"You might learn something"

by Adam Michlin
Oct. 17, 2002

David Huffman was not your average computer science professor. Walking into his office in the 1990s at the height of the Internet craze one saw none of the usual trappings. No fancy ten thousand dollar computer workstation, no shelf full of computer books. He was the only faculty member I ever knew as an undergraduate computer science major who didn't have (or at least, didn't advertise) an email address. What you would have found in his office was a wall full of math books and an old Macintosh Plus. To most people in the 90s that Macintosh Plus would have been more useful as a paperweight than a computer.

I think Professor Huffman really saw himself more as a mathematician and a teacher than a computer scientist. He was one of the rare people who excelled in his field and also seemed to truly love teaching rather than seeing teaching as a necessary evil in the pursuit of research. He was probably the best teacher I had at U.C. Santa Cruz (UCSC), but also the hardest.

In case you don't already know, David Huffman was one of most famous computer scientists of his time. He served on the faculty of MIT and then founded the computer science department at UCSC. He served as an active professor until 1994 and continued to teach as an emeritus faculty member until his death in 1999. One of the advantages of being both famous and retired was that he could teach the classes he wanted, the way he wanted.

I entered the picture as a third year computer engineering major at UCSC. As both a musician and a computer-type, I decided to go into Digital Signal Processing. If you who don't speak "tech", digital signal processing is the heart of using computers to shape all the sounds and sights we've come to expect in modern life. Whether it comes from a stereo, television, movie or even a lowly cell phone, it was certainly shaped by this technology.

I learned long ago that my own half-hearted efforts were often better than most people's full efforts. Don't get me wrong, I excelled in my classes when I was interested. But if I wasn't interested, I put in the minimal effort and was always able to get by. I squeaked by my college physics and advanced math classes not by mastering the material but by being among the average of the worst. I distinctly remember getting a 25/125 on a physics midterm and earning a `C'. "Curves are wonderful things", I thought at the time.

I wanted to take the Digital Signal Processing class, but I first had to take Professor Huffman's computer science (really math) class as a prerequisite. Being as cocky as most college students, I assumed I would have no problem because, as badly as I was trained, I was way beyond most of my fellow students. I assumed I could always squeak by on the curve. It was common knowledge among the undergraduates that Professor Huffman's class was very hard, and most people failed it on the first try. But I was sure I wasn't "most people".

Then we got our first assignment. I can't remember the topic or how well I understood it, but I remember my grade: zero. A zero wouldn't be so traumatic if I had blown off the assignment, but I had worked hard, with the requisite sleep-deprivation as proof. No partial credit? No points for trying? I was bewildered.

This continued on for a few weeks. Sometimes I got one point, but usually zero points. Finally, when I felt I had lost all hope, I set up an appointment to meet with Professor Huffman. He politely explained to me that he considered most of his students undisciplined in math and not properly prepared for his class. I don't want to put words in his mouth, so I will just say that the essence of the conversation was that he was trying to bring some real world discipline into his classroom. In his class, it was either right or wrong, and there were no points for trying. I would, of course, later realize this is exactly how the real world works, but at the time it was perplexing to me.

Professor Huffman and I had a few other conversations. While my math was undisciplined, it turned out that I wasn't doing terribly. I would later realize that I essentially understood the material, but tended to make simple math errors. We also talked about the educational system, which is probably how I first got involved in education reform. He was frustrated that so many students came to his class unprepared by their previous math classes. He would, literally, get students who had taken advanced calculus and didn't really understand the graph of a sine curve. In the math of digital signal processing, that's the equivalent of understanding that 2 + 2 = 4. There is nothing simpler and nothing more fundamental.

At the time, of course, I didn't think he was talking about me. I worked harder and harder and managed to eke out some points here and there. I never did well, but I did improve. At this point you're probably expecting the typical happy ending where I madly prepare for the final and pull off a miracle comeback to pass the class.

Well, no. A week before the final I met with Professor Huffman for the last time. I suspected I didn't have enough points to pass the class even with a 100% on the final, and I asked him if that was the case. He responded that no, I couldn't pass the class. With a large amount of frustration I responded, "Then why should I bother showing up?" His response to me was simply, "You might learn something."

I didn't get it at the time, and I didn't show up for the final. As predicted, I failed the class and moved on to my next quarter. I did start to realize that I wasn't cut out to be a Computer Engineering major and switched a few quarters later to Computer Science. After graduating, I lasted, literally, one day working in the field and decided I really wanted to be a professional musician.

The story of how I became a professional musician is best saved for another time, but I will say that as I learned to play professionally I realized that what Professor Huffman said about the real world was exactly right. Partial credit and curves simply don't exist in the real world. In hindsight, by failing me, Professor Huffman probably did the best thing he could have done for me. If I had passed, I would have continued to fake my way through engineering school and eventually dropped out. I didn't know it at the time, but he made me realize I had faked my way into his class and that faking wasn't going to work in the real world.

I like to say that all my best lessons were learned as a result of failure. I try to instill in my students the confidence to know that failure is a good thing (except in skydiving!) as long as one learns from the experience. I can't think of a single teacher who changed my life by telling me I was doing well. Yet I have been privileged to have had many teachers who changed my life by telling me how badly I was doing and I try to thank each and every one of them as often as I can. Professor Huffman passed away before I could tell him, but I'm fairly sure he knew exactly what he was doing.

Unfortunately, many educators in modern times are making great efforts to eliminate all possibility of failure in the classroom. To them I only ask: What's harsher? Letting students fail and learn in the controlled school environment, or telling them how great they are until they are crushed in the far less forgiving real world? I see the latter approach with music students on a regular basis, and it breaks my heart.

When I study and work now, I no longer do just enough to get by. I no longer expect the curve to save me or expect people to be impressed because I'm the best of the worst. I try to do things to the best of my ability. I try to approach every new situation, no matter how much of a waste of time it might seem, with the simple idea that I "might learn something." I try to instill these values in all of my students. Thank you, Professor Huffman.