In many fields, first authors on scientific papers represent the person who’s performed the bulk of the research. Sometimes, that determination can be difficult to make, so we’ve seen many papers that list multiple first authors, noting that each contributed equally to the work. But is it possible — or ethical — to claim six authors all deserve top billing on a paper?
In a recent letter in Science and Engineering Ethics, Govindasamy Agoramoorthy — at Sengamala Thayaar Educational Trust Women’s College in India and Tajen University in Taiwan — flags a 2014 paper in The Plant Journal that lists six first authors, noting all “contributed equally to this work.”
As Agoramoorthy notes in “Multiple First Authors as Equal Contributors: Is It Ethical?“:
If this trend continues, it’s a matter of time before all authors may claim first authorship in all papers. To make the ethical crisis worse, some argue that it’s time to recognize the co-first authorship in journals as a reality facing the scientific journal industry since many researchers desperately depend on it to get promotions, grants, awards and other monetary benefits.
But the question is, would it be possible and practical for several authors to equally contribute in a single research paper? If so, what are the scientifically testable measures and variables that journals can use to authenticate this far-fetched claim?
Tell us what you think.
[polldaddy poll=9523118]
Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our daily digest. Click here to review our Comments Policy. For a sneak peek at what we’re working on, click here.
I’m not into micro-managing other people’s decisions.
Recognizing multiple first authorships is nothing but a survival technique, in a science world that poses increasing pressure on researchers forcing them to “publish or perish” and taking into account only papers in which you are first or last author. And nothing else.
I don’t agree with Agoramoorthy when he asks for “scientifically testable measures and variables” to test the authorship. Science is based on collaboration and solidarity among researchers. Insisting on rankings will just result in mining and compromising the very fundamentals of scientific research. Until we will not relinquish from such insane bibliometric attitudes, “all the authors contributed equally to the study” should be a must have in any manuscript. Of course provided that complimentary/honorary authorships will be inexorably suppressed. I’m much much more concerned with this latter kind of serious misconduct, rather than with ranking the true and genuine authors of a study.
it’s either joint work — consider the cake being the joint work of flour, eggs, sugar, et cet. — or it ain’t. If it ain’t joint work, they should write separate papers. And then randomize the sequence of authors, who cares about the sequence of authors, it’s the scientific contribution that should matter. first authors, my left foot.
Another trick I’ve noticed which I don’t think you’ve written about here is citing other papers which have nothing much to do with the work/research and don’t provide supporting evidence or arguments. Basically citing as many works of your friends as possible. Maybe for quid pro quo, in the hope they’ll all somehow manage to cite your paper too in their future references.
Is there a name for this trick?
It is known as a citation circle when academics deliberately cite each other but the citations are still relevant. The term could be extended to include any citations.
” Should journals allow authors to list more than two co-first authors?”
Two? Journals shouldn’t list more than one first author.
The letter implies that the sanctity of the “first author” is some kind of moral imperative that scientists must preserve at all costs, then proceeds to make the slippery slope argument that soon every paper will claim every author as first author.
What is immoral is assigning scientific merit badges based on something so arbitrary as the order of author names on a piece of paper. In reality, no one but the authors of the manuscript themselves will be able to judge the true contributions of each other. In some fields (e.g. mathematics) there is no first author as the usual convention is alphabetical ordering of authors, and therefore such nonsensical “ethical” dilemmas are cleanly avoided.
Totally agree. Anyone that thinks that high end research can still be done alone (so only one first author) is definetely out of reality. In fields like genomics, there are so many people doing important parts of the work that the first author system is very unfair.
Agreed. As I’ve argued before (http://dx.doi.org/10.1126/science.327.5965.523-a; paywalled), authorship is too blunt an instrument for determining credit – as this situation points out. Some are working on difference methods of determining credit (http://neurodojo.blogspot.com/2015/10/badges-for-scientific-paper-contributors.html) However, given that it is used for credit now, it’s not silly to want to prevent authors from gaming the system until better systems are in place for assigning credit.
I can see the equality of work implied by first authorship occurring in a non-university setting – such as in an R&D unit of a for-profit entity. But in universities, there should be only one first author – the graduate student into whose thesis this will go, or the post-doc who will claim this paper as a personal step in the ladder of seeking to join the self-perpetuating system known as academia.
In my experience, there’s no 1:1 mapping of individual PhD theses to papers. Many papers result from collaborative work that involved substantial, critical effort from multiple people, often with different skills.
In my mind, author contribution statements, when written honestly and correctly, are more important and yes I do read them when asked to judge someone’s body of work. Sometimes a “non-first” author was a key part of the team, and sometimes a “non-first” author contributed a minor reagent.
Verbatim email sent to authors and The Plant Journal editors, on Wednesday, July 27, 2016 3:35 PM:
“Dear Professors Takashi Gojobori, Naoki Osada, and Tzen-Yuh Chiang,
Dear Plant Journal editors
REF: http://link.springer.com/article/10.1007%2Fs11948-016-9794-x
I would like to ask your opinion about a comment made by Dr. Govindasamy Agoramoorthy (see PDF attached) that basically characterizes your 6 first / co-corresponding authors as being an unethical act: “I was shocked when I saw a paper published in the Plant Journal in 2014 (impact factor 5.972) where six authors (out of total 14) had claimed the first authorship with a cross sign asterisk that referred to a statement that ‘these authors contributed equally to this work’ (http://onlinelibrary.wiley.com/doi/10.1111/tpj.12676/pdf). The cross-signed statement is so small that readers may need a magnifying glass to have a good look at it clearly. It indicates that even the journal may not like readers to see this statement at first sight.” The author also takes a swipe at The Plant Journal claiming that it is trying to hide this fact.
What is rather curious, and surprising, is that there is no author contribution statement explaining what each author did and what weighting those contributions had to the overall experiment and final paper. There is also no conflict of interest statement. Is perhaps Dr. Agoramoorthy suggesting guest authorship or inappropriate authorship with his criticisms?
I think that your opinions would be very important, specifically to address the following:
a) how is it possible that six individuals can share first authorship? Did they all require this position, as is often the case in China and Japan, to guarantee their PhD degrees?
b) Does The Plant Journal have specific guidelines regarding authorship that limits first authorship to only two authors?
c) Do the authors and the journal agree with the statements and criticisms made by Dr. Agoramoorthy, and if not, would they be willing to challenge Science and Engineering Ethics editors to request a correction, or clarification?
Dr. Bird, who is a co-EIC of SEE, is on record defending the right of authors to define authorship and to select authors, thus embracing the concept of co-first authorship, which would in essence allow for 6 or 100 scientists to be co-first authors, if they so deemed necessary. This posture, which contradicts that stated by Dr. Agoramoorthy, merits additional exploration so that the plant science community can better understand how future cases should be viewed, and handled.
I look forward to a fruitful discussion on this topic.
Sincerely,
Jaime A. Teixeira da Silva
https://www.researchgate.net/profile/Jaime_Teixeira_Da_Silva “
Even with just two co-equal first authors, awkwardnesses can crop up. As a graduate student, I was pressured by an adviser to add the “equal contribution” statement to the author note of the publication that resulted from my dissertation. Given it was my PhD work, I thought I deserved to be first author alone, but I knew I would need letters of recommendation from my advisor in the future. So I acquiesced. In the journal, I was listed first, but for all these years my former mentor has listed the publication on his CV with his name first–an inaccurate citation. Not a big deal to me. To all the world I am the first author, and few people read the author note. But how do we reconcile the “equal contribution” idea with the strictly linear nature of reference citations? Is there need for some kind of mathematical notation—brackets or parentheses—that indicates the subset of equally contributing authors?
I think the question is less on joint co-authorship and more on transparency on contributions. As a managing editor I have seen some authors in second, third or fourth position that could have easily claim to be first authors (or joint first author at the very least). Increasing transparency of the contributions that each author is doing is not always possible, but with cloud file sharing and the chance to collaborate with online tools, attribution would be more easily established for great parts of the work that needs to be done. I have no problem with multiple first authors, as far as they all work equally. We should promote more collaborative ways to work (that I personally think reduce the risk or temptation for misconduct). Unequal work (like the example on the previous comment of a supervisor requesting to be joint first author) is a lot more problematic, and I think raise more ethical concerns.
I agree with your comment. I am a recent PhD graduate. This has opened my eye to how much the level of research has deteriorated in academia and academic scientist no longer wears a noble mark as it used to be. One major issue is indeed in publications. I strongly recommend that journals impose the requirement to have a contribution statement for each author listed as an author. This will give a clear impression of the contributions from each co-author. Also, funding institutions should not rate groups only through the number of publications but on the percentage of impact each publication has in the area. Reviewers in publishing companies could help by additional rating the impact level of the work. This could further push for novel contributions to the scientific world.
Also, I hope this method will avoid academic supervisors/co from being super concern with papers and oblige all their students to co-author them (I am ok if the person is the PI to a certain extent) in every paper although they have not contributed anything. I hope this reduces the level of neglect of PhD students as they are not self-generating publication vending machines who doesn’t eat nor sleep. It takes a lot of hard work and sacrifice, even on the physical and mental health. Students are there to learn, contribute and work together as a team, just how an employee in a company is trained.
In mathematics, authors are always listed by alphabet; hence we don’t know the concept of first author. For us the question posed in your poll (fortunately) doesn’t make any sense!
“what are the scientifically testable measures and variables that journals can use to authenticate” any contribution to a paper? None whatsoever. How do you know that the last author really supervised or at all contributed to the research? What about the, sometimes, multitudes of authors in the middle? Even with a single first author it’s impossible to tell if this author actually performed the bulk of the research or took credit for somebody else’s work.
Maybe the true question is when journals will evolve beyond simple authorship lists and give credits for the real role of each contributor.
A paper may have one or more project leaders, a project manager, and a scientific director; all with distinct, but important roles for the paper; all deserving full recognition to count for their career progression. But the whole system—journals only reflect that—is trapped in the romantic ideal of intelectual authorship.
I’ve discussed this point in more details in https://mariobarbatti.wordpress.com/2016/03/27/when-science-goes-wrong/
I think up to 3 co-first authors is OK. And I think if you are second or third of these, it is OK to permute the order between them – after all they are all co-first authors; or indicate this by three asterisks with an explanation.
How the magic number three ??
I can see reasons for multiple first authors, for instance with one person contributing most of the experimental work and another contributing most of the computational work. Six seems ridiculous, and since the first six are not in alphabetical order, I assume there is some internal idea of who contributed more or less to the design and analysis. At least that is how it has worked when I’ve been one of three “contributing equally” authors.
While based on relative contribution we could decide the sequence of equal contributors, but not necessarily.
Just for your information, in my recent 4 first authored paper, we did a lucky draw for the first four authors. I will not go by alphabetical because why should be anyone not have an earlier order in a sequence for a name he/she did not choose.
If three is fine and why six is ridiculous ?
Some background reading of relevance:
Teixeira da Silva, J.A., Dobránszki, J. (2016) Multiple authorship in scientific manuscripts: ethical challenges, ghost and guest/gift authorship, and the cultural/disciplinary perspective. Science and Engineering Ethics (in press)
http://link.springer.com/article/10.1007/s11948-015-9716-3
DOI: 10.1007/s11948-015-9716-3
How regressive. In the ear of multidisciplinary research and ever increasing demands of data by Editors and reviewers, how on earth it is possible to have a single first author ? Soon we will be entering era of multiple corresponding authors as well.
I feel surprised seeing articles having more than 100 figures (including supplementary) and a single first author. I do understand there are some green fingered postdocs and PhDs but most cant manage that quantum. Me as a PI have to take data from several students and postdocs to make a single manuscript and the decision should be best left to me of the numbers of the first authors. I dont want this moral policing please.
We “will” be entering the era of multiple corresponding authors? That ship has SAILED: http://neurodojo.blogspot.com/2015/09/everybody-gets-to-be-corresponding.html
Four corresponding authors already.
Thanks for the entry, Alison. Speaking of multi-authorship, I and other 98 authors from around the world have co-authored an 11.000-word manuscript. This is the first article in the social sciences to be written by such a large number of international academics. It is entitled Academic Domains as Political Battlegrounds. Although the process of crowd-authoring was, in itself, successful in the sense that a crowd of authors managed to get together and produced a crowd-authored manuscript, the act of getting editors to accept and publish the manuscript proved exceptionally difficult. That is, the manuscript was submitted to more than 50 accredited journals, which all rejected it even without the consultation of peer reviewers. After all these rejections, the manuscript was eventually accepted for publication in a respected SAGE journal [Impact Factor: 0.491]. Here is a free copy of the article:
http://idv.sagepub.com/cgi/reprint/0266666916646415v3.pdf?ijkey=QtsF7QaLOhMzxrt&keytype=finite
For an audio copy of the article, simply click on the link below and listen:
https://drive.google.com/file/d/0B-Ek4c_UE37WNDA2OW5fZG9jZWs/view?usp=sharing
There is a website dedicated to the Crowd-Authoring approach, which reports on and critically reflect on this authoring experience. Here it is: https://crowdauthoring.wordpress.com/
I think the alphabetic order isn’t perfect too. The whole idea behind the author’s list is to give credit where credit is due (btw. http://www.phdcomics.com/comics/archive.php/archive_print.php?comicid=562)
so that people who quote the work in the future acknowledge who did the job.
The authors of the textbook “Molecular Biology of the Cell” listed themselves in alphabetical order, which does not stop students from referring to “Alberts”, instead of the title.
So, if the Latin alphabet were any different, would we speak of “Johnson”, “Lewis” or anyone else?
I’m actually wondering how do you quote these mathematical papers in the text of your paper? You know, in all these “Jones et al. contributed the key idea that…”
Perhaps this works if there aren’t many authors. You can write then “Jones, Smith and Wang showed that…” However, if the paper has 20 authors, giving appropriate credit to them is a tough job no matter the underlying system.
How are papers cited in mathematics? Typically, there is a numbered reference list (listed by alphabet according to the first-named (sic!) author); let’s say the Jones et al. paper is number 11. Then you might refer to it in the text saying something like, “It was proved in [11] that a reflexive Banach space etc.”. Incidentally, it is extremely rare for a mathematical paper to have more than 4 authors; I reckon the average to be 2.
Dr. Lily, could you kindly clarify the following doubts, please.
https://crowdauthoring.wordpress.com/2014/11/13/definition/
In this Elsevier paper, you describe 101 authors:
Al Lily, A.E.A. (2016). Crowd-authoring: The art and politics of engaging 101 authors of educational technology. International Journal of Information Management, 36(6), 1053-61. Elsevier Publisher; Impact Factor 2.692.
http://www.sciencedirect.com/science/article/pii/S0268401216300275
In this SAGE paper, you describe 99 authors:
http://idv.sagepub.com/content/early/2015/12/07/0266666915622044.abstract
Are both papers referring to the same set of authors, and study?
A few things are unclear to me: in essence, in the SAGE paper, your “idea” for the paper was to develop a paper on crowd-authoring, is that correct? Then, to achieve this, you sent an invitation to almost 2000 individuals by email, requesting them to develop your idea further, and then through three rounds of edits, you incorporated their edits and additions until the final paper was achieved (and accepted in ISLS). Would this be an accurate simplified description of your objective and methodology?
Dear Anonymous,
Thank you so much for your comment on my entry and for your interest in the work. These two papers come from the same project – at first, we were 102, then one withdrew, then one withdrew, then one withdrew, thereby ending up being 99. Hope this makes sense to you. About your second question (your “idea” for the paper was to develop a paper on crowd-authoring, is that correct?), the idea was to develop a _crowd-authored_ paper on a subject in the field of educational technology. But your idea is also interesting – to have a crowd-authored paper on crowd-authoring. Hope I addressed your questions.
Warmest wishes,
Abdul
Dear Prof. Lily, thank you for your kind explanation. However, the topic that was developed by 99 authors is very unclear to me. And it is also very unclear exactly what each one contributed. I understand that you served as the “mediator”, but some may question the validity of adding 99 names to an 11 page paper when it is unclear exactly what each one wrote, or contributed. Your papers carry no conflicts of interest statement, or authorship contributor statement, which is surprising given that these are three famous publishers.
Can you indicate the reason why three participants dropped out?
Thanks again for your interest and engagement. I have written a book that details this experience and explains in detail each phase. Here is the book:
https://www.amazon.com/Crowd-Authoring-Social-Sciences-Abdulrahman-Essa/dp/1537735675/ref=sr_1_1?s=books&ie=UTF8&qid=1474196972&sr=1-1&keywords=Crowd-Authoring+in+the+Social+Sciences
Hopefully, the book will address all your questions.
Warmest wishes,
Abdul
My sincerest apologies, Dr. Lily, I forgot to add one more query. In your third paper, you describe 100 authors:
http://link.springer.com/article/10.1007/s12109-016-9451-x
Once again, it is unclear if the 99, 100, and 101 authors are representing the “same” group. If yes, could you please explain the discrepancies in the numbers. Thank you.
Hi again. All these numbers go back to the same project but in different times – at first, we were 102, then one withdrew, then one withdrew, then one withdrew, thereby ending up being 99. Hope I answered your question.
I always thought it is harming science to have just one first author on the paper. All should be listed as equally contributed. Otherwise it is like devoiding people of the possibility to truly work together if they want = getting results faster, publishing faster, having somebody else to discuss in-depth details of the project (motivation spike!), getting valuable results more often as a group. I think it does make a difference, especially in such a low success/failure rate field like biology to see positive results more often. Nowadays people do not like to devoid their time into other people’s projects because they simply will have less time for their own and will be regarded as less successful and won’t get grants. So who of you is still contributing to other projects without at least a bit thinking that you’re wasting your precious time?
Also, devaluating “first author” positions could bring better quality of research as one can focus more on the science and not on publishing as fast as possible… Of course the question then remains how to decide who gets the grant money…. So the new system of grant-receiving would need to be devolped. And this seems to me the hardest problem here. I am curious what you think.
In a perfect world the first author should be the person that coordinated, planned and took responsibility for the work as a whole. Of course many people will have contributed to any given paper, but the “brain” should be one person. Of course those “secondary contributors” should be that person for their own respective projects.
The problem is, that “first authorship” is a currency. You need it to get jobs or acquire funding, so it is basically inevitable that papers will have more and more first authors, because that little asterix practically amounts to production of currency ex nihil.
When everyone is a first author, nobody is.
My personal rule is that co-first is only appropriate if the paper could have been more than one manuscript. Sometimes independent, but related, studies get bundled together to tell a bigger story.
Dirk Werner, from my understanding, the mathematical sciences use/used the Hardy-Littlewood axiom. Here some more information:
Teixeira da Silva, J.A., Dobránszki, J. (2013) Should the Hardy-Littlewood axioms of collaboration be used for collaborative authorship? The Asian and Australasian Journal of Plant Science and Biotechnology 7(Special Issue 1): 72-75.
http://www.globalsciencebooks.info/Online/GSBOnline/images/2013/AAJPSB_7(SI1)/AAJPSB_7(SI1)72-75o.pdf
Encoding researcher contributions in an author sequence is a relic of old times when labs were tiny and collaborations unnecessary. It’s short-sighted and half-assed. These petty discussions about who goes first should make it obvious to everybody, how flawed the convention is. I feel embarrassed for the whole profession.
If I could re-invent the convention, I’d put a credits section at the end of the paper that lists the contributions without indication of hierarchies. Instead of authors, I’d name the involved laboratories alphabetically as sources.
This would solve this ridiculous and childish debate we force ourselves into.
Maybe we should embrace that.
Here is a recent paper with THREE FIRST and THREE LAST authors with equal contribution, respectively. Still a few others in between, but I can see how one could create a paper with an arbitrary number of authors who all are EITHER FIRST OR LAST.
http://jem.rupress.org/content/213/8/1429.abstract
I believe most comments missed the point: are this articles with too many authors (and now too many first authors) a device to artificially improve the groups CV (and receive money for this trick)? At my university we have a research group were all articles are authored by all researchers and hence all articles go into the whole group CV.
It’s shocking to me that most of the comments are really confortable with this status and look at this as something “natural”.
But isn’t that worry just an artifact of just using the number of papers (of an individual or a research group) as the main metric for money distribution or as a proxy for productivity in general? IMO this ignores all the other issues around first authorship, and maybe valid reasons for this practice you noticed.
This discussion is a real eye-opener for me. Thank you RW.
I think ordering should not matter, and if it doesn’t- alphabetize.
What should matter is the contribution section, to make it more transparent- each area of contribution should be marked. Think a sidebar, like track changes. In a systematic review each extracted study should include the initials of the extractor, forest plots the same…..
Make each author answer a correspondent to their section.
I can see situations in which more than two first authors could be appropriate, but six would be a lot for the types of papers published in my field. Some journals now have a section that lists the specific contributions of each author, and this seems like good practice.
This one has 15 authors with equal contributions. So 15 first authors? Or 15 corresponding authors?http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0163716
It is also equally burserk to have up to three or five co-corresponding authors that are commonly seen in many esteemed journals.
Pls use this topic as a subject of discussion.
To be frank, multiple co-first authors are for young investigators to mutliply use the credits for graduation, jobs applications, promotion, etc.
And multiple co-corresponding authors are for senior invesigators to multiply use the credits for monetary awards, projects/grants, credentials, etc.
Soon I will be a corresponding author on a paper with three corresponding authors (and two first authors). We did this, because the work involves three independent investigations into the same aspect. There is therefore not one PI, but there are three, which each individually take responsibility(!) for the work performed in their specific labs. All three of the PIs have so many papers already that one extra in an OK journal (IF 3-point-something) really doesn’t matter on bit. At least two of us will not get any monetary award for having a paper as corresponding author – a paper is a paper, whether as first, last, or just one of the many authors. I am not 100% certain about the third, but don’t think so either. My university database does not even allow me to indicate the corresponding author, so it doesn’t matter there, either.
The situation is a bit different for the two first authors. Both are junior scientists, and many evaluators still look at how many first-author papers they have when considering hiring them or granting them funding. So, what do you do if you have two people who have done more or less the same amount of work? Only one can be put as the very first. The appropriate thing to do is to credit both these junior scientists equally. Not offering one of them co-first-authorship is denying their equal contribution. As long as these young scientists are being judged on their first-author papers, I, as a PI responsible for their development as scientists, must make sure they are treated equally. That means co-first-authorships in certain papers. That is, in my opinion, the only ethical choice to make.
I am sure there are places where such ethical considerations are not taken into account, but it would be nice if people would not just shove everything in the same corner of “nefarious intent”.
Couldn’t agree more.
http://www.nature.com/nsmb/journal/v21/n8/abs/nsmb.2846.html
This has 6 authors with equal contribution and five co-corresponding authors.
To credit students and postdocs who take part in true team work, where each person’s contribution is fundamental, and interdisciplinary projects it should of course be possible to share first authorships. Otherwise there is no incentive to work as teams, and science will be biased towards the somewhat old-fashioned 1 person 1 project style that is not always the right mean to provide progress in an area.
Is there a standard/unbiased approach to evaluate the contributions in order to give someone an ‘asterisk’ for equally contributing 1st author?