In my role as junior hiring chair, I’ve been thinking a little bit about how a (R1) institution can best serve its postdocs. Many find the transition from graduate student life to being a postdoc somewhat of a rude shock. At the same time as the intellectual support structure of your advisor and fellow graduate students is taken away, while at the same time you have to take on significant teaching responsibilities. Even for those will a fellowship to offset their teaching, it can be a little daunting to figure out exactly how to interact with your new research group.
What should the expectations of a new postdoc be? Many universities assign research mentors to new postdocs, but (in practice) this is essentially meaningless unless it carries with it certain expectations for mentor and mentee to interact. How much of the role should senior faculty help in suggesting problems for postdocs to work on? No doubt the answer to many of these questions is “it depends on the postdoc” but I would love to hear personal stories (positive and negative) about your postdoc experiences, especially as it relates to practical steps that an institution can make to improve the experience.
Feel free to leave your comment anonymously (well, people feel free to do that anyway). I don’t particularly trust my own experience since I feel that I was probably more independent than most as a graduate student, and was fairly happy working alone in my office (not to mention already having a number of collaborations ongoing with Kevin Buzzard and Matthew Emerton). Harvard was a welcoming and friendly place (to me), but my best interactions happened serendipitously more often than not. The initial seeds of my collaboration with Barry started by joining in conversations he was having with Romyar Sharifi and William Stein in front of their offices (all on the 5th floor I believe) discussing (early forms of) Sharifiology in the context of Barry’s paper on the Eisenstein ideal. I had a few lunches with Richard Taylor at the law school (I have a vague memory that I realized this was possible from Toby — could that be right?). Richard is definitely generous with his time, and (in this context) he was ideal for bouncing off ideas. On the other hand, I don’t think Richard’s style in mathematical conversation is to be very speculative; he certainly never suggested any particular problem to me but nor did I ask. My collaboration with Nathan surely started out by virtue of the fact that we would chat socially at tea time.
I can’t quite distill from my own experiences either any recommendations for new postdocs or specific recommendations for institutions (particularly the University of Chicago) to put things in place to improve the lives of postdocs. But perhaps you can help!
Here at Illinois, we have tried to strengthen the mentor-postdoc connection by making the choice of mentor an integral part of the postdoc hiring process. That is, the committee considers nominations from individual faculty that say (roughly), “I would like to mentor this postdoc, this is why they are awesome, here are some projects I would like to work on with them, and here is my track record of outcomes of former postdocs”. Other faculty are allowed to chime in in support of a candidate, but a single mentor must be named in advance, and the committee is supposed to look carefully at that aspect when making its ranked shortlist. NB: We have about 1 postdoc for every 4 faculty (15-20 and almost 70 respectively).
At Wisconsin postdocs are hired by a specific faculty member with the expectation that the two people will meet regularly and probably collaborate on at least one paper during the postdoctoral term. It doesn’t always work but there’s at least some concentration of responsibility.
This was very much the opposite of my experience at Princeton, where people were hired if they seemed to be interesting and no one in particular was tasked with mentoring them. Princeton worked out great for me but I saw it work out less well for plenty of others, and I think a lot of chance is involved in that system.
The Wisconsin system has downside too — it makes it harder for us to hire postdocs who’d bring a lot of interesting math to the department but to whom no individual faculty member is close enough to commit.
Most postdocs at Imperial are attached to grants for an individual faculty member, so the situation is quite different; that said, when hiring there is always the balance (as I’m sure there is everywhere) between hiring the strongest person versus the person closest to your own interests.
As for tips for postdocs: anecdotes aren’t data, but at Harvard I got a lot out of talking to/working with the PhD students. PhD students have the most time of anyone in the department, and of course there is not much difference between a final year PhD student and a first year postdoc anyway.
I also found the fortnightly law school lunches with Richard very useful, because if I’d been stuck for a week on something and could explain the problem, it was likely that if he thought about it for five minutes he could get me unstuck. Of course it helped that I was spending my whole time thinking about how to prove generalisations of [CHT].
(I would have said that the chronology was off for you to have learned about the law school lunches from me, but maybe not – I guess we first met when I was visiting as a grad student for fall of 2002, and sometimes I would tag along when Richard and Kevin went to lunch, maybe you came along too or something like that?)
Stanford requires postdocs (at least in the sciences, maybe also more broadly than that) to write up an annual “individual development plan” and meet with a faculty member to discuss it — here’s a link https://postdocs.stanford.edu/current-postdocs/navigating-your-individual-development-plan-idp/your-individual-development-plan
There’s not much enforcement (just a button somebody clicks to say they met with you), and people don’t always take it seriously, but I felt like I got a lot out of having a formal framework to reflect seriously about my goals during my first year as a postdoc.
One big and simple thing that faculty can and should do is to invite the postdocs they know / are in their area (but to whom they are not necessarily assigned formally as mentors) to meet to chat about math, say at least once a quarter. Of course it’s better if a postdoc has the confidence to ask for meetings themselves, but it’s always going to be easier for the more senior person to ask. As faculty are busy and forgetful too, it could be helpful, especially in large departments, to have a more formal system for this.
Another great thing that faculty can do is to arrange time for visitors (e.g., seminar speakers, but also longer visitors in town for collaborations, etc.) to meet individually or in small groups with postdocs or graduate students in the same area. Seminar lunches and dinners I think are meant to provide a bridge for this, but in my experience they’re not very effective –it’s much better to have a smaller, less intimidating group, meeting in a situation where the expectation is that math will be discussed.
I’ll also second the comment that interactions between postdocs and graduate students are very important, but again, faculty can play an important role in identifying groups of people that will talk fruitfully.
To summarize what I think might be my point: Faculty, as the only pseudo-permanent fixtures in a department, are uniquely well-placed to facilitate interactions because they know a large chunk of the other faculty, graduate students, visitors, and postdocs, none of whom may know each other. The best thing faculty can do with this knowledge might be to literally to set up formal times for specific people to meet and talk about math — this sidesteps all of the social nuances of making introductions, or encouraging grad students/postdocs to set up meetings when they may be too intimidated or shy to do it on their own.