Teaching Heuristics

In this section, Professor Patrick Winston shares some of the heuristics that shape his teaching. He notes that he has learned many of these strategies from colleagues. Central ideas include beginning a lecture with a promise, inspiring students by conveying passion about what he's teaching, moving beyond teaching skills toward teaching powerful ideas, encouraging note-taking during lectures, and always ending a lecture with something fun.

Begin with a Promise

I’ve learned a lot about teaching by observing and talking with colleagues. My colleague Randall Davis told me that you always want to start a lecture with a promise of how students will be empowered by what they will learn in the lecture. In other words, you want to tell them what they’ll be able to do at the end of the lecture that they couldn't do in the beginning. I try to start every lecture with a promise, every time.

In one lecture I show a program that will, depending on how you arrange it, complete in anywhere from 2 to 20 billion years. And then I show students how we can adjust it slightly and it will complete in the blink of an eye. I say, “By the end of this lecture, you'll know how to write the faster arrangement.” It’s very motivating!

Inspire Students by Conveying Passion

A few years ago, my department chairman asked me to give a talk to the incoming faculty in the School of Engineering about how to be inspiring. I'd never thought about that before, so I did an informal poll. What I discovered as the common theme—from the entering freshmen to the senior faculty—is that people who inspire others express passion for what they are teaching or the ideas they are conveying.

In my own teaching, when I give a demonstration, I frequently tell students I think it’s “really cool,” as a way to be explicit about the fact that I’m passionate about it. That was something I’d always done intuitively, but didn’t realize the purpose of until I did the poll about what people found inspiring.

I was talking with a colleague of mine about the subject of inspiration, and he said, “You know, some of the stuff I teach is boring beyond description. I have to convince myself before I talk about it that it’s interesting and that I’m passionate about it, otherwise it’s a horrible lecture.” What he said is true, because there are things people need to know, but you can’t say they’re very exciting. You have to pretend they’re exciting. Somehow. Otherwise, your passion won’t come across and your teaching won’t be inspiring.

Move Beyond Skills with Stories and Powerful Ideas

I tell a lot of stories during my lectures. In my own research, I concluded some time ago that the distinguishing characteristic of human intelligence is our story competence. We tell stories, we listen to stories, and we make up new stories by blending old ones together. That’s really what education is all about, if you think about it. We start with fairy tales and folk tales, and then we're exposed to history and literature. Our culture has stories. Religion has stories. Then we go to school, professional school especially, and what do we have but case studies?

That’s why I think stories are an important element of education, and if you strip them out, you don't have much left that can possibly be inspiring. You have only recipes when you strip out the stories.

I value stories to a greater degree now than I did when I first began teaching. When I started out, I was more concerned about whether students learned the skills I was trying to teach. I'm still concerned about whether or not somebody learns the skills, but I think if that's all they do, then my teaching wouldn't be as fun for me or as interesting to my students. I think sharing the stories, the opinions, the asides, and understanding how a person solved a particular problem, what they were thinking of when they did that, what they were motivated by, etc. is just as, and probably more, important than teaching the actual skills.

In a recent lecture, for example, I was talking about the history of support vector machines, which were developed in the PhD thesis of Vladimir Vapnik in Russia in the 1960s, but remained unknown in the West until sometime in the early 1990s when Vapnik moved to AT&T Bell Labs. Now they are the standard element in the tool kit of anyone who claims to do machine learning.

The interesting story is that Vapnik further developed support vector machines when he got to the United States, in part, because he was irritated that several of his original papers were turned down. The power of showing that you're right and conquering doubters is enormously powerful motivation.

My friend Edward Roberts in the Sloan School studied entrepreneurship for much of his career. He will tell you that many, maybe most, entrepreneurs are not interested in the money. They're interested in showing that the people who doubted them or their ideas are wrong and that they will, in fact, be successful.

I include ideas like this in my lectures. I call them powerful ideas. If all you’re teaching is skills, the educational experience you offer students is okay, but if you can accompany the skills with some big-picture, powerful ideas, the educational experience becomes more impactful, more important.

Encourage Note-Taking

I like to think that I'm one of the first MIT faculty members to forbid laptops and cell phones during lectures. The reason I do this is because there's a lot of evidence that we only have one language processor in our heads, and it's easily jammed. If you jam it by reading your email, texting, or doing something else, you're not actually going to be able to pay attention to what’s going on in the lecture.

I encourage my students to take notes because it forces engagement. You can't take a note without deploying your language apparatus and your drawing apparatus. And that's the reason for taking notes. It has nothing really to do with looking at the notes again; it has everything to do with forcing concentration.

When I was an undergraduate at MIT, I was on the timid side. I came from a small Midwestern town. I went to a good high school and had inspiring teachers, but nevertheless, it was a small town. When I got to MIT, there were a lot of people who looked like they had been much better prepared than me. What I ended up doing, after every lecture my freshman year, was to go to the library and attempt to recreate what was said during the lecture. When I got stuck, I'd look at my notes. I would work that way until I could go through the whole lecture without looking at my notes. I learned to explain the content of the lecture in such a way that I could clearly understand the concepts.

I think the accident of being a little timid actually threw me into a way of thinking that was extremely useful. It forced me to be able to explain stuff to myself. It was probably one of the factors that helped me become good at explaining research to potential sponsors. I was director of the MIT Computer Science and Artificial Intelligence Laboratory for about 25 years and my main function was to explain what people in the laboratory were doing in clear and attractive ways. The early experience I had in reproducing the lectures from my notes was good training for that role because it forced me to think about how things could be explained clearly.

End with Something Funny

One of my colleagues told me that he always ends his lectures with something fun so that people feel like they've enjoyed the class the whole time. It could be a joke, or an historical anecdote, or an intriguing demo. I do that now, too. I always try to end with something fun.