Those contexts include any field of study or expertise that involves making distinctions. Is that a rhombus or a trapezoid? An oak tree or a maple? The Chinese symbol for “family” or “house”? A positive sloping line or a negative sloping one? Computer PLMs as Kellman and others have designed them are visual, fast-paced, and focused on classifying images (do the elevated bumps in that rash show shingles, eczema, or psoriasis?) or problems rather than solving them outright (does that graph match x—3y = 8, or x + 12 y + 32?). The modules are intended to sharpen snap judgments—perceptual skills—so that you “know” what you’re looking at without having to explain why, at least not right away.
In effect, the PLMs build perceptual intuition—when they work. And they have, mostly, in several recent studies. In one, at the University of Virginia, researchers used a perceptual learning module to train medical students studying gallbladder removal. For most of the twentieth century, doctors had removed gallbladders by making a long cut in the abdomen and performing open surgery. But since the 1980s many doctors have been doing the surgery with a laparoscope, a slender tube that can be threaded into the abdominal cavity through a small incision. The scope is equipped with a tiny camera, and the surgeon must navigate through the cavity based on the images the scope transmits. All sorts of injuries can occur if the doctor misreads those images, and it usually takes hundreds of observed surgeries to master the skill. In the experiment, half the students practiced on a computer module that showed short videos from real surgeries and had to decide quickly which stage of the surgery was pictured. The other half—the control group—studied the same videos as they pleased, rewinding if they wanted. The practice session lasted about thirty minutes. On a final test, the perceptual learning group trounced their equally experienced peers, scoring four times higher.
Kellman has found that his PLMs can accelerate dermatology students’ ability to identify skin lesions and rashes, which come in enormous varieties and often look indistinguishable to the untrained eye. He and Sally Krasne at UCLA Medical School have found similar results in radiology, as well as in reading echocardiograms (ECGs). Working with other colleagues, Kellman has also achieved good results with a module that prompts chemistry students to categorize chemical bonds between molecules.
True, this is all advanced, technical stuff for people who’ve already done just fine in school. What about the kid watching the clock in math class, trying to figure out what on earth “slope” means or how to graph 3(x + 1) = y?
Here, too, perceptual modules have shown great promise. At a school in Santa Monica, Kellman tested a module that works just like the instrument panel trainer, only with equations and graphs. A graph of a line pops up on the computer screen, and below it are three equations to choose from (or an equation with three choices of graphs beneath; it alternates). Again, students have to work fast: make a choice and move on; make another choice, and another, through dozens of screens. With enough training, the student begins to feel the right answer, “and then they can figure out why it’s right afterwards, if they need to,” as Joe Wise, the high school teacher working with Kellman, told me.
Scientists have a lot more work to do before they figure out how, and for which subjects, PLMs are most effective. You can play computer games all you want, but you still have to fly the plane or operate on a living human being. It’s a supplement to experience, not a substitute. That’s one reason perceptual learning remains a backwater in psychology and education. It’s hardly a reason to ignore it, though. Perceptual learning is happening all the time, after all, and automatically—and it’s now clear that it can be exploited to speed up acquisition of specific skills.
• • •
The promise of this book was to describe techniques that could help us learn more effectively without demanding more effort. The goal is to find more leisure, not less. I’m now about to break that promise, but not shatter it into little pieces.
We’re going to make a slide show together.
I know, I know. But look: I once made my own flashcards in high school with old-fashioned paper and No. 2 pencils. It’s just as easy to create a PLM, right here, right now, to show how it can be done, and what it can and can’t do. I was determined to be as lazy as possible about this. I subcontracted the work. I hired my sixteen-year-old daughter to design the module for me, because I’m a busy professional writer, but also because, like many kids, she’s digitally fluent. She’s perfectly capable of making her own digital slide shows, PowerPoint presentations, or videos, downloading images off the Internet. And that’s what I told her to do.
I also poached the subject matter, or at least the idea. I decided to do exactly what Kornell and Bjork did in their interleaving study of painting styles described in the last chapter, with a few small changes. Those two used interleaving to teach students to distinguish individual styles among landscape artists. I changed that. My module would focus on famous artistic movements, like Impressionism. This wasn’t a random choice. My motives here were selfish: I’d been embarrassed on a recent visit to the Museum of Modern Art by how little I knew of art history. I recognized a piece here and there but had zero sense of the artistic and cultural currents running through them. Van Gogh’s Starry Night holds the eye with its swimming, blurred sky, but what did it mean for him, for his contemporaries, for the evolution of “modern” art? I sure didn’t know.
Fine. I didn’t have to know all that right away. I just wanted to know how to tell the difference between the pieces. I wanted a good eye. I could fill in the other stuff later.
What kind of perceptual module did I need? This took a little thinking but not much. I had my daughter choose a dozen artistic movements and download ten paintings from each. That was the raw material, 120 paintings. The movements she chose were (inhale, hold): Impressionism, Post-Impressionism, Romanticism, Expressionism, Abstract Expressionism, Abstract Impressionism, Dadaism, Constructivism, Minimalism, Suprematism, Futurism, and Fauvism. Got all that? You don’t have to. The point is that there are many distinctions to make, and I couldn’t make any of them. I came into the project with a thick pair of beginner’s goggles on: I knew Monet and Renoir were Impressionists, and that was about it.
Kornell and Bjork had presented their landscape paintings in mixed sets, and of course that’s what I had my daughter do, too. The order was random, not blocked by style. She made a PLM and rigged it just as Kellman did. A painting appears on the screen, with a choice of twelve styles below it. If I chose right, a bell rang and the check symbol flashed on the screen. If I guessed wrong, a black “X” appeared and the correct answer was highlighted.
I trained for as long as I could stand it in a single sitting: about ten minutes, maybe sixty screens. The first session was almost all guessing. As I said, I had a feel for the Impressionist pieces and nothing else. In the second ten-minute session I began to zero in on Minimalism and Futurism; baby steps. By session four I had Expressionism and Dadaism pretty well pegged. What were the distinguishing features, exactly? Couldn’t say. What was the meaning of the unnatural tones in the Fauvist pieces? No idea. I wasn’t stopping to find out. I was giving myself a few seconds on each slide, and moving on. This was perceptual learning, not art history.
Eventually I had to take a test on all this, and here, too, I borrowed from Kornell and Bjork. Remember, they’d tested participants at the end of their study on paintings (by the same artists) that they’d not studied. The idea is that, if you can spot Braque’s touch, then you ought to be able to peg any Braque. That was my goal, too. I wanted to reach a place where I could correctly ID a Dadaist piece, even if it was one I hadn’t studied in the PLM.
Henri Matisse, Portrait of Madame Matisse (The Green Line), 1905, 2014 Succession H. Matisse/Artists Rights Society (ARS), New York.
After a half dozen sessions, I took a test—no thinking allowed—and did well: thirty out of thirty-six correct, 80 percent. I was glancing at the paintings and hitting the button, fast. I learned nothing
about art history, it’s true, not one whit about the cultural contexts of the pieces, the artistic statements, the uses of color or perspective. But I’ll say this: I now know a Fauvist from a Post-Impressionist painting, cold. Not bad for an hour’s work.
The biggest difference between my approach and Kornell and Bjork’s is that interleaving may involve more conscious deliberation. Perceptual modules tend to be faster-paced, working the visual (perceptual) systems as well as the cognitive, thinking ones. The two techniques are complementary, each one honing the other.
What I’ll remember most, though, was that it was fun, from start to finish—the way learning is supposed to be. Of course, I had no exam looming, no pressure to jack up my grades, no competition to prepare for. I’ve given this example only to illustrate that self-administered perceptual training is possible with minimal effort. Most important, I’ve used it to show that PLMs are meant for a certain kind of target: discriminating or classifying things that look the same to the untrained eye but are not. To me it’s absolutely worth the extra time if there’s one specific perceptual knot that’s giving you a migraine. The difference between sine, cosine, tangent, cotangent. Intervals and cadences in music. Between types of chemical bonds. Between financing strategies, or annual report numbers. Even between simple things, like whether the sum of two fractions (3/5 and 1/3) is greater or less than 1. Run through a bunch of examples—fast—and let the sensory areas of your brain do the rest.
This is no gimmick. In time, perceptual learning is going to transform training in many areas of study and expertise, and it’s easy enough to design modules to target material you want to build an instinct for quickly. Native trees, for example, or wildflowers. Different makes of fuel injectors. Baroque composers or French wines. Remember, all the senses hone themselves, not only vision. As a parent I often wish I’d known the dinosaurs better by sight (there are way more types than you might know, and categories, too), or had a bead on fish species before aquarium visits.
The best part is, as Eleanor Gibson said, perceptual learning is automatic, and self-correcting. You’re learning without thinking.
* * *
* “Chunking,” in psychology, is the facility to store studied items in meaningful clusters based on prior knowledge. Take the sequence of letters Y, N, B; C, B, B; C, E; F, I, F; A, C, I; A M, B; A, Y. Study those for a few minutes, then cover your eyes and try to remember as many as you can. The typical number most of us can remember is about seven. Now try it again after grouping the letters in this way: Y, NBC, BBC, FIFA, CIA, MBA, Y. You remember more, because you’ve stored the letters in meaningful groups.
Chapter Ten
You Snooze, You Win
The Consolidating Role of Sleep
The giant rabbit hole in our lives, the dark kingdom we all visit regularly, is sleep. Sleep is a perfect mystery for most of us. We need it, we want more of it, and we long for it to be of a deeper, richer quality. On one hand, we know it can betray us on any given night. On the other, we know that there’s some alchemy going on during those unconscious, dream-filled hours, some mixing of fact, fantasy, and feeling that can turn our daytime struggles to master new skills into that most precious thing—understanding.
You don’t have to be a New Age dream therapist to believe that the brain makes connections during sleep that it doesn’t while awake. Who hasn’t sat upright in bed now and then at 3 A.M. and thought, Oh, of course!, suddenly remembering where you stashed your keys, or visualizing how to alter your golf swing, or to refinger a piece by Albéniz. Countless times I’ve gone to sleep in a state of self-pitying frustration—held hostage by some story I can’t outflank—only to rouse myself in the middle of the night, grab the pen on my nightstand, and scribble out some thoughts that had bubbled to the surface between dreams. In the morning I wake to find a scrawl of partial sentences that, if legible, often help me write my way out.
It’s not just me, either. The history of scientific discovery is salted with hints that sleep fosters profound intellectual leaps. The nineteenth-century German chemist Friedrich August Kekulé, for example, claimed that he stumbled upon the chemical structure of benzene—in which the molecule curls into a ring shape—after dreaming of snakes biting their tails. The Russian scientist Dmitri Mendeleev reportedly pulled several all-nighters, to no avail, trying to piece together what would become his famous periodic table of the elements, but it was only after nodding off, he told a colleague, that he saw “a table where all the elements fell into place.” These kinds of stories always remind me of the Grimms’ fairy tale “The Golden Bird,” in which a young man on a mission to find a magic bird with golden feathers falls in love with a princess, whose father the king will grant her hand on one condition: that the young man dig away the hill that stops the view from his window in eight days. The only complication? This is no hill, it’s a mountain, and after seven days of digging, the young man collapses in defeat. That’s when his friend the fox whispers, “Lie down and go to sleep; I will work for you.” And in the morning, the mountain is gone.
Sleep is the stuff of legends and fairy tales precisely because it’s so unknown, a blank screen onto which we can project our anxieties and hopes. If the darkroom is locked, we can only guess at what images are being developed in there. All of which raises the question: What is the sleeping brain doing, exactly?
For that matter, why do we sleep at all?
The truth is, no one knows. Or, to be more precise, there’s no single, agreed-upon scientific explanation for it. We spend fully a third of our existence unconscious, so any theory about sleep’s central purpose has to be a big one. Doesn’t the body need regular downtime to heal? To relieve stress? To manage moods, make muscle, restore mental clarity? Yes to all of the above. We know that sleep deprivation makes us more reckless, more emotionally fragile, less able to concentrate and possibly more vulnerable to infection. None of those amounts to an encompassing theory, though, because none explains the vast variations in sleep times and schedules. Just think of how dramatically sleep habits differ from person to person. Some people thrive on as little as three hours a night, while others feel helpless without eight; some function best awake all night and out most of the day; others need their daily nap. A truly comprehensive theory of sleep, then, would have to explain such differences. It would also need to account for the sleep-wake cycles in animals, which is breathtaking in its diversity. Female killer whales can be mobile and alert for upward of three weeks when looking after a newborn calf—nearly a month without sleep. Migrating birds fly for weeks without stopping to rest.
Two new theories have emerged that make sense of this chaos.
One is that sleep is essentially a time-management adaptation. Our body’s internal clock evolved to keep us out of circulation when there’s not much of a living to be made—at 3 A.M., for instance—and awake when there is. Consider the brown bat, perhaps the longest-sleeping mammal of them all. It sleeps twenty hours a day and spends the other four, at dusk, hunting mosquitoes and moths. Why only four hours at dusk? Because that’s when food is plentiful. But also because, as Jerome Siegel, a neuroscientist at UCLA, says, “increased waking time would seem to be highly maladaptive for this animal, since it would expend energy and be exposed to predatory birds with better vision and better flight abilities.” Siegel argues that our obsession with sleep quality and duration is, in a sense, backward. “We spend a third of our life sleeping, which seems so maladaptive—‘the biggest mistake nature has made,’ scientists often call it,” he told me. “Another way of looking at it is that unnecessary wakefulness is a bigger mistake.”
When there’s hay to be made, we make it, whether the sun is shining or not. And when there’s none—or too little, given the risks of being out and about—we bed down. In short: Sleeping and waking adjust themselves to the demands and risks of our life, not according to what the health manuals say.
The other theory is that sleep’s primary purpose is memory consolidation. Learning. In r
ecent years, brain scientists have published an array of findings suggesting that sleep plays a critical role in flagging and storing important memories, intellectual and physical. Also (yes) in making subtle connections—a new way to solve a tricky math problem, for example, or to play a particularly difficult sequence of notes on the viola—that were invisible during waking. Think about what we described back in chapter 1, all those streaming sensations, the sheer, insane volume of neural connections the brain has to make in the course of any given day. At some point, we have to decide which of these connections are worth holding on to, and which can be ignored. That’s an easy choice sometimes, and we make it immediately: a new colleague’s name; the pickup time at day care; which house on the street has the angry Dobermans. Other choices are not obvious at all. Some of the most critical perceptions we register in a day contain subtle clues—shrugs, sideways glances, suggestions, red herrings. A world of impressions swirls in our heads when we turn the lights out and, according to this theory, that’s when the brain begins to sort out the meaningful from the trivial.
In the contentious field of sleep research, these two theories are typically set in opposition, one trumping the other as the primary function of our unconscious lives. In reality, they are hardly mutually exclusive. Only by putting them together, in fact, can we begin to understand how sleep aids learning—and to use that understanding to our advantage.
How We Learn Page 19