I'm thinking about leaving my current lab, since it's moving to a field that I'm not really interested in. I am thinking of contacting other PIs that work on topics that are a closer match to my research interests, to see if they are interested in starting a project together.

I am concerned that if I talk to my current PI before taking any steps, she could bad mouth me to prolong my staying in the lab, and to prevent leakage of the lab's knowledge. So from this point of view it would be my best interest to inform her as a fait accompli when I find a new lab (while still giving proper notice).

I still have some reservations about doing things behind her back though. I would like to leave on good terms, and since academia is such a small community doing this could affect my reputation.

  • 1
    As a practical matter, depending on academic culture in your part of the world, it might be very difficult to get another postdoc without a letter of recommendation from your current PI. I'm not sure why she'd be worried about "leakage" though; after all, a postdoc isn't a permanent position, so you're going to leave eventually anyway. Commented Aug 23, 2016 at 15:29
  • @NateEldredge - the 'leakage' issue would seem to be an impression from the OP, not an actual fact. How long has the OP been a post-doc? How long will it take for the lab to 'move' to a new field (what does that mean, anyway)? Why do post-docs seem to have such bad relationships with their supervisor? More at eleven...
    – Jon Custer
    Commented Aug 23, 2016 at 18:04

2 Answers 2


In my experience, people who have been in a field a long time all know each other, even if only peripherally. If you do apply to another position, there is a real possibility that the the new PI will simply call up your current PI and ask about your qualifications etc.

To me, it seems completely reasonable to tell your PI exactly what you say in your first sentence. You could even ask for her help -- if you are this uninterested in your current position I doubt she will go out of her way to try to retain you. (Also, if you are indeed looking to move to a lab that does not compete with the current one, can she legitimately be concerned about intellectual property "leakage"?)


Three major points:

First, either you are very paranoid, or you are working in a very unhealthy situation. I want to highlight three things that shouldn't be considered normal: 1) You're worried about your advisor lying in your recommendations, 2) Your advisor is trying to restrict, not disseminate their techniques, and 3) You seem to think you won't be supported in your interests when the lab goes in a new direction. All of these argue that you're not in a healthy spot.

Second, what you are proposing is possible. People can get postdoc offers without advisor recommendation letters. [disclaimer: in many cases, I don't know if there was an informal call - but I don't think so.] It helps to have a strong CV/case for what work you want to do, and how it links to your new PI's plan. Some of it is also luck - some PIs just always ask for letters, and if they do, your explanation is not going to sound great unless your current advisor is well-known as an asshole. (The other possibility is to lean on your own network: people who already know you as a scientist will be willing to skip an advisor's letter.)

Third, it is possible you will burn some bridges by leaving. Having seen this from the side of people left behind in the group, the less notice you give and the more projects you leave unfinished, the more likely you are to burn bridges. However, unless your advisor is particularly spiteful and well-known in the field you want to work in, this probably won't follow you around: academia is a small world, but it is also insular, and big wheels in one field are unknown in another.

Basically: assess if you are really in as unhealthy a situation as it seems, and if you are, it might be worth burning a few bridges to get out of there.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .