11 May 2015

When does authorship stop meaning anything useful?

“My god... it’s full of authors.”

Prof-Like Substance drew my attention to this fruit fly paper, which may be a new record setter. I have never before seen a paper with over one thousand authors. One thousand and fourteen, if my count is correct. (Image at right. Click to enlarge... if you dare. That took some image stitching, let me tell you)

The highest I’d seen before this was a paltry 816.

I was curious what you had to have done to be listed as an author. With that many, it seemed like the criteria for authorship might have been, “Have you ever seen a fruit fly?” I went looking for statement about author contributions, which some journals have. There is not any such declaration in the paper.

The PDF of the paper gives a bit of a clue as to what’s going one. The author list is more modest on the title page, which lists the authors as, “Wilson Leung and Participating Students and Faculty of the Genomics Education Partnership.” So a lot of these authors are students who took a class, and probably completed part of the analysis as a course assignment.

Digging into the acknowledgements, though, suggests that the inclusion for authorship was marginally higher than being an data monkey:

The authors also thank additional students who contributed data analysis to this project, but for various reasons did not participate in reviewing the manuscript.

This suggests that all thousand or so authors at least looked at the paper and signed off on it. But judging from the course listings in the Acknowledgments section, it seems that many of these were undergraduate students, and I wonder whether any of them had any substantive opportunity to have input into the text and interpretations of the paper. And can everyone stand behind, and vouchsafe, the data here? Some guidelines require that of authors, and I think that’s a pretty good guideline.

I am all for engaging students in research, and crediting them. But this is a bad way of doing it.

Papers like this render the concept of “authorship” of a scientific paper meaningless. This feel more like a participation award than authorship. A possible solution, as I’ve suggested before (also here; paywalled), is that we need to give up “authorship” and focus on “credits” that are clearer descriptions of the contribution individuals make. Call the students “contributors” rather than authors. Put it in a supplemental file.

Additional: Okay, this paper has 1,446 authors, and this one has 2,932 authors. (Hat tip to Jens Foell for pointing those out.) Both are particle physics papers, though, a field which has been dealing with large author numbers for a long time. The paper under discussion here, the fruit fly paper, may be a record for biology. Even the draft human genome got ‘er done with “just” 272 authors.

New rule! If the number of authors on your paper can be measured in “kiloauthors,” having your name on the paper will not count for tenure and promotion purposes.

Update, 12 May 2015: The journal’s blog describes how this paper came to have over 1,000 authors, with over 900 of them being undergraduate students.

I will point out that there are guidelines for who gets to be an author. These are not perfect (Drugmonkey hates them), and not often followed in the trenches. But they do represent an attempt to spell out what authorship should mean, by a fairly substantial number of people working in scientific publishing.

I doubt that every undergraduates on this paper truly helped draft or revise the paper (criterion 2) or can be truly accountable for everything related to the paper (criterion 4). To their credit, author number 1,014 says on the blog:

“Actually we got some important comments back from students,” says Elgin.

I’m pleased that some students made important comments, but I have doubts that all the students genuinely met the “draft and revise” criterion. Reading a paper and saying, “Okay,” doesn’t cut it for me.

Update, 13 May 2015: This story has bubbled over to Nature’s website, with some comments from the non-student authors. Warning: contains me.

Update, 21 July 2016: This blog post discusses a highly cited technician (mostly so I can find it again later).

Related posts

Letter in Science

External links

Class projects as publishable research?
Undergrads power genomics research
Who is an author? (ICMJE “Vancouver guidelines”) 
Fruit-fly paper has 1,000 authors

6 comments:

Unknown said...

Including undergraduates as co-authors in an interesting trend that I've noticed a fair amount recently.

You may be interested in this about the http://www.ncbi.nlm.nih.gov/pubmed/25740935 citation: Massively parallel biology students
http://news.wustl.edu/news/Pages/very-large-lab.aspx

This eLife article http://www.ncbi.nlm.nih.gov/pubmed/25919952 lists 2853 "collaborators" when you click on the article & author info tab. I think quite a few of those are undergrads also but not sure.

Andy said...

I have high school student co-authors in some cases...but in all cases they are substantively contributing to the work (usually in working on some component of the overall project, and writing drafts of portions of the manuscript). So, I don't think it's necessarily a bad thing to have student authors, but it requires some thought and effort from senior supervisors.

David Roberts said...

"this one has 2,932 authors."

which paper was that?

Anonymous said...

Even if every author fixed a few typos , imagine how long it would take to enter all of those edits! Or discard them because you've already corrected it. Even with tracking in Word that would take forever.

Matt said...

I read this blog post. I'm a co-author! :)

Anonymous said...

how do you add such a paper to Mendeley?!