Marvin Minsky, MIT professor and artificial intelligence pioneer, passed away last Sunday (January 24, 2016). Minsky is widely celebrated for his diverse scientific contributions to artificial intelligence, computer science, mathematics and even microscopy. Less well known are Minsky’s radical views on education.
Some mathematicians think that learning mathematics is about “reaching the summit and then swiftly burying the path.” Minsky didn’t. He was interested in why mathematics is hard for children to learn and what can be done about it. He thought that the difficulty was partly “caused by starting with the practice and drill of a bunch of skills called Arithmetic”. As a result, “instead of promoting inventiveness, we focus on preventing mistakes.” Some children, frustrated by this negativity, dismiss mathematics as repetitive, boring and punitive.
The obsession with preventing mistakes marked a greater flaw that Minsky diagnosed in the education system. We’re fixated on a view of expertise as knowing how to do things right. Isn’t it more accurate, he suggested, to think of acquiring expertise as “having learned to avoid the most common bugs”?
One of Minsky’s insights, which applies to fields beyond mathematics, is that learners need a “cognitive map” of the subject they’re exploring. Cognitive maps are hard to make; they’re conceptual by definition and require synthesis of multiple seemingly disparate strands in a field. But without them, learners would be left with “several collections of scripts and facts, without good ways to know which of them to use, and when” — and with no strategies for finding alternatives when getting stuck.
According to Minsky, we don’t understand something unless we understand it in at least two ways. Cognitive maps help us see multiple angles on a topic. And if so, why stop at the cognitive map of the subject? A cognitive map of the learner would help too. Minsky proposed that we “provide our children with ideas they could use to invent their own theories about themselves!” Minsky saw the value in the intellectual challenge of theorizing about ourselves. Beyond intrinsic value, it isn’t crazy to think that “meta-learning” — learning how we each learn — would help us organize materials for ourselves, develop better cognitive maps, and so on.
We’re nowhere near implementing Minsky’s ideas in the school system. Much of education is still “practice and drill.” Even in universities, cognitive maps of fields aren’t generally taught. These cognitive maps may be crucial to fostering truly interdisciplinary work, of the sort many institutions are apparently committed to promoting.
Minsky’s views may be dismissed as out of touch, impractical, or naive. They certainly don’t fit into an education system like ours that is based on calculated risk, where an early misstep in a student’s education can have far reaching penalties. It’s interesting to consider what an education system that could support Minsky’s ideas would look like.
Yarden Katz is a fellow in the Dept. of Systems Biology at Harvard Medical School and a graduate of MIT.