Authorship is weird. As an instrument to attribute of credit, it’s far too coarse whenever the number of authors is greater than one.
Authorship is a slippery concept, because you can’t really define what constitutes a substantial contribution, in a way that can apply generally.
What merits authorship? That’s a big question, with a necessarily complicated and messy answer. Lots of folks have come up with their own answers trying to make this as clear cut as possible, coming up with definitions and matrices and essays to argue for a particular set of criteria. Regardless, it appears that people don’t follow these criteria even when specified by a particular professional society. The more people study how authorship gets decided, the more they realize that it’s a nebulous process.
Defining who gets authorship is a serious and emotional issue for scientists, because this is the currency of success that makes or breaks job applications and tenure files. Depriving a scientist of an earned authorship isn’t just about ego — to junior scientists, it’s an actual harm. Likewise, when authorship is unearned, it’s because power has been levied to take advantage of other individuals.
The way universities evaluate scientists for tenure, a lot of folks feel that they need to rack up their names on as many publications as they can in their tenure-track years. And “senior authorship” (even though it’s not a convention that everybody understands the same way) is often critical, depending on your field and where you work.
This has led to a culture where some people expect that a the advisor of a graduate student should, by default, be an author of every paper published by a student in their lab. I said a thing the other day which provoked some strong reactions:
“I just saw a sole-authored manuscript written by a grad student. I can’t remember the last time I saw one of those. It’s so nice when advisors don’t add their name to papers when they didn’t earn authorship.”
In response, some guys were telling me that it wasn’t possible for a grad student to do a science independently enough of their dissertation advisor to be a sole author. Which as a simple matter of fact is wrong. Of course a student can do work independently enough so that they should be entitled to sole authorship. (If you wanted to fish through the replies, you’ll see plenty of examples.) So, why are these reply guys getting it wrong? They assumed their own limited experiences were universal. They projected their own expectations, and priorities onto others, who are working in different environments, in a community with different norms. Sure, I can imagine that in some specialities, and in some labs, all of the science might be so heavily collaborative that even a single paper would require a joint effort and large resources. But there are a lot of fields where this isn’t the case. And, of course, in any field, a student could write a meta analysis or a review without requiring enough contributions from their advisor that would amount to authorship.
I know far too many grad students who have been expected or require to include their PIs as senior authors on their papers even though these people made the slightest of contributions. In my field, if you look through journals from a few decades ago, it’s clear that authorship by the advisor was not the default at all, and that dissertation papers were often sole-authored by the person who did the work. Things are different nowadays.
Many have said that this is because science is more collaborative than it used to be. I do think that dissertations involve more collaboration than they used to, but I don’t think that’s entirely responsible for the cultural shift in the expectation of senior authorship by dissertation advisors. Senior authorship used to not be a thing in my field, and now it is. It’s still not a fair assumption that the last author on a multi-author paper is the “senior author,” though I regularly see folks jumping to this conclusion. Collaborations have many manifestations, and when you see a paper with a lot of authors, you can’t really know who did what unless you see a statement about this in the paper or you talk to the authors.
So, how can a grad student’s advisor could earn authorship? Let’s consider the different ways this might happen.
1. Just being the head of the lab? Should someone be the senior author just because the person who did the work is a member of their lab? Without doing anything else? That sounds more like a divine right of PIship, rather authorship. Which means that you had an actual hand in creating something. If you’re one of those people who think that PIs are entitled to authorship simply because it happened by someone who is under their authority, then I imagine you must love feudalism. I’ve known labs where the PI didn’t even know about a manuscript until it was ready to be submitted. Because they weren’t involved in it at all. And they expected to be an author? That’s absurd.
2. Paying for it? Should the advisor be a senior author because they paid for it? Well, that depends on what you mean by “paying for it.” Did the PI write a grant that directly paid for expenses associated with the work, and the work is conceptually tied to the work described in the grant, then yeah, that sounds like senior authorship might make sense. If the student conducted the research using the time funded by a research assistantship that was provided by their PI, then perhaps. If the student is being funded by a fellowship, or a TAship, and “paying for it” doesn’t make sense in terms of salary. If the PI simply is the one is simply who brings in the dough for a lab, and the only way the project was connected to the lab was that the student who did the work is a member of the lab, then that’s not paying for the work. That’s feudalism.
3. Providing guidance and advice? If the advisor helped steer the student’s project by giving them ideas, guidance, feedback, and support, shouldn’t they be senior author? Well, it depends. If the actual idea for the project came from the PI, then yeah, they should be an author. If the PI wrote sections of the manuscript, did some non-trivial analyses, provided huge insights that altered the paper, designed the figures, or stuff like that? Then yeah, they should be author. But what if the PI simply gave some helpful tips at the outset, and then provided remarks on the draft like any other friendly reviewer would do? That’s not enough to earn coauthorship? What if they provided some good ideas when the project was in development? What if they let the student use equipment or supplies? Well, this is the subjective grey area. There are a lot of projects where the PI is there for helpful advice, but really doesn’t have a hand in directing the project or doing any of the real work. If the student developed the idea, and the student did the work, and the student wrote the manuscript, and the cost of the project was principally managed by the student, and all the student got from their advisor was some helpful feedback? I don’t think that the student’s advisor should be a coauthor.
4. Being an actual collaborator? Did the student’s advisor actually collaborate on the project by making clearly substantial contributions in conceptualization, design, implementation, analysis, and/or writing? Well, then, yeah, coauthorship for the advisor makes sense.
Do you think it’s really not possible for a grad student to develop their own project, get it done, and then write it up, all without substantial contributions from their advisor? Then you need to expand your worldview.
I thought doing a PhD was supposed to be about developing scholars who are capable of independent work. PhD students are supposed to be capable of writing papers on their own, right? There are many circumstances where contributions of a PI are on par with what you would expect from a peer reviewer. Yet it’s very rare for peer reviewers to be added to the manuscript as authors!
When I raised this issue, I was very surprised to discover some downright unethical conduct by journal editors with respect to PI authorship. Apparently, when some editors receive a manuscript, the editor will google the first author to discover that they are graduate students and then find the identity of their PI. And then they will contact the author of the paper and ask them to add their advisor as an author — as if the editor would know who deserves authorship! And then, the journal editors will directly contact the advisors of the author of the paper, and require them to affirm with a signature that the students worked independently before sending the paper out for review! I think this is outrageous gatekeeping. Sure, I suppose it will rarely happen that a rogue student will submit a paper from a collaboration to steal credit away from those who legitimately earned authorship. But for every time that happens, I imagine there are many more times where PIs rob junior scientists of credit. Nevertheless, I don’t see journal editors writing all the postdocs and students and collaborators of every PI to makes sure that they were unfairly excluded from a manuscript! It’s wrong to assume that a graduate student would be depriving their advisor of credit, while giving PIs the benefit of the doubt, even though we know that it’s power that corrupts. With editors attempting to force authors to gift authorship to senior faculty, how is this not racketeering?
Let me raise a hypothetical scenario: A grad student has a side project, and they work on it on occasional evenings and weekends. It’s not related to their dissertation. The work for it is not done in the laboratory at the university, and no university supplies are used. The laptop they used was purchased with their own money. They never told their advisor about this project. Right before they are about to submit it, they hand a copy of the complete manuscript to their advisor just to keep them informed. Should the advisor ask or insist to be the author? I see absolutely no valid grounds for them to do so.
From this hypothetical base scenario, what kinds of contributions from the advisor you would need to add, for them to earn authorship? I think everybody draws a different line. Some of those lines are downright exploitative, and some are irrational. Please be sure to think consciously about how you draw those lines — with respect to funding, intellectual contribution, and work contributed — and communicate this with people who work with you.
Having a clear conversation about criteria for authorship before the start of any project should be standard practice. If you’re about to join a lab, it’s a good idea to talk to the PI about their authorship practices. If you’re a PI, you should have your expectations about authorship expressed in writing for all lab members. And you need to make sure that these practices conform to the code of ethics for your professional society.
If you’re counting on publications by members of your laboratory to carry you over the line for tenure, the onus is on you to make sure that you contribute enough to these papers to makes sure that you earn your role. Are you actually collaborating? Or are you just providing a roof for people who are doing their own work?
Terry, I agree 100%
On my first paper (Klawinski et al. 1994. Journal of Herpetology 28:225-230) a group of friends and I did a study on competition in introduced geckos in Galveston, TX. We were one undergraduate student, 2 MS students and one Ph.D. Student spread across three universities. We asked both of the research advisors (Fred Rainwater [Stephen F. Austin State University] and Jim Dixon [Texas A&M]) if they wanted to be listed as co-authors because they gave us a ton of advice and help and we might not have gotten it in shape for publication without their guidance. Both men said that we had done all the lab work and field work, analyses and writing and that they felt they had just served as advisors and asked to be included in the acknowledgements instead. These two mentors served as examples that I think all of us who were coauthors on that first paper still try to emulate to this day. Sadly, Jim is now gone but I know a number of his students and his mentorship is recognized as being exemplary. I just talked to Fred about a week and a half ago and he is still the same thoughtful, generous mentor he was back in the early 1990s in spite of the fact that he is now long-retired.
My comments from Twitter:
http://twitter.com/ManuSaunders/status/1173379749193740289
I would be interested in hearing your thoughts regarding this scenario:
In a desperate move for submitting tenure packet, advisor copy/pastes over 50% of grad student’s thesis and credits themselves as 1st author for journal publication. I’m addition, advisor gives authorship credit to significant other even though significant did not contribute to research. Grad student, isn’t aware publication exists until receiving an email from journal for final revisions. Should graduate student feel grateful for a publication as 2nd author or upset that most of their research was plagiarized by advisor?
In my opinion, it’s legitimate for the graduate student to be outraged at being exploited by their advisor, who perpetrated an act of academic misconduct that, in a just world, would result in substantial repercussions for the perpetrator.
Excellent point about the contribution of peer reviewers being similar to the contribution of PIs. It’s sad that anyone would consider performing expected service to the field without getting special credit for it as being “generous.”
I think this raises a lot of crucial and important points but I also think it requires an institutional level change because otherwise you run the risk of cutting the feet out from under junior faculty. If everyone expects that the lab PI is on every paper that comes out of their lab, if that’s what constitutes “authorship”–and I think that’s unfortunately rather common, although I am admittedly currently just a postdoc–then if you have junior faculty who are following different criteria, their tenure package will not be competitive. Therefore this is something that needs to be spearheaded by departments and senior faculty (not that I think it’s something that shouldn’t be done, just that it has to be done at an institutional level.)
There were a number of interesting points, concerns, or critiques that have popped up on twitter, that I thought would be of broader interest to readers here, so I’ll do my best to recapitulate those points here and respond to them. I don’t want to drag anybody’s identity involving a critique into the comments here (if they wanted to leave a comment on the site, they would have), so I’ll not name names, though this is all public on twitter if you’d like to search it up:
Dr. Saunders mentioned (as she pasted above, thank you!) that this discussion could have a negative consequence on grad students, with an implication that being a sole author or not having a senior author would be something that they might feel a need to do.
-I didn’t intend this in the slightest! I think it’s entirely fine for students do have genuinely collaborative relationships with their PIs, and that’s super cool. I did not intend to imply that not authoring your own papers without your PI is some sort of shortcoming. I don’t think anybody holds this expectation, and because it’s so customary for PIs to be included as authors now regardless of their contribution, I don’t think think this comes with any positive or negative evaluation of a student, as this is just the way things are working now. I wasn’t intending to express concern about incidents where students are being exploited by having names added to their paper that don’t belong on them. When PIs are so hands off they don’t even earn authorship, but insist on it anyway. If your project is an actual collaboration with your PI, that presumably is a good thing for you!
That I’m ‘victim-blaming’ and that PIs who are receiving de facto authorship on papers are not merely feeling a need to rack up pubs, they have an actual need to do so to advance their careers.
-I think this is a good an important point that I didn’t mention. I think the solution to this problem for PIs in such circumstances is not to remove their names from papers, but to build a laboratory where their student projects are clearly a part of their research agenda and they’re involved enough to earn authorship. I think every biomedical lab that I’m aware of operates this way, for example.
That because I said, “Things are different nowadays,” that I think things are better the way they were, and that we should go back to the way it was.
-I don’t think that. I think more collaborations are better. I think in a culture where collaborations are more common, it makes it easier to sneak names into papers that don’t belong there. But going back to the way things were wouldn’t make it better for anybody.
That the code of ethics for professional societies is problematic if these codes are inconsistent with current practices, and asking people to follow these ethical guidelines without looking at the consequences for their careers is costly to these people.
-If we are to agree that these codes of ethics are proper, and that current practices are not in line with the ethical guidelines, then we have a problem that is not easily fixed. I think if PIs are looking to behave ethically and rack up the papers they need, then they need to take on students who are working closely enough with them so that authorship is merited. I think the PIs who are the ones insisting on having unearned authorship are the ones who are also failing to serve their students in a variety of ways, and the thing I’m discussing in the post, when it happens, it a symptom of bigger cultural problems.
That I’m complaining about “an apparently widespread problem.”
-When I wrote this Sunday morning, I wasn’t really imagining it was a widespread problem. I didn’t intend to claim in this blog post that it was widespread. but an occasional one. But looking at the feedback from junior scientists in my inbox and my DMs this morning, I’m beginning to think it’s more common than I thought. And, in particular, postdocs are concerned that their current employers are insisting on authorship on papers from work that was done before joining the lab. That might be a bigger problem?
That I made an unsubstantiated claim of unethical conduct by journal editors who violated confidentiality, and that I should name and shame the journal.
-The post has a link to public claims of this. Because singling out the journal would put a marginalized scientist in a position of further precariousness, that’s not my decision to name the journal or to ask this person to do so.
That I should specify in the blog post the number of times that I am aware of the form of misconduct that I’m discussing in the post, and convince y’all that I performed due diligence to validate the stories of the students or former students who told me about what happened to them.
-I’ve heard so many of these stories over the years that I don’t even remember how many times. In some of these cases, I was involved enough with the parties that I was confident that the PI was in the wrong. I could name plenty of names here, including people who have already interacted with me since the post came out. But it’s not my place to do so because that’s putting the victim at risk, and that’s not my choice to make.
-I should note that I’ve written on here often about more pervasive problems in our field (such as PIs that sexually harass their own students, and professors that inappropriately hit on students at conferences, and peer reviewers that intentional sink manuscripts from competing labs.) But in those cases, I haven’t been asked to name and shame, or to count up the number of incidents. I’m not sure why one would go out of the way to expect a high standard of evidence for certain kinds of misconduct but not for others. You can believe me, or you can disbelieve me, it’s your call. This is just my blog, and it’s at the reader’s discretion to invest whatever level of credence into my words as they feel is appropriate. I wouldn’t say this kind of request for information is sealioning, but I’m not sure it’s entirely in good faith. Because it’s pretty clear that I won’t provide evidence when it means throwing people under the bus. In general, the PIs who have bullied their way into unearned authorship are nasty pieces of work, and definitely the scientists who had the misfortune of being in their labs don’t want to anger them by outing their misconduct. I’m not ready to turn this blog into a tell-all gossip rag, and I rely on my history of sincerity and credibility to help people to decide whether I’m just making shit up.
I have one better for you. My postdoc adviser, the senior editor of a significant scientific journal, regularly gets senior authorship for work he does not conduct or often even understand. He also adds other senior scientists’ names to author lists, for contributions as small as proofreading the manuscript. Whose papers he gets listed on in return.
That’s all relatively ordinary. What is remarkable is that at the same time, he routinely removes the names of junior scientists who did legitimately fulfill authorship criteria, saying that the author list is “getting too long”. Even when the papers first author defends their authorship, junior scientists have to continually (and exhaustingly) fight for due authorship. The moment you leave, you’re no longer physically there to fight so forget due credit. A paper without my name was published recently in the journal this pi is senior editor of. I had done 100% of the data collection and most of the analysis prior to leaving. The first author of the paper is their new postdoc hired 6 months after I left and well after all the data was collected. My lab notebooks must have been used to write methods sections at a minimum, and there’s plenty of hard copy evidence of figure drafts that look nearly identical to those published. They say they re-ran the stats, but the results look identical.
They did not let me know this paper was being drafted or submitted. Now that it’s published and I privately tell them that my name needs to be added to the author list, they justify my name not being on the paper as I didn’t help write, edit or approve the manuscript I didn’t know existed. They acknowledge I did the data collection and they “forgot” my contributions to the “early stages of the project”, but then subtly threatened me with legal action if I make any official complaints… nice, hmm?
And so an editor of a journal he self-publishes in is apparently above the ICMJE authirship criteria his own journal ascribes to. And I can hardly ask the journal editor to investigate. Conflict of interest, anyone?
I appreciate these posts on authorship, and outraged, but unfortunately not surprised at some of these unethical practices described. Fortunately I have been lucky to have supportive and ethical Ph.D. and postdoc supervisors. This includes one sole-author dissertation publication, and another one that excluded my advisor (with his support).
The main reason for me commenting is a bit off topic, but as a new (2nd yr) Asst. Prof at a SLAC, I’d be interested in reading a future post about your thoughts, Terry, on specifically navigating authorship considerations with undergraduate researchers.
Its hard to imagine a situation at a PUI where a student where I wouldn’t be involved in the study enough to be an author. But my challenge is figuring out when a student has earned the right to be a first author vs myself. I can imagine clear cut examples either way… but the trick is navigating that middle ground. My current projects that heavily draw from data collection/analysis/study design done before I started working with these students, so it seems a little simpler of a solution, but as time progresses the situations will become more difficult.
Adam, you’re in luck:
https://smallpondscience.com/2013/02/13/undergraduate-first-authorship/
https://smallpondscience.com/2013/03/01/undergraduate-authorship-a-formal-model-exists/