Seán Dunphy explores how the structures of academic funding and publishing have shaped modern scientific culture.
Scotsman Alexander Fleming’s discovery of penicillin is one of the most impactful in science, but Fleming’s intention was never to study antibiotics; rather, it came as an accident while studying Influenza. Like Fleming, neither Jared Leadbetter nor Joseph Mougous initially intended to study the topics of their headline-grabbing papers published this week. Manganese metabolising bacteria, which have been long-hypothesised but thus far undiscovered, were uncovered by Leadbetter after he noticed a colour change in a manganese stained jar which had been left in a sink to soak. A fortunate accident, but nonetheless required his scientific curiosity and instinct. Similarly, Mougous’ discovery of a method for genetically editing mitochondrial DNA developed tangentially from a study of bacterial toxins. The discovery followed an observation that unlike other bacterial toxins, which act on single stranded DNA, DddA mutates double stranded DNA.
Both microbiologists’ discoveries could be attributed to happy accidents – or more appropriately, the patience, drive and freedom to follow an aside – but unplanned discoveries like these are becoming more and more uncommon in research. The restrictive nature of academic funding, and its dependence on regular publishing, is a growing obstacle preventing researchers from following their instincts and curiosity in favour of safer pursuits.
Publish or perish: the measure of a scientist
Academic funding is mostly available in the form of grants which rarely exceed 4 years. This short time frame forces fast science, in order to generate flashy results on which to base the next grant proposal. This high-pressured constraint leaves little scope for long term projects; no flexibility to pursue tangents; and above all else, no room for failure. In this environment, the pressure to follow the safest path to publication surpasses the instinct to ask truly engaging questions. “Publish or perish” – words which weigh heaviest on the base of the pyramid. Ever conscious in the minds of early career researchers, this weight creates an ever more cautious, short-sighted scientific community.
Science, as an endeavour to improve human life, is mostly funded by government bodies. This places a responsibility on the grant holder to use this funding efficiently, thus requiring the cooperation of the entirety of an academic field to avoid, for example, needless repetition of the same study by different research groups – but academia is incredibly competitive. The number of successful PhD candidates increasingly exceeds the availability of faculty positions, and the number of principal investigators increasingly exceeds the availability of funding. This disequilibrium in the academic system is exacerbating the challenge of securing funding. Since 2004, there has been a 33% decrease in the proportion of successful NIH proposals. We cannot allow this trend to continue; the volume of grant applications required of a researcher demands an already excessive portion of their time.
In this hyper-competitive world, publication records become a quick gauge of a researcher’s worth, aggravating the need for multiple publications within a single 4-year grant cycle. Intense competition between rivals for publication space manifests as a reluctance to collaborate or present ideas. The fear of being “scooped” wastes resources by creating redundancy and slowing scientific progress. The well-established use of publication as currency drives division, which not only impedes research and neglects science’s responsibility to the public who fund it, but also forces rushed negligent research and prevents any deviation from a predetermined research path.
Loud science or good science?
To stand out, scientists favour hype and the overuse of positive vocabulary. Use of words like ‘novel’, ‘pioneering’, or ‘significant’ are viewed by many of those preparing manuscripts as a requirement for publication. In an environment where failure is not an option, and hype is a necessity, the only viable strategy for protecting an academic career is pursuing safe hypotheses with a high probability of producing “significant” data. This perspective has caused a crisis in biomedical publishing. Statistical significance has been used as a means of protecting the impact of a paper, creating rampant misuse of p-values as an inference of statistical significance. In the age of clickbait, in which social media platforms are vital tools for improving the impact of a paper, this crisis requires immediate intervention. If not, scientists will continue to play it safe with their research goals, and the epidemic of hype in science will worsen.
Competition for publication leads to a strong affinity amongst researchers for publication trends, creating research bubbles; the direct result of the fanfare placed on emerging, often implausible reports. Unable to sustain the unwarranted excitement, these bubbles collapse – most often upon an initially over-stated study which lacks sufficient rigour – leaving those who followed it in a baseless field with little hope of publishing. The exaggerated findings of these initial reports are often blamed on the competition between publications for the greatest impact factor, but inadequate peer review as a result of a saturated peer review system is equally as likely.
In addition to exaggeration, a wider publication bias exists favouring positive results over those that are passive or negative. This bias creates a high-risk environment which deters against the inherent unpredictability of truly ground-breaking research. The current reward system in science is failing to incentivise well designed rigorous science, in favour of a fast-flamboyant version of research. In fact, in a tribute to twice Nobel prize winner Fred Sanger, Nobel laureate Sydney Brenner speculated that “a Fred Sanger would not survive today’s world of science”. His slower, methodical approach to research resulted in periods of sparse publication which would have led to him being branded unproductive, preventing acquisition of further funding.
If left unchecked, the evolution of the scientific environment further in favour of fast science will prevent remarkable discoveries such as that of penicillin, or drive researchers with the potential and fore sight of Fred Sanger from academia.
In favour of fast science
Despite the restriction imposed upon curiosity by the current environment, there is no doubt that we are currently in the most productive period of scientific research in history. This may suggest that faster science must have its strengths.
Through the practice of regularly writing grant applications, scientists gain the routine opportunity to reassess a hypothesis, breaking longer projects into more regular milestones, each step therefore requiring further critique of the hypothesis and experimental design.
The competition pushing fast science naturally selects for scientists who constantly innovate and are efficient in their use of funding. Since sharing resources through collaboration contributes to the stretching of funding, those with a strong ability to collaborate excel, creating a more integrated and dynamic scientific community. This interdisciplinary approach brings a fresh perspective to many disciplines, evolving the fields and creating further technological advancement.
Similarly, competition in publishing forces researchers to re-think their approach to publishing. The fear of being beaten to publication has contributed to the emergence of more efficient forms of sharing ideas. Presenting data to the greater community through preprint publication on platforms like arxiv is ever more common, significantly increasing the overall speed at which science progresses.
Each of these advantages are seen in the greatest achievement of the current scientific environment – how it has reacted to the COVID-19 pandemic. The scientific community has proven its ability to adapt to the ever changing needs of the public who fund it. Within the first 5 months more than 2,500 studies on COVID-19 were published, a truly astonishing feat of fast science.
Demand rigor and encourage curiosity
Modern structures of scientific funding and publishing have created a frantic, highly competitive environment that drives productivity and innovation, but also an environment with no room for failure or deviation from the predetermined plan. Since science by its nature strives for efficiency and innovation, and this system of funding does little to incentivise rigorous research, it is no longer fit for purpose.
Researchers have proposed many solutions to the many shortcomings of this system. There is ubiquity in the suggestion that funding systems should be redesigned to take the emphasis off publication history in favour of collaborative skills and rigorous experimental design. However, the design and implementation of a system effective at a scale relevant to the volume of applicants is a daunting task. Other suggestions would be more easily implemented, for example limiting the number of grants an individual can hold, or funding individuals for longer periods instead of funding short-term projects, but there is no doubt that these ideas would garner abundant controversy; particularly among long-established academics with formidable political clout.
Similar to longer grants which better incentivise collaboration of rivals, Brian Martinson’s suggestion of a lifetime word limit as a means of incentivising research conducted and communicated with utmost care would result in fewer, but more rigorous, papers. This could reduce the competition in publishing and ultimately, the use of exaggeration and hype. It is critical that this is addressed now as publication adapts to the open access movement, since traditional publications may attempt to maintain their relevance by focusing on their ability to advertise papers, leading to further exaggeration of scientific findings.
Science has entered a new internet age in which print media, in its traditional sense, is redundant, and the sharing of information is less regulated. The traditional systems for incentivising and sharing research are quickly becoming inept, generating an excess of publication, placing an unsustainable burden on peer review. Science must return its focus to the basics of the scientific method, and the reward system must adapt to incentivise patient, rigorous science, which allows remarkable tangential studies, like those uncovering manganese digesting bacteria, tools for editing mitochondrial DNA, or even penicillin.
Written by Seán Dunphy and edited by Ailie McWhinnie.
Ailie’s thoughts… In lieu of an author’s comment, as we have been lucky enough to hear an article on Seán’s thoughts already, I was eager to throw in my own two cents as an editor’s comment in it’s place.
In just one year – one – of working in a lab, my image of science was flipped. Of a world limited only by imagination and willing, to one of paperwork and planning. Of course, planning is key to efficiency, but I wasn’t prepared for the limits on exploration. As our results started to come in, new avenues – big, important questions that we needed answers to to get a meaningful picture of our work – opened up. Questions we could not have foreseen when initially planning our work, therefore ones we could not pursue as we had not laid the methods out in our original road map.
I can see, in theory, that such regulations seem sensible. In a world where we are racing to combat so many crises – be that medical, agricultural, environmental – there is simply too much to do for time and resources to be wasted.
But surely it also seems sensible to allow researchers the freedom to change course if something else emerges en route that seems more important, with more potential, and more relevant – whether to the intended field, or something entirely different but equally pressing.
Science is unpredictable, so why on earth should we predict our discovery methods?
My exception, and perhaps one of the biggest limitations in research in the UK, is when animals are involved. In these cases, we must indeed be cautious and I think Seán’s points about collaborations to avoid redundancy are more important than ever in these cases. I hope that in the future we can rely more heavily on technologies such as stem cell models, which will bring a double bonus of reducing animal usage and making research more flexible by by-passing the strict planning needed for such experiments.