Summers on evolving higher ed

In the Times, Larry Summers opines about the structure of college learning and future trends, with plenty of provocative perspectives. It's hard to disagree with the notion that teaching collaborative analysis rather than individual demonstration of knowledge would probably be a better fit of the needs and the technology. But under conditions of large student-to-instructor ratios, the free-rider problem would be made a lot worse by such endeavors. I wouldn't suggest that the current passive-learning model really gets this particularly right either, but it's hard to see how assigning credit based solely on group projects could ever get the incentives right for students without a lot of monitoring.