Not Good Enough.
Ed Yong is one of the English language’s better science writers. He often weighs in on issues facing academia, in addition to reporting science. One of the biggest issues facing academia right now is the pipeline problem: too many PhDs are graduating compared with the number of professorships that will ever be available for them. This has led to the creation of intermediate steps like “post-doctoral scholar”, and in some places “instructor” or “assistant researcher”, which each have their own bars to hurdle, and their own attrition mechanisms. At each level some people leave. Too often, those who leave are considered, and called, failures by those who advance. Especially those who advanced long ago, before the pipeline narrowed as much as it does today.
This has resulted in people being considered “junior” investigators for nearly half their careers, usually. Because universities and funding agencies are so risk-averse, they expect scientists to be mentored and coached and coddled until “young” can be used only in jest. Or sarcastically. In many ways, it’s degrading and infantilizing to tell a scholar, trained for decades and published many times over, that they must still be buttressed against failure by crouching obsequiously beneath the umbrella of a grey-haired magister. Failure is incredibly useful. A topic for another day.
But on the other side of it, perverse incentives lead to the graduation of so many PhDs. Grant money is scarce. It’s much less expensive to pay a graduate student or a postdoc to do lab work than it is to hire full-time accomplished technicians or full-time researchers to collaborate. So more and more cheap labor is hired. More PhDs are trained. And the wide end of the pipe gets wider. Meanwhile, universities continue to divest themselves of tenure-track and equivalent positions. The narrow end gets narrower. I’ve read (but don’t recall the source – treat as speculative) that fewer than 10% of graduating PhDs can expect to end up in tenure-track positions. There are massive structural reasons that people cannot and do not advance.
And people write about it. Many people, when they leave, describe these structural reasons. Or family reasons. Or any number of reasons that they don’t advance. But something it seems we rarely read is, some people surely must not advance because they just aren’t good enough. Ed Yong put it this way:
Funny how you never see confessional posts from people who left academia because they just weren’t very good.
— Ed Yong (@edyong209) April 12, 2014
Well, I haven’t exactly left academia. But I trained to be a professor of systems engineering. I have a Doctor of Science in Electrical and Systems Engineering from one of those fancy, elite universities. I had good connections and a prominent advisor. But I am not, and I will never be, a professor of systems engineering. And the reason is, I’m not good enough.
Studying systems engineering at the graduate level means doing a lot of theoretical mathematics. I spent five years doing proofs. Mostly, systems engineering revolves around being able to model and control how large numbers of objects interact with one another in complicated ways with respect to time. Usually, this means doing vast systems of nonlinear, time-varying, partial differential equations. Now, a lot is known about this field. In fact, it is provable that most such systems cannot be solved with what we call “closed form solutions”. Meaning, it is impossible to simply solve the equations and use them to calculate how the future state of a system will unfold. We have to manipulate. Approximate. Linearize.
I took an entire class on control systems on free-floating locally-Euclidean manifolds. So, imagine being on the surface of a doughnut, and you want to negotiate a spiral and end up where you started. That kind of thing. Your point-mass vehicle weighs X and has control functions Y and Z. You want to get from A to B in minimum time. What do your controls need to be and for how long? How do you stitch together locally-Euclidean reference frames that allow you to numerically solve the equations of motion for the brief period you’re located in each one? Before the non-linear effects overwhelm the linear approximations. This is the kind of work that put Curiosity on Mars. I got a B+ in that class.
I got a lot of B+’s in graduate school. B+’s are just a step above failing in graduate school. I even got a C in my class on Linear Dynamic Systems. Once we added in stochastic noise that needed to be filtered out, I got very confused. I needed to take the class a second time to understand it. The second time I got an A. And I deserved it. I worked hard for it.
Now, a real theoretical mathematician will read the above things and say, “that’s not theoretical math”! And they’d be right. It’s not. It’s applied math. Very, very difficult applied math. And I could do it. At least, I could follow along while the professor did the math on the chalkboard. Remember chalkboards? God I miss them. I could do that math well enough to understand the proofs and do most of my homework. But doing that applied math isn’t the real job of a professor of systems engineering. Sure, they do a lot of it, and solve problems and consult for NASA and other such organizations who need people who are really good at applied math.
The real job of a professor of systems engineering is to invent new math. That lets us solve new engineering problems. Or solve problems that we can’t currently solve because they’re too big, or too non-linear, or happen too quickly. A professor of systems engineering, a good one, isn’t so much dedicated to solving problems. They’re dedicated to building tools. That allow us to dream new problems to solve. A professor of systems engineering is a theoretical mathematician.
I am not. I’m not good enough. And I learned that pretty rapidly. Today, I use fairly simple math, and reasonably cool computer science, to solve huge, interesting, and relevant problems. I am a practicing engineer, not a theoretical engineer. I publish. I teach sometimes. I am an adjunct professor in a department of emergency medicine. I am a principal investigator at a hospital. But I’m not really an academic. I’m not a full-time professor. I’m not a full-time researcher. Mostly, I solve the problems my hospital asks me to solve. I’m good at it. And I’m happy at it.
I didn’t fail at academia. And the academy didn’t fail me. I’m a success story. While training to be a professor, I discovered I wasn’t very good at doing the things a professor in my field is expected to do. So, like the engineer I am, I built something. I built a career that didn’t exist when I started: a professional simulator of health care systems. I can’t do the theoretical math of a systems engineer. I can’t do the theoretical computer science of computer scientist. But I can use these tools to solve problems in healthcare delivery that no one has looked at in this way before. And that’s of interest to both the practical world, and to the academy. Just, not the same academy as the one I trained in.
To call what I did a failure, either of me, or of the system, is absurd! I’m doing interesting work, publishing it. I’m employed and my employer is happy with my work. But it is completely fair to say that I am not a professor of systems engineering because I am simply not very good at it. I confess.