Negotiate authorship before collecting data


Authorship disputes are not uncommon. Even when there are no actual disputes over who did what on a project, there may be lots of authorship resentments. That’s because a lot of folks — by no mere coincidence, junior scientists more often — end up not getting as much credit as they think they deserve when a paper comes out.

I’ve never been involved in an authorship dispute — at least not to my knowledge. I was once on a paper with many many people and other authors had concerns about the sequence and why it wasn’t entirely alphabetical, but I didn’t pay much attention.

There is no way that your position in a series of names can appropriately represent what you actually did for a paper. The way we assign credit in science is outdated. Careers depend on the credit associated with authorship, and with the exception of single-authored papers, any list of authors is bound to be unfair to someone. (And, come to think of it, I bet many single-authored papers probably have credit disputes, too.)

But this system is the one we have for the time being. So if you want credit for your work, then authorship is where you get it.

One way to minimize disputes is to have a conversation about authorship and roles upfront and negotiate an agreement before any serious work happens.

Negotiation sometimes gets a bad rep. A lot of people think it means fighting for as much as you can get, and then seeing who wins the argument. Certainly, if you approach negotiation positionally, it can end badly, perhaps for both sides.

Negotiation is more successful as a principled discussion, in which both parties create an arrangement that mutually meets their interests. (The 1981 classic, Getting to Yes: Negotiating Agreement Without Giving In, describes principled negotiation and explains why it is better than the positional kind. The book is a valuable read even if you don’t have anything that you plan to negotiate immediately.) You and your collaborators need to discuss your mutual interests as well as how the work will be conducted and what authorship will look like. It’s a bad idea to start working on a project without this agreement. If you decide the authorship arrangement won’t be worth the amount of trouble you would be putting in, drop out of the project.

Of course, there are imbalances of power in any collaboration. Powerful people may attempt to take more credit than they are due, and less powerful people can get the other end of that stick. The negotiations may be unfair, or subsequent unfair decisions may alter a negotiated agreement. Upfront negotiations can’t eliminate the possibility of malfeasance but can establish a clear arrangement, so that any violation can be clearly documented.

A case in point is one of my own paper I’m working on now. It has about five authors:

  • Me, as the one who came up with the idea and got the grant.
  • A grad student in another lab who collected some samples.
  • The PI for that student’s lab.
  • My student who did lab work for the samples and some bibliographic work.
  • And a researcher who did some of the analyses.

There are probably several different authorship sequences that we could propose, all of which might seem fair. We found one that we all agreed to. If any one of us ends up having to do more work than that person initially agreed to do, then we can renegotiate.

It’s OK to change things by mutual agreement as a project progresses. For example, I’m working on another paper with a coauthor. The project was my idea, and I did the labwork and the fieldwork, I got the funding, and I was going to write it up. I brought in a collaborator, who has the expertise to analyze the data in a particular way. As the project evolved, we learned new cool things, and decided to expand the scope, and my collaborator spent a lot more time doing more analyses, building the figures, and assembling the results. At that point, I mentioned, “You should be first author on this.” He definitely could have brought it up himself, saying, “Hey, if I’m going to be doing any more on this, then, could we revisit authorship? I’ve been done enough work to be second-author, but I’m about to pass that point, so what do you think?” I suppose we could go co-first-author. But that’s a bunch of silly nonsense so I’m fine not being listed first.

It might seem silly — or like you’re getting ahead of yourself — to have a formal discussion about authorship before any data exist. But really, that’s the only sensible time to have that conversation, if you have any hope of avoiding unfair surprises. If you’re working with less-experienced scientists, they might not be aware that these authorship discussions are standard practice. Negotiating authorship at the start of a project is something that senior scientists pick up from experience. We need to make sure that new scientists learn this early on in their career — and not as the result of a bad experience.

If you’re working on a project, and you don’t know the plans for authorship, then now is the time to have that discussion. If you’re working with people who do not want to be transparent about authorship, you probably shouldn’t be working with those people. If someone doesn’t want to talk about authorship, the most parsimonious explanation is that they’re trying to hide something from you.

If you’re a senior scientist and there are junior scientists in your group, does everybody know what the authorship arrangements are? My students and collaborators will (or at least, should) have zero surprises when it comes to who will be on the papers related to the work that they’re doing. Can you say the same thing about what happens in your lab?

Note: This piece ran in the Vitae section of the Chronicle of Higher Education last week.


10 thoughts on “Negotiate authorship before collecting data

  1. Why was the PI of the student who collected the samples on the list ? – just being someone’s boss seems pretty tenuous grounds for an authorship.

  2. Couldn’t agree more, Terry. People really shy away from these discussions, but a little bit of awkwardness of front can save a ton of awkwardness later on. One of my own students brought me up short about authorship expectations a while back, and I’m glad she did – I’d rather be on the same page right away.

    For what it’s worth, I have a whole chapter on “Managing Coauthorships” in The Scientist’s Guide to Writing ( It includes some tips on making the authorship conversation a bit less awkward, especially for early-career folk.

  3. To answer your question, Peter: Because the study design was built on a framework developed by this person, because of intellectual contribution, analysis, writing, and editing. If I am in a position where I have to choose whether or not this person deserves credit, this person deserves credit, a no brainer.

    • Fair enough – his inputs were a lot more than just being the PI.

  4. A PI is doing actual research (e.g., conceptualization the problem prior to data collection), otherwise it sould not be called a PI and of course not be an author.

  5. “A PI is doing actual research (e.g., conceptualization the problem prior to data collection), otherwise it sould not be called a PI and of course not be an author.”

    Not always according to many students and postdocs who complain about their PIs not being involved in their work.

    To clarify why I commented earlier; four out of five of the authors had their contribution explicitely mentioned, only the PI was (apparently) there just for being a PI. Since the post was about who gets authorships and where they go on the list, it struck me as odd. Terry has subsequently expanded on the PI’s role, and obviously he/she did more than enough to be on the list.


  6. I consider this as an easy example. But remember that data and analytical tools may be reused and ideas expanded in a new project and this may be not planned and therefore negotiated beforehand. The most problems with coauthorship I experienced in such cases. Sometimes people supposed that they deserve more credit sometimes people refused to be part of new project… It was always solved but bitterness persisted as well

  7. “If you decide the authorship arrangement won’t be worth the amount of trouble you would be putting in, drop out of the project.”
    That’s an excellent bit of advice for nearly everyone. BUT. What about a very inexperienced, very junior person? I’m thinking of a new Master’s student, invited onto a project after initial discussions, near the beginning of their degree. “I’ve got a great project for you to get involved with! We’re having a meeting on Monday.”

    This hypothetical junior person is in an awkward position. They don’t have the experience to recognize too much work for too little gain, they probably don’t know much about the other people involved, and how are they going to say “I would rather not participate in this” to their new boss? Who has just given them what looks like a pretty good opportunity? And what else is that person going to do if they say “no” to this project?

    Professors are famously busy, so a prof saying “no” to a project is doing so with another, better use of the prof’s time in mind. A really new grad student might not have anything else of comparable value to work on (besides “read until your eyes bleed, then read some more”), so it’s not like this hypothetical person will have a ready answer for the question “Why not?”.

    I’ve never read Getting to Yes, does it cover instances of large power imbalances like that?

    • This hypothetical new MS student should be able to ask, “if I do X and Y, then this would mean a middle coauthorship, I imagine?” And if that person’s PI doesn’t have their back, that’s a huge red flag.

  8. I’d be curious to know what percentage of scientists have actually had a serious authorship dispute. It seems like I know more people who have encountered research fraud than I know who have had a serious authorship dispute. The way I was trained, authorship decisions should always rest with the first author (who should always be the person who did most of the physical work) and the senior author (who should almost always be the funding author) together. Other people involved are free to make authorship suggestions, but with the understanding that the final decisions are up to those two people, with the wishes of the funding author usually weighted just a little more heavily. With this approach, I’ve published 60 papers and 10 chapters and can count on one hand the number of discussions I’ve had about authorship, and every one has actually been positive (e.g someone requesting that someone else be made an author instead of or in addition to them personally). But maybe I’ve just been lucky?

Leave a Reply