The April commentary on multiauthor publications is a timely discussion of an important problem. I, too, am disappointed that most journals have failed to provide clear guidelines about who should be included as an author. My PhD adviser set out a simple test: An author is someone who has made significant intellectual contributions to the research presented in the paper. Today authorship is used in other ways, perhaps to get technical assistance from another lab or to garner larger visibility for the paper by recruiting a prominent scientist as a coauthor.
As author Philip Wyatt notes, the journal Science requires an unpublished statement of the contribution that each author has made to the paper. The journal has also published a thoughtful editorial 1on the issue of authorship. Nature’s policy is a bit more aggressive, in that a statement of author contributions is published with each paper. This approach trades on our innate desire to do the right thing, at least while in public. I have already adopted this approach for papers that I coauthor, and I encourage others to do so.
Let me end with one word of caution about Wyatt’s analysis. Yes, it is useful to compare our current practices with those used in the past. But past practices were not necessarily fairer or better. A good example is the oil-drop experiments published by Robert Millikan in 1913 (see Physics Today, May 2007, page 56, and reference 2). Harvey Fletcher did a significant amount of the lab work for those experiments as part of his PhD with Millikan. Under current practice, the paper would certainly have included Fletcher as a coauthor and perhaps as first author as well. If the work had been published that way, then Millikan may have had to share his 1923 Nobel Prize in Physics with Fletcher.
