Ecologists disagree on whether co-authors should agree

Last week I asked what should be done if co-authors disagree on what their paper should say.* My own view is that all co-authors should agree with and stand behind everything in their paper, so that in the event of a serious, irresolvable disagreement, some co-authors would have to withdraw from the paper. An alternative view is that authorship just indicates that you contributed to the ms in some appropriately-substantial way, not that you agree with everything in it. And there might be other views on the matter as well.

I was curious to get a sense of the range of views out there, so I included a little poll in the post. Most of the votes likely are in at this point. As of this writing, about 24 h after the poll went up, we have 149 votes, distributed as follows:

  • 57% think all co-authors should agree on everything in the paper, and withdraw their names if they don’t
  • 33% think authorship doesn’t mean that you agree with everything in the ms, just that you contributed to it
  • 10% have some other view

I suspected there’d be a lot of disagreement on this–and I was right! Obviously, the poll respondents aren’t a random sample from any well-defined population. But the poll certainly suggests that there’s no consensus on this issue among ecologists.

One consequence of this disagreement is that it makes authorship a little hard to interpret. When I see someone’s name on a paper, I assume that they agree with everything in it and stand behind it. Apparently I shouldn’t assume that! Indeed, I can imagine a situation in which some of the co-authors on a paper assume that the other co-authors agree with everything in the paper, when in fact they don’t. I have no idea if that’s ever happened, but it seems possible.

I’m curious if the poll results would’ve been different if you could go back in time a couple of decades. I wonder if the view that co-authors should agree on everything in the paper is declining in prevalence. That is, as collaboration becomes more common (for both good and bad reasons), norms of appropriate author behavior are shifting in ways that facilitate collaboration. Wildly speculating, I really have no idea.

Formal statements of author contribution are becoming increasingly common, as an antidote to changing authorship practices that make ordered lists of authors less informative summaries of author contributions. Perhaps we also need formal statements of which bits of the paper each author agrees with? After all, back when most papers had just one author, you could safely assume that the author agreed with and stood behind everything in the ms. Nowadays, apparently that’s no longer the case. I’m still deciding if or how much I’m kidding about this…

*I actually have no idea how common it is for co-authors to seriously disagree, and how often those disagreements are resolved in one way vs. another. Maybe I should do a poll on this?

13 thoughts on “Ecologists disagree on whether co-authors should agree

  1. These results surprise me too, but for the opposite reason: I admit to being part of that 33% who think authorship merely indicates a substantial contribution.

    I think my choice here stems from how absolutely option 1 is framed. “… agree on everything …” Everything is a lot! I personally think that agreement should be considered quantitatively, like contribution. Just as research teams decide to invite/exclude laborers based on how much / how important their contribution to the research was, individuals laborers should decide to accept or decline the invitations based on how much in a paper they disagree with and how important those disagreements are. I never agree with all the grammatical choices of my co-authors, but that doesn’t prevent me from accepting authorship. Similarly, when I’m writing a paper, I don’t totally agree with all the recommendations of coauthors and reviews, but I choose my battles. My sense is that my tolerance in all these respects will change as I transition through my career, which is now decidedly “early”.

    The biggest problem seems to be that contribution and agreement can be at odds, potentially with very unsavory consequences. Suppose you provide a crucial component of a paper, say, resolving a phylogeny for some group. Since phylogenies can be used for dozens of purposes, it very well may be that you don’t agree a main point of the paper, even if your work was absolutely integral to make it. Suppose you disagree so much that you want to pull your name from the paper. That’s your prerogative, but should you insist the phylogeny be removed as well? If not, perhaps you’re not receiving appropriate recognition for your work. If so, it could be perceived as intellectual blackmail: “change the title, or else!”. Thankfully this has never happened to me, and should hope this approach is not widespread.

    • “The biggest problem seems to be that contribution and agreement can be at odds, potentially with very unsavory consequences. Suppose you provide a crucial component of a paper, say, resolving a phylogeny for some group. Since phylogenies can be used for dozens of purposes, it very well may be that you don’t agree a main point of the paper, even if your work was absolutely integral to make it. Suppose you disagree so much that you want to pull your name from the paper. That’s your prerogative, but should you insist the phylogeny be removed as well? If not, perhaps you’re not receiving appropriate recognition for your work. If so, it could be perceived as intellectual blackmail”

      That’s a great point. I suspect that’s the thinking of many folks who voted as you did. I definitely see where you’re coming from even if I don’t entirely agree. In the hypothetical situation you describe, one resolution would be for you to pull your name from the paper, and get acknowledged for doing the phylogeny. Or else, yeah, you put your foot down and insist that the paper only say things on which all co-authors agree. I wouldn’t call that intellectual blackmail, I’d call that taking responsibility for the paper. But clearly this is an issue on which there’s a range of defensible views.

    • Some journals express a preference on this. I had a collaboration: the lead group wanted to use my data to argue for one thing, when the data actually said the opposite, which wouldn’t have made such a big story (they’d used the wrong stats on it). They submitted without my approval (to IF ~10 journals), so I wrote to the editor to point out the error and the manuscript was pulled. Main problem was that two sets of data together were much better than the data apart: collaborators suggested that I pull my name if I wasn’t happy; I didn’t see why I should effectively give them a couple of years of my career for free.

      I suggested that the disagreement be highlighted in the discussion, but the journal said they wanted all authors to agree on all aspects, so we ended up in a death spiral. All got very nasty very fast.

      Paper was eventually published by neutering the discussion so that the data were effectively presented with no conclusions drawn. Horrible all round, really.

      • I bet your story is more common than we realize. I once had a collaborator that apparently either knew little about stats, or was just unwilling to spend anytime scrutinizing them. I had a project separate from her, as well as one we teamed up on. She served as the PI on that one and made the final calls. Despite repeated warnings that her data were not suitable for my approaches, she kept applying them anyway. It was complete nonsense, really. The most fundamental of concepts, like sample size were just thrown out the window, making her outcomes gibberish. That situation also experienced a “death spiral” and ended with me walking away from the entire project- after investing of myself very heavily in it. Despite the emotional turmoil and blow-back, I can see in hindsight I made the right calls, because having my name attached to any of that crap would have been far more debilitating.

  2. I would hope that all co authors agree on the methods that were used and the results that were reported. It isn’t clear to me that everyone should have to agree on interpretation, implications, significance, and other more subjective issues.

  3. I once had a significant disagreement with two co-authors concerning whether or not a particular figure should have been included in a paper. I thought it beautifully illustrated the expression of Fox-A1 in the adult mouse urogenital system. They thought it was ghoulish, because it was a photograph of a completely dissected urogenital tract. It was a disagreement that did not go away swiftly. As we worked on the manuscript- I kept putting it in, and they kept taking it out. I felt like I was being “picked on” in a way- because it was my experimental work, & I was darned proud of it. The backin’ and forthin’ went on for several months. We finally reached a compromise when I included a photo of just the region where the ureter attached to the bladder. Case closed- we all went home happy.

    I’ve been involved with other pubs where the debate became quite intense, & I think that is a very good process. So in my experience, disagreement is common at the early stages of ms development, but in the end, everyone reaches a point of agreement. If not, then I think you should not put your name on it. The risks can be high.

    Case in point: Fairly recently, I had to INSIST that a prior colleague ELIMINATE my name from all aspects of a project. No authorship, no acknowledgement, no nothing. When she did not comply, I had to go up the food chain to the Provost & DEMAND my name be retracted from EVERYTHING. Thank goodness I pursued this, because she went on to repeatedly publish reams of fabricated data. Can you say OUCH???

    An ounce of prevention is worth a pound of cure…

  4. You might be interested in a famous story from paleontology about THE “consensus” paper, Sepkoski et al. (1981). You can read a nice retrospective by Arnie Miller on the origins of that paper and how/why “consensus” was forged. The relevant sub-section is appropriate titled “forging a consensus”.

    Miller’s story actually understates the level of disagreement. My understanding is that even to this day Raup remains skeptical of any Cenozoic increase in marine diversity, while Valentine still believes in an order of magnitude increase – though don’t quote me.

    Anyway, that paper dominated everyone’s view of marine biodiversity for almost thirty years, but neither of the key authors fully agreed with it!

    • Very interesting!

      I can share a similar story from ecology. Back in 2002 there was a major consensus paper on some at-the-time controversies in biodiversity-ecosystem function research. I have it on good authority that at least one of the authors of this paper later backslid, giving talks that proceeded as if the consensus paper had never been written, much less that he’d co-written it.

      And there’s now debate between Peter Abrams and Lev Ginzburg as to what their “here’s where we agree, here’s where we disagree) paper on ratio-dependent predation actually agreed on.

      In light of these examples, I’m now wondering if there are *any* examples of “consensus” papers for which all the authors actually agreed with the consensus!🙂

  5. Are people willing to compromise on author control over content in order to publish in certain journals? Would your views differ depending on whether the manuscript was going to Nature or Naturwissenschaften?

  6. I’ve served on a couple of high-profile committees where we had to publish book-length reports that reflected the consensus views of a large and diverse committee.

    It’s not always easy! At one level, you simply have to trust colleagues who know things you don’t know. For example, on the anthrax-letters committee, our report covered approaches, evidence, and interpretations from both the physical and biological sciences.

    On these committees, we were also given the option of writing dissenting opinions that might be signed by one or more of the committee members. Though we were strongly encouraged to avoid taking that path, I think it provided an important option. Even short of using that option, it gave everyone the opportunity to argue for the inclusion of, say, some alternative perspective or hypothesis.

    In the same way, I would think that coauthors should, in many cases, be able to find ways to express (at least some) disagreements by writing, for example: “On balance, we think our data support the hypothesis that … On the other hand, we also cannot exclude the possibility that … because, with hindsight, our experiments lacked a control for the effect of …”

    In extreme cases, perhaps editors might even allow authors to have short sections where one or more co-authors briefly and clearly articulate their dissenting opinions. That’s obviously a last resort (as it was with the committees on which I served), but in these days of open science, it seems like an option that might — occasionally — be used. It would allow both the recognition of all the authors contributions to the main body of the paper yet, at the same time, the reasons that some authors might disagree with some critical interpretation or implication.

  7. Thanks JF for addressing this important question. I am about to enter this exact situation and have been discussing with numerous coauthors what we are going to do when the shit hits the fan.

    Background: a large international group put together a massive regional dataset over several years. The team leader is highly influential and stubborn and stories abound from past coauthors of various Science papers of his refusal to change anything in the paper. After the fact, i.e., after receiving authorship on a Science paper, they claim to not agree with “all that was written”. I’ve been tough on these coauthors, but am now about to experience the same situation. Do I stay or do I go? We already published a report from the dataset and literally hundreds of pages of corrections, edits, suggested changes to analyses, inferences, etc. were ignored.

    Basically, the lead is using our data and project as a platform to push the same paradigm he’s been pushing for decades (as old-white-dudes-in-power tend to do), even though the data clearly don’t support it. He has told us he is about to send us the “Science manuscript” and that we will have two weeks to make edits and agree to be authors. The reasons many of us haven’t already pulled out are that: 1) We provided all the data! why should we bail? 2) We hope (unrealistically) to change his mind and publish some sound science with our awesome data, 3) Hey, it’s gonna be a Science paper…

    I tend to still lean towards Jeremy’s views, but it would be a pisser if I have to withdraw after contributing a major and essential portion of the data, working for years on the project, and also because the paper is on a topic that I’ve done tons of work on and the lead has basically never touched (which is part of the problem). To make matters worse, and this is what I really want help with – maybe a new poll? – the lead author has so far refused to share the data and code with us coauthors! Yes, you read that right. And recent queries from several team members indicate he NEVER plans to release the data EVEN TO THE COAUTHORS!!! Totally crazy right? If I read Science’s data sharing policy correctly, I believe you have to share the data, i.e. all the raw data not just summary data, with ANYBODY who asks. I kind of want to stay on board to see how the Science editors handle this. And the reasons we want the data set are: 1) that was an explicit agreement when we develop the project and agreed to share our own data, 2) we suspect an error has been made in the analyses, 3) we need to see the r code so we can figure out what the heck was actually done.

    • “it would be a pisser if I have to withdraw after contributing a major and essential portion of the data, working for years on the project,”

      This raises an important general point that hasn’t come up yet in this thread. One important guard against this sort of awkward situation is good communication among co-authors from the get-go, including up-front discussion (before anybody’s invested very much in the project) about what constitutes authorship and how any potential disagreements on ms content will be resolved. Precisely because it’s much more difficult to resolve this sort of situation after everyone concerned has put in a lot of time and effort. Not saying that would’ve necessarily prevented your collaboration from ending up where it’s at now; just making a general comment.

      Yes, it’s very odd for a co-author not to share data and code with other co-authors. Never heard of that. Personally, and emphasizing that I don’t have any other information besides what’s in your comment, that would make me very nervous as a co-author. Especially if it’s contrary to what was previously agreed or if I had reason to suspect an error. Just speaking generally and personally, I wouldn’t want my name on something I didn’t completely trust.

      Sounds like you’re doing the right thing staying in close coordination with your co-authors.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s