Learning English: Walking versus Sitting

A Chinese friend of mine learned about my discovery that it was much easier to study Chinese while walking on a treadmill than sitting. This led her to buy a treadmill. She began to study English (e.g., GRE vocabulary words) while walking on the treadmill. “It worked very well,” she told me. She found that if she studied words while walking, she could remember them four days later. If she studied them sitting down, she could remember them only a day later. With Anki, the default settings assume you can only remember what you’ve studied for a day — the first time you learn a word, you will be tested a day later.

This reminds me of Allen Neuringer’s finding of better memory for material learned while moving, but the size of the effect my friend observed is still shocking. If you can remember words four times longer before you need to review them, you can learn four times as fast. The effect that Jeremy Howard and I observed was of similar size. We could only study 10 minutes sitting down but could easily study for 40 minutes or more while walking.

Walking Meeting Update

In a recent post, I described how much easier it was to meet with students while walking than while sitting. The content of the meetings stayed the same. I met with them after my Academic Writing class to help them with their writing.

I asked my students what they thought of these meetings. They had three complaints:

1. I walked too fast.

2. It wasn’t so easy to avoid bicycles and listen to me at the same time.

3. It was cold.

I told them that I found the meetings less tiring. They did not notice this.

I considered walking inside the teaching building but it turned out to be too dark. Instead, we walked outside the building in a nearly-deserted alley (solving Complaint #2). I walked more slowly  (colving Complaint #1). I couldn’t do anything about Complaint #3.

The students could choose the length of the meetings. Last week all five of them chose 15 minutes. After 1 hour and 15 minutes of walking meetings, I felt entirely refreshed. As if I had done no work at all.


Walking Meetings Much Better than Seated Meetings

In an interview about his new book The JFK Assassination Diary, Edward Jay Epstein was asked how he, a Cornell undergraduate, managed to talk to the people who did the research behind the Warren Commission Report. “It was a different age,” he said. “People actually communicated by sitting across a desk from one another and talking.” When I heard this, I was amused. I had just discovered that it was much better to meet with students walking than seated. What Epstein considered the good old-fashioned way (seated meetings) was to me the crazy new-fangled way.

As I’ve blogged, this semester I am teaching a class about academic writing. I am trying to apply my no grading/no lecturing method that worked well last year in a much different class (Frontiers of Psychology). In the writing class, my plan was/is to meet with students one-on-one right after class, in the same  room. They choose the meeting length. During the meeting they show me what they’ve written and I make comments. During the next class they give a brief talk (e.g., 10 minutes) in which they tell the rest of the students what I told them. The course is much easier to teach than usual: no lecture, no grading, no written comments. Yet the students get as much one-on-one feedback as they want. I think spoken (face to face) comments are much better than written ones because they allow the recipient to ask questions. Continue reading “Walking Meetings Much Better than Seated Meetings”

Walking and Learning: GRE Words

Most of my earlier examples of the benefits of walking while studying involved treadmills and learning a foreign language. A Stanford student named Govind writes:

I found I was able to memorize GRE words very effectively while walking [compared to sitting]. It not only made the process much more enjoyable, but since I walked outside (around Oxford [England]), I also was able to associate words with physical cues. The difference between propitiate and propitious is now inextricably linked to Cowley. (I am now a memory palace convert.)

At Berkeley, I once assigned my intro psych students to do self-experiments. One of them measured how many French words she could study before falling asleep. She tried three body positions: sitting at a desk, lying on her side on her bed, lying on her stomach on her bed. She also tried three audio environments: silence, classical music, heavy metal. Best combination: lying on stomach, heavy metal. Worst combination: sitting at desk, silence. This amused me, but I now see that the real lesson of her experiment is that she didn’t try walking. It shows how little-known the walking-helps-memorize idea is, even though the effect is easy to notice, as Govind’s story shows.

More on the Synergy of Walking and Learning

A few years ago, I discovered that walking made studying Chinese more pleasant and studying Chinese made walking more pleasant. It’s a big effect. While walking on a treadmill I could easily study Chinese for 40 minutes; while sitting or standing still, 5-10 minutes. The general idea seems to be that walking creates a thirst for novelty, for dry information. An evolutionary explanation is that this effect caused us to  better explore our surroundings. Such exploration paid off too rarely and/or with too-long delays to be supported by the usual reward-action mechanism.

Jeremy Howard, the president of Kaggle, discovered the same effect independently while studying Chinese. A few days ago, I heard from Patrick Roach, a medical student in the Midwest, who also discovered the same effect independently — in his case, studying anatomy rather than Chinese. He blogged about the Anki/treadmill combination. I asked him if walking on a treadmill made it easier to study Anki? He replied:

Absolutely.  I originally tried this with a 3100 card deck I created while studying anatomy in med school.  The format (Image/Name) was perfect for reviewing while walking, as there wasn’t too much text to read.  I imagine your experience with learning a new language was similar.  Anyways, Treadmill + Anki (+Music) along with my Tablet / Wiimote combo was much more productive than either task alone.  I could easily spend 1-2 hours and not notice the time passing in the same way it dragged on when trying to study endless flashcards sitting in a quiet room.  Getting tired or losing focus was less of an issue as well – I noticed I had less distractions/extra attention to spare while walking.

Thanks for getting in touch, Patrick. As Lewis Carroll said, “What I tell you three times is true.”




Independent Discovery That Walking Catalyzes Learning

Two years ago I discovered that if I walked while studying Chinese flashcards (using Anki), both activities — walking and studying — became easier. I could walk much longer on my treadmill and I could study much longer. Walking made studying more pleasant and vice-versa. Around the same time, Jeremy Howard, the president of Kaggle, made the same discovery independently. In an email to me, he writes:

I came up with the idea accidentally a couple of years ago – I needed to go to the gym every day, and that included 30 minutes on a cross-trainer (but I only managed to do 15 min most days). I needed something to do to keep me amused, so I brought along my PC and started doing my Anki whilst on the cross-trainer. I discovered I could do my cross-trainer for at least twice as long, and my Anki results were better too. Later I added treadmill walking to my Anki study too.

He says more about it, including how much it helped him, in a QS talk.

As Nabokov says in Pale Fire,

If on some nameless island Captain Schmidt
Sees a new animal and captures it,
And if, a little later, Captain Smith
Brings back a skin, that island is no myth.

Learning methods that use this effect are going to have a big advantage over learning methods that don’t.

New Walking-Catalyzes-Learning Results

Two years ago, I discovered that if I walked on a treadmill while studying Chinese flash cards, it became much easier. Without walking, I could barely study 10 minutes without getting exhausted and stopping. If I walked at the same time, however, I could study much longer — say, 60 minutes. Huge difference. Walking on a treadmill made studying Chinese pleasant. This was stunning because walking on a treadmill by itself was boring and studying Chinese (or any other dry knowledge) is supposed to be boring. I concluded that walking created a thirst for dry knowledge, which studying Chinese satisfied. My evolutionary explanation was that this linkage evolved to push us to explore our surroundings. My posts about this.

In an April 2012 QS talk, Jeremy Howard reported the same thing.

I discovered that if I am walking on a treadmill at 1.2 miles per hour at a 1 degree incline I have an error rate of about 5%. Whereas if I don’t [walk on a treadmill] it’s about 8%. I also know that I can do that for an hour. Whereas normally if I’m just sitting down I can just do it 20 minutes. . . . And at the end of that hour I was ready to do something else. Whereas at the end of 20 minutes, normally I’d [audience member: “Take a nap”] Yeah, I’d be totally ready for a rest . . I also discovered I was 40% faster [at learning].

He added, “I love my Chinese every day.” More recently, someone named Adam posted on the QS forums that he’d had a similar experience:

As Jeremy Howard mentioned in his talk, SRSing (is that a word?) is exhausting. Like him, after a period of about 20 minutes, I often reach a level of fatigue that makes it difficult to continue studying. I first read about the “treadmill method” on Seth Roberts’s blog & found it highly effective. Like Mr. Howard, I could study for hours without become bored . . . The only problem here is that I don’t have easy access to a treadmill. My gym is quite far & it is impractical to go there every day, while I desire to SRS every day.

That two other people noticed such a big effect is good reason to think that it will be true for most people.

My Treadmill Desk

In 1996 I put a treadmill in my office so that I could work standing up. My goal was better sleep (the more I stood, the better I slept), not weight loss (the usual reason for a treadmill desk). It was hard to walk a lot. Mostly I stood still. It was noisy, too — my neighbors complained. When the treadmill broke I didn’t replace it.

Now I walk on a treadmill for different reasons: to lower blood sugar and learn Chinese. Above is my current setup. I use the laptop to study Chinese (using Anki) or watch TV or movies. Studying Chinese while walking is much easier than studying Chinese while standing still or sitting. I have used flashcards but Anki (shown on the computer screen) helps space repetitions optimally. The headphones (Bose noise-reduction) are for TV and movies. I don’t need them for Anki.

Memory Palaces and the Walking/Learning Connection

In this excellent article, Joshua Foer describes how he got really good at competitive memory tests, such as remembering the order of a deck of cards. He competed in the national championships.

Foer writes a lot about using “memory palaces” to remember stuff. You take a familiar building or neighborhood and vividly imagine what you want to remember at different places within it. To retrieve the memories, you mentally visit each place.

This is an ancient and famous method. I knew about it but had not realized until I read the article that it sheds light on my discovery that treadmill walking makes learning Chinese pleasant. (A commenter named Tom also noticed the connection.) Foer gives the obvious evolutionary explanation for why the memory palace method works so well: long ago, we needed to remember where to find important stuff (water, food, special plants, useful materials). So we evolved a memory system well-suited for doing so.

Less obvious is another evolutionary idea: why stop there? It’s a system. When you design a car for a certain sort of driving, you don’t stop with the engine. You adjust the drive train, the tires, and so on. If evolution shaped our brains for a certain sort of data (things in places), surely it also shaped our brains to collect that data. Pointless to design a car no one drives.

Two more changes would help make use of the system:

1. Hedonic. Make it pleasant to fill the system with data. This is what I noticed — dry knowledge (such as the order of cards in a deck) became pleasant to learn. Long ago, the hedonic change I noticed would have pushed people to walk in new places rather than old ones.

2. Efficiency.  Make learning more efficient (= more learning per unit time). Several confounded comparisons point in this direction. For example, I found that 15 minutes studying flashcards while riding the subway was a lot less help than spending 15 minutes while walking on my treadmill. Of course there are many differences between the two situations. Likewise, using Anki is working much better now than in the past, when I used it sitting down. I will try to study this more carefully.

Years ago, evolutionary explanations such as these were mocked as “just so stories” by prominent scientists, such as Stephen Jay Gould, Noam Chomsky, and Richard Lewontin. It’s now clear they were wrong.

Walking and Learning

A new study supports my idea that walking and learning are connected. Normally I found it boring to study Chinese flash cards. While walking, I found it pleasant. You could say walking made me more curious. Just standing on the treadmill didn’t have this effect.

The study divided men and women in their 60’s into two groups: (a) walking for 40 minutes/day and (b) stretching. At the end of the study, for persons in the walking group, part of the hippocampus — which is associated with learning — had grown. For persons in the other group, that part of the hippocampus got smaller. Several other parts of the brain, not associated with learning, did not differ between the groups.

Three Observations About Walking and Learning

1. Studying Chinese-character flashcards while walking on a treadmill is as pleasant as drinking something when thirsty. Unlike actual thirst and drinking, the pleasure lasts a long time and the desire is under your control (to turn it on, you start walking; to turn it off, you stop).

2. What is the opposite of betrayal? There is no antonym. The opposite is so rare it isn’t even obvious what it is. Betrayal is when your friend becomes your enemy; the opposite is when your enemy becomes your friend. Living in China and not knowing Chinese was not exactly my enemy but it was certainly negative. This treadmill discovery turns it into a positive: Chinese becomes an inexhaustible source of dry knowledge that I can enjoy learning.

3. Learning is the central theme of experimental psychology and perhaps all academic psychology. Psychology professors have done more experiments about learning than anything else. Practically all of those experiments have been about efficiency of learning: The amount of learning (e.g., percent correct) in Condition A is compared with the amount of learning in Condition B, where A and B “cost” about the same. As a result, we know a great deal about what controls efficiency of learning, at least in laboratory tasks. I think many psychologists are surprised and disappointed that this research has had little effect outside academia. I have never heard a good answer to the question of why. If you’d asked me a month ago I would have said it’s because they haven’t discovered large non-obvious effects. That’s true, but says nothing about how to discover them.

My treadmill experience suggests a more helpful answer: Hedonics matter.  Learning exactly the same material can be more or less pleasant. When Learning X is pleasant, it is learned easily; when Learning X is unpleasant, it is learned with difficulty or not at all. In the real world, hedonic differences matter more than efficiency differences. If they want to improve real-world learning, psychologists have been measuring the wrong thing. It is a hundred times easier and ten times more “objective” (= “scientific”) to study how much has been learned than to study how pleasant was the experience. But that doesn’t mean it is better to study.

Michel Cabanac, a physiologist, strikes me as someone on the right path. Cabanac has studied how the pleasantness of this or that experience goes up or down to help us properly self-regulate. A simple example is that cold water feels more pleasant when we feel hot than when we feel cold. A common example is that exactly the same food becomes less pleasant during a meal. The food doesn’t change; we change.

Walking and Learning in Rats

Yesterday I blogged that walking on a treadmill made studying flashcards enjoyable. I also felt my retention was noticeably better than when I studied sitting or standing in one place.

Thanks to Matt Weber I learned of a rat experiment that supports the idea I was more retentive. Long-term potentiation (LTP) is a long-lasting (hours) change in synapse properties caused by a certain type of electrical stimulation from electrodes. Leung et al. measured the amount of LTP produced by the electrodes when rats were in one of four states: (a) walking, (b) immobilized, (c) short-wave sleep, (d) rapid-eye-movement sleep. They found clear LTP in all four states, but the LTP was much larger (50% larger?) when the rats were walking during the stimulation. During the other three states the LTP was about the same.

The walking and immobilization conditions must have differed in many ways. Perhaps immobilization was uncomfortable. Perhaps it required more handling. And so on. Comparing just those two states, you might wonder if (a) walking produces changes that cause things to be remembered better or (b) any of the other walking/immobilization differences made things worse (e.g., the shock of handling reduces learning). The fact that immobilization and the two sleep states produced similar results argues against the second sort of explanation.

Walking Creates A Thirst For Dry Knowledge

A few weeks ago I got a treadmill for my Beijing apartment. Two days ago I was walking on it (I try to walk 1 hr/day) while watching Leverage to make the activity more palatable. But Leverage bored me. It was too simple. So I took out some Chinese flashcards (character on one side, English and pinyin on the other) and started studying them. I was astonished how pleasant it was. An hour of walking and studying went by . . . uh, in a flash. In my entire life I have never had such a pleasant hour studying. The next day it happened again! The experience appears infinitely repeatable. I’ve previously mentioned the man who memorized Paradise Lost while walking on a treadmill.

I’ve noticed before that treadmill walking (by itself boring) and Chinese-character learning (by itself boring) become pleasant when combined. So why was I astonished? Because the increase in enjoyment was larger. The whole activity was really pleasant, like drinking water when thirsty. When an hour was up, I could have kept going. I wanted to do it again. When I noticed it earlier, I was using Anki to learn Chinese characters. Now I am using flashcards in blocks of ten (study 10 until learned, get a new set of 10, study them until learned . . . ). The flashcards provide much more sense of accomplishment and completion, which I thinks makes the activity more pleasant.

My progress with Chinese characters has been so slow that during the latest attempt (putting them on my wall) I didn’t even try to learn both the pinyin and the meaning at the same time; I had retreated to just trying to learn the meaning. That was hard enough. I have had about 100 character cards on the walls of my apartment for a month but I’ve only learned the meaning of about half of them. No pinyin at all. In contrast, in two one-hour treadmill sessions I’ve gotten through 60 cards  . . . including pinyin. For me, learning pinyin is much harder than learning meaning.

It’s like drinking water when you’re thirsty versus when you’re not thirsty. The walking turns a kind of switch that makes it pleasant to learn dry knowledge, just as lack of water creates thirst. Not only did studying dry materials become much more pleasant I suspect I also became more efficient — more retentive. I was surprised how fast I managed to reach a criterion of zero mistakes.

I had previously studied flashcards while walking around Tsinghua. This did not produce an oh-my-god experience. I can think of three reasons why the effect is now much stronger: 1. Ordinary walking is distracting. You have to watch where you’re going, there are other people, cars, trees, and so on. Distraction reduces learning. If the distractions are boring — and they usually are –  the experience becomes less pleasant. 2. Ordinary walking provides more information than treadmill walking (which provides no information at all — you’re staring at a wall). The non-flashcard info reduces desire to learn what’s on the flashcards. 3. On these Tsinghua walks I had about 100 flashcards which I cycled through. Using sets of 10, as I said, provides more sense of accomplishment. I’ve also had about 20 Chinese-speaking lessons while walking around. The walking made the lessons more pleasant, yes, but it wasn’t nearly as enjoyable as the treadmill/flashcard combination. And because lessons with a tutor are intrinsically more enjoyable than studying flashcards, the increase in enjoyment was less dramatic.

As I said earlier I think there’s an evolutionary reason for this effect: The thirst for knowledge (= novelty) created by walking pushed us to explore and learn about our surroundings. One interesting feature of my discovery about treadmill and flashcards is that it may take better advantage of this mechanism than did ordinary Stone-Age life — better in the sense that more pleasure/minute can be derived. In the Stone Age, novelty, new dry knowledge, was hard to come by. You could only walk so fast. After a while, it was hard to walk far enough away to be in a new place. Whereas I can easily switch from flashcards I’ve learned to new ones. An example of a supranormal stimulus.

Boring + Boring = Pleasant!?

Fact 1: For the last few weeks, I’ve been studying Chinese using a flashcard program called Anki. It’s an excellent program but boring. I’ve never liked studying — maybe no one does. Fact 2: I’ve had a treadmill for a very long time. Walking on a treadmill is boring so I always combine it with something pleasant — like watching American Idol. That makes it bearable. I don’t think listening to music would be enough.

Two days ago I discovered something that stunned me: Using Anki WHILE walking on my treadmill was enjoyable. I easily did it for an hour and the next day (yesterday) did it for an hour again. The time goes by quickly. Two boring activities, done together, became pleasant. Anki alone I can do maybe ten minutes. Treadmill alone I can do only a few minutes before I want to stop. In both cases I’d have to be pushed to do it at all. Yet the combination I want to do; 60 minutes feels like a good length of time.

I’ve noticed several related things: 1. I could easily study flashcards while walking. This was less mysterious because I coded walking as pleasant. 2. I can’ t bear to watch TV sitting down. Walking on a treadmill makes it bearable. This didn’t puzzle me because I coded TV watching as pleasant and sitting as unpleasant (although I sit by choice while doing many other things). 3. I have Pimsler Chinese lessons (audio). I can painlessly listen to them while walking. While stationary (sitting or standing), it’s hard to listen to them. 4. When writing (during which I sit), it’s very effective to work for 40 minutes and then walk on my treadmill watching something enjoyable for 20 minutes. I can repeat that cycle many times. 5. Allen Neuringer found he was better at memorization while moving than while stationary. 6. There’s some sort of movement/thinking connection — we move our arms when we talk, we may like to walk while we talk, maybe walking makes it easier to think, and so on.

You could say that walking causes a “thirst” for learning or learning causes a “thirst” for walking. Except that the “thirst” is so hidden I discovered it only by accident. Whereas actual thirst is obvious. The usual idea is that what’s pleasant shows what’s good for us — e.g., water is pleasant when we are thirsty. Yet if walking is good for us — a common idea — why isn’t it pleasant all by itself? And if Anki is good for us, why isn’t it pleasant all by itself? The Anki/treadmill symmetry is odd because lots of people think we need exercise to be healthy but I’ve never heard someone say we need to study to be healthy.

The evolutionary reason for this might be to push people to walk in new places (which provide something to learn) rather than old places (which don’t). To push them to explore. David Owen noticed it was much more fun  for both him and his small daughter to walk in the city than in the country. He was surprised. When I drive somewhere, and am not listening to a book or something, I prefer a new route over a familiar one. If I am listening to a book I prefer the familiar route because it makes it easier to understand the book.

Maybe the practical lesson is that we enjoy learning dry stuff when walking but not when stationary. Pity the 99.9% of students who study stationary. Ideally you’d listen to a lecture while walking somewhere, perhaps around a track. Now and then I’ve interviewed people while walking; it worked much better than the usual interview format (seated). The old reason was I disliked sitting. Now I have a better reason.