Reluctant Philosopher Kings? Leadership in Academia

Tags: academia, musings

Published on
« Previous post: We Are All Nerds: The Literary Works of … — Next post: The Machine Learning Scientist as … »

Academia is fraught with irrational, inconsistent rules and outright weird phenomena, despite many of its ‘inhabitants’ wanting to pretend otherwise. One such weird phenomenon is the onslaught of ’non-research’ work, in particular administrative duties, as one attains more supervision responsibilities. Somewhat paradoxically, we ostensibly get rewarded for being good scientists,1 but as we progress in our career, we get fewer and fewer opportunities to actually do science.2

Some people suspect that there is some kind of Peter principle at work here, but I think the origin of this phenomenon is rather a systemic one: first, administrative work is often left to the group leaders and principal investigators, despite them receiving no actual training for this. This already creates an inefficient situation since the skills that make you an excellent scientist are not necessarily the ones that make you an excellent organiser. Second, most institutions do not provide a whole lot of administrative support for research groups. At your typical European university, administrative positions are vastly underpaid and, unfortunately, underappreciated. This is not a great combination, as it creates a culture of institutional resentment, manifesting itself in a proliferation of administrative processes, extra documentation, and bureaucratic obstacles. If you have not encountered this so far in your career, you may want to ask your more seasoned colleagues about how hard it is to, say, organise an in-person workshop and get budget for conference dinners.3

When I look at my colleagues and their colleagues, I feel that some of them are like ‘reluctant philosopher kings,’ i.e. they understand the necessity of shouldering the non-research tasks, giving up what they probably enjoy most, so that others can continue doing the research. To me, this always felt a little bit like a noble, albeit maybe unnecessary, sacrifice. However, I was floored to hear senior professors at my alma mater confessing that they have about three hours (!) for research in a good week. The rest of their time is spent with other tasks. I am not quite sure what this does to a person over the long run, but as a very harmful immediate consequence of this, we lose a lot of talent: some people that are passionate about doing research are just not incentivised to stay in this system. And why would they? At most German universities, there are few, if any, permanent researcher positions. You either exit as a researcher or stay long enough to not have any time for research any more.

This also creates massive issues with academic leadership since no time is left to supervise and train the next generation of researchers. Everyone probably encountered that one very senior professor with an infinite number of Ph.D. students who, if the stars are aligned correctly, get to talk to their supervisor for 15 minutes once a year. I am pretty sure that more time is needed to discuss and, well, guide trainees in research. This situation is indeed unfortunate for both parties, because I am pretty sure that the professor would have a lot to teach and the student would like to learn a lot from an experienced person. In most large research groups, there is a special tier of postdoctoral researchers for assisting in supervision duties, but their employment status is tenuous in many cases, with some of them already applying for the next position the minute they get their temporary contract for the next one. Hence, in many cases, students are left alone in figuring out all the things their supervisor already figured out—while this, at best, creates independent thinkers (I am putting this somewhat optimistically here), it also wastes a lot of resources: as any reinforcement learning expert will tell you, you need a lot of iterations to learn things in a random fashion. Plus, there is no guarantee what you will end up learning.4

So, what can we do? Admittedly, I have no perfect solution to this.5 I understand that, with increasing responsibilities, someone has to actually manage the resources of a research group and be in charge. Hence, having a professor with 0% administrative duties would strike me as equally weird as having a research professor who spends 100% of their time managing everything. Perhaps the ideal position is a fiction and cannot be obtained easily. In my mind, I picture something like Star Trek: while Captain Picard has to think about the well-being of the crew and the ship, there are still times where he gets to follow his own research passion, archaeology.

Now, real life is not like Star Trek—unfortunately—but I believe that we can strive to improve academic leadership. At many research institutes, for instance, a better research versus management balance can be achieved, even though the fact that these institutes are situated outside the university system also creates its own complexities at times. Whatever your path, let me give you this cri de cœur: never forget that everything in academia can be dismantled, removed, or rebuilt! Keeping Chesterton’s Fence in mind, not everything should be stripped down, of course. But we need to constantly remind ourselves that this is a human-made system, built originally with a certain purpose in mind. If certain aspects of the system do not serve the original purpose any more, change is not only necessary but imperative.

  1. Let us not start a discussion about meritocracy. ↩︎

  2. It seems that, mutatis mutandis, the same phenomenon can be found in other domains as well. I shall resign myself to commenting on academia, though. ↩︎

  3. Interestingly, learning about the difficulties of these seemingly-innocuous activities has left me with a greater appreciation for being invited to such events. Understanding how hard it is to wrangle some extra money for food out of an apparently uncaring administrative body makes me enjoy such events much more. ↩︎

  4. Looking at certain aspects of machine learning research, for instance, one might be tempted to say that our ‘best practices’ are actually the ‘worst practices,’ but this is maybe best kept for a future post. ↩︎

  5. Maybe it would be better to say that I do not have the perfect solution yet↩︎