather walk and discuss than sit and discuss like old Socrates – although the Socratic maieutic is a superior way of learning philosophy… Ah well, good insight Seth…

  15. Seth Roberts Says:

    Evelyn, that’s very interesting about seeing people walking while studying. You saw this yourself?

    Kevin, that’s a good reference. I prefer to talk with students while walking rather than sitting in my office.

  16. Evelyn Majidi Says:

    I lived in Tehran most of the time between 1963 and 1980. Seeing high school and college students studying while walking in public parks and narrow streets was a very ordinary sight during final-exam season. A vital part of the process was repeating the material out loud, over and over, usually to the rhythm of the steps. It reminded me of how easily children learn nursery rhymes, even when they do not understand the meaning of the words, especially when they accompany them with specific gestures or movements. My mother always said that the more senses are involved, the better one learns.