My Cup of Tea

Over lunch the other day, in the company of three fellow graduate students in theoretical physics (one particle physicist, one solid state physicist and one quantum chaos theorist), the conversation as so often drifted into the subject of the various frustrations of being a grad student. Since one of us has just handed in his doctoral thesis for printing, and two others (one being myself) have their dissertations coming up in a few months’ time, we are all starting to become a bit retrospective. What have we actually accomplished in these four years? Do we really know anything about what we are doing?

My girlfriend has had to suffer quite a few complaints from me that I have really learned anything at all, that I have absolutely no clue about anything really fundamental underlying my own research, let alone other topics and models of condesed matter theory and many-body physics. Even Wikipedia knows about as much about density functional theory as I do, and I am after all supposed to defend a doctoral thesis pretty much based on this method a few months from now. Quite a depressing thought. To add inslut to injury, quite a bit of the software I have been using in my research was not written by me, but years ago by some of our collaborators (since modified back and forth). Did I actually learn anything at all in all these years? Sometimes I really have my doubts.

Well, it turns out I haven’t exactly cornered the market for such concerns. Quite to the contrary, they are probably shared to some extent by every theoretical physics grad student that ever set foot at a university. Our discussion started by the particle physicist wondering what strategy to use when reading up on things in preparation for his upcoming disertation, being increasingly frustrated by not having a clue about fundamental proofs in the dervation of the theory employed in his reasearch. Solid state physicist expressed similar concerns regarding her research, and let out a forceul "yes!" as I voiced my frustration at not knowing the first thing about anything even slightly outside my own research. Meanwhile, quantum chaos theorist (who still has some time to go to his disertation) revealed that he now spends evenings and weekends trying to understand a particularly obscure piece of mathematics related to his research (thereby adding to my own bad conscience about spending evenings and weekends on everything except physics, but also confirming my suspicion that grad students in general don’t feel they really know half the things they should).

Very likely, one of the greatest sources of frustration for grad students is this feeling of never ever being on top of anything. As an undergrad physics major, you take one class at a time, spending all your time mastering its specific contents. After the exam, you move on to master the next class, quickly forgetting everything except possibly the barest essentials of the previous class. In grad school, you have no such luck. No matter what you do, there is always an infinite amount of stuff you should master. You finish reading one paper, having hopefully understood something about the main points, only to find the next paper equally incomprehensible. At the same time, you’re supposed to work diligently on your own research, a constant frustration of its own as weeks and months go by and you still haven’t come up with one single really new idea (though it seems like everyone else produces a handful of them every day), and most of your concerns have nothing to do with the physics you’re supposed to study, but with programming that refuses work, with some pesky integral that looks like something out of calculus one but is impossible to work, or simply with the sheer boredom of collecting seemingly meaningless data from some piece of software. Every hour spent looking at the same twenty lines of old Fortran code feels like two hours wasted, that you could have used to actually read up on that advanced statistical physics class you never took as an undergrad, and every hour spent banging your head against the wall trying to understand some review paper about some other theory feels like two hours that could have been better used producing another few megabytes of data, or writing that piece of code that was supposed to be "just a simple modification." This not to mention the time wasted just looking out the window, or doing whatever else simply to try to clear your mind, or because you were again just too easily distracted from that exceedingly boring task you’re supposed to do, or because suddenly you don’t know what to do next, despite having four or five things that do need your attention — but you’re equally stuck on all problems.

This, perhaps more than anyting else, is where the thesis advisor has his/her most important task: To ease some of the shock of going from the ordered class environment of undergrad studies, to the great sea of disordered knowledge and petty tasks of everyday research work. What one most sorely needs as a grad student is someone to serve as a guide to what to study, what graduate classes to take, what to read up on, and to constantly show an interest in the work the student is doing. It is all to easy for a grad student, coming with a burning interest in physics into the world where neat theories belong to the textbooks in the library and actual research consists of getting a fortran code to calculate some matrix element somewhere that probably no-one in the world will ever be interested in, to lose interest and motivation.

And for the grad student, the most important lesson to learn is that you will always be frustrated about all the thing you’ll never find the time to learn, and that one will always have to look for inspiration, unless you are lucky enough to be the type who finds joy and inspiration in the details of computer programming, electronics or whatever your research tool may be.