When I visit Matthew Carpenter’s math class, I peer over his shoulder at his laptop and see on-screen the question he’s tackling:
cos-1(1) = ?
It’s a scrap of inverse trigonometry. I’ve long forgotten much of my trig, so I shrug. Matthew, however, is undaunted. Squinting in concentration, he clicks one of the four possible multiple-choice answers: 0 degrees. Ding: the software informs him that he’s gotten it right. It throws another question at him, which he also answers correctly; then another, and another, until he’s aced ten questions in a row. “This is my favorite exercise right now,” he tells me. He’s certainly practiced a lot. He points to a section of the screen that shows he’s tackled 642 inverse trig problems. “It took a while for me to get it,” he admits sheepishly, but he’s plugged away at it in class and at home for hours.
Matthew shouldn’t be doing work remotely this advanced. He’s ten years old, and this is only the fifth grade. Matthew is a student at Santa Rita Elementary, a public school in Los Altos, California, where his sun-drenched classroom is festooned with a giant paper X-wing fighter, student paintings of trees, and racks of kids’ books. Normally grade five math is simpler fare—basic fractions, decimals, and percentages. You don’t reach inverse trig until high school.
But Matthew’s class isn’t typical. For the last year, they’ve been using the Khan Academy, a free online site filled with thousands of instructional videos that cover subjects in math, science, and economics. The videos are lo-fi, even crude: about five to fifteen minutes long, they consist of a voice-over by Khan describing a mathematical concept or explaining how to solve a problem while hand-scribbled formulas appear on-screen. The site also includes software that generates practice problems, then rewards hard work with badges—for answering a “streak” of questions right, say.
Matthew has amassed fifty-two Earth badges, one of the more desirable awards on the site. When a girl in a pink jumpsuit wanders by and peeks at Matthew’s recent streak of inverse trig, she groans: “Oh, great. I need to catch up with you now!” She flops down at a laptop and begins pecking away at her own trigonometry problems. Her first one asks her to divine the slope of y = −1x3 + 4y2; scribbling on a piece of paper as I watch, she figures it out in a few seconds and clicks the right answer on screen. So far, she tells me, she’s watched dozens of hours of math videos, mostly at home.
How did these elementary school kids zoom ahead to high-school-level material?
In part because the site lets them learn at their own pace—allowing their teacher, Kami Thordarson, to offer much more customized instruction. The problem with traditional classroom dynamics, Thordarson tells me, is that they don’t easily account for the way kids learn at different rates. When she stands up at the chalkboard lecturing on a subject, there’s a predictable pattern that takes hold: one quarter of the kids quickly fall behind, so they tune out. Another quarter already know the material, so they tune out. At best, “you’re teaching to this middle group of students.” Thordarson sighs.
What works better? Personalized, one-on-one tutoring. Back in 1984, the educational scholar Benjamin Bloom compared students taught in regular classrooms—one teacher lecturing to the assembled class—to students who got months of one-on-one attention or instruction in small groups. These tutored students did far better; two standard deviations better, in fact. To get a sense of how much of an improvement that is, think of it this way: If you took a regular-classroom kid who was performing in the middle of the pack and gave her one-on-one instruction for a few months, she’d leap to the ninety-eighth percentile. This became known as the “Two Sigma” phenomenon, and in the decades since, public-school teachers have struggled to give students more one-on-one time. This isn’t easy, given that the average class in the United States has roughly twenty-five children. (Worse, after years of slightly falling, that number is now rising again, due to budget cuts.) Until the government decides it’s willing to subsidize smaller classes, how can teachers get more personal time?
One way is by using new-media tools to invert the logic of instruction. Instead of delivering all her math lessons to the entire class, Thordarson has them watch Khan videos and work on the online problems. This way, the students who quickly “get it” can blast ahead—and Thordarson can focus more of her class time on helping the students who need coaching. Other teachers are even more aggressive about inverting their classes: They assign videos to be watched at home, then have the students do the homework in class, flipping their instruction inside out.
This makes curious psychological sense. A video can often be a better way to deliver a lecture-style lesson, because students can pause and rewind when they get confused—impossible with a live classroom lesson. In contrast, homework is better done in a classroom, because that’s when you’re likely to need to ask the teacher for extra help. (Or to ask another student: Thordarson and her colleagues noticed students helping one another, sharing what they’d learned, and tutoring each other.)
“Kids get to work in their place where they’re most comfortable,” says Thordarson as we wander around her class. “They’re allowed to jump ahead. It gives kids who are above grade level a chance to just soar! And for kids who struggle, it gives them a chance to work through some of those issues without everybody watching.”
Still, as Thordarson quickly points out, the Khan Academy isn’t enough on its own. You can’t just plunk kids in front of laptops and say, “Go.” The point isn’t to replace teachers. It’s to help them reshape their classes in new ways—and spend more time directly guiding students. You can’t even say it makes the teacher’s job easier. If anything, it has made Thordarson’s job more challenging: there’s more noise, more kids talking, and she’s constantly darting around the room to help out. One U.S. federal study found that students learned best in classrooms with precisely this sort of “blended” learning—traditional teachers augmented with online instruction. But the increase in learning wasn’t because of any magic in the medium. It’s just that online tools helped students and teachers spend more time on the material.
Judging by Thordarson’s success, though, it works. She’s seen particularly strong improvements at the low end: Only three percent of her students were classified as average or lower in end-of-year tests, down from thirteen percent at midyear—and other math teachers at Santa Rita have seen similar results. The kids who need help have been getting more of it; the kids who want to push ahead are pushing ridiculously far.
“It’s like having thirty math tutors in my room,” Thordarson says.
• • •
The classroom hasn’t changed much over the years. Over the centuries, actually. In the 1350s, artist Laurentius de Voltolina painted a scene of a university lecture in Bologna that looks quite like a present-day classroom: The professor sits at a podium at the front, pontificating to twenty-four seated students, one of whom is keeling over in apparent boredom, four of whom are ignoring the lecture while talking, and one of whom appears to be completely asleep. As various educational analysts have joked, if you brought a bunch of surgeons from a hundred years ago into today’s hospitals, they would have no idea what was going on, because everything about their craft had evolved: antibiotics, laparoscopic devices, MRIs. But time-traveling teachers would have no trouble walking into an elementary school (or even Harvard) and going to work, because schools are nearly identical. Walk to the front of the class, pick up the chalk, and start lecturing.
These days, a huge debate rages in the United States about how schools ought to modernize. The school-reform movement argues that schools are “failing” because they’re hobbled by union-coddled teachers who block change. Only by rigorously testing kids and firing teachers who can’t produce rising scores—while offering merit pay to teachers who can—will schools improve. Critics of the reform movement counter that these standardized tests not only fail to measure actual learning but deform education, as teachers drill children in meaningless test-prep skills. To really improve schools, goes the counterargument, you’d need to seriously train teachers (as they do in Finland), tackle the poverty that haunts many students’ home lives, and pony up public funds for smaller class sizes.
This debate extends cosmically beyond this book, so I’m certainly not going to resolve it here. (Though I personally agree more with critics of the school-reform movement.) I’m also not going to delve deeply into the question of how frequently young students ought to be sitting in front of screens, another fraught area of debate. There’s little clear data on the subject as yet, though the work I’ve seen suggests that children in early grades learn best with hands-on, tactile experimentation—which suggests digital tools ought to be used very sparingly early on. What interests me here is the unique roles that digital tools might have in the later grades: Can they help students learn, and if so, how?
Of course, classroom technology has a long history of hype that has rarely delivered. In the nineteenth century, George Parsons Lathrop predicted that movies would let children experience “the majestic tumult of Niagara” or “the animated presence of far-off peoples.” Soon after, radio became the hot new thing; surely it would enrich students with high-quality lectures. A couple of decades later, the journal Nation’s Business declared that TV would become “the biggest classroom the world has ever seen.” Each technology barely made a dent.
Of these newfangled devices, computers have been the most successful in infiltrating classrooms—certainly in terms of dollars spent. Schools have spent an estimated sixty billion dollars in the last twenty years on computers. Indeed, manufacturers and software firms love to target schools, since, rather like the military, they are lucrative customers, an enormous sector loaded with tax dollars to spend. (Apple computers are omnipresent in schools not just because they’re well made, but because, as New York Times reporter Matt Richtel has documented, Apple salespeople actively woo school officials to visit their Cupertino headquarters, paying for hotel stays and pricey dinners.) As a result of this high-tech spending, the ratio of computers to kids has shifted from 1:12 in 1998 to less than 1:4 today.
Despite all that spending, computers are not often used to do anything that couldn’t be done as easily—and more cheaply and effectively—with old-fashioned books, pencils, paper, and chalk. “Interactive” whiteboards are too frequently used merely for displaying text or pictures to a dutifully seated class, not much different from what you can do with an overhead projector or chalkboard. Computers are used for mere word processing, to display cognitively bleak commercial software packages like PowerPoint, or to play dull educational games of dubious merit. Computer “art” packages get used as bland substitutes for paint-and-paper art.
But some educators are realizing that this is a dead end. They’re the ones like Thordarson, who are doing something different with technology in the classroom. These teachers have realized that the point isn’t to simply replicate more expensively what they’re already doing quite well with paper, pencils, and books. It’s to do new things that they currently can’t do. It’s to teach kids by using the peculiar abilities of networked devices—like public thinking, new literacies, and the powerful insights that come from not just using, but programming, the machine.
• • •
Consider what’s happening beneath the hood of the Khan Academy. In one sense, Khan’s videos are the most prominent part of the system. But they’re also the least innovative one. They’re still pretty much just traditional lessons and lectures, albeit ones that can be consulted and reconsulted worldwide, at any time.
What’s new is how teachers use the Khan Academy to track progress. The system offers a dashboard that displays nuanced information about each student: which videos they’ve looked at, which problems they’ve tackled, how many times they had to work at a problem before they solved it. This data offers pragmatic insight into whether a student is struggling or not—in real time, whether the child is working in the classroom or at home. It takes the invisible and sometimes mysterious development inside a student’s head and makes it visible.
One afternoon I drop by the seventh-grade classroom of Courtney Cadwell, a math teacher at Egan Junior High in Los Altos, California, across town from Thordarson’s elementary school. Cadwell is a high-energy former Texan who was in line for NASA’s astronaut training program before becoming a teacher (and who has since become the principal of another Los Altos school). She needs that energy because her job is about as hard as they come: Her students are remedial, and many come from poor immigrant families in which the parents speak little or no English, so they’re rarely able to help their children with homework. When the children arrived at the beginning of this year, some were functioning at a third grade level in math, she tells me.
The dashboard helped her zero in on what kind of help each student needed, often in real time, letting her tailor her instruction or offer extra guidance. “Usually we get to the end of a lesson and we get a quiz or a test and then you realize, Oh my gosh, they have all these gaps!” Cadwell says. “With this, I can spot it sooner and fill those gaps.”
Being able to target students has paid off dramatically. In her first six months using the Khan Academy, her class’s scores had improved more than 106 percent. One girl advanced an astonishing 366 percent. Whereas the class average used to be three years behind, now it’s a bit over a year behind. At this rate, by the end of the year, Cadwell will have them caught up. She pulls up the dashboard to show me the charts. Interestingly, their progress goes in bursts: They seem to grapple with a concept for days or weeks, then suddenly get it and improve abruptly in performance.
Cadwell has been teaching for eighteen years but she’s never been able to so quickly identify and address students’ areas of weakness. “It’s just incredible, these gains, you know!” she gushes. “It’s insane.”
I hang out and observe the students on their laptops as they watch videos and blast through problems while Cadwell darts around, coaching one and then another. I talk to the student who’d advanced by 366 percent, a smiling, long-haired girl in purple-rimmed glasses. “I hated math,” the girl tells me cheerfully as she plows through a long-division problem. “But now it’s actually fun.” Like many students, she also found it encouraging to see her own improvement—to watch the chart line move upward. Normally students have only a crude sense of the rate at which they’re learning, but when software is charting their progress, it’s motivating simply to see the data. “When you’re collecting badges for getting a streak of questions right, it’s instant feedback,” one of them tells me. “You want to keep going, doing more, more, more.” One Santa Rita teacher told me she’d noticed that a kid had worked on problems from midnight until 2 a.m. during a bout of insomnia.
Collecting such fine-grained data is likely to have other payoffs. When I talk to Sal Khan, who runs the Khan Academy as a nonprofit supported by donations, he points out that students have answered more than a billion questions on his system, and the videos have been viewed over 230 million times. “So we can start looking for trend lines that help us figure out, What types of things are students likely to get stuck on? If someone breezes through trigonometry but gets stuck on the intro to statistics, can we predict what other things they’ll find hard or easy? Can we help give more information to teachers to help them teach?” This is, of course, one of those things that computers are uniquely good at: finding patterns that we can’t see ourselves.
• • •
The Khan Academy can work for math and sciences, where problem sets can be autogenerated and automatically graded. But what about teaching kids to read and write?
These skills lags behind math, believe it or not. One of the most reliable measures of how kids are learning in the United States is the National Assessment of Educational Progress, a testing program run by the federal government. The tests are reliable mostly because they’re not used to assess or rank individual schools and teachers; since nobody’s worried about getting fired if their kids don’t excel on NAEP tests, the results aren’t deformed by test prep. And what do they show? News both good and bad. The good part is the math scores, which have slowly and slightly risen over the last thirty-five years. The bad news: the reading and writing scores. Only the lower grades have gained; amazingly, the average reading ability of seventeen-year-olds in 2008 was nearly identical to what it was in 1971. As for writing, the NAEP’s data on this isn’t complete, since it was only collected for a decade beginning in the mid-eighties. But in that window, things didn’t look great: Writing skills remained generally flat and got worse for kids in eleventh grade.
Studies show that the main way to help kids read is to give them more reading time in school, offer them books they’re interested in, and teach them the mental strategies of good readers—like summarizing what they’re understanding and what they’re not.
But research also shows something else: Writing is curiously pivotal to reading. As I noted when I first discussed public thinking, writing about the stuff we’re reading activates the generation effect. We internalize our reading more deeply. Indeed, literacy scholar Steve Graham recently crunched dozens of reading studies and found that “writing about a text proved to be better than just reading it, reading and rereading it, reading and studying it, reading and discussing it, and receiving reading instruction.”
But as any teacher knows, getting students to write is unbelievably difficult, in part because writing assignments feel so artificial. After all, only the teacher is going to read your paper. Why bother working hard on it?
Dorothy Burt, a literacy project facilitator in Point England, New Zealand, knew all about these problems. Like Cadwell’s school in California, Burt’s is located in a low-income cluster of schools, with high illiteracy rates and many students for whom English is a second language. Back in 2007, the writing skills of students in that cluster languished far below the national average. The teachers wanted their kids to practice writing more but could only get a few sentences out of them.
They decided that the students were right. Composing essays is meaningless. Teachers are an inauthentic audience. They aren’t necessarily interested in what their students have to say; they’re just reading as a part of their jobs. A writer is being forced to write for an audience that’s being forced to read. No wonder they think it’s pointless, Burt thought.
So the New Zealand teachers decided to try using the power of public thinking. Instead of having assignments filed in paper and dropped on the teachers’ desks, students would post on public blogs. Anyone could read the posts and comment on them, and the school encouraged parents and friends to do so, including those overseas.
At first, nothing changed. The students still grumbled. But soon, when comments started appearing, they realized that they were writing for a real audience. Occasionally the comment would come from a complete stranger—and that caught the students’ attention. “How is someone from Germany reading what I’m writing?” one student said. When another student posted a review of a book, the book’s author heard about it and showed up to comment, too.
The students were electrified. They began writing posts far more frequently, waiting patiently for computer time and groaning when teachers told them to stop. The ones who had computers at home posted in their own time, on weekends and vacation. Previously reluctant readers began doing more careful research. Students began paying closer attention to grammar and punctuation: They’d read their entries aloud to notice that commas were missing or used too often.
Students also began critiquing each other’s writing, demanding they clarify their points. “I don’t get what you mean,” one told another. “We can’t read your mind.” Some began editing one another’s posts to remove local jargon, reasoning that foreigners wouldn’t understand the references to, for instance, New Zealand’s national rugby team: “People in America won’t know what the ‘All Blacks’ are,” one student admonished another. “They were writing for a global audience,” one local literacy researcher, Colleen Gleeson, tells me. These are acts of self-awareness that professional writers struggle with: forming a theory of mind of one’s audience, the better to communicate with it. Or as Burt adds when I interviewed her: “The blogging environment gave the students an audience that had a choice not to read. So if they do choose to read it it’s because they want to.”
A year into the program, the New Zealand schools decided to expand the experiment dramatically, by finding a way to get every student a netbook. It cost each family about twelve U.S. dollars per month per student—significant, but affordable—and after three years the student would permanently own the laptop. The schools even began building neighborhood-wide WiFi networks to share their connection freely.
Did the experiment actually improve learning? When they looked around, Burt and the teachers could see dozens of positive effects. Students were more excited to write and research, and saw school as newly relevant—they were, in the au courant term, engaged. But “engagement” is a fuzzy, catch-all rationale for using technology. Sure, having lots of shiny new devices around might confer a sense of novelty, improving “engagement”—but parents and politicians aren’t going to care unless you can actually prove the kids are learning. They want test scores to go up. (And while test scores are a crude way to measure actual education, they’re the current yardstick, for better and for worse.)
It turns out the scores did improve, and dramatically. By the end of the second year, the scores of those New Zealand schools for reading and writing were advancing at an astonishing clip—some posting improvements ten to thirteen times larger than the national average increase. In 2008, they were dramatically below the nation at large. By 2011, some grades had reached parity, and some were above.
The motivational force of public thinking and online collaboration for kids is even easier to see if you look outside the classroom—because that’s where it’s mostly happening. As fan culture scholar Henry Jenkins once pointed out to me, children and young adults who bridle at writing a few paragraphs in school will cheerfully spend months on sprawling works of online prose, from fan fiction to TV show recaps to video game guides. One student I spoke to, Eric Davey, wrote a patiently detailed thirty-eight-page guide to a Star Trek game that clocked in at thirteen thousand words—while enrolled in a high school where the longest assignment was two to four pages. Over at the Lostpedia, one of the pioneers was a thirteen-year-old boy named Sam McPherson, who’d spent long hours after school and on the weekends writing and painstakingly editing the work of others. When I called him to talk, Sam (by now seventeen and in college) told me the project gave him not only hundreds of hours of writing practice but also the difficult but valuable experience of working with others.
“If there was an argument about an article, we would set up a discussion, and it could go for a week,” he noted. The show provoked him to do academic reading in his spare time. He became intrigued by the show’s recurring Egyptian symbolism and eventually wrote a four-thousand-word chapter on the subject for a book of collected essays on Lost. “I wound up getting pretty into it,” he joked.
Now, simply turning kids loose online won’t necessarily improve their formal writing. Teachers are still crucial. Kids know this: when high school students were surveyed about their writing by the Pew Research Center’s Internet & American Life Project, more than half said “the writing instruction you have gotten in school” was the main reason their writing improved, and over 80 percent thought their writing would improve if teachers spent more time teaching formal writing. Nor should all writing be public. Some types of psychological and intellectual exploration require privacy, as any old-fashioned pen-and-paper journal keeper knows. But when harnessed well, digital forums can give students a reason to write more frequently, while teachers can help them to write better.
This also gives them a chance to learn digital citizenship. Heidi Siwak, an Ontario elementary school teacher, began encouraging her students to use Twitter to create ongoing class conversations. Siwak quickly realized this was a golden opportunity to learn online civics—how to respond politely and intelligently not just to friends but strangers. In a sense, she’s teaching tummeling. Siwak set up several clever Twitter projects, including a daylong event where her class posted their thoughts on Hana’s Suitcase, a book about a teenager killed in the Holocaust; as the stream got retweeted, interested strangers from Asia and Europe chimed in. Soon Siwak’s class began using Twitter for other research, contacting experts on subjects like the environmental impact of Arctic exploration. “They’re learning how to conduct themselves online—how to have productive conversations and exchanges,” Siwak tells me.
• • •
Most attempts to use digital technology in education focus on having students learn programs: word processing programs, presentation programs, microblogging tools, search engines.
But truly clever teachers go one step further: They teach their students programming itself—how to write code. This isn’t just about imparting geeky skills that will be useful on the job market. The teachers know that programming has deeper effects: For children, it becomes a philosophic act, a way of learning about learning.
This was the epiphany of Seymour Papert, an MIT mathematician and computer scientist. Back in the 1960s, Papert became interested in why so many people become “mathophobic,” and he argued that it was because students are taught these subjects as dry, abstract rules. When you’re exposed to high school calculus, it doesn’t seem to map onto anything real. You can’t easily apply the knowledge. It’s like learning French in the United States, where there’s nowhere to speak it. To really learn the language, you have to go to a place where it’s spoken all day long, such as France, where kids pick it up in everyday life. “If we had to base our opinions on observation of how poorly children learned French in American schools, we would have to conclude that most people were incapable of mastering it,” Papert wrote.
Math, geometry, and logic, Papert figured, suffer from the same problem. We assume they’re hard to learn, but maybe that’s because we don’t live in a place where they’re spoken as an everyday language. To really learn them, you can’t just take lessons in school. You need inhabit a land where they’re spoken every day, a “mathland.”
Computer programming, Papert realized, is just such a mathland. At MIT, where Papert and his colleagues were pioneers on early computers, he’d learned that programming is more than just a technical skill. It shapes your thought and allows you to think in new ways.
Among other things, programming requires you to think logically. Computers are obedient but they’re dumb; they do what you say, but only what you say. If you ask a human friend to get you a glass of milk, he doesn’t need much more instruction. Humans implicitly know what they have to do: Go to the fridge, get a carton of milk, open it up, check that the milk is still fresh, pour it into a glass, keep the glass upright, and put the glass on a flat surface near you so it doesn’t spill. But to program a computer to do the same thing, you’d have to meticulously spell out every teensy step. (If you forget to tell the program to get a glass, it’ll pour the milk on the floor.) Programming requires an attention to detail and an ability to think about everything as a series of processes. What’s more, computers “think” in numbers and procedures, so they’re an environment where math is the native language. And because they have the brute-force ability to follow your instructions over and over, dozens or hundreds or millions of times, they never get tired. You can conduct what-if thought experiments that are impossible with pen and paper.
To give children a toehold in mathland, Papert and his colleagues created Logo, a very simple computer language. In Logo, the child controls a little turtle on-screen, issuing it commands to make it move around. The turtle draws a line wherever it goes, so it’s kind of like using a computerized Etch A Sketch. To draw a square, a child would tell the turtle to go forward thirty steps, turn right ninety degrees, then do the same thing three more times. Children quickly got the hang of it, using Logo to write programs that would draw all manner of things, like houses or cars. They’d laboriously write one instruction for each step of the picture, almost the way you’d set up the dots for a connect-the-dots drawing. To draw a bird, they’d connect two quarter circles together.
But pretty soon the kids began to discover something even more fun: You could take a simple command, repeat it endlessly, and produce something unexpectedly beautiful. If you took that little bird program and adapted it to have the computer draw twenty copies of the quarter-circle while rotating it a bit each time, presto: The resulting picture looked like a flower. Or you could pick a seemingly simple instruction—go forward ten steps, turn right ninety degrees, increase the number of steps by five, then repeat over and over—and discover it produced something unexpected: a square spiral, growing eternally larger. The children began to grasp the concept of recursion, the idea that complexity emerges from repeating a simple procedure over and over. They also began to intuit the butterfly effect: how changing one tiny part of a program can radically alter the outcome. If you tweak one element in that square-spiral program, making the angle ninety-five degrees instead of ninety, surprise: The squares will shift slightly, producing a new creation, looking like a spiral galaxy. And ninety-seven degrees looks different, too.
This idea—that very small alterations can produce wildly different results—is something that many adults often fail to grasp, leading to massive failures in corporations, governments, teams, and families: The people at the top think that making a little change won’t make much difference, but that little change spirals out of control. “Our culture,” Papert wrote, “is relatively poor in models of systematic procedures.” The turtle let the students think about math, and the world around them, as a series of systems: This is computational thinking.
The students also intuited many deep concepts of math on their own, merely by being able to experiment with geometry in a practical, sensual fashion. While trying to get the turtle to move in a square, they’d realize that the sum of the angles always adds up to 360 degrees. While trying to get the turtle to walk in a circle, they figured out that the trick was to have it take a tiny step, make a slight turn, and do it again and again—in essence, that a circle was nothing but a series of straight lines so short that they look like a curve. That’s a foundational concept in calculus. Yet elementary school children discovered it, completely on their own, merely because they could mess around inside a world that encouraged those thought experiments.
Computational thinking isn’t limited to math. Papert also had students craft poetry-generating programs. As in Mad Libs, the children would feed the program verbs, adjectives, and nouns, and the computer would combine them into lines of poetry: “MAD WOLF HATES BECAUSE INSANE WOLF SKIPS” or “UGLY MAN LOVES BECAUSE UGLY DOG HATES.” The process of trying to get the program to work lent students startling insights into language. Jenny, a thirteen-year-old girl who had previously earned only mediocre grades, came in one day and announced, “Now I know why we have nouns and verbs.” She’d been taught grammar and the parts of speech for years without understanding them. But as Jenny struggled to get the program to work, she realized she had to classify words into categories—all the verbs in one bucket, all the nouns in another—or else the sentences the computer spat out wouldn’t make sense. Grammar suddenly became deeply meaningful. This produced an immediate spillover effect: Jenny began getting As in her language classes. “She not only ‘understood’ grammar, she changed her relationship to it,” Papert noted.
Indeed, teaching programming can invert one’s sense of which students are good and bad at math and logic. Gary Stager, a disciple of Papert’s, told me about teaching Logo Microworlds—a variant of Papert’s language—to a group of young students at an international school in South Korea. Most of the kids were eagerly working on crafting programs to solve a problem. But “a little girl of about six years old, who was publicly identified as highly gifted, quickly burst into tears after realizing that the programming problem confronting her didn’t have a single immediate answer,” Stager noted in an e-mail. Kids who are good at traditional school—repeating rote concepts and facts on a test—can fall apart in a situation where that isn’t enough. Programming rewards the experimental, curious mind.
Most important, students learn about learning itself. Computer programming is about trial and error: Few programs work the first time. Usually you’ve omitted an instruction or perhaps made a typo. The process of figuring out what’s wrong and fixing it is exhilarating—it works! It’s also a powerful lesson: It proves that you learn by experimenting and making mistakes, not by trying to be perfect the first time. As Papert wrote, “Many children are held back in their learning because they have a model of learning in which you have either ‘got it’ or ‘got it wrong.’ But when you learn to program a computer you almost never get it right the first time. . . . If this way of looking at intellectual products were generalized to how the larger culture thinks about knowledge and its acquisition, we all might be less intimidated by our fears of ‘being wrong.’”
One of the most popular modern descendants of Logo is Scratch, a programming language for children created by MIT and distributed for free online. Using Scratch, children can create quite sophisticated games and animations. One morning I visit the class of Lou Lahana, a technology coordinator at PS/MS 188, a public school in a low-income neighborhood in New York. As the bell goes off, twenty students file into the computer lab—a modest room tucked into the corner of the building—then grab their school laptops and begin working.
Two eighth-grade kids sit next to each other, peering at one screen and puzzling over a game. They’re Ruben Purrone and Esmil Sanchez, and they’re working on Ruben Invaders—their clone of the famous arcade game Space Invaders. They’ve been fiddling with it for a few days, but it’s still filled with bugs that keep it from working. For starters, when the player fires a missile at the descending aliens, it passes harmlessly through them. In technical terms, this is what’s known as a “collision detection” problem.
“I don’t get it—what’s up with the missile?” wonders Ruben as he fires projectile after projectile.
Esmil grabs control of the trackpad and begins clicking through the game’s code. It’s extensive and complex: There are scores, possibly hundreds of commands in their program. “It gets hard to keep track,” Esmil mutters, half to himself.
Still, they doggedly bang away at it for ten minutes until Ruben suddenly spots the problem: The missile doesn’t have the right bit of code to detect the presence of the aliens. It doesn’t “know” when it’s made contact with them. Esmil pounces; he knows how to fix this. “We gotta set the color,” he says, mousing over to the code and fiddling with its settings.
Then he hits “fire” and presto: The missiles now work. Aliens are exploding!
As I watch, the two spend the next half hour poring over the code, tweaking parameters and brainstorming solutions. (Ruben: “How do you switch the background?” Esmil: “Make the background a sprite!”)
This is another side benefit of learning programming: It’s a collaborative activity. Like most Scratch creators, Esmil and Ruben learned their skills not just from Lahana but also from reading guides and online forums where other kids (and adults) offer one another advice and debug one another’s code. MIT hosts a site where any student can post their project for anyone to play, see, and download. Often students will download someone else’s Scratch game to reverse engineer it, then tinker and upload their own version for public scrutiny. This has produced a rich culture of public thinking via code. Some Scratch players have formed their own international groups, collaborating on games remotely—with kids in Poland working alongside students in the United States, the United Kingdom, and India to create programming, music, and art. Sometimes students will discover another kid using their code, prompting debates on plagiarism that have scholarly dimensions. How much should you credit someone else if you use a bit of their code? Or a “sprite” from their game? What constitutes a contribution so creative that you can put your name on a remix?
Learning how to collaborate—particularly on a meaningful real-life project, not just a piece of schoolwork—is a crucial skill. In real life, people rarely learn and solve problems in isolation. They do it together. “Learning from others is neither new nor revolutionary; it has just been ignored by most of our educational institutions,” as Douglas Thomas and John Seely Brown write in their book A New Culture of Learning. Yet this type of work happens very seldom in classrooms. A 2007 U.S. government survey of 737 fifth-grade classrooms found that the students spent over 90 percent of their class time working solo or listening as a class to teacher lectures, while spending almost no time—not even 5 percent of the day—working collaboratively. Obviously, not all work should be collaborative. Students need to be self-reliant, and some forms of creativity require deep solitude. (Just ask a novelist.) But current school is far too weighted toward the type of solitary work that is, for good reason, rare in daily life.
There’s one last, unexpected benefit to learning programming: like public thinking, it has civic dimensions.
Young people live in a world where their daily activities are channeled by digital tools—from Facebook to photo-sharing apps to Google—yet few understand how these tools work. If you learn even a bit of programming, it is, as media theorist Douglas Rushkoff argues, like gaining X-ray vision into the digital world around you. You might begin to realize that if Facebook’s much-criticized privacy settings are overly complicated, they were designed that way—and that they could just as easily have been designed a different way, less favorable to advertisers and more favorable to users. Or you’d realize that black-box electronic voting machines are inherently untrustworthy, or how digital-rights copyright protection on e-books devalues the book by transforming it from a piece of personal property to a revokable license. Much as learning law gives students tools to think about justice, learning programming gives them tools to think critically about digital life.
As Rushkoff puts it, “You gain access to the control panel of civilization. . . . Programming is the sweet spot, the high leverage point in a digital society. If we don’t learn to program, we risk being programmed ourselves.”
• • •
Programming games opens up new ways to learn and think. But even playing them can also do so, as Constance Steinkuehler has discovered. Steinkuehler is a professor at the University of Wisconsin who studies how and why young people play video games. This means she plays a lot herself, including games like Lineage—a “massively multiplayer” online world that’s particularly huge in Asia. Much like the popular World of Warcraft, you pick a character and go on quests to beat monsters, often banding together with other players in a guild to defeat the biggest, baddest enemies—the “bosses”—which wins you the richest treasures.
Steinkuehler had joined a guild in which many players were teenage boys. They were unusually good at defeating the really hard bosses. Guild members spend a huge amount of time chatting via the in-game messaging system; despite the brief about games being “isolating,” Lineage is as social as a pub, with kids showing up as much to talk as to play. One day Steinkuehler asked her guild members how they’d gotten so good at beating the bosses.
It turns out a group of the teenagers had built Excel spreadsheets into which they dumped all the information they’d gathered about how each boss behaved: what magical potions and weapons wounded it the most, what counterattacks the boss would employ, and how much damage each attack would cause. Like many video games, Lineage is quite numeric—each attack shows a number toting up the damage done. After carefully collecting all their data, the teenagers used Excel to build a mathematical model that explained how the boss worked. Then they’d use the model to predict which attacks would be most likely to beat him.
That’s when it hit Steinkuehler: the kids were using the scientific method. They’d think of a hypothesis, like “This boss is really susceptible to fire spells.” They’d collect evidence to see if the hypothesis was correct. If it wasn’t, they’d improve it until their hypothesis accounted for the observed data.
“My head was spinning,” she tells me. When she met up with one of the kids, she asked him, “Do you realize that what you’re doing is the essence of science?”
Steinkuehler began researching conversations between players in World of Warcraft discussion boards and found that a shockingly high percentage of them involved “scientific” activity. Fully 86 percent of every posting was devoted to “knowledge construction,” players offering hypotheses about how the game works. Meanwhile, 37 percent of the posts involved players thinking collaboratively, building on one another’s ideas, and just as frequently the posts would offer counterarguments. And 58 percent displayed “systems-based reasoning”—thinking about the game as a complex environment and meditating on the rules that govern it. Here’s a sample of the type of talk she found in a discussion where a player was outlining the relative powers of a mage and a priest:
By intuition, you should notice a problem . . . but I’ll give you the numbers anyways
For Mindflay, SW:P, and presumpably VT [3 priest spells]:
Damage = (base_spell_damage ? modifier * damage_gear) * darkness * weaving * shadowform * misery
For Frostbolt [mage spell]
Average Damage = (base_spell_damage ? (Modifier ? empowered frost) * damage_gear) * (1 * (1—critrate—winter’s chill—empowered frost) ?
(1.5 ? ice shards) * (critrate ? winter’s chill ? empowered frost)) * piercing ice
mindflay = (426 ? 0.45 * dam) * 1.1 * 1.15 * 1.15 * 1.05
650.7 ? 0.687 * dam
frostbolt = (530 ? (0.814 ? 0.10) * dam) * ((1—crit—0.10—0.05) ? (1.5 ? 0.5) * (crit ? 0.10 ? 0.05)) * 1.06
(530 ? 0.914 * dam) * ((0.85—crit) ? 2 * (crit ? 0.15)) * 1.06
0.968 * (dam ? 579.7) * (crit ? 1.15)
Please notice the 0.687 versus the 0.968. That’s the scaling factor.
This is, as the academic James Paul Gee describes it, “algebra talk.” Kids who normally couldn’t care less about science were conducting university-level analysis as part of their hobby.
Steinkuehler now argues that video games are one of the best modern conduits to teach kids about the scientific method—why and how it works.
As she points out, many kids hate science because it’s taught as a set of facts. Indeed, that’s how most adults see science: a bunch of guys in lab coats solemnly delivering information about How the World Works. But science isn’t about facts. It’s about the quest for facts—the process by which we hash through confusing thickets of ignorance. It’s dynamic, argumentative, collaborative, competitive, filled with flashes of crazy excitement and hours of drudgework, and driven by ego: our desire to be the one who figures it out, at least for now. Viewed this way, the scientific method is deeply relevant to everyday life, because it describes how to approach and solve problems. But in school, students are rarely asked to actually use the scientific method.
Games, Steinkuehler says, are an ideal native environment for teaching the power of scientific rigor. If science seeks to uncover the invisible rules that govern the world around us, video games are simulated worlds with invisible rule sets just waiting to be uncovered. Teachers should bring games into the classroom, she argues, so they can use them to help explain how science works.
These are fighting words. Educationally, video games are derided as a supreme waste of time and a detriment to literacy, sucking up teenagers’ hours that could be devoted to reading or presumably more productive hobbies. These concerns can be valid, as I can attest; I’ve played video games avidly for thirty years and am painfully aware how compulsive they can become. (I had to almost completely avoid my Xbox to get this book written.) As former Wired editor Chris Anderson once semijoked to me, “My kids would rather play games than breathe.” Helping students learn to moderate how much they play ought to be a crucial piece of teaching and parenting.
But Steinkuehler is also right. Games evoke modes of thinking that can be enormously valuable in education. They teach you that complex things are interesting because of their complexity. The trick is to learn how to use them in the right way.
Kurt Squire is figuring that out. Another professor in the University of Wisconsin’s games department, he has used the game Civilization 3 to teach low-performing kids about history, geography, and politics. In Civilization 3, the player picks a country and, beginning in 4000 BC, guides it through history. Players have to figure out how to devote the scarce resources of their country: Should they focus on improving agriculture? Building armies? Developing technology or artistic culture? If they’re landlocked, how will they get access to water? The player interacts with other countries, which might try to invade or offer to trade. Civ 3 is renowned for its difficulty, something like chess played with geopolitics. Squire hoped that playing would inspire the students to think about the processes that drive history, like economic development and geography and war.
He had his work cut out for him. He was invited to oversee a group of struggling students in a Boston high school. Some had absentee rates of 50 percent; all had flunked ninth grade, and one seventeen-year-old was repeating ninth grade for the third time. Most had woefully little geopolitical education. Amazingly, the school had eliminated world history as a course, because the subject wasn’t part of Massachusetts’s high-stakes tests. (Why teach it if the politicians weren’t testing for it?) At first, the class was chaotic, the students aggressive and misbehaving. Squire tried to lecture about history to prepare the students for the game, but they had little interest. This is going to be a disaster, he thought.
But when they started playing, things changed. The students quickly discovered that Civilization 3 was hard. Most often they’d start off frantically building armies, hoping to defend themselves against invasions. But the invading armies were always stronger, because those countries hadn’t focused solely on militarization—they’d also cultivated agriculture and technology and infrastructure. The students began realizing that a country needs to thrive on many fronts; you needed guns and butter.
So they demanded education. They’d pester Squire with questions about history and economics and geography. One girl, Andrea, began reading encyclopedia entries and interrogating Squire about naval warfare, to try to devise a strategy that would beat her dreaded Roman neighbors. Chris, a boy struggling to play as the Iroquois, realized he needed to know more about agricultural development, so he, too, began pestering Squire for lectures (“and he wanted details,” Squire marveled). Another student, Dwayne, applied his reading of Sun Tzu’s Art of War to divine deeper military strategy.
The students would also collaborate, often with a sophistication that astonished the observing teachers. One day, Tristan and Tony argued about diplomacy versus war, with Tristan trying to goad Tony into building a bigger military. Tony wasn’t buying it.
TONY: Tanks? I don’t need tanks.
TRISTAN: Tony, Tony, Tony. Why don’t we go by America’s principles? Build as many weapons as you can even though you don’t need it, just in case war breaks out.
TONY: Isn’t that overkill?
TRISTAN: What was the Cold War about? Building as many weapons as you can, just in case Russia starts something. Build enough weapons to destroy the Earth ten times over.
This political chat was coming from kids who normally were thrown out of the classroom up to 20 percent of the time for poor behavior.
What’s more, they were behaving like scientists. Failure didn’t kill their enthusiasm. It motivated them. When their countries collapsed, they didn’t huff and complain; they wondered why, gathered more information, and ran another experiment, trying a different strategy. This was hypothesis testing, and it’s the opposite of what often happens in school, where failure is punished (with a bad grade), demotivating kids. What’s more, playing if-then experiments gave students an almost tactile feel for the reality of geopolitical facts that had previously seemed meaningless and dry. “I always knew that certain locations helped certain people,” Tony said, but now “I have a better understanding of it.” Sure, he’d been told that river valleys grew more food, that a location near the ocean helped a country grow, while also exposing it to invasion. But now he’d experienced that knowledge and prodded its dimensions. The abstract became concrete.
They also weren’t gulled by the game. Squire had worried that the students would think Civ 3 was realistic. It isn’t, of course. It doesn’t model the historical impact of disease, slavery, or religion. But as it turns out, the students understood that. Indeed, they’d sit around arguing about the biases built into the game—such as its bias toward conflict. War broke out far more frequently than in actual history, they discovered, because the game was trying to be dramatic.
Educationally, Civ 3 is a rare game, insofar as its content dovetails nicely with the goal of teaching history, geography, and politics. Not many games have that property. (Though good teachers are adept at ferreting out the often-unexpected educational properties of lightweight games: One physics teacher uses Angry Birds to teach his students about gravity and projectile motion.) But Squire, and a new movement of “educational games” theorists, are finding that it’s also possible to create games specifically to educate children, for a cost comparable to making a textbook. To try to teach the physics of charged particles to eighth graders, Squire, working with a team of designers led by Henry Jenkins, created Supercharged!, a game in which students assume the role of a particle flying between electric fields. The behavior of particles in these situations is paradoxical and hard to grasp; you can show a student formulas, but they don’t give the student a visceral sense of what they mean.
But after playing the game, students’ intuitive understanding of electrophysics bounced upward. They scored 20 percent better on a test of the concepts. When asked to draw what an electric field looks like, they created drawings more nuanced and accurate; when asked to describe it, they produced remarkably more detailed explanations. Like the Civ 3 player who deeply got the role of geography in nation building, they got electrophysics because they’d been able to play around in a system that made it concrete and manipulable.
Video games can’t do everything, of course; they’re only one (new) arrow in a teacher’s quiver. But their side effects can be unexpectedly powerful, including—surprisingly—their ability to improve students’ ability to read.
Squire found that the Civ 3 students began voraciously reading to try to improve their game. This makes sense, of course: When you’re trying to solve an immediate problem, you’re deeply motivated. (This is what the reading scholar Louise Rosenblatt calls purpose-driven “efferent” reading, in contrast to the “aesthetic” pleasures of losing yourself in a novel.) School rarely motivates students to read in this urgent, engaged way because school rarely offers children any problems they find particularly urgent. Games, in contrast, are designed to provide you with problems so urgent and tantalizing you can’t stop thinking about them.
Steinkuehler also noticed that players were reading to improve their play—everything from discussion forums to wikis to walk-throughs. She wondered: Was this reading actually challenging? Indeed it was. She took a group of boys aged twelve to eighteen who played World of Warcaft and who all read, at best, at grade level, which is pretty much in keeping with the lackluster reading performance of young men on average. But then Steinkuehler asked them to pick a problem they were currently grappling with in World of Warcraft. The researchers selected a text for each boy to read that related to the game problem he was trying to solve, specifically picking texts that were lexically complicated and challenging. Yet the students were able to read those texts with ease: The nonstruggling readers tackled texts that were 3.5 grade levels higher than their typical abilities, and boys classified as struggling improved even more, reading texts fully 6.2 grade levels higher.
What happened? Why did their ability to read suddenly increase so dramatically? Because they were interested in the subject, making them willing to ponder and deduce the meaning of more complicated language and unfamiliar, even academic, terms.
As Steinkuehler writes, “Interest does matter”: The kids were trying to solve a problem they cared about. Games can provide a pathway for teachers to reveal what students are capable of. And as Squire has shown, they can hook students into reading deeply and excitedly in everything from history to economics.
• • •
Ever since Marc Prensky coined the term “digital natives,” we’ve been told that young people have an innate edge in using the technology. They’re comfortable with it; they get it, effortlessly, in a way older people don’t. But this, alas, isn’t really true.
That’s what Bing Pan, a business professor at the College of Charleston, discovered in a clever experiment. He wanted to test students’ facility on an omnipresent digital skill: How adept are they at using Google? So Pan asked a group of them to use the search engine to answer several questions. As you might expect, the students favored the top few links that Google returned.
Then Pan artificially inverted the results on the first page Google returned, putting the tenth result in the number-one slot and so on. More often than not, the students took the bait and again favored the first links—even though they’d been put there falsely. As Pan realized, the students were not actively evaluating the actual relevance of the results. They just trusted the machine.
Other studies have found similarly dismal results. A study of 102 Northwestern University undergraduates found that none ever bothered to check authors’ credentials on a Web site. Another found that more than a third of college students were unaware that search engines include paid-for links in their results. These were students who’d been using the Internet, on average, for seven years. In other words, digital natives might feel like they’ve mastered their tools, but that doesn’t mean they truly understand how they work. This ignorance is intellectually crippling, because the results on Google (and all search engines) are prone to all manner of artificial gaming and corporate juking. The upside of public thinking—that anyone can publish—is matched by its perfectly inverted downside, which is that anyone can publish, leaving the online environment devoid of the marks of hierarchical authority on which students for centuries have relied. When I was in elementary school in the 1970s, the biggest resource we had was a couple of sets of encyclopedias. We weren’t asked to judge whether they were accurate or not; the school system and librarians took care of verifying that. (Certainly, the encyclopedias had their own deficits; they became quickly out of date and were, compared to today’s online resources, woefully narrow.) In the 1950s social critics pondered “Why Johnny Can’t Read.” Now they should ponder “Why Johnny Can’t Search.”
Whose fault is it? Not the students’. If they’re unable to navigate online information, it’s because, rather amazingly, they’re almost never taught search literacy in schools. It ought to be a core part of what kids learn in school (and new common core standards in the United States are beginning to emphasize it), but for years it was barely touched upon. This is surpassingly ironic, because teaching search literacy is a golden opportunity to teach critical thinking: What am I being told? What motivations does this person have for telling me this? Does the information match other things I know? Is it even checkable or is it speculation? These are the skills that adept adults deploy, often unconsciously, when they search for information online.
The thing is, it’s quite possible to train kids in search literacy. Indeed, librarians worldwide are the heroes in this story. They’re frantically working on teaching those skills, picking up the ball that the curriculum has thus far dropped.
Consider the efforts of Frances Harris, librarian at the magnet University Laboratory High School in Urbana, Illinois. Harris takes eighth and ninth graders and puts them through a search boot camp, showing them how use advanced settings and alternative search engines that feature curated content and don’t track their users. She steers them away from raw Google searches, pushing them toward academic and news databases, too. And crucially, she trains them to check the credibility of sites they find—whether it was written by an academic, an advocacy group, or a hobbyist. She trains them not to be fooled by professional-looking design; plenty of corporate sites look flashy while peddling self-serving infojunk.
It works. Within a few weeks, the students can pass several adroit tests. For example, they begin to detect when Google is serving up pure crap (like a sham Martin Luther King site that’s actually run by white supremacists) or, more commonly, the gray-area stuff like content farms, Web firms that flood the Internet with dull, minimally informative articles just to sell ads.
“It’s not the outright lies that you have to teach them to watch out for,” Harris tells me. “It’s just this vast sea of mediocre stuff. But I see them start to get really paranoid. They keep on asking, ‘Wait, wait, is this a content farm?’ And this is what you want. Most people in their lives aren’t going to be writing term papers, but they’re going to be looking for information their whole lives.”
Crap detection, to use Howard Rheingold’s phrase, isn’t easy. Among other things, it’s easier to do if you already know about the world. For instance, Harris found that students had difficulty distinguishing a left-wing parody of the World Trade Organization’s Web site from the real WTO site. Why? Because you need to understand why someone would want to parody the WTO in the first place—knowledge the average eighth grader does not yet possess. In other words, Google makes broad-based knowledge of the world more important, not less.
Many pundits, myself included, argue that kids should spend more time with books, since they’re a time-tested vehicle for deep thought. Librarians do push kids toward books—but, brilliantly, they extend the critical thinking there, too. As Buffy Hamilton, a librarian at Creekview High School in Canton, Georgia, points outs, books also frequently contain flat-out factual errors, flaccid arguments, and loopy biases. And unlike digital documents, these errors can’t easily be fixed or called out in comment threads. “We can all think of situations where there are factual problems in printed books. Or that academic paper that was peer reviewed but then, whoops, it turns out to be wrong,” she tells me. It’s the dirty secret of the book world that virtually no publishers vet facts before they’re printed. “Let’s not bash on the Web and Wikipedia. Let’s check everything,” Hamilton says.
Training students to use a new tool for thought is never easy, and historically it’s been done rarely and unevenly. Consider that even an old-fashioned research library is bewildering if you’re not shown how to use it, and a lot of students never are. When I arrived at the University of Toronto in 1987, I—being an apple-polishing nerd—signed up for an orientation session for the university’s sprawling Robarts Research Library. Only a handful of students showed up. Assuming the other sessions were as sparsely attended, I figured that barely a sliver of the four thousand new students each year were actually trained to use the library. It showed: My professors complained that students barely sampled any secondary sources and never consulted academic journals. When CD-ROM collections of journals emerged in the 1980s, librarians complained that even professors weren’t bothering to learn to use them correctly. In a prefiguring of today’s Google illiteracy, they’d type in a single word for a search, print out the top five hundred results regardless of whether they were relevant—then happily walk away, feeling like assiduous researchers simply because they had generated a massive pile of something. (They were the “inept and satisfied end user,” as University of Alabama librarian Scott Plutchak called them.)
Our new tools are powerful, but only if we’re taught to use them. The lucky part is that students are in an environment—school—where society has a shot at instilling its intellectual values, skills, and culture. It’s up to us to make sure students are taught to use these new tools, instead of being used by them.