Last week I was in Dallas, at the annual March meeting of the American Physical Society, attended by members of the Division of Condensed Matter and Computational physics. For different reasons I ended up spending quite a bit of time talking to postdoctoral fellows (postdocs). Some of these individuals are former graduate students of mine; others are current collaborators, but the majority I had never met before.
They simply introduced themselves to me, more or less as follows “Hello, my name is so-and-so … I am a postdoc, read your blog, and I just wanted to say that, while I generally agree with you…”, and then politely proceeded to articulate their reservations over some of the views that I have expounded in my posts. In most cases we argued our points and eventually “agreed to disagree”, but I did hear a number of intelligent observations, concerning problems that postdocs face, that definitely made me pause and think. I was not aware of some of them, or perhaps I was but had not appreciated in full their seriousness.
I am going to try and summarize here the most interesting points that were made, which in my opinion merit some in-depth discussion.
The Nature/NaturePhysics/Science/PRL-obsessed PI
A number of postdocs to whom I have spoken, expressed to me their frustration with Principal Investigators (PIs) who lose interest in the work of their postdocs as soon as they become convinced that it will not lead to a high profile publication, i.e., in a journal like Nature Physics. In some cases, postdocs are pressured  to abandon their “uninteresting” project to start working on a brand new one, which the PI deems more likely to yield some spectacular outcome.
For a postdoc who has spent a substantial amount of time working on a project, to be told more or less explicitly that that work is no longer worth pursuing, that it would only result in an article in an “ordinary” journal (e.g., Physical Review B for a condensed matter physicist), as if that were somehow a mark of infamy, can be quite demoralizing. More importantly however, a postdoc can ill afford to spend time working on anything not resulting in something publishable . While PIs are definitely right in trying to direct their postdocs to research potentially resulting in a breakthrough, a significant advance (it is the best way for junior scientists to make a name for themselves), at the same time a noticeable gap in the publication record can deal a serious blow to one’s chance of landing an academic job later on. Want it or not, productivity is expected of an aspiring tenure-track faculty, and responsible PIs, genuinely concerned with the future of their advisees, had better make sure that no hiatus occur while a trainee is under their supervision. Contingency publication plans are a must, in case a project cannot be brought to fruition as initially hoped, or if it should fail to yield the expected results or conclusions.
What is a postdoc to do if the PI is not interested in the work anymore ? Should she wrap it up and publish it anyway, possibly without including the PI as an author ? Most emphatically: yes.
My recommendation is that a postdoc in such a predicament, politely but firmly make the case for publication to her PI. A reasonable PI will agree that, even though the final outcome of the research may not warrant the type of publication initially envisioned by the two of them, it is important that some tangible permanent record of that work remain on the postdoc’s research portfolio — search committees pressed for time, noticing a two-year period without any publication in an applicant’s CV, may well discard that person’s application based on that reason alone. That harsh reality must override any consideration of glamour of the publication.
A PI may still resist the idea, but it is important to remember that a PI is not entitled to co-author all articles written by her advisees — proper funding acknowledgment at the end of the article suffices for grant renewal purposes. If the PI cannot be convinced that publication is still worthwhile, while the postdoc still regards the work as publishable, she should request access to the resources needed to complete it (if necessary), and submit it on her own.
The revenge of the angry PI
One thing that is clear, to anyone talking to a postdoc or reading a blog written by one, is that the person is generally worried about keeping a good personal relationship with her PI. This is not just to ensure a smooth, trouble-free day to day interaction with the person, but also (and more importantly) in order to receive later on all the support that the PI can provide, aimed at promoting her junior colleague among her peers, for a permanent position. Postdocs worry that a PI harbouring resentment toward them, over some scientific or personal disagreement, may undermine their professional future by writing harsh, disparaging letters of recommendation, or offering negative (possibly unsolicited) assessments of their ability in informal settings to individuals in charge of hiring.
On the one hand, there is no denying that an embittered, vindictive PI has plenty of opportunities to derail the career of a current or former junior associate. In fact, there is not even a need for openly criticizing the person on a letter — simply using a subdued, neutral tone is enough, these days, given the fierce competition that exist for entry-level research jobs. A lukewarm letter, conveying anything less than unconditional, whole-hearted support, is often interpreted as lack of endorsement for the applicant.
However, one should not make too much of the above scenario. Postdocs should not be constantly walking on eggshells, “put up and shut up” and do whatever the PI says, even when it makes no sense and/or does not ostensibly serve their best interest, for fear of future reprisal (“What will they say about me ?”). My personal experience (I am in this business for over two decades) is that most PIs are better than that, and when it comes to assessing someone’s promise or talent, they do so objectively and dispassionately, without letting any possible personal animosity toward the person get in the way of acting professionally and responsibly .
Clearly, like there are unreasonable bosses in industry, there are also boneheaded PIs, but in general a postdoc should not hesitate to express her disagreement, or a difference of opinion with her PI, whenever the situation warrants it — especially if the difference in opinion is backed by solid arguments and in-depth knowledge. I think most of us appreciate being challenged by intelligent colleagues with ideas and opinions, and have little or no interest (much less need) for “bureaucrats” who will simply do what they are told.
Postdoc evaluation forms
One of the most interesting remarks that I have heard, is that there is currently no formal venue for a postdoctoral fellow to express her opinion and evaluation of her PI, possibly after the end of the appointment. I am not sure how one would go about doing it, but it seems as if, much like we have undergraduate students evaluate us as course instructors, postdocs should be able to communicate to granting agencies their opinions on the PI’s effectiveness as a mentor. In turn, granting agencies should keep a record of evaluations offered by former postdocs, and in part base upon them their decision to renew grants and/or increase or cut down funds assigned to that PI, earmarked for support of future postdoctoral associates. Obviously, confidentiality would be an important part of the process, and it seems that ensuring the anonymity of the postdocs whose opinion is sought would be the most serious problem to overcome, given the relatively small number of postdocs that a PI supervises during a period of, say, ten years. Still, it is an issue with which the private sector deals on a daily basis, and there seems to be no reason why it would not be workable in academia.
Conceivably, funding agencies could make available to PIs the content of anonymous opinions of former postdoctoral advisees, in order to provide feedback to the PIs, only after a sufficient number thereof have been collected (i.e., five or more), in order to make it more difficult for the PI to identify the author of a particularly negative comment. Moreover, anonymous evaluation forms could be returned to PIs a few years after the end of the postdoctoral appointment, making it less likely for the former postdoc to be still in a vulnerable position. I am sure that there are other potentially delicate aspects of which one should think, but I do feel that some “official” mechanism should be put in place, for postdocs to help PIs be better mentors, as well as junior colleagues to make more informed decision (yes, I do believe that evaluations should be made accessible to a young scientist considering a postdoctoral offer from a given PI).
 PIs can apply pressure in more ways than just verbally. A postdoc told me that his access to resources needed to complete his research project would simply be denied, after that project had suddenly been “unendorsed” by the PI.
 For the sake of clarity, it is important to stress that I am not talking here about a case in which the PI comes to the realization that a research project that may have seemed promising early on, is in fact based on a fundamentally flawed approach, resulting in data of dubious accuracy, or in any case not publishable, not adding anything new or useful to the state of knowledge on that subject. That state of affairs is unfortunate, but this is science — it can and does happen. If the PI is genuinely, honestly convinced of the fallaciousness of the project, (s)he should insist that the postdoc not waste any more time on it, much less publish. No, what I have in mind here is different, I am referring to a situation in which the methodology is sound, the results correct and novel, but just not interesting enough to be published in some important venue, in the opinion of the PI.
 A PI delivering a blistering critique, going over the top in lashing out at an advisee (whether that be in public or in private, verbally or in writing), is more likely to bring ridicule on to herself than cause damage to the young colleague, especially if the PI’s opinion should be at variance with that of several other colleagues, or if the advisee’s record should appear to contradict what stated by the PI. If anything, I have witnessed much more often the opposite, namely reticence on the part of PIs, who omit mentioning in their letters or in conversation, noticeable shortcomings of their trainees — shortcomings that become evident only at interview time. The reason is simple, namely, it is ultimately in the PI’s best interest for advisees to succeed. No one wants to be identified as the “PI from hell”, the person who makes the life of her postdocs a nightmare, badmouths them with colleagues, and ends the careers of promising young scientists.