Science first, scientists later

How and why we should get rid of author lists in scientific publications

About a month ago, Steven Burgess, conducted a Twitter poll asking if biologists would consider making the lists of authors that accompanies every scientific publication, alphabetical. Unsurprisingly, I and 670 other scientists voted against this proposal.

i. the problem

In the biological sciences, authorship of scientific, peer-reviewed articles is perhaps the single biggest determinant of career success, recognition and grant funding. However, merely being one author out of, say 5, on an article is not enough, where you are on that list matters too.

A very slightly exaggerated version of how scientists read author lists.

Closer to the front and you’re probably someone who did most of the lab work. If you’re first on the list, you probably wrote the entire paper too. If you’re last or second last, you are likely the professor or Principal Investigator who brought in the money, played some role in designing the experiments, and edited the paper.

So far, so good; the system seems perfectly logical. The problem is that author lists form an important part of any academic CV where they are present in the “Publications” section. For example, a snippet from my CV:

Author lists in an academic CV

Reviewers on hiring committees and grant funding applications usually don’t have the time to read each publication and assess author contributions, and hence rely, to different degrees, on the position of you name on the author list for each publication. The same applies for other scientists reading papers, many of whom remember papers purely based on the first author’s name (as in “Did you read that paper by Li et al., 2017?”) or based on the last author (like: “Doudna’s lab just published a new Cas9 paper!”).

This equation of author list order with relative “importance/contribution” then, obviously leads to differing perceptions among each author and each reader:

Perceptions of importance. ©Devang Mehta 2017

Consequently, author lists often cause problems in many laboratories, leading to tension, stress and even outright fights. Some labs have tried to deal with this by instituting metrics to determine author order, or at least some rough criteria.

Another trend that’s picked up steam recently is the inclusion of explicit “Author Contribution” statements at the end of scientific journal articles, describing (even using standardised language) the roles of each other.

The author list, however, stubbornly sticks around.

ii. a solution

After the Twitter poll, Steven asked and received feedback which was then published in an excellent article on Medium:

Toward’s the end of the article he presented us with a revolutionary new format for scientific manuscripts (inspired by a tweet from State Space).

Film Credit Model by Steven Burgess

What struck me immediately was the complete absence of an author list.

This is an incredibly simple solution to an entrenched problem in scientific publishing. And it comes with some far-reaching, and in my opinion, positive, consequences:

Gift authorships and that whole debate will end!

There’s been some talk lately about how some labs are ‘generous’, giving authorship to all members, how some authors are given authorship just because of the impact big names carry, and about how some labs place too high a threshold for authorship. A mandatory contribution statement cuts through all this. If someone is listed as an author just for giving ‘helpful comments’, it’s fine. As long as that is made clear in the respective contribution statement.

Academic CV’s will change.

My CV with no author lists and with CRediT contributions.

Reference lists will change.

Out with the old…

but not by much:

… in with the new.

Citations in academic papers will use numbered lists and hyperlinks instead of “author-date”.

Science will go on, but authorship will stop mattering so much.

In the long run, I expect that the Film Credit model will steadily erode the primacy of author names vs. paper titles. We will no longer refer to papers by the first author’s name but rather by the actual science in there. And maybe, someday, we will start reading papers based on their titles, not by how famous or how white the author’s name is.

iii. how do we get there

To my mind the Film Credit model is undoubtedly better than any current publication format. It avoids making arbitrary lists of author importance and does away with the stressful process of ranking your colleagues while trying to be objective.

However, scientists don’t usually like to rock any boats and I expect some serious difficulties with getting them and publishers to adopt the Film-Credit model.

Here’s my brief two-step outline of how we can get this done though.

STEP 1: Talk to DORA

The San Francisco Declaration on Research Assessment sets guidelines for funding agencies and hiring committees to assess research outputs fairly. DORA’s goals align perfectly with solving the author list issue and it can convince a large number of grant agencies to ask applicants to remove author lists from CVs and proposals. Once fellow scientists a familiar with the idea of eliminating author lists in their own CVs, they will probably find it less risky to do so in publications.

STEP 2: Pilot the model with a major publisher/major labs

Like it or not, most of us try to emulate standards set by high-impact journals and the scientists who publish there regularly. Once a journal like Nature, Science, Cell, PNAS or eLife makes eliminating author lists a mandatory policy, along with the support of leading scientists in the field, others will follow.

Eventually, the only way this model becomes reality is if there is sustained interest among Early Career Researchers and some publisher(s) to try it out.

As a first step, vote for the Film Credit model in this 1 minute survey and spread the word.

In the meanwhile, I firmly retract my response to this tweet: