Academic Autonomy: How Much Freedom Do Post-Docs Have?

I'm not entirely sure why I keep responding to this, but Bruce Charlton left another comment about the supposed dullness of modern science that has me wondering about academic:

The key point is that a few decades ago an average scientist would start working on the problem of his choice in his mid- to late-twenties - now it is more likely to be early forties or never.

In the UK most people got a 'tenured' university lectureship straight after their PhD (or before) - created a lot of 'dead wood' but also gave people time and security to be ambitious.

Longer time spent as a doctoral student plus an extra added delay of six, ten, maybe fifteen years; spent doing short chunks of post-docs and temporary fellowships, here and there under supervision of various other people.

There are still a whole host of unsupported assumptions (bordering on assertions) in his piece, but the start of this comment is one of those things that makes me wonder how typical my experience of academia really is.

While it's true that post-docs are not fully in charge of the labs where they work, my experience as a grad student and post-doc myself does not suggest that they are mere wage slaves toiling for tenured professors who allow them no freedom at all. This might be one of those areas where my experience as a grad student at NIST doesn't generalize well, but the post-docs I knew there had a great deal of autonomy.

There was one guy, now in a tenured position at a major university, who came in to work on one experiment, and ran with something that was really a tangent to what the experiment was supposed to do. He spent six months plugging away on it, and nobody really had any idea what he was up to until he presented the final results to the group.

That's an extreme case, but essentially all of the operational decisions-- what to work on when, how to modify the apparatus to get a given result, etc.-- were made by post-docs. Hell, even as a graduate student, I got to make a few calls-- one of my grad school papers was an experiment that we did basically because I didn't feel like rebuilding a Ti:Sapph laser one day.

Now, it's true that neither post-docs nor graduate students were able to completely re-invent the projects to look at problems wholly of their own invention, but that's an unavoidable constraint due to experimental resources. It takes a lot of time and money to set up a whole new experiment, and that's not something that can be done every two years. My experience, though, is that within the parameters of an existing experiment, post-docs have a great deal of autonomy to work on what they want to.

And, moreover, it's not like post-docs are assigned on a totally random basis. The decision of where to go (or at least where to apply) gives a prospective post-doc a fair bit of choice when it comes to the general sort of project they'll be on.

I hear complaints about the post-doc situation fairly frequently from the disgruntled academic sector, though, so this may be one of those areas where the oddness of my academic career path prevents wide generalization. It may be that AMO physics is generally nicer than other fields, or that the NIST experience is even more anomalous than I thought. Or it may be that the people who leave grumpy comments on the Internet are selected from a naturally grumpy subset of researchers.

So, what is the lot of a post-doc in your experience? Is having to work for a more senior researcher a soul-crushing, creativity-stifling experience, or is it a useful and important step in the process of academic training?

More like this

Is having to work for a more senior researcher a soul-crushing, creativity-stifling experience, or is it a useful and important step in the process of academic training?

Both.

As an academic postdoc, my experience has been much different than yours. I think that in academic the freedom given to postdocs varies widely depending on the advisor, the project, and the advisor-postdoc relationship, but is generally in the range of the freedom given to grad students. This is frustrating to a lot of people, particularly more senior postdocs, and especially when grad students (because of their project or their special and sometimes rather unfair relationship with the PI) are given more freedom. (It also sometimes leads to worse science, when the opinions of the people actually doing the project are ignored, wasting months or years of valuable time... not that I'm bitter.)

But I think it is also a good thing. Few newly-minted PhD's are ready for total independence - especially theorists (of which I am one). There's not as far as I know a surefire way to turn people into mature scientists, but perhaps self-definition by opposition is one of the more reliable routes. I may think my advisor's orders are arbitrary or unhelpful or flat-out wrong, but they do have the effect of forcing me to think about why I think my way is better and exactly how I would proceed if my advisor were mysteriously detained from his email for the next six months. (again, not that I'm bitter)

By anonymous (not verified) on 29 Jun 2009 #permalink

The situation in particle theory and mathematics is typically at an opposite extreme from the complaints of those working in labs. In these fields postdocs are often hired by an institute or group, and don't have an identifiable "advisor" to tell them what to do. My experience during my post-doc years was one of being left alone to do whatever I felt like, and I think that's not uncommon. The only pressure to work on certain topics came from the fact that it was a three-year postdoc, so being on the job market again relatively soon loomed large.

My experience in quantum information and quantum foundations is that there are basically two types of postdoc: the type where you are closely supervised by an advisor on a specific project and the type where you are given almost complete freedom. I think that both types are needed, especially since people are spending longer as postdocs these days so there are varying degrees of research maturity involved.

Personally, I have only ever done the "complete freedom" type of postdoc. It was tough going at the beginning, especially because I came out of the short and sweet UK Ph.D. system, but I wouldn't want to change my experience for the world.

I wouldn't say that the "complete freedom" postdocs are in short supply either. The only real problem with them is that you end up spending about half the duration of your contract worrying about where the next job is coming from. This means that you have to work on things that you know are going to be finished quickly and lead to publications, which can lead to shelving the more ambitious and risky ideas. Otherwise you risk spending years in academic limbo as the postdoc positions get harder and harder to obtain. In my view, unless lots of new faculty positions are suddenly going to appear, more longer contracts (about 3-5 years or so) is the best we can hope for to solve this.

I think the main reason why one might have to stick to a given theme as a young postdoc is that working on something really new often means a higher and higher level of specialization. This is not a new trend [see: JC Slater, Physics Today 22, 35 (1969)].

Generally, I think funding can be impose constraints, while supervision usually doesn't.

Charlton is right on one point: The trend toward longer time-to-degree and serial postdoctoral appointments is a problem. But it's not a question of lacking autonomy (for the most part; I'm sure that toxically domineering postdoctoral advisors exist). You do have constraints as a postdoc to work on something related to what the lab/group is doing, but as you pointed out, most people choose their postdoc group based on the sort of work the group does (again, there are exceptions, usually involving a two-body problem).

The real problem with the postdoc system is that it has evolved into proliferation of soft-money positions. The notion that you have to get a tenure-track position at a Major Research University to be considered a success still holds sway in large swaths of academia, particularly at Major Research Universities. But there are not, and can never be, enough tenure-track positions, so some people take a third or even fourth postdoc in hopes of qualifying (these people probably damage their chances to get a tenure-track job at a liberal arts university in the process), with the result that one postdoc is seldom enough to win a tenure-track job at a Major Research University anymore. I don't know how it is in AMO physics, but in my area there are several people who have been career soft money scientists (availability of NASA funding is a factor; they are not constrained by the educational mission NSF has to focus resources on students and postdocs). I'm perilously close to becoming one of them myself.

By Eric Lund (not verified) on 29 Jun 2009 #permalink

After having heard a lot of friend's stories I tend to think that in the AMO community, there really could be fewer slave-drivers among bosses, than in other areas, such as surface science. If this is true, I wonder why...

P.S. greetings from Atomic Gordon Conference!

The notion that you have to get a tenure-track position at a Major Research University to be considered a success still holds sway in large swaths of academia, particularly at Major Research Universities. But there are not, and can never be, enough tenure-track positions, so some people take a third or even fourth postdoc in hopes of qualifying (these people probably damage their chances to get a tenure-track job at a liberal arts university in the process), with the result that one postdoc is seldom enough to win a tenure-track job at a Major Research University anymore. I don't know how it is in AMO physics, but in my area there are several people who have been career soft money scientists (availability of NASA funding is a factor; they are not constrained by the educational mission NSF has to focus resources on students and postdocs).

The post-doc "requirement" is a funny sort of thing. It helps to have one (and we've had candidates fail to make the short list in some job searches primarily because they had no post-doc experience), but it doesn't really help to have more than one or two.

In AMO physics, my anecdotal impression of junior-faculty hires at Major Research Universities has been that the jobs tend to go to people who have done only one post-doc. The few I know who were hired after two post-docs generally had done one of the two (usually the first) in a foreign country.

I wouldn't count this as data-- it's my impression based on people I happen to know in the field-- but I haven't seen much indication that taking a third postdoc improves anybody's chances at a Research I job.

I'm currently a particle theory postdoc, and both now and in grad school I've been pretty much free to do whatever I wanted. In fact, only my first paper in grad school was something I would characterize as my advisor's project; other papers I wrote with him were either my ideas or areas of mutual interest. If there are constraints, they're not imposed by anyone I work for or with, so much as by needing to publish in areas that are popular enough that I can guarantee I still have a job beyond the short-term.

Experimentalists obviously have somewhat more constraints, given that people who obtained expensive equipment tend to have more say in how it's used. But from what I hear and see of experimental particle physicists, postdocs have a great deal of freedom, and students do as well to some extent. The large collaborations offer a lot of room to find a niche one is interested in (though people do, I think, tend to stick relatively close to something their group is expert in).

anonymous above said "Few newly-minted PhD's are ready for total independence - especially theorists (of which I am one)." That's not very consistent with my experience, but I suppose different branches of theory can work differently.

As I am in the last weeks of a postdoc in geology (and having a lot of friends in the same boat), I can say that at least in my field, postdocs have a lot of latitude in what they want to do. There has been very little slave-master tendencies in my postdoc and the same can be said for most geologists I know. I have a pile of projects that I've started since I started my postdoc - and heck, I wrote the proposal for my postdoc that got NSF funding with my mentor here, so I've always been able to work on the parts of the research in which I am interested.

However, there is definitely a strong sense that you need to do a postdoc (or two) before you can get hired a faculty. The average for most people I know is 3-4 years of postdocing before heading off to faculty-land in general. Now, I can understand the need to do a postdoc if you plan to work at an R1, but if you're heading off to a small liberal arts school (like I am), I find it a little more perplexing as I've been able to do much less teaching in my postdoc than I was doing as a grad student. That being said, the postdoc was a great way to get out of the "grad student" mentality and into a faculty one without being plunged headfirst into it.

Like grad school, but not to the same extent, it's mentor-dependent; the good news is that you *have* the PhD and can leave.

That was the only bright spot as my first postdoc, the content of which was primarily dictated to me** wound up. When I got to the point of considering driving into a tree rather than get to the lab.. not good. Then I did a second postdoc where I essentially did whatever the heck I wanted. My experience since then suggests that the latter is more common. Now I have two of my own; they get to do anything that they can convince me is scientifically worthwhile and at least tangentially related to our funding (or work on something part-time in order to be able to get additional funding for it, if worthwhile but not funded).

[**to be fair, I'd gone there to learn a specific technique. But the only thing that came out of it productively was my own idea and off the lab's track then.]

During my postdoc(s) (in biophysics), I had to work on something that fit within the overall focus of the lab, but other than that I was pretty much left to develop things on my own.

By AcademicLurker (not verified) on 29 Jun 2009 #permalink

I considered my post-doctoral years to be among the most rewarding and exciting of my scientific career. My mentor and I developed the projects I worked on in a completely collaborative manner.

One important factor in the biomedical sciences in terms of intellectual freedom is whether the post-doc is paid from the mentor's research funding, or rather has her own fellowship. The latter can dramatically increase the freedom of the post-doc.