How to survive in science, Part 1
Let's first define the problems you face, of which there are many
In a previous article I mounted a modest defence of science academia. I was not arguing that academia is perfect and all of its detractors should shut up; simply that much social-media criticism has not only taken off from the surface of reality, but has by now blasted into the stratosphere of batshit craziness. It could be, of course, that I am entirely wrong, and all of those strident advocates for radical change understand everything far better than I do, and if so, I look forward to the time when they have concluded the listing of their grievances and let us know what unsinkable plans they have to bring about research utopia. In the mean time, since I am not nearly so wise nor visionary, I will restrict myself to humbly relating my own limited understanding of how to survive academia as it currently exists.
Before doing that, I need to explain a crucial point about science: it is an apprenticeship trade. This was discussed recently by my fellow stackmate Erik Hoel1, but is otherwise under-appreciated and rarely mentioned. All successful scientists learn from other successful scientists. Do I really mean all? Don’t some emerge, self-created, out of nowhere? Ok, maybe a tiny few — and I bet that's the group your mind has focussed on. Stop it! It's enough to have deluded yourself that you could be the next Einstein, you don't need to go full looney and believe you can pull it off by hiding for five years alone in the corner of a library2.
I am not calling science an apprenticeship trade to make it sound ye-olde and romantic, like being a 17th-century violin maker. Nor am I trying to portray scientists as an elite class of the intellectual chosen few3. I'm just telling you how it works. Good research requires years, and years, and years, to acquire the necessary specialist knowledge in a particular topic. It's not a previously defined topic — you can't look up “specialist topic 732” at the library and set yourself up for seven years of reading. It is a very particular sampling of lots of specialist knowledge, which you will get only from an experienced practitioner, or more likely a series of experienced practitioners as you slowly build up a collection of supervisors, colleagues, collaborators, mentors and rivals.
If you thought it was like being a spy — ace a special test and an enigmatic figure shows you the secret door into a world of mystery — forget it.
Maybe that's why the apprenticeship model gets so little attention: it makes success seem hopeless. I joked in an earlier piece that one of the keys to success is to ignore good advice, but I also believe (and could there be a clearer definition of the scientist mentality?) that you're always better off if you know how things really work.
So, to continue.
Let's take this apart, starting backwards from the top. The goal of any research faculty hire is to identify someone at the top of a cutting-edge field. Cutting-edge fields are defined partly by the vicissitudes of recent scientific progress, but also by the political machinations of the currently anointed top researchers. If you want to be hired, your best chance comes from having worked in one of those fields, and ideally for or with one of the relatively small number of science stars that still shine brightly.
This is not as impossibly difficult as it sounds. There are a lot of sub-sub-sub-sub-fields of science, and lots of hot topics within them, and there are a lot of university departments that would love to hire someone to take them in one of those exciting new directions. Identifying the very best people is so difficult as to be often indistinguishable from a lottery, and even if there was a clear ranking, most universities don't have the prestige to hire from the top 10%.
On the other hand, a significant fraction of those young hot shots fizzle out very quickly, if they were even as good as advertised to start with. On top of that, all research topics slowly die, except for those that quickly die. These two effects are the main reasons why most of the faculty in any department — look around, you know it's true! — are useless. Well, not completely useless: they are still great at pumping out hype about themselves. So your chances of correctly identifying a brilliant supervisor who will mentor you to science greatness have just taken another hit. Especially since you have your own peculiar preferences for what you find interesting, and they are almost certainly not what you think they are.
Does that sound grim? I haven't finished yet.
There is almost no way to identify a promising research topic. Actual research topics are highly specialised, and will be incomprehensible and impossible to assess for an undergraduate; you cannot place them in the context of the wider field, you cannot judge their level of difficulty or suitability to your particular strengths, which, in addition to your interests, you have probably not yet fully identified, and you cannot predict some topic's future impact on the field, or indeed the future direction of the field itself, either towards a succession of breakthroughs or a stumbling into obscurity.
Consider an example. You may have heard that the nature of dark matter and dark energy are two of the biggest mysteries in physics. Indeed they are. But they are not research problems. A research problem is, “Some people think that dark matter is X. The observational consequences of X in fifteen different upcoming experiments and observatories have been calculated by 107 different researchers in 342 different papers. Prof. A (your potential supervisor) disagrees with one set of those results, and has come up with a novel independent way to check them.” Is that an interesting and worthwhile problem? Who knows! Maybe it's pointless. Or maybe the student who does that calculation will attract the attention of all 107 other researchers on this topic (now swelled to 468). How hard is the problem? It could be a trivially easy game-changer, or a fiendishly difficult dead end.
This sounds like I'm saying don't bother. Not at all. All I'm saying is that if you want to do science, the most important thing is to work on a promising topic with an excellent supervisor, and you have severely limited means to identify either. Most people can't give you solid advice, because they just got lucky. That includes me; I also got lucky.
In this environment there is no recipe for success, only guesses at likely tactics for survival. It could take a long time to find a topic you consider interesting, and is currently active, and in which you can make a contribution, and to actually make those contributions. It can also take a long time to find the people who can help you find the people who can get you there.
I have posed the problem as one of getting a good PhD topic and supervisor, but I could re-frame it for each step between undergraduate and faculty member. All these steps make it even harder, but they also come with opportunities. A poor project or a poor supervisor (or even both) are not the end, if you land a lucky postdoc position. Or during your PhD find someone to work with on a promising side project. Or talk to someone at a conference who points out a useful project hiding inside your terrible project. Or... the possibilities are endless.
So: “survival” is a fine way to think about it.
I have put together some suggested survival tactics. I don't know how good they are, only that they make sense to me, given what I've seen in the more than a quarter of a century since I started my PhD. This piece has set the scene, so that in the next part I can just rattle them off.
Hoel also has some academia-survival advice, which I may say more about later.
An earlier version phrased this as “All great scientists are trained by other great scientists”. Pretentious rubbish! What the hell is a “great scientist”? I’ve never met one.
Obviously they are.
I did not "survive" as a researcher, but I also maintain that my research training taught me to think like nothing else I've ever done.
Terrence Howard could benefit from reading this. But then again, he’d say you’re one of those gatekeepers in mathematics who are suppressing the truth. 😉