Larry Summers Calls Higher Education Stubborn and Anachronistic, Offers Suggestions

I squirmed a lot reading Larry Summers’ recent piece in the New York Times on where he thinks and hopes higher education will go in the future. Here’s a point by point analysis:

He begins by undermining his own credibility:

A paradox of American higher education is this: The expectations of leading universities do much to define what secondary schools teach, and much to establish a template for what it means to be an educated man or woman.

REALLY? Have you paid attention to any of the K-12 school reform of the administration you have been a part of? The encouraged emphasis on basic reading and math skills at the cost of social studies, science, physical education and extracurricular activities runs exactly counter to the template of colleges and universities in which diverse offerings, and choices of majors proliferate. But I’ll forgive this vague handwaving and move on. Summers’ point is that colleges are seen as cutting edge, but in fact offer stale education which is stuck in the past because tenured faculty (who are often in charge of the curriculum) are stubborn. Dismissed college president says faculty are stubborn and old-fashioned, the Times is ON IT!

The paragraph in which he lays out the reasons that colleges are old fashioned seemed to me to be amazingly disingenuous. Colleges are staid and stuck in the past because… departments and courses have the same names as they did 50 years ago? Students take four classes and exams in blue books? Students pick a major? So the biology major is the same as it was 50 years ago because it is still called biology? Really?

Summers wants higher education to better reflect how the mind and world works. But as someone with expertise in mental processes who works in higher education, Summers’ understanding of both the current state of higher education and the science of cognitive psychology are simplistic and off base. As a result, we shouldn’t take his six “guesses and hopes” seriously except as a warning of the perils of breezy theorizing by famous intellectuals.

1) College curriculum will become “more about how to process information and less about imparting it”.

This is the standard: “You don’t need to know any facts because you can Google them, you just need critical thinking skills of finding and evaluating facts.” It is so tempting. Information is everywhere, it is at our fingertips, and the ubiquity of this information will spare us from keeping any of it in our heads, just like we don’t have to remember phone numbers, or directions anymore. Unfortunately, this is not how the brain works. As Daniel Willingham reminds us in his book “Why Don’t Students Like School?” “Factual knowledge precedes skill.” Whenever cognitive psychologists look closely at critical thinking, we find that it is tightly integrated with background knowledge. Any definition of critical thinking involves the creative and rigorous application of a network of facts. It is impossible to think critically about neuroscience unless you know dopamine from acetylcholine, and dorsolateral prefrontal cortex from the occipitotemporal junction. Not remembering phone numbers is not the same as facts which we will sometime need to recruit to do our thinking. Summers shows he doesn’t have certain facts about language education and cognitive psychology, which he could easily look up, but which undermine the validity of his “critical thinking.”

2) Collaboration is the rule in the world, but individual evaluation persists in college. I agree with this point, but there are a whole host of reasons that one would rather have an individually tailored (and evaluated) learning experience before a collaborative job experience. This doesn’t necessarily require grades, but being able to track one’s own progress, independent of group work, has value, and will continue to be a part of education. But I agree that more group work should be expected and acceptable. But then he proceeds with this bit of absurdity:

One leading investment bank has a hiring process in which a candidate must interview with upward of 60 senior members of the firm before receiving an offer. What is the most important attribute they’re looking for? Not GMAT scores or college transcripts, but the ability to work with others. As greater value is placed on collaboration, surely it should be practiced more in our nation’s classrooms.

This is absurd, because from what I read (from sociologist Lauren Rivera) credentials still matter more than anything. Maybe what Summers is referring to is that investment banks want to ensure that Harvard recruits can get along with their future colleagues from Yale and Princeton.

3) Technology will change how information is conveyed. Just like no one has to memorize Canterbury Tales anymore, and textbooks can be written by the experts, the next phase of technology will change how education happens. Why should anyone bother to lecture about Intro to Philosophy, when we could all just watch Michael Sandel teach Justice? Who needs me to create a good lecture for Intro to Psychology if I could just have my students watch Paul Bloom at Yale. I have a lot of thoughts about this, but the short answer is this: On the one hand, I agree that maybe we shouldn’t have 200 state universities all having 500 people lecturing at the same time. Why not pick the most amazing one, and have all 10,000 intro psych students watch, and then come in and talk about it. On the other hand, this has at least two negative consequences; first, the lecture isn’t just “the best” but it is also meant to reach the widest possible audience. In doing so, it can lose some pedagogical punch. Yes, 18-year-olds are 18-year-olds, and I don’t think Yale students are that different from my students at Randolph-Macon. But student populations do vary, and the lectures should reflect that. Second, a good lecture potentially can involve some questions or interaction, a video can’t do that. When I lecture, I think of my students and their interests, I ask them questions and they ask me questions, but they are often unmistakably lectures. An intro psych class for engineers (yes, these exist) would be different than an intro psych class at an art focused school. There is no platonic lecture, just as there is no platonic textbook. This is a good thing.

4) Summers says we know more about learning and name-checks Daniel Kahneman which leads us to: “Not everyone learns most effectively in the same way. And yet in the face of all evidence, we rely almost entirely on passive learning.” Ok, whoa there. This sounds suspiciously like learning styles. Which I am pretty sure, given that it is wrong, is not in Kahneman’s book. A big point of Kahneman’s research is not that people have different ways of learning, but that certain things are harder to learn than others.

Second, The distinction between active learning and passive learning is not that easy. I can give a ten minute lecture, answer two student questions, show a demonstration from YouTube or even one of my amazing magic tricks, then ask students to apply what they have learned in a quiz the next morning. Did they learn anything? I don’t know, it depends on the quality of the lecture, how well the demonstration matched the lecture, and how much the quiz made them study. They could be sitting back, listening and thinking this whole time, but Summers calls this passive.

Students listen to lectures or they read and then are evaluated on the basis of their ability to demonstrate content mastery. They aren’t asked to actively use the knowledge they are acquiring.

But there is a fine line between “demonstrating content mastery” and “actively using the knowledge.” For example, I can ask “What do we call the three different kinds of cone receptors and what distinguishes them from each other?” Or I can ask “If I have special sunglasses that block out electromagnetic energy which has a wavelength below 1000 nanometers, what will my visual experience be like when I put these on?” Despite not having any extra technology, and not putting them into groups, the second question can be an incredibly active process, because it asks students to apply a concept that they have learned (our photoreceptors are only sensitive to electromagnetic energy with wavelengths between 400 and 700 nanometers) to a (relatively) new situation. If no wavelengths below 1000 pass through the sunglasses, none would enter your eyes, your photoreceptors would not activate, and you would see nothing, it would be totally dark. Would students learn more if I made that an interactive clicker question rather than on a hard copy of a quiz? (I have tried both, not noticed too much of a difference). I am an early adopter who loves gadgets, but I am deeply skeptical of their ability to improve education without the right content. If the content is not accurate, appropriate for the student’s level, and organized, no technology can help. Likewise, if students don’t have any background knowledge, no amount of active learning group discussion is going to help them teach each other anything. In psychology, where myths and misconceptions abound, students can simply reinforce each others errors. Don’t get me wrong, active learning classrooms can be great, but they can also mistake student engagement for student learning. Engagement and motivation are necessary, but not sufficient, for learning.

5) The world is more open, according to Summers, so students should have more international experiences, but shouldn’t necessarily bother with more language, because English is taking over. I mostly agree that colleges should have more international options, but I haven’t heard that colleges are short of travel courses or options for international study. But what about the case for the cognitive benefits of bilingualism? According to the research of Ellen Bialystock (summarized in this New York Times interview) learning two languages has far wider effects than simply adding another language. For example, bilinguals are better at multitasking.

6) Summers wants every student to take more probability and statistics. I agree, with no qualifications. Ok, maybe one. I think more statistical literacy should be required for all students. But we shouldn’t mistake this for being able to understand any data and any statistics. There is some basic statistics knowledge necessary to understand medical statistics, but also some knowledge of medical phenomena. Statistics is rarely divorced from its subject. Moneyball was successful not just because Bill James was a statistics god, but he had a better understanding of the connection of his sabermetrics to team wins than the typical batting average/RBI/ERA stats. Summers, expert in math pedagogy, thinks we should drop trigonometry (i.e. that old-fashioned math that was used when kids needed to survey land) and teach a lot more statistics and probability. That makes sense until you consider how important calculus is for scientists and engineers (and even economists) and how trigonometry lays the foundation for calculus.

Diminishing trigonometry as merely “math for surveying” shows me that Summers doesn’t really understand math education. He also doesn’t acknowledge that part of higher education’s strength is its diversity. While each of these suggestions does have the undercurrent of “Larry Summers knows more about international relations and language than you do because he thought about it a little this weekend,” each also presents a view of a world of higher education that is monolithic and growing more so. By suggesting these broader trends, he ignores those schools that already follow his suggestions, but also ignores that some schools do not, but for good reasons. Some schools (such as most business schools) are already incredibly collaborative, with most work done in teams (which is why MBAs have such a strong reputation for cooperation and collaboration). Some programs, such as Kalamazoo College, are incredibly focused on the world and travel. But some are not. Some students take language skills into communities closer to home. Many departments (such as mine) require a semester of training in probability and statistics as well as another entire semester in which students design their own project, carry it out, and analyze the data. But some departments do not, and some schools do not. My history of science degree did not require probability and statistics, so I didn’t take it. I wanted to take Frank Lloyd Wright, and Computer Science, and Poetry instead. Which were all awesome.
I am no expert on math curricula, or the relative costs and benefits of languages vs. travel courses, but I know enough to see someone being breezy and lazy in their sweeping generalizations. I have been in enough meetings and discussions about college curricula, with intelligent and diligent people considering how we weigh the “template of an educated human” with “what this set of 18-year-olds in front of me wants to do with their time” that I know that these aren’t easy decisions, and our colleges aren’t nearly as filled with arbitrary anachronisms as everyone seems to think. Funny thing is, Larry Summers has been in those meetings too. Maybe he just wasn’t using his critical thinking skills because he was a passive learner and they were just people lecturing.