You can't actually learn by osmosis while sleeping

The CS department recently decided to overlook the mess I made of grading in my new post-parenthood life (for which I'm grateful. :) and gave me a teaching award.  What they didn't mention was that it came with two caveats:

  1. I'd have to pay income tax on the cute glass apple that came with it.
  2. I had to write something smart-sounding about teaching (!).

The problem with #2, of course, is that like most tenure-track faculty at a research institution, I don't actually know anything about teaching.  As any of you who're Ph.D. students choosing to go into academia will discover, we do an absolutely terrible job of preparing you to teach college classes as part of your Ph.D. education.  Actually, that's too strong:  We don't do the job at all, not even terribly!  (In fairness, Hari actually gave me some great advice when I TA'd for him.  But that was luck, not the structure of the Ph.D. program.)  But when I protested, the powers that be that run the department told me to suck it up and write something anyway.  Which got me to thinking a bit.  Here's what I wrote:


You can't actually learn by osmosis while sleeping

When I entered the academic job market, one of my mentors, John Guttag, read my naive first-draft teaching statement, and commented, "Dave - you're thinking about this all wrong. You can't teach people; but you can help them learn." More than anything, the comments arrayed for the nomination for the Herb Simon award have convinced me even more that John was right.
Instead of trying to sound wiser than I am, I'll simply quote from evolutionary developmental psychologist Peter Gray's book Free to Learn:
"Children are biologically predisposed to take charge of their own education. When they are provided with the freedom and means to pursue their own interests, in safe settings, they bloom and develop along diverse and unpredictable paths, and they acquire the skills and confidence required to meet life's challenges. In such an environment, children ask for any help they may need from adults."
This doesn't just work for kids – the same holds for college students, graduate students, and professors to boot.
So, teaching, in three parts:
First, show how and why the subject matter is important, relevant, and interesting. Fortunately, computer science is an easy sell – it's the best field in the world, with an amazing mix of intellectual depth and complexity and practical, world-changing applications. If you don't believe me, think about everything that happens, from systems and networking, to machine learning, to programming languages and algorithms, just so you can search for stupid cat videos on YouTube.
Second, to be a resource for the students to learn from. My only comment here is that this is as much or more a skill of listening as it is of talking. You don't help people learn by blabbing the same thing at someone ten times: a textbook can do that. You teach first by listening carefully to understand where and why they're misunderstanding the material. Teaching is a conversation, not a monologue.
And third, keep the students awake. Despite their many other wonderful characteristics, Carnegie Mellon undergraduates are notoriously over-subscribed with classes and credit hours and projects and ... well, everything. You can't learn when you're sleeping in class. I hypothesize that it's better to be laughed at by wakeful, engaged students for being, er, excessively bouncy or a little weird, than it is to face a room of students dozing off their food-truck lunches.
The last parts of great teaching are easy. Be lucky enough to have incredibly smart, motivated, and creative students in your class; be in an environment that deeply values education and is committed to the success of its students; be paired with excellent co-teachers; and have astounding TAs who rescue the course when the absent-minded professor forgets to grade projects on time. Which reminds me that some of the most important aspects of learning have little to do with the professor.

Popular posts from this blog

22 Months of Voice-Based Assistants

Reflecting on CS Graduate Admissions

Minting Money with Monero ... and CPU vector intrinsics