Teaching

The TU Berlin Skyscaper

Posted by James

So today I want to talk about teaching. The "official" reason I came to Berlin this summer was to teach an Internet Routing course in the Internet Architecture department at the Berlin Institute of Technology (TU Berlin). Naturally, there were more reasons than that, but teaching is a large part of what I did for the first two thirds of the 6 months I've been here.

I didn't do much teaching in graduate school, just one course on linear systems. At the time, I was more interested in research. After that course, I co-wrote some research grant applications with my major professor, which were funded, and so I didn't have to teach again. Like many PhD students in the US, I really didn't enjoy teaching. The students were just a couple years younger than I was, and I found having to decide their fate in terms of grading uncomfortable. When a student came to me and pleaded that in effect "the dog ate my homework" I was inclined to give them the benefit of the doubt. I also didn't find much interesting and engaging in teaching. It seemed like I was just there to transfer information to the students, and they really could get that through reading the text book in the end.

In the fall of 1996 while I was working at Sun Microsystems, one of my co-workers who was an alumnus of Mills College, a woman's school in Oakland, sent out an email saying that two professors had quit from the Mills CS department and the department urgently needed someone to teach a course in operating systems in the spring. Mills had an old style combined CS/Math department, and there were no other professors who felt qualified to teach operating systems. Since I had just finished a research project developing a new object-oriented operating system, I volunteered, and Sun agreed to let me take one morning off a week to teach the course.

The course was senior level course (Mills at that time had no graduate program) so I expected most of the students already knew a great deal about how programs are developed and deployed on an operating system like Unix. But around half way through the course, I was driving home pondering the kinds of questions I was getting from the students when I realized that questions indicated a basic misconception about how program development and execution work together on an operating system, i.e. the compile, link, and load process. So I modified my lecture for the next class and explained how the program development process works. This was somewhat of a revelation to me: that the job of the professor was not to transfer information but rather to dispel the students' misconceptions about how something works. Perhaps not all the time, though, the students might actually not have enough of an idea about how something works to even have a misconception.

So that was the extent of my teaching experience when I started at TU Berlin in April. One issue I quickly faced is that the university system, especially at the graduate level, is very different in Germany and the US. The courses at the graduate level at TU Berlin are taught in English, but there are different courses with different kinds of activities. Lecture courses (which is what the Internet Routing course is), seminars in which papers are discussed, and lab courses in which the students participate in hands on activities are all separate courses rather than being combined. In addition, for the lecture course I was teaching, there were no homework assignments and no exams except for the final, and the final was an oral exam. The students also didn't have to come to the class because there was no grade for class attendance, and about a quarter of them didn't. Of course, in the US, lab and lecture courses go together, and professors often assign readings in papers or textbooks that are then discussed in class. During my lecture class, I would get a few questions each class, but not many, and they were usually from the same students.

These differences in the organization of learning in the US and Germany reflect fundamentally different philosophies about the role of the student in the two countries. The German system puts the responsibility for learning the material onto the student. The system isn't designed to encourage the students to compete with each other, and the students don't really expect that. In contrast, most US students would be unhappy with the German system. They want to know where they stand with respect to their peers, a kind of unsubtle competition, and for that they need constant grades on exams, homework, and lab assignments, together with a system for deciding how their final grade will be calculated from their grades on the various assignments during the semester. American students feel perfectly comfortable with the professor constantly judging their knowledge of the material and would feel uneasy if that weren't the case. Then too, many US students are paying a lot of money for tuition, even going deeply into debt, so they expect some kind of forcing function on their learning. In contrast, there is no or very little tuition at most German universities, and a degree of maturity is demanded from the students that most US students would be hard put to show, especially at the undergraduate level.

I ended up spending a lot of time developing material for the course, mostly on data center networking which is something new that has only come into prominence within the last 5 years, and enjoyed researching the material and presenting the results to the students very much. I was actually very familiar with the topic, having spent most of the last five years working on it myself. In the end, all the students passed the final, and only a few exhibited signs of not knowing the material sufficiently to pass with the US equivalent of a B. The only regret I have is that I didn't have much direct contact with the students. Because the students aren't constantly having to study for mid-term exams, complete homework assignments, or do lab assignments, they can wait until the end of the semester and cram for the exam, so they don't have much of an incentive to come by the professor's office and ask questions. But I suppose, if I were to go into teaching full time, I'd have an opportunity to work more closely with them. Most of the PhD students at the TU that are my colleagues (the PhD students do almost all the teaching in Germany) work closely with students, on Masters' projects and other research. I don't know whether I'll get a chance to teach again, but I would certainly like to.