Being Digital Learners

What happens when MIT and Boston schoolkids get together

By Lars Kongshem

Lars Kongshem is assistant editor and webmaster of Electronic School.

Boston's underserved Jamaica Plain neighborhood might seem an improbable setting for something billed as the "School of the Future." Quietly deteriorating Victorian houses -- not gleaming office towers -- dot the landscape of this economically depressed community. Yet judging by how soundly today's conventional wisdom is being trashed at the James Hennigan Elementary School, the future is right here.

At the Hennigan School, unspoken assumptions about the use of computers in education have been turned upside down. Here, computers don't regurgitate answers; they're used to formulate questions. Inside, you won't find kids on the receiving end of automated drill-and-practice programs or glitzy "edutainment" software. Instead, students use personal computers much like carpenters use hammers -- as tools to construct authentic, personal, and meaningful projects.

What does that look like? Taking a seat in front of one of the 72 networked IBM PS/2 computers arranged cluster-fashion in a common area in the school, fifth-grader Nebiyu Elias proudly fires up one of his inventions: a grand tour of the solar system. He's written a computer program that allows users to navigate an on-screen spaceship from one planet to the next with the keyboard's cursor keys. The program rewards successful arrival at each port of call by displaying interesting facts about the planet and then launching the ship into space with an eye-catching animation sequence. Nebiyu wrote this interactive adventure in Logo -- a programming language for kids -- over a period of three months.

Nebiyu is smart, but his sophisticated computer program is typical of the work kids at this school are doing. In fact, says Hennigan teacher Joanne Ronkin, when the kids sit down in front of a computer and start programming, "you can't tell who's the smartest child and who [has] special needs. It's a great equalizer."

Four computer periods per week is the norm at Hennigan. On this particular day, some of the fifth-graders are working on a long-term assignment to create educational computer games for third-graders. The fifth-graders choose topics such as multiplication or spelling, and as their programs take shape, the third-graders try out the games and offer their feedback. These homemade games tend to be much more popular among the kids than commercial ones, Ronkin says.

Students help one another out on their projects, too, using a custom on-line conferencing system. Here the kids can exchange programming tips and ask each other questions on thorny Logo topics. The system has been designed so that questions can't be directed toward any specific person or group but must be asked of everybody. As a result, the forum has broken down on-line some social barriers -- between girls and boys and between ethnic cliques -- that are evident in the children's face-to-face interactions.

"In real life, girls don't ask boys questions about how to do something in Logo, but on-line it's different," says Michele Evard, the Massachusetts Institute of Technology researcher who designed the conferencing system. In fact, girls feel comfortable speaking up and even answering boys' questions on-line, she says.

One thing's for sure: At this school, students are taking ownership and control of computer technology and using it as an extension of their creativity -- and themselves.

The high church of high tech

The motive force behind this experiment is the Massachusetts Institute of Technology's Media Laboratory, a fashionable temple of high-technology gurus whose broad aim is to investigate how computational media can be used to help people communicate, learn, and live in new ways. Some of the way-new technologies being invented at the Media Lab include wearable computer networks, computers that know when you're angry, and software agents that find strangers who share your interests and introduce them to you.

The Hennigan School project, funded by corporate donations and supported by researchers from the Media Lab's Epistemology and Learning Group, is the brainchild of celebrated Media Lab professor Seymour Papert, the inventor of Logo. Formally known as Project Headlight, it's been in place since the birth of the Media Lab over a decade ago.

On the MIT campus in nearby Cambridge, the Media Lab occupies a striking white-tile building designed by architect I.M. Pei, which bears the unmistakable mark of 1980s-era futurism. Walk inside its austere atrium, and a barely audible whir of indeterminate origin hints at the proximity of a vast amount of computational machinery. Up on the third floor, graduate students in the Epistemology and Learning Group work in a terminal garden of powerful computer workstations, with a loose assortment of brightly colored LEGO building bricks scattered around. A sense of serious play is palpable here.

Does what goes on at the Media Lab have any relevance to the challenges you face in implementing technology in the classroom? You bet. "Our mission is to develop computational tools to help kids learn in new ways," says Mitchel Resnick, the contagiously enthusiastic head of the Media Lab's Learning and Common Sense section, which includes the Epistemology and Learning Group. "Most importantly, we're using technology to rethink the nature of learning and the practice of education," he says. With researchers drawing on backgrounds in technology, psychology, and education, the Media Lab is well-positioned to do that.

Given the engineering emphasis of the Media Lab, it's not surprising that the guiding principle for its work in education is constructionism -- the theory that kids learn best when they're making things. (Think of it as an engineer's version of the learning theory known as constructivism.) "We're enabling and engaging children to learn through design and invention opportunities," Resnick says. The idea is for students to take charge of their own knowledge-building as they construct personally meaningful artifacts -- such as computer programs, animations, LEGO-based robots, simulated microworlds, and virtual homes in on-line communities.

The Media Lab's programmable brick project is a good example of constructionism theory in action. Embedding computational ability in everyday objects -- such as doorknobs or clothing -- is big at the Media Lab, and the programmable-brick concept implements this idea of "ubiquitous computing" for kids by putting a small computer inside a LEGO brick. By wiring the brick to various sensor devices and motors, kids can instruct the brick -- through computer programs they write in the Logo language -- to respond to and act on its environment. The result is a robot that can be programmed to do any number of amusing and instructive things.

Resnick picks up a programmable brick-based robot that has been sitting on his desk. This prototype has wheels driven by an electric motor; its input device is a light sensor. "One kid wrote a program for this called Afraid of the Dark," Resnick says, smiling. "If you turned off the lights, it would run away."

Future versions of the programmable brick will be even smaller and more portable, so students can easily use them to collect data from their everyday activities and their environment. For example, one boy wired a brick to his bicycle and used it to collect data on his bicycle riding. By later dumping the data into a desktop computer, he was able to analyze how far he typically rode in a day, what times of day he rode the most, and so on. Resnick envisions students dropping future programmable bricks into a pond to collect and analyze data on its ecosystem.

Tools to think with

Future advances aside, are schools taking full advantage of the computers they already have? Probably not. In many classrooms, computers are not used very productively, Resnick says, because they typically are asked simply to improve upon other media. For example, a CD-ROM encyclopedia might be faster or more compelling to use than a traditional printed encyclopedia, yet it's just an incremental improvement, not a fundamental change. (In some cases, computers can even be counterproductive to learning, Resnick points out, such as when drill-and-practice programs give students the impression that "the computer has the answers.")

Schools often fail to take advantage of the fact that computers' significant number-crunching ability can be used to allow kids to explore, play with, and gain an understanding of complex phenomena and concepts that otherwise would be far beyond their grasp, Media Lab researchers say. "What we like to do is see how the new medium of the computer can do things that weren't possible before," Resnick says. "For example, computational media allow young children now to gain an understanding of group behavior and complex systems without having to know differential equations."

To that end, Resnick has developed StarLogo, a programmable modeling environment for the Macintosh. High school students in the Boston area have used StarLogo to explore and learn about the behavior of decentralized systems -- such as bird flocks, ant colonies, termite communities, and traffic jams -- in which orderly patterns emerge despite the absence of leadership and centralized control. This is a difficult concept for even adults to grasp, but by building their own computer models and tweaking the simple rules of behavior that each participant in the system follows, the students can see immediately how the changes they make affect the behavior of the system as a whole.

To demonstrate, Resnick runs a computer simulation in which termites build piles out of randomly scattered wood chips, even though their explicit instructions -- "pick up a chip if you come across one and put it down when you come across another chip" -- say nothing about creating piles. As the simulation runs, small piles quickly begin to emerge, some of which disappear while others grow, and soon only two or three large piles of wood chips are left on the screen.

"Traditionally, you had to reach a high level of understanding in math before you could play with concepts, but with computers, you can play around with something before you understand it," says David Williamson Shaffer, a former math teacher who is now a graduate student researcher at the Media Lab. Shaffer's current research project is Escher's World, which explores the integration of mathematics and art education using computational media. "The question we're asking is, 'What if you could learn mathematics in an art studio?'" Shaffer says.

One of the tools Shaffer introduces the high school students in his project to is Geometer's Sketchpad, a software program that allows the students to investigate and play with the concept of mirror symmetry. By combining deep mathematical ideas with arresting graphic design, the students are learning a way of thinking mathematically with an expressive component, Shaffer says. The computer is a more forgiving environment in which to learn either discipline, Shaffer points out, because the "undo" function frees students to make -- and learn from -- their mistakes. By making this multidisciplinary approach possible, the computer also is encouraging educators to re-examine outcomes, Shaffer says: "When you put art and mathematics together, you're forced to be explicit about what your goals are."

What's next for the Media Lab's work with schools? The answer: more Internet. Resnick foresees a project involving "distributed constructionism," in which classrooms scattered across the globe participate in an active computer model, sharing "dynamic constructions or computational objects." For example, if each participating classroom "made" a fish by describing its behavior, the result could be a living ecosystem, glued together by the network.

The obvious route to making this happen is to develop a kid's version of Java, the hot new programming language that is poised to bring unprecedented degrees of interactivity to the World Wide Web. The Media Lab's working name for this Logo-like scripting language for joining dynamic constructions on the Web is, appropriately, Cocoa.

No lab rats

Considering the Media Lab's experience in making cutting-edge technology work for students in schools, it might be reasonable to assume that if anyone has found proof of technology's impact on test scores, this would be the place. Not so. "The research we do to evaluate projects at the Media Lab is loosely ethnographic and qualitative," Shaffer says. In fact, trying to find a direct relationship between test scores and technology may be a fool's quest.

"I don't think we can prove that computers help kids get better test scores," says Joanne Ronkin at the Hennigan School. "There's no way I could show you a piece of paper that proves there's a causal relationship between computers and achievement scores. There are just too many variables, even between teachers," she adds. Or as Media Lab researcher Amy Bruckman puts it: "People aren't lab rats."

Yet it's obvious to anyone who's looking that exposure to computers -- and to the mindset that anyone can take control of computer technology and use it to explore, experiment, and express -- is an incredible addition to these kids' lives.

"It's meaningless to ask the question, 'Is technology a good thing?'" Resnick points out. "There are two concurrent revolutions happening today -- a digital revolution and a learning revolution. The digital revolution necessitates a learning revolution, because it makes things change so fast. At the same time, it makes possible a learning revolution, one that emphasizes learning as a lifelong endeavor in which the individual takes charge."

Photography by the author.