Ask Early, Ask Often: "How do you see the authorship on this?"

Jul 09 2008 Published by under Science Publication

The essential currency of the modern biomedical science career is the research publication. By which I mean your name listed as an author on a peer-reviewed journal article. This has always been the case and is not selective to science, the publish-or-perish mantra is common to many academic disciplines.
Nevertheless. In bioscience the current realities of career, funding and advancement put a huge pressure on scientists to accumulate as many authorship credits as possible.
Which means you must ignore conventional or traditional concepts of how things are done, what is polite behavior, whether you are being a selfish snot, etc and discuss authorship early, often and continually with your labmates, mentors, collaborators and trainees.


MsPhD has a post up over on YoungFemaleScientist in which she ponders some of the issues:

...I noticed something interesting: despite having no new first-author publications, MrPhD has now surpassed me in the grand total of publications, because he has more papers on which he is a middle author.
For almost every project in his lab that he has contributed to, MrPhD was made an author. He says they usually tell him up front that he'll be an author.
I, on the other hand, have learned to ask, or risk wasting my time, when I should be working on my own projects. This is true regardless of whose lab wants my help (my own or someone else's).
I am often told I'm being greedy when I ask up front whether I'll be an author.

YHN has the back of just about anyone who wants to discuss authorship in a legitimate and collegial fashion. When someone tells you that you are being too demanding or otherwise socially awkward for bringing up authorship, I am very suspicious that they are trying to hose you for their own interests. Nobody should be embarrassed to discuss authorship in collaborations or long-term mentoring associations. It is just too important. And I feel quite strongly that a little discussion along the way can to a LOOOOONNNNG way to heading off the furor of authorship nastiness that has been known to accompany manuscript submission.
Make no mistake. I should make myself quite clear that people need to be engaged in the authorship discussion in a professional and legitimate manner. This is not permission to grub for authorships that you do not deserve, not to alter my stance that ultimately the PI decides authorship disputes. Above all, you must be aware that you do not necessarily deserve authorship on any basis other than what the eventual manuscript has become. Not for "ownership" of a technique or domain, not for length of time served, not for amount of "effort" expended.
All I am saying is that it is a legitimate and helpful behavior to have discussions about how particular experiments, lines of experiments, collaborations, etc may contribute to an eventual paper in an ongoing and open manner.
These ongoing discussions within lab can, perhaps, help with intra-lab competition. Admittedly, no amount of discussion can overcome a legitimate lab-cancer person who permeates every interaction with selfishness. Nevertheless, if people are feeling a little miffed about always contributing to someone else's first author publication as a middle author it is important to discuss how/when s/he will likely get their own first author manuscript together including which of the aforementioned other lab members will be contributing. Worrying about publishing is a legitimate, but time and motivation sapping, tradition of the training process. Charting a path to each trainees' eventual manuscript submission can go a long (and painless/cost free) ways towards improving morale.
The mentor / PI should always be attentive to the authorship needs of each trainee's career whilst s/he is pumping out the best papers possible from the group as a whole. However, things can get lost in the shuffle. So it is important to have those conversations, even if trainees need to work up the courage to say "So, PI, I'd really like to attain a first author paper, here's what I've been doing and I think we need to refocus a bit on my paper now".

29 responses so far

  • Couldn't agree more. Right before I joined my Grad lab, a paper had come out in which one of the postdocs spent several months doing nothing but writing analysis code for another person's project. He never asked about authorship, and the PI didn't realize how much effort he'd put in. The first author did know about the effort, but since the topic never came up, it never occurred to him.
    After the paper was published, both the first author and the PI were confronted by another student, and they both felt really bad and continue to feel guilty to this day. They almost put this postdoc on another paper to which he'd contributed next to nothing as a "make-up" but ultimately decided that it was slightly dishonest to do so.
    The postdoc still seems to not mind the situation, but in my view the "sitting quietly and not minding" is exactly what caused the problem.

  • scicurious says:

    Thanks for the post, it's really informative for people just entering the field. Authorships can get pretty fuzzy in our lab, as everyone lends a hand on everyone else's projects. Every once in a while we sit down and talk about who is exactly doing what, but even so, people easily get forgotten.

  • Lou says:

    I totally agree with you.
    Would being a "local expert" i.e. the best person in the lab to do a certain technique, make you more likely to being a middle-author?
    I guess that in some labs, when some people are better at (for e.g.) doing western blots than others, you might as well give them the sample to run to get a nice figure - rather than wait weeks for the perfect blot to be produced by someone (may be first author) who doesn't regularly use the technique, get someone else to do it and publish the paper pronto.

  • lylebot says:

    I do wonder if there's some latent sexism in YFS's case (she seems to be hinting at it). Are women who want to talk about authorship more likely to be seen as over-demanding or greedy? My girlfriend runs into these problems, too, in cases where it seems obvious to me that she should be a co-author.

  • JSinger says:

    Thanks for the post, it's really informative for people just entering the field.
    I'd gently suggest that, while this is great advice for everyone, junior grad students might want to apply it more conservatively than would people like MsPhD, until they get a clearer sense of what the lab's practices are and what constitutes substantive contribution.
    I guess that in some labs, when some people are better at (for e.g.) doing western blots than others, you might as well give them the sample to run to get a nice figure...
    I'd question whether that merits authorship. But that's why it's worth asking first!
    Charting a path to each trainees' eventual manuscript submission can go a long (and painless/cost free) ways towards improving morale.
    Problem is, people always interpret that "path" as a commitment, even after they've left the lab, and the reviewers wanted another year's worth of experiments (see previous post), and someone still in the lab had to do them and rewrite the paper, and the PI is afraid of the departed person even if he's a thousand miles away (see the lab hoarding / fistfight discussion of a month ago)...

  • JSinger says:

    I do wonder if there's some latent sexism in YFS's case (she seems to be hinting at it).
    This larger topic has come up before in her comments, from different commenters: male postdocs watch each other's backs, the females won't cooperate with each other, and the complainant can't work with guys for some reason or another. (Sometimes the person says they won't talk to her, sometimes the problem is murky, sometimes it seems like she's snubbing them.)
    I've never noticed any of this, but my work style is more like MsPhD's. It's my wife who raked in all the fifth-authorships. How much those mean in practice is another matter, but they're definitely psychologically important to junior people.

  • DrugMonkey says:

    Problem is, people always interpret that "path" as a commitment, even after they've left the lab, and the reviewers wanted another year's worth of experiments (see previous post), and someone still in the lab had to do them and rewrite the paper
    Yeah but are a collection of personal unspoken assumptions about the way things are going to go with authorship lists any better? I emphasize the ongoing nature of discussions as a project moves along. Also, the more open the group is about just how much work has gone into wrapping up each "ready to submit" story from the departed post-doc, the better, in terms of the expectations of those who may plan to depart prior to submission (or post-review revision)
    Good point about new lab members coming in and stomping around like they own the place. This is why I couch this more in terms of asking about how someone views the eventual paper authorships rather than in terms of demands for services rendered from the start of any discussion.

  • PhysioProf says:

    YHN has the back of just about anyone who wants to discuss authorship in a legitimate and collegial fashion. When someone tells you that you are being too demanding or otherwise socially awkward for bringing up authorship, I am very suspicious that they are trying to hose you for their own interests. Nobody should be embarrassed to discuss authorship in collaborations or long-term mentoring associations.

    I agree completely with this. It is also important to document any understandings that arise in written form, such as by e-mails on which all relevant parties are copied.

    I guess that in some labs, when some people are better at (for e.g.) doing western blots than others, you might as well give them the sample to run to get a nice figure - rather than wait weeks for the perfect blot to be produced by someone (may be first author) who doesn't regularly use the technique, get someone else to do it and publish the paper pronto.

    This is definitely what PIs would prefer: to be able to easily deploy whoever is best at a particular technique when that technique is perceived as necessary to complete a paper. However, trainees have a somewhat divergent interest in getting first-authorships rather than middle-authorships, because the latter are rightly perceived as nearly valueless when job searching for PI positions.
    I have literally had a post-doc say to me that he refuses to do any work for any papers in the lab that he will not be a first author on. And he had had other lab members already provide key experimental results for his first author papers. While it would be best for him individually to have other lab members working on his first-author papers without him working on others', it is of course totally fucking delusional to think that might be possible.
    Let's just say that our ensuing discussion was "interesting".
    And there are whole other dimensions to this general issue. One is the situation where the PI smells something big brewing--potentially C/N/S-worthy--and chooses to shift people's efforts around in a very explicit way to move as quickly as possible to the complete set of experiments for a manuscript. Can a PI simply order trainees to drop what they are doing and participate in this other project, with the reward being only a low-value middle authorship?
    Another dimension when deploying effort in the lab via top-down PI decision-making involves the source of funding of a particular trainee. If a PI is providing a trainee's salary from NIH research grants or other funds, does the PI have the right to unilaterally direct the trainee to do what the PI says to do? What about if the trainee is paid on a training grant? What about if the trainee is paid on an individual felowship? What if the PI put a huge amount of effort into the trainee's individual fellowship application versus the trainee essentially doing everything with minimal PI input?
    That is all highly-laden and fascinating shit for PIs and trainees to deal with, the former as a manager, the latter as the managed.

  • Can a PI simply order trainees to drop what they are doing and participate in this other project, with the reward being only a low-value middle authorship?
    This one's certainly worth more discussion...
    One thing my GradAdvisor was always good at emphasizing was that her recommendation of me would include mention of situations in which, for example, I put in a lot of work but didn't have an authorship; or where I provided the crucial start for a project but have nth authorship because someone else had time to pick it up and run with it. Although the contents of her recommendation letter don't mean crap if no one gets so far as looking at it, I think (hope?) they're important in explaining my contributions, even if not first-author level.

  • James F says:

    Don't forget that many journals list author contributions, which can help the reader sort out who did what.
    Doctor Achilles @1: Was there ever consideration of submitting an erratum in which the postdoc was inserted into the author list? Unorthodox, but I've seen titles and affiliations revised via erratum, why not authorship?
    PhysioProf @8: Wow, that postdoc should find future collaborations...interesting as well.

  • Mad Hatter says:

    I agree that authorship discussions should be had early and often. I've found the "often" part to be much more useful than the "early", because many projects to which I contributed ended up being rather different in content when written into manuscripts from how they were originally conceived.
    I have literally had a post-doc say to me that he refuses to do any work for any papers in the lab that he will not be a first author on.
    Okay, that's a very extreme position to take, but I don't think pushing back on being asked over and over again to do work for middle authorship is out of line. I am the "resident expert" on a really work-intensive, time-consuming and shitty assay and am often asked to "collaborate" by people who just don't want to have to do the assay themselves. Problem is, this assay never generates sufficient data for first-authorship, which has resulted in my doing a hell of a lot of work for largely worthless (as in, not useful to my career) publications.
    And wouldn't you know it, once I started offering to teach would-be collaborators how to do the assay (no authorship expected, in this case) instead of agreeing to actually do it, many of them decided that getting data from that assay wasn't so critical to their projects after all.

  • whimple says:

    I have literally had a post-doc say to me that he refuses to do any work for any papers in the lab that he will not be a first author on.
    This is such a poisonous attitude that I would tell that post-doc to look for another lab. Seen many two-authors papers lately?

  • rb says:

    I have given away more authorships because it wasn't worth the battle to me than I have authorships. Maybe thats why I spend most of the time teaching and not doing research. I did the work because I enjoyed the science not the authorship. Not the path to "success", but I enjoy life.

  • TreeFish says:

    I am an author on more than 20 papers and the only time I had an issue with author-jockeying was when, every time I spoke to a Dr. Post Doc about getting mice, she reminded me that she expected to be an author.
    Every time, I said, "Of course." Then, once the Dr. Post Doc had this cookie, she brought up sharing first authorship. My reaction, of course, was to smile and remind her that she was only genotyping the mice for me...hardly a contribution worthy of shared first authorship. I heard from a graduate student (who was running the behavioral portion of the collaborative experiment), that Dr. Post Doc was trying to insert herself into the behavioral training portion of the experiment to supercede said grad student's contribution.
    Dr. Post Doc is no longer in science, but it was clear that her motivation and lack of decorum got her snubbed from a collaboration with me (I got the mice from another student in the lab).
    I always tell people with whom I collaborate that, if you aren't getting your hands as dirty as mine, you will not share first authorship. If you contribute to data analysis or your ideas helped inspire certain experiments, then hell yes you'll be an author. But I also make it clear that it doesn't matter where you are in the list, unless your first or last.
    I applied this same approach to a recent collaborative project with a grad student. Though I trained this person to do a relatively difficult technique, and did much of the first-half of the experiments, I said he/she clearly was the first author and that it didn't matter where the hell he/she put me.
    Author-jockeys drive me nuts. Perhaps I am up front about it from the beginning, and that's why I haven't encountered it very often.
    Sage advice, given to me by a big dog in the field of the neuroscience disease: "Be up front. Tell people what you want. Tell people what they'll get."

  • Pinus says:

    Treefish said "I always tell people with whom I collaborate that, if you aren't getting your hands as dirty as mine, you will not share first authorship.
    I really like that!

  • yogachick says:

    Treefish said "I always tell people with whom I collaborate that, if you aren't getting your hands as dirty as mine, you will not share first authorship.
    That's a nice idea, but what happens when the first author won't give credit to co-authors for getting knee deep in it?

  • DrugMonkey says:

    Lou @#3 Would being a "local expert" i.e. the best person in the lab to do a certain technique, make you more likely to being a middle-author? and http://labcoats.blogspot.com/2008/07/last-few-posts.html
    Mad Hatter@#11:... I don't think pushing back on being asked over and over again to do work for middle authorship is out of line. I am the "resident expert" on a really work-intensive, time-consuming and shitty assay and am often asked to "collaborate" by people who just don't want to have to do the assay themselves. Problem is, this assay never generates sufficient data for first-authorship, which has resulted in my doing a hell of a lot of work for largely worthless (as in, not useful to my career) publications.
    PP @#8: trainees have a somewhat divergent interest in getting first-authorships rather than middle-authorships, because the latter are rightly perceived as nearly valueless when job searching for PI positions.
    Middle authorships earned by providing some a key assay or figure for which you are expert are not completely worthless. While it is true that trainees need first-author and PIs need last-author pubs, middle authorships do count for something. This is speaking as someone for whom middle authorships have not come easily and yes, I've had critiques on some of my first grant apps that ran along the lines of "can't the PI just sort of get on some pubs as middle author?".
    The point being that total pubs do create an impression and even if the reviewer of your candidacy or grant is isolating first-author papers. The person with 5 first-author pubs and nothing else is going to look worse than the person with 4 first-author pubs and 10 middle author pubs.
    And I'm all about the advice to never turn down a pub when you have a chance at it...with a reasonable amount of effort of course. That's the rub, innit? What's "reasonable"? It sounded to me that the PP was describing a case in which the lab-cancer postdoc was getting his first-author pubs just fine- that's a pretty clear call. If your service work for someone else in the lab isn't hindering you from getting a first-author pub then you are being a selfish bastige to refuse to help others.

  • juniorprof says:

    If your service work for someone else in the lab isn't hindering you from getting a first-author pub then you are being a selfish bastige to refuse to help others.
    Exactly, which is why I also am not getting this middle author not useful thing. I can understand for the grad student who needs to graduate or the junior postdoc who is trying to establish a career path but in general I see middle author pubs as a clear indication that one plays well with others and contributes when the crunch is on. Moreover, being able to clearly demonstrate that you are keenly aware of all the details as a middle author shows that you are engaged beyond the scope of your primary experimental attention. Both of these things can be demonstrative of someone who will contribute to a department culture or broad research goal (for the academic setting) or to a site or project oriented goal (for the industry setting). I suppose that what I am trying to say is that while obtaining authorship is clearly paramount, a variety of authorship positions can be "spun" to your advantage in a variety of situations where one's career advancement is at stake. Lack of authorship is not "spinable" in any manner.

  • Lorax says:

    While middle authorships are often important and justified, they are also often padding. How many papers have more authors than experiments? Plating some cells or contributing in some miniscule way is NOT sufficient for an authorship, although some PIs are happy to put every Tom Dick and Henrietta who let the first author borrow some pipet tips on as an author. I am also disgusted by PIs who insist on authorship for providing a published reagent. If you make a mouse and publish it, you have earned your publication, if someone want to use this mouse for an experiment afterwards you didn't contribute in any more substantial way the local vending rep. Obviously, mice and other reagents can be expensive so recompense is warranted, but authorship is supposed to go to those who were directly involved in the work, not who footed the bill.

  • Arlenna says:

    "That is all highly-laden and fascinating shit for PIs and trainees to deal with, the former as a manager, the latter as the managed."
    Hah, as the trainee however, your life can be much more under your control if you can also learn to reverse that and manage your manager productively. Not that this has much to do with authorship (although it would if you were able to manage your PI as your advocate in that context).

  • okham says:

    I have for many years now routinely responded to requests for help, where I could see myself spending a substantial fraction of my time working on that project, with something along the lines:
    I shall be happy to collaborate with you on this, but need to make it clear right away that I am pressed for time these days, at this particular juncture in my career. Thus I am in no position to take on a commitment of this size and importance, which will require a substantial investment of time and energy on my part, without something to show for in the end. For this reason, before I even get started we need to discuss authorship.
    (Be sure to use the word "collaborate", never "help" -- and the "career juncture" is always "particular", all the time). Does this work ? Yes, in the sense that either people say "well, of course you'll be Nth author, goes without saying" (at which point I can decide if I want to bother or not) or, they will say "Oh... I see... OK, then , see you later", in which case I have wasted a few minutes only. YMMV

  • Lora says:

    Umm...This might be a dumb question...Who the hell can't run a decently publishable Western blot? I know undergrads who can barely spell their own names who nevertheless can carry out any protocol they're handed. What in the heck are these PIs doing if they have to break up the technical work amongst several trainees just to get a decent figure? Aren't students and postdocs there to, you know, learn new techniques and stuff? Wouldn't just, I dunno, training people well in the first place, or having a ring binder full of idiot-proofed protocols for everyone, solve this issue of "gosh, how do we credit the person who did three tail bleeds for us?"
    Anyway. I have seen some of the sexism that YFS is alluding to. The idea is that women are nurturing and therefore will help any fool who asks, for no credit. "Difficult" == "bitch" in these instances, and most women strongly object to being tagged as a bitch. In that case your best option is to find someone else to be a mentor and collaborate with, even if it's outside your department, because that shows that it's not that you are a horrible person, it's that that particular instance was not productive for you.
    You can be nicey-nice about how you want to be credited. "I'm interested in this subject, but I don't know how much time I can spend on it. My priorities are XYZ, and I really need to get a paper out" is always a good opening gambit. Negotiate from there, make the person asking put some cards on the table.

  • bayman says:

    Wouldn't just, I dunno, training people well in the first place, or having a ring binder full of idiot-proofed protocols for everyone, solve this issue
    Exactly ... divvying up research into ultra-specialized tasks according to each lab member's preexisting "expertise" is how lazy PIs manage their way out of their training responsibilities.

  • PhysioProf says:

    Exactly ... divvying up research into ultra-specialized tasks according to each lab member's preexisting "expertise" is how lazy PIs manage their way out of their training responsibilities.

    This may make sense for trivial cookbook "kit" shit like molecular biology, but some techniques take years of training to get good at. If a particular project requires use of such a technique to come to fruition, you aren't going to wait years for someone to achieve competence when there is someone sitting right there who can do it today.

  • Lora says:

    This may make sense for trivial cookbook "kit" shit like molecular biology, but some techniques take years of training to get good at.
    PP, without knowing exactly which techniques you're referring to, and without getting even drifty-er than we already are, in my personal experience there are ways of making the 2+ years it can take to master certain skills a lot more like 3-4 months. Explaining exactly how one does that is a bit outside the scope of a blog comment, but if you'd like me to explain how we went from a 2-year training period to mastery of large-scale reactor development to a 3-4 month training period, I will be more than happy to email you or post it somewhere for anyone's edification. I would imagine that this is useful info to someone out there, because academic techs, grad students and postdocs have a comparably high turnover rate to the development engineers where I work--that is, we found that after we trained folks thoroughly, they'd only work for a year or two more before moving on to bigger and better things. This wasn't, yanno, sustainable, so we figured out the training issue. It's not magic, it just takes some time and thought.
    It's quite helpful for succession planning, so that it's not such a catastrophe when the person with the magic hands gets a better job elsewhere. Other nice benefits, in addition to the career growth for the trainee: You don't feel so panicked or rushed because you don't have to juggle priorities quite so much because anyone can run the assay (i.e. less priority juggling and last-minute scheduling), you can be more efficient with how you distribute projects amongst people, students can see multiple projects from start to finish so they learn the thought processes (this also helps them anticipate what you'll ask for next, saving you the time of explanations).
    Like I said, be happy to write it up for you if you're interested, it's not magic.

  • juniorprof says:

    Lora,
    I can imagine that PP may be thinking of things like electrophysiology. While you may be able to run an experiment effectively in 3-4 months it take years to be able to do this well and the only way to get there is via trial and error and extensive experience. In vivo electrophysiology is a particularly good example as the health of the subject is paramount to all results. The lit is full of artifacts due to hypoxia, high BP and other abnormal physiological states that anyone with much experience could notice quickly. A fresh trainee will not be able to make such determinations in the early stages of training.

  • TreeFish says:

    Also, Lora, using techniques like electron microscopy, dendritic patch-clamp recording, and 2-photon glutamate uncaging will most definitely take longer than 3-4 months...even if the grad student is hopped up on goof balls and doesn't sleep. The fact of the matter is that many techniques used by leading neuroscientists take a long time to master; and it takes experience in an established lab to master them because these places are host to a cauldron of insider-secrets regarding the methodologies that could make your neck mold. Do you think it took Noah Gray 3-4 months to master 2-photon laser scanning microscopy of in vivo spine and synaptic protein dynamics? Nope, it took a heck of a lot longer than that.

  • Odyssey says:

    In my field (molecular biophysics) many techniques are pretty straightforward. Heck, I can teach the most incompetent of grad students to collect data on a CD spectrapolarimeter or a fluorimeter in an hour or less. As with all branches of science, it's the knowing what experiments to do and how to analyze the data that always have the steepest learning curves. For those things there's no substitute for experience. If anyone has any ideas on how to accelerate those processes I'd love to hear them.
    And Lora, I'd be happy to hear how you compressed training. There are plenty of techniques out there somewhere between "simple" spectroscopy and electrophysiology.

  • Lora says:

    OK, here you go:
    http://reasonableexplanation.blogspot.com/
    How to compress training time. Takes some energy investment at first, but pays off in the end IMHO.

Leave a Reply