I’m known for being stingy with my authorship lists. I don’t believe collecting samples, doing a couple PCRs, or otherwise being present, or even bouncing ideas off of, justifies you name as a author on scientific paper. I am skeptical of long author lists in top-tiered journals. I do like to promote my hard-working undergraduate assistants by giving them authorship, but require them make presentations for lab meetings, read relevant background papers and do substantial labwork. If you worked hard for a semester but produced no usable results, sorry I can’t put you down as an author.
Science, Nature and PNAS also seem to be cracking down on “honorary authorships”. Science editor-in-chief Bruce Alberts describes in a recent editorial:
“to discourage “honorary authorships,” we [Science, Nature and PNAS] agreed that before acceptance, each author will be required to identify his or her contribution to the research (see www.sciencemag.org/about/authors). Science‘s policy is specifically designed to support the authorship requirements presented in On Being a Scientist: Third Edition, published by the U.S. National Academy of Sciences. That report emphasizes the importance of an intellectual contribution for authorship and states that “Just providing the laboratory space for a project or furnishing a sample used in the research is not sufficient to be included as an author.”
While I applaud this step, for someone that describes these journals (rather pompously) as an “influential gatekeeper to the peer-reviewed literature” whose articles are chosen to “help to define scientific excellence for scientists” they are a little late the game. Many journals, PNAS included, have already required author justifications. Others, like PLoS journals, go so far as to make authors declare any competing interests. If they regard themselves so highly, why was Science not at the forefront of publishing ethics?
Additionally, the editorial places the onus of data quality on senior lab personnel and encourages senior faculty to mentor junior faculty. I’m not sure how this solves any ethical problems. It merely places the blame elsewhere. In fact, some unethical practices were knowingly conducted by a senior author. I’m not saying that it is a worthless endeavor, everyone on the author list should have seen discussed data at every step of the process and it is good to get multiple confirmations of one’s interpretation of that data. Rather these rules are only putting a couple stitches in a much larger gash.
Tenure and promotion needs to be redefined to really make a dent at healing the ethical scars in science. Don’t give people an incentive to cheat the system! It is not really a problem that publishing companies can solve. Rather, they tend to aggravate the system by some of their self-aggrandizing and editorial practices. Place the emphasis on publication quality, not the illusory journal quality. If the tenure committee has to actually skim through publications and count citations, so be it. Let’s be honest and invest in our researchers and not look for easy ways out.
I agree with much of what you say here. One thing to note though – it is not the length of the author list that should be the issue. As someone who has worked on massive collaborations involving genome sequencing – and someone who is EXTREMELY careful with author lists (to the point of having many people hate me for not giving them honorary authorship) – there are perfectly reasonable examples where long author lists are valid. And there are many cases of short author lists w/ people who did not contribute in a real way. The issue is that we should recognize contribution to work and contribution to science, in every way possible.
A for publication quality vs. journal quality – you are dead right there. That is all that should matter in tenure, job application, etc reviews.
I guess I was a bit snarky about long author lists. You are certainly correct that they can be justified. I would expect everyone on that author list to be able to describe the paper and its interpretations though. Can you say that for all the authors on your massive collaboration papers? Point also taken about short author lists.
As someone who is going back for PhD (hopefully starting next fall) and very goal-oriented in getting to a real job as efficiently as possible, I am aware of the games that must be played. My publications so far are low impact but have had glowing reviews. Out of 5 publications I have published or are in press none have been rejected (well, except *cough* one open access editorial *cough*), but my job interviewers and potential tenure committees won’t know what my reviewers say about me, just judge me based on my IF. That infuriates me because I am not aiming low, but targeting the audience that will be most interested (and most likely to cite) my publications! Maybe I can submit my reviews in my applications?
I’ve always believed that “authorship” clearly should be confined to those who literally write significant portions of a given manuscript (just doing a lot of the hands-on work is NOT enough — it’s fine if you want to acknowledge in footnotes or citations 200 other contributors/associates to the work, but “authorship” is a different matter).
And on the flip side, once you include your name as an author, then you better be prepared to take heat if any parts of the paper are later found to be fraudulent or deliberately misrepresented, even if they weren’t your specific parts (yeah, there are some exceptions, but in general, if your name goes on, it means you stand by the work as a whole).
Standards vary among groups, fields, and journals as to what can determine authorship. In general, most take the stand that authors should make a significant contribution to some portion of the work being presented and should approve of the manuscript. There is some disagreement about whether individuals need to actually be involved in writing the manuscript itself as opposed to viewing and approving the manuscript (as long as they made some significant contribution).
For example if you look at PLoS One guidelines (http://www.plosone.org/static/guidelines.action) they take the position of the Uniform Requirements for Manuscripts Submitted to Biomedical Journals (http://www.icmje.org/):
“”Authorship credit should be based on
1. substantial contribution to conception and design, or acquisition of data, or analysis and interpretation of data;
2. drafting the article or revising it critically for important intellectual content; and
3. final approval of the version to be published.
Authors should meet conditions 1, 2, and 3.”
I think this is pretty reasonable. But other groups/journals emphasize only #1 and #3 and leave #2 much more open to interpretation.
Kevin – as for my own large collaborations, when I am in charge I have always been on the side of fewer authors, more acknowledgements. For example, when I was at TIGR sequencing genomes (for eight years …) I was the first person to NOT put the heads of facilities and even the head of the institute (Claire Fraser) on genome papers. However, for large projects, I think it is OK to not require all participants to help write what ends up being the final paper – in fact it is almost impossible to do so. For my recent “GEBA” paper in Nature, for example, I tried to get everyone to contribute material on a Google Doc and it worked OK for some period but in the end became very cumbersome with so many contributors. But I liked the Google Doc approach b/c I could track actual contributions to the manuscript and then better justify who should be authors and who should not.
Hmmm….
(warning, thanks to Grrlscientist’s best efforts I am not totally sober)
(1) the guest authorship is one specific problem. I’ve heard about in it a couple of places, that the departmental boss gets their name on any paper. I think this is what they’re targeting.
(2) Competing interests is a medical thing – it can be a problem if the trial you’re reporting was paid for by the Evil Pharma Company that makes the drug you’re testing.
(3) I don’t believe there is any objective way of determining paper quality. Which means whatever we do will be imperfect.
(4) I need more alcohol.