Diana Kimball

This year—my sixth on campus, and my last—I decided it was finally time to take CS50, Harvard’s introductory computer science course. CS50 lectures happen twice a week in Sanders Theatre, a ridiculously majestic setting. Yet for all intents and purposes (and completely by choice), I’m taking the course online. 

I’ve tried taking courses online before—I’ve even tried taking computer science courses specifically. But they’ve never quite worked for me. I’ve learned all kinds of things in my time by plundering Wikipedia and following links with abandon, but I’d all but given up on the traditional course format transposed online. CS50 is changing my mind. It’s extraordinary.

With over 700 students enrolled (and many, many more signed up to take the course through edX starting October 15), CS50 serves a staggering number of concurrent learners. In the interest of satisfying as many different learning modes and styles as possible, the course is exquisitely well-documented.

Every lecture is videotaped and posted online within 24 hours, available for streaming or download. Every shred of source code used in class is posted. Thorough notes (taken by one of the class’s many undergraduate teaching assistants) are posted to the course’s website soon afterward, too. There’s a custom-built discussion board staffed seemingly around the clock by teaching assistants. Another teaching assistant leads a walkthrough of each week’s problem set—also videotaped—with just enough information to set you on the right path, but not so much that the thrill of the solution is taken away. There are short videos covering topics like functions, binary, and ASCII. And most of this information is freely available to the public: try poking around http://cs50.net and you’ll see what I mean.

Over the past few weeks, I’ve streamed lectures on my iPad at the gym and downloaded others to watch on airplanes; I’ve watched walkthroughs in a browser window tucked in the corner of my screen as I struggle through a problem set. All of this, and I haven’t yet set foot in Sanders Theatre. But I feel like I’m getting the full “classroom experience” and more. What’s going on here?

In reading about the future prospects of online courses, I’ve seen the importance of extraordinary teachers, the challenges of accreditation, and the benefits of a community of learners come up again and again. 

But right now, I’m in the middle of a course that’s working remarkably well. If we’re ever to unbundle the classroom experience, we first need to know what it’s made of. So before my view onto this particular bundle dissolves at the end of the semester, I wanted to record some of its component parts.

Teaching as Scalable Performance

While at XOXO last weekend, I saw Tim O’Reilly post a fragment from a conversation with Greg Borenstein:

“Teaching is a performing art.” – @atduskgreg explains why massively open online courses will favor great teachers #xoxofest hallway track

In response, Jesse Farmer posted a link to an article by Michael Strong on “Scaling Quality through the ‘Missing Institution’” in which he argues:

Teaching is fundamentally a performance art – real time interactions  in chaotic and complex human situations.  There are no institutions in our society that provide for an environment in which master practitioners of this performance art systematically transfer their expertise

In the context of CS50, this line of thinking resonates for me.

David Malan is CS50’s instructor, and his verve makes all the difference in the course’s strength and scalability. He’s been teaching the course for several years now, iteratively improving it each time, and it shows.

Malan’s examples are well-considered and often accompanied by a short clip from a movie. (This week, it was Spaceballs.) His stunts—like sitting down on stage and putting on socks according to students’ yelled-out instructions to demonstrate how computers do exactly what we tell them to and nothing more—are well-rehearsed. His explanations are clear and energetic.

Thanks to Malan’s presence, the lecture videos aren’t a slog at all. They move at a good clip and are filled with humor and enthusiasm. And thanks to the audiovisual resources justified by the course’s large in-person enrollment and even larger online reach, the production values of the lecture videos are excellent.

CS50 is explicitly designed to be good on tape. The students who attend lecture serve as a kind of live studio audience, laughing and applauding and asking questions. Everyone who watches the videos learns through their vicarious experience of the lively atmosphere Malan creates in dialogue with the students who are there. Just as you don’t need to be at each taping of the Daily Show to feel connected to Jon Stewart, you don’t have to be in Sanders Theatre to be fan of David Malan.

Teachers are expected to be many things, and it seems like a tall order to expect them to be confident, talented, entertaining performers on top of everything else. But over the next few years, I do think we’ll see a handful of star performer-teachers emerge as the leading lights of online learning—as we’ve already seen, to some extent, with Sal Khan of Khan Academy.

Is this a good thing? On balance, I think: yes—for students, at least. Finding the teacher whose delivery connects can make the difference between a student embracing a subject and discarding it permanently. For a subject as fundamental and tricky as computer science, elevating and magnifying the reach of a teacher like Malan gives more students more of a chance of getting hooked.

I worry about the cost to teachers of effective celebrity (think of all those emails from students!), and I worry about the cost to students of not having in-person access to strong role models. I worry about the side-effects of this spinning up an apparatus for identifying, celebrating, and videotaping “great teachers” without attendant acknowledgment of the extent to which teaching is a performance art.

But given the extent to which teachers at the university level are still judged by the strength of their research more so than the strength of their classroom delivery, I don’t think it’s a bad thing to create some external recognition (in the form of YouTube likes?) for teachers who do manage to scale up their message through online courses. Many teachers I know dream of the chance to get more students to fall in love with their subject. Online courses can make that process the main event.

Commitment and Sacrifice

It’s easy and alluring to declare that you’re “starting” an online course, just as it’s tempting to make grand New Year’s resolutions. Declarations and resolutions are typically precipitated by real interest and desire; communicating that desire is a way to sound out your support network, and a first step toward making the interest feel real.

But publicly declaring an intention doesn’t make it so. On the contrary: in my experience, if I announce a time-consuming intention (without knowing how, exactly, I’m going to follow through), my initial, sincere interest gets replaced with guilt more often than not. 

I experienced some of this with Codecademy’s Code Year. Inspired by a flurry of friends resolving to make 2012 the year they learned to code—and encouraged by what at the time felt like slow and lurching but real progress on that front over the past few years—I took the plunge and “signed up.” Over the next nine months, I received weekly emails from Codecademy imploring me to “Start a new coding streak!” and to “Keep learning!”

Ironically, spring of 2012 was when coding really took off for me: I spent the semester building a full-on web app for a field project at Harvard Business School.

And so it was that on February 28, 2012 I got an email from Codecademy with the subject line: “Still want to learn to code? We’ll help  you catch up!”—the very same day I pushed four separate commits to a private GitHub repository. I remember getting that email and thinking: huh? And also, weirdly, feeling simultaneously triumphant and strangely guilty.

I felt like Codecademy was (very, very gently) chastising me, yet Codecademy had no sense of the scope of my life and my learning goals. Codecademy didn’t know about my private GitHub repo. Yet I’d still invited it into my life to bug me on a weekly basis, and it was dutifully doing so. That was the whole point! But even when I ignored the emails, even when I knew how out of step they were with reality, I still felt pretty bad for not doing the thing I’d “committed” to doing—even though that “commitment” was just typing my email address into a big white box.

This is all to say: commitment is good, but lightweight commitment is dangerous. Make declaration of learning intention too easy, and you’ll set people up for guilt—guilt about exactly the thing they were once interested in and excited about.

In thinking about what’s powerful for me about CS50, I realized that part of it is sacrifice. I have two semesters left at Harvard Business School. Each semester, I can take five courses. CS50 is one of them. To take it, I opted not to take any of the 150+ other courses available at HBS alone. In fact, I went further: I organized my whole schedule around creating big blocks of time for working on problem sets. CS50 is my highest priority this semester, and it got that way because I went through the exercise of stack-ranking all the components of this semester’s reality at a single point in time. In order to take CS50, I consciously sacrificed other opportunities to learn and rest. 

In the working world, this kind of conscious sacrifice in favor of outside learning is difficult to pull off. Reality is continuous; projects and obligations pile onto each other and when they end, they rarely end all at once. And so online courses just get added to the heap. Yet unlike salsa dancing classes at the local gym, online courses can be shifted around and taken anywhere. Because they require little logistical planning and lots of focused effort, they start to slip down the priority list. And once you stop paying attention to an online course’s rhythm (as I did with Codecademy’s Code Year), the weight of “catching up” starts to feel heavier than the original desire to learn.

I’m not sure what the solution to this conundrum is. My mind has been spinning on it for the past half-hour, hoping that I can come up with some morsel of structural insight. I still don’t have an integrated answer; I guess it would be pretty amazing if I did. I’m hoping that Kio Stark’s upcoming handbook on lifelong learning, Don’t Go Back to School, will help me think through these questions further. But in the meantime, I will say this:

School is a place where it’s expected that you’re bad at something in September and then, after a lot of discomfort and effort and epiphanies, way better at it come December. If we can set those expectations for ourselves and each other outside the context of school, we’ll have a much better shot at learning for the rest of our lives.

Do you know of any other posts or research I should read along these lines, or have your own experiences to share? Please don’t hesitate to email me at diana.kimball@gmail.com.