A multitude of authors
Raf Aerts, on his Nature Network blog African Blog of Ecology, writes about the relentless increase in the number of multi-authored manuscripts. He reproduces a graph from the magazine Science Watch (November/December 2007 edition) which tracks papers grouped into four tranches of more than 50, 100, 200 and 500 authors. Between 1993 and 2003, the numbers of papers with these large numbers of authors were fairly stable, but after that time, the numbers in all categories increased significantly (to 2007). Raf Aerts writes:
"But then came the cracker: in 2000 there was a report with 918 authors, and the current record holder is physics paper published in 2006… with 2512 authors. Imagine all these authors track-changing the manuscript!"
At Nature, and many other journals, the editors ask one author to coordinate revisions and changes between all coauthors, and convey those to the journal on behalf of all of them. The challenge for the poor corresponding author must be quite significant on occasion.
Why do papers need so many authors? Modern, "big" science means that a paper can take years to gestate, involving researchers at many centres, international facilities (for expensive equipment, for example), and complex software. Martin Fenner, in a comment to Raf Aert's post, writes: "In my last published paper I have 82 coauthors, my personal record. The paper is the result of a consensus conference on the management of testicular cancer."
The papers tracked by Science Watch do not cover the "ten-author" paper in which a couple of professors might use their seniority muscle to add their names to an author list even though they made no contribution to the intellectual or physical effort of creating the paper. Is this type of paper making more of a mockery of the concept of "authorship" than the several-hundred author collaboration? How else would these researchers receive deserved credit for their work?
The Nature journals appreciate the problems, but wish authors to be transparent with their readers -- as we do not want to support the practice of honorary authorship, while being sympathetic to genuine collaborations. We encourage coauthors of a paper to specify their contributions to the work, in a statement in the acknowledgements. Our policies, as well as (free-access) editorials in our journals on this topic, are gathered at the Author and Reviewers' website; and our discussions of authorship on Nautilus are gathered under the tag "authorship", which you can see listed by clicking on this link. We welcome your views, which can be made online as a comment to any of these posts.