The last four decades have seen a steady increase in the number of authors on scientific publications. Since 1975, when there was on average 1.9 authors per paper, we have seen increases each decade to 3.12, 3.76, 4.61 and finally 5.12 authors per paper in the period 2010-2013. It is clear that science has become a team sport, yet our rewards mechanisms are still almost completely based on individual achievements.
Individuals get papers. Individuals get jobs. Individuals even get Nobel prizes. But… individuals don’t do science by themselves.
Many of my colleagues will know that I have a mountain/valley analogy for the authorship list of a life-science paper. The first and last author peaks are given the highest importance and the valley exists in between where the “others” live – higher up on either side is even marginally better.
But what are the consequences of this misaligned reward system?
Ambition trumps science
Everyone who is in a medical science lab (and probably others!) will be familiar with authorship discussions/battles. Horror stories emerge from research groups about “stealing the credit” or choices being made for “political/career reasons.” These incidents are not uncommon, but they are also subject to each person’s opinion. I often joke with my colleagues about how every research paper is actually 200% work because if you total each author’s perceived percentage contribution it would always exceed 100%. But the important first/last places are always claimed by the “leaders” – real or perceived – and this raises a whole range of sociological issues for how that is determined.
No reward for helping others
For me, this is by far the biggest problem. Imagine you are in research to reach a particular goal (e.g., cure breast cancer) and you work toward that goal by getting really good at a particular job that is essential to advancing cancer biology. Let’s imagine you develop a technique that lots of people use but one that rarely forms the main thrust of the “publishable story.” You have helped advance cancer research, but your academic career could be at risk (e.g., no first-author papers). If you do this at the postdoctoral stage of your academic career – you’re toast.
The logical career path therefore is to only work on projects that have the potential of being first-author papers. In other words, we actively discourage working together on a project. One possible solution is listing people as equal contributors (also known as joint first authorship), but this does not always work out so well for the second name listed. Indeed, this makes me wonder how things play out with joint first authorship between the sexes – my guess is that men more often lay claim to the pole position of a “joint” first author paper…
Academics could take some lessons from companies
Industry works almost completely differently when it comes to this – a defined goal is set (e.g., develop product X that can do job Y) and a team of people is assigned to achieve it. If there is a particular person that does an outstanding job, they get rewarded (maybe a promotion or a bonus), but you rarely see “wonder drug, designed by Jon.” While I can certainly appreciate the unpredictability of scientific research, there are still broad goals. Notably, industry also manages to figure out how to recruit and retain people as well making me wonder what metrics are used – they probably do not rely on first/ last author publications.
So, yes, there is a problem – but how do we solve it? Journals have started to request individual author contribution statements which is a good thing, but the vast majority of these simply say “did research” or “performed experiments” – hardly informative (or noticeable!) should you be on a hiring committee judging a scientist’s merit. A very interesting article in 2007 from PLoS Biology pitched the idea of splitting the impact factor of a paper across the authors by percentage contribution. This would have the added benefit of policing the number of and/or subsequent credit given to “ghost” contributors (those who gift a reagent or patient sample that is important to the paper, but really had no involvement in the paper itself).
At the end of the day – the public via government and charitable funding provides academics with a huge pot of money to help advance society through ideas, healthcare solutions, and technological breakthroughs. Our current rewards system is designed to get a single person working on each problem – and if it’s a big problem, many single people working individually with the person that gets there first getting the reward. Am I saying that an individual cannot have an excellent idea and pursue it through to completion? Absolutely not – but the reality of the situation is that society ends up better if we work together to win the world cup, not individually to get the golden boot.
Couldn’t agree more Dave, perfectly written!!
In my area (but that’s pure math, very far away from life sciences) these debates simply don’t exist, as multiple authors are always considered to have contributed more or less the same to the paper (or more precisely, I don’t think we really believe that trying to quantify “I contributed 30% to this paper” can possibly make any sense), and are always listed alphabetically. I’ve been careful in paying attention whenever I read a paper, and I have yet to find a counterexample to this. I do believe this kind of mindset is very fruitful, and may be one of the main reasons behind the success of a few highly collaborative collective discoveries (if you want examples, just Google “the polymath project”).
Great article but just one caveat-
I won’t say there is no reward for helping others, as long as it leads to a co-authorship on a paper/presentation. Given the increasingly inter-disciplinary nature of biomedical research, I would expect more and more people to pursue collaborations and co-operations in the future.