I had a bit of a gap in blogging over the last few weeks, as we scrambled to pack up our house (we’re renting out it while we’re away), and then of course, the roadtrip to Colorado to start the first of my three studies of software development processes at climate modeling centres. This week, I’m at the CCSM workshop, and will post some notes about the workshop in the next few days. But first, a chance for some reflection.
Ten years ago, when I quit NASA, I was offered a faculty position in Toronto with immediate tenure. The offer was too good to turn down: it’s a great department, with a bunch of people I really wanted to work with. I was fed up of the NASA bureaucracy, the short term-ism of the annual budget cycle, and (most importantly) a new boss I couldn’t work with. A tenured academic post was the perfect antidote – I could focus on long-term research problems that interested me most, without anyone telling me what to study.
(Note: Lest any non-academics think this is an easy life, think again. I spend far more time chasing research funding than actually doing research, and I’m in constant competition with an entire community of workaholics with brilliant minds. It’s bloody hard work)
Tenure is an interesting beast. It’s designed to protect a professor’s independence and ability to pursue long term research objectives. It also preserves the integrity of academic researchers: if university administrators, politicians, funders, etc find a particular set of research results to be inconvenient, they cannot fire, or threaten to fire the professors responsible. But it’s also limited. While it ought to protect curiosity-driven research from the whims of political fashions, it only protects the professor’s position (and salary), not the research funding needed for equipment, travel, students, etc. But the important thing is that tenure gives the professor the freedom to direct her own research programme and the freedom to decide what research questions to tackle.
Achieving tenure is often a trial by fire, especially in the top universities. After demonstrating your research potential by getting a PhD, you then compete with other PhDs to get a tenure-track position. You have to maintain a sustained research program over six to seven years as a junior professor, publishing regularly in the top journals in your field, and gaining the attention of the top people in your field who might be asked to write letters of support for your tenure case. In judging tenure cases, the trajectory and sustainability of the research programme is taken into account – a publication record that appears to be slowing down over the pre-tenure period is a big problem; if you have several papers in a row rejected, especially towards the end of the pre-tenure period, it might be hard to put together a strong tenure case. The least risky route is to stick with the same topic you studied in your PhD, where you already have the necessary background and where you presumably have also ‘found’ your community.
The ‘finding your community’ part is crucial. Scientific research is very much a community endeavor; the myth of the lone scientist in the lab is dead wrong. You have to figure out early in your research career which subfield you belong in, and get to know the other researchers in that subfield, in order to have your own research achievements recognized. Moving around between communities, or having research results scattered across different communities might mean there is no-one who is familiar enough with your entire body of research to write you a strong letter of support for tenure.
The problem is, of course, that this system trains professors to pick a subfield and stick with it. It tends to stifle innovation, and means that many professors then just continue to work on the same problems throughout the rest of their careers. There’s a positive side to this: some hard scientific problems really do need decades of study to master. On the other hand, most of the good ideas come from new researchers – especially grad students and postdocs; many distinguished scientists did their best work when they were in their twenties, when they were new to the field, and were willing to try out new approaches and question conventions.
To get the most value out of tenure, professors should really use it to take risks: to change fields, to tackle new problems, and especially to do research they they couldn’t do when they were chasing tenure. A good example is inter-disciplinary research. It’s hard to do work that spans several recognizable disciplines when you’re chasing tenure – you have to get tenure in a single university department, which usually means you have to be well established in a single discipline. Junior researchers interested in inter-disciplinary research are always at a disadvantage compared to their mono-disciplinary colleagues. But once you make tenure, this shouldn’t matter any more.
The problem is that changing your research direction once you’re an established professor is incredibly hard. This was my experience when I decided a few years ago to switch my research from traditional software engineering questions to the issue of climate change. It meant walking away from an established set of research funding sources, and an established research community, and most especially from an established set of collaborative relationships. The latter I think was particularly hard – colleagues with whom I’ve worked closely for many years still assume I’m interested in the same problems that we’ve always worked on (and, in many ways I still am – I’m trained to be interested in them!). I’m continually invited to co-author papers, to review papers and research proposals, to participate in grant proposals, and to join conference committees in my old field. But to give myself the space to do something very different, I’ve had to be hardheaded and say no to nearly all such invitations. It’s hard to do this without also offending people (“what do you mean you’re no longer interested in this work we’ve devoted our careers to?”). And it’s hard to start over, especially as I need to find new sources of funding, and new collaborators.
One of the things I’ve had to think carefully about is how to change research areas without entirely cutting off my previous work. After many years working on the same set of problems, I believe I know a lot about them, and that knowledge and experience ought to be useful. So I’ve tried to carve out a new research area that allows me to apply ideas that I’ve studied before to an entirely new challenge problem – a change of direction if you like, rather than a complete jump. But it’s enough of a change that I’ve had to find a new community to collaborate with. And different venues to publish in.
Personally, I think this is what the tenure system is made for. Tenured professors should make use of the protection that tenure offers to take risks, and to change their research direction from time to time. And most importantly, to take the opportunity to tackle societal grand challenge problems – the big issues where inter-disciplinary research is needed.
And unfortunately, just about everything about the tenure system and the way university departments and scientific communities operate discourages such moves. I’ve been trying to get many of my old colleagues to apply themselves to climate change, as I believe we need many more brains devoted to the problem. But very few of my colleagues are interested in switching direction like this. Tenure should facilitate it, but in practice, the tenure system actively discourages it.
Our specialissue of IEEE Software, for Nov/Dec 2011, is out! The title for the issue is 