I do not have any particular situation at hand but I am about to start in a new short-term (research assistant) position. The thing is, I will be working on a topic that is very controversial. There are papers (appearing in top journals) explaining how my supervisors and their colleagues got it all wrong. On the other hand, there are reviews (in top journals also), summarizing how many important insights have been learned, and how things are progressing very nicely. It appears to be a kind of battlefield.

I have not yet started in the position, but I have been reading several articles about the research topic. I already know that I will disagree with my supervisor about many things. There will probably be difficult conversations. If we get everything right, maybe we could resolve some of the major issues, by discussing both perspectives thoroughly.

I think I will need very good negotiation skills. If we have exact opposite views with my supervisor, I must do what he wants, but I also want to do what I want. I want us to succeed. How do I prepare for this? How does one convince her supervisor that he is wrong about something? Of course, I will be wrong many times as well. So I should also think how to quickly realize that I am wrong about something.

  • 27
    Maybe you shouldn't have accepted a position which didn't align with your views? Arguing with your supervisor about many things sounds very counterproductive. Commented Aug 23, 2016 at 9:29
  • 2
    Could you maybe say a little more about the field you work in and what it is you're disagreeing about? If this is a scientific field and the disagreement is over some theory/method that your soon-to-be supervisor is a proponent of, then having someone involved in the project who is more sceptical can be an advantage
    – Ian_Fin
    Commented Aug 23, 2016 at 10:18
  • @101010111100 as long as I don't think my supervisor is morally wrong, I don't think I should avoid working with him on disagreements alone. In fact, the combination of our skills and perspectives can be much more productive.
    – Davidmh
    Commented Aug 23, 2016 at 11:44
  • 2
    @Davidmh Nobody is talking about morally wrong. It's about working with a person on a project with which you don't agree academically (for one reason or another). Disagreements are fine, but arguing, and often, with your supervisor, who is also supposed to be your mentor, is a recipe for a bad time. Commented Aug 23, 2016 at 12:09
  • 4
    @101010111100 Depends on what you mean by "arguing". If by "arguing" you mean "at loggerheads, going at it hammer and tongs", then yes, it's a bad time. But if you mean "present rational arguments in the spirit of open academic discourse", then it can arguably (cough) be better than blindly agreeing. It's all about how you approach the disagreement. Is it an "us vs. them" situation, where the goal is to prove your opponent wrong? Or is it a cooperative situation, where the goal is to discover which perspective is right? If either side is pursing the former, it's not going to work.
    – R.M.
    Commented Aug 23, 2016 at 20:17

3 Answers 3


The whole point of scientific debate is to refine our knowledge. Go into the discussion with a mentality to learn, not to win. Keeping this in mind will help you keep the discussion not adversarial and not emotional. On unresolved disagreements, try to find a way to falsify either hypothesis. If it is not possible, consider that which option is correct is probably not that important.

Listen to what they have to say about the criticisms, maybe they disagree with them, maybe they agree (and maybe even that is why the brought new people in). Probably, it is somewhere in the middle.

If (or when) your relationship is good, and you have a good understanding of their position, you can even propose playing devil's advocate.

Of course, I am assuming that they are decent people and not tyrannic professors. You should assume the same until proven otherwise (and if so, consider if you want to shut up and obey, or walk away).

  • 15
    +1 excellent point: "to learn, not to win"! Some supervisors like critical students, some consider them a burden, as they take time. Which one is yours? Frankly, a good supervisor will not like a yes-woman, but sometimes progress needs to be focused on and discussions wrapped up. The discipline of switching from controversy to operational research (with or against the mainstream) and then, on reflection, to switch back to controversy when required, is difficult to master, both for student and supervisor. Commented Aug 23, 2016 at 15:39

Do not go in head-first, do not argue the main point. Your professor is comitted to his side and the more you argue logically, the more he'll dig his heels in.

Look for small commitments. Don't ask him to jump into the water, just ask him to dip his toe in.

Here's an example: in the 90s, many people were vehemently anti-Apple. There was a huge Microsoft/Apple divide, and there were great religious debates going on. How do you convince someone to switch? Not by extolling the virtues of the system. Because the harder you argue, the more they'll dig in. The way to convince them is to buy them an iPod shuffle for their birthday. That's a small commitment. A tiny little step that forces them out of their black and white worldview. They can keep their Windows laptop, and its ecosystem, but one tiny device in their life now comes from Apple. And suddenly the idea of owning Apple products doesn't seem so inconceivable anymore.

This is easiest to achieve if you don't have the religious debate beforehand, so they won't be suspicious of you. You have the luxury of planning your strategy before your professor knows your position. So avoid the religious debates at all cost, and take the position of not having a position. Then look for the tiniest little commitment you can get him to make: maybe that the research is not conclusive yet, that in certain cases, his method is not optimal, that it would be good to have a constructive discussion with the other side, etc.

I actually had this with my supervisor, and we managed to change his view towards the end of the project . Here are some more tips that will make your life easier:

  • Maintain an agnostic position. Without undermining him, discuss your view of the research at all times. Let him make his ambitious promises, but don't go along if you don't agree.
  • Be up front about your agnostic position. You are going into the research with an open mind, and you may not end up seeing eye to eye.
  • Find an ally. If it's you versus him, the power differential will make it difficult to maintain your position. If there's a postdoc arounds who can also put a dissenting position forward, the whole thing becomes a open discussion.

Basically, you're in a difficult position. You also have a PhD to finish, and that's a difficult task even in supportive surroundings. However, you're going in with the right attitude. Your aim should be to take no sides, and to investigate the question fairly, and to ultimately resolve the debate. If you manage that, you will have made a solid contribution to the research, and your career will be off to a good start.

  • 2
    You won't make a MS Guy happy with an IPod! I hated having to install iTunes, just to get some Musik on a MP3-Player. Commented Aug 24, 2016 at 8:21
  • @MichaelStoll Well, the aim is not to make people happy, the aim is to get a toe across the line. You did install iTunes, didn't you? Commented Aug 27, 2016 at 12:27
  • Sure, I did ;-) Commented Aug 28, 2016 at 12:41
  • @Peter: And I didn't! M$ is bad enough. iThings? Hmph. =P
    – user21820
    Commented Apr 19, 2018 at 8:36

You might want to have a look at the Toulmin Model of argumentation: https://web.cn.edu/kwheeler/documents/Toulmin.pdf. While it's not necessarily a good tool to use during conversation, it will (over time) allow you to recognize when arguments are lacking, and be able to pinpoint what's missing (both in your own arguments and those of others).

You must log in to answer this question.

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