Self-Censorship and the Cult of Productivity in Academic Research

This is the English translation of an article that was originally published in German as part of the annual essay collection of Laborjournal (publication date Jul 7, 2020). 

Science finds itself exposed to an increasingly anti-intellectual and post-factual social climate. Few people realise, however, that the foundations of academic research are also threatened from within, by an unhealthy cult of productivity and spreading career-oriented self-censorship. Here I present a quick diagnosis with a few preliminary suggestions on how to tackle these problems.

In Raphael's "School of Athens" (above) we see the ideal of the ancient Academy: philosophers of various persuasions think and argue passionately but rationally about the deep and existential problems of our world. With Hypatia, there is even a woman present at this boy's club (center left). These thinkers are protected by an impressive vault from the trivialities of the outside world, while the blue sky in the background opens up a space for daring flights of fancy. The establishment of modern universities — beginning in the early 19th century in Berlin — was very much inspired by this lofty vision.

THE RESEARCH FACTORY

Unfortunately, we couldn't be further from this ideal today. Modern academic research resembles an automated factory more than the illustrious discussion circle depicted by Raphael. Over the past few decades, science has been trimmed for efficiency according to the principles of the free-market economy. This is not only happening in the natural sciences, by the way, but also increasingly in the social sciences and the humanities. The more money the taxpayer invests in academia, the higher the expectation of rapid returns. The outcomes of scientific projects should have social impact and provide practical solutions to concrete problems. Even evolutionary theorists must fill out the corresponding section in their grant applications. Science is seen as a "deus ex machina" for solving our societal and technological problems. Just like we go to the doctor to get instant pain relief, we expect science to provide instant solutions to complex problems, or at the very least, a steady stream of publications, which are supposed to eventually lead to such solutions. The more money goes into the system, the more applied wisdom is expected to flow from the other end of the research pipeline.

Or so the story goes. Unfortunately, basic research doesn't work that way at all. And, regrettably, applied science will get stuck quickly if we no longer do any real basic science. As Louis Pasteur once said: there is no applied research, only research and its practical applications. There are no short cuts to innovation. Just think about the history of the laser, theoretically predicted by Albert Einstein in 1917. The first functional ruby laser was constructed in 1960, and mass market applications of laser technology only began in the 1980s. A similar story can be told for Paul Dirac's 1928 prediction of the positron, which was confirmed experimentally in 1932. The first PET-scanner came to market in the 1970s. Or let's take PCR, of Covid-19 test fame. The polymerase chain reaction goes back to the serendipitous discovery of a high-temperature polymerase from a thermophilic bacterium first described by microbiologists Thomas Brock and Hudson Freeze (no joke!) in the hot springs of Yellowstone Park in the 1960s. PCR wasn't widely used in the laboratory until the 1990s.

A study from 2013 by William H. Press — then a science advisor to Barack Obama — presents studies by economist and Nobel-laureate Robert Solow, which look at the positive feedback between innovation, technology, and the wealth of various nations. Solow draws two key conclusions from his work. First, technological innovation is responsible for about 85% of U.S. economic growth over the past hundred years or so. Second, the richest countries today are those that had first set up a strong tradition in basic research.

Press argues, building on Solow's insights, that basic research must be generously funded by the state. One reason is that it is impossible to predict which fundamental discoveries will lead to technological innovations. Second, the path to application can take decades, as the examples above illustrate. Finally, breakthroughs in basic science often have a low appropiability, that is, money gained from their application rarely flows back to the original investor. Think of Asian CD and DVD players equipped with lasers based on U.S. research and development, which yielded massive profits while outcompeting more expensive (and less good) products of American make. This is the economic argument why state-funded basic research is more important than ever.

EFFICIENCY OR DIVERSITY? 

But here exactly lies the problem: basic research simply does not work according to the rules of the free market. Nevertheless, we have an academic research system that is increasingly dominated by these rules. Mathematicians Donald and Stuart Geman note that the focus of fundamental breakthroughs in science has shifted during the 20th century from conceptual to technological advances: from the radical revolution in our worldview brought about by quantum and relativity theory to the sequencing of the human genome which, in the end, yielded disappointingly few medical advances or new insights into human nature. A whole variety of complex historical reasons are responsible for this shift. One of these is undoubtedly the massive transformation in the incentive structure for researchers. We have established a monoculture. A monoculture of efficiency and accountability, which leads to an impoverished intellectual environment that is no longer able to nourish innovative research ideas, even though there is more money available for science than ever before. Isn't it ironic that this money would be more efficiently invested if there was less pressure for efficiency in research?

Researchers that need to be constantly productive to progress in their careers, must constantly appear busy. This is absolutely fatal, particularly for theoretically and philosophically oriented projects. First of all, good theory requires creativity which needs time, inspiration, and a certain kind of productive leisure. Second, the most important and radical intellectual breakthroughs are far ahead of their time, without immediately obvious practical application, and generally associated with a high level of risk. Who tackles complex problems will fail more often. Some breakthroughs are only recognised in hindsight, long after they have been made. Few researchers today can muster the time and courage to devote themselves to projects with such uncertain outcomes. The time of the romantics is over; now the pragmatists are in charge. Those who want to be successful in current-day academia — especially at an early stage of their careers — must focus on tractable problems in established fields, the low-hanging fruit. This optimises personal productivity and chances of success, but in turn diminishes diversity and originality of thinking in academic research overall, and wastes the best years of too many intrepid young explorers. Unfortunately, originality cannot be measured, while productivity can. Originality often leads to noteworthy conceptual innovations, but productivity on its own rarely does.

Goodhart's Law — named after a British economist — says that a measure of success ceases to be useful once it has become an incentive. This is happening in almost all areas of society at the moment, as pointedly described by U.S. historian Jerry Z. Muller in his excellent book "The Tyranny of Metrics." In science, Goodhart's Law leads to increased self-citations, a flood of ever shorter publications (approaching what is called the minimal publishing unit) with an ever increasing number of co-authors, as well as more and more academic clickbait — sensational titles in glossy journals — that deliver less and less substance. Put succinctly: successful researchers are more concerned about their public image and their professional networks today than ever before, a tendency which is hardly conducive to depth of insight.

What follows from all this is widespread career-oriented self-censorship among academics. If you want to be successful in science, you need to adapt to the system. Nowhere (with the potential exception of the arts) is this more harmful than in basic research. It leads to shallowness, it fosters narcissism and opportunism, and it produces more appearance than substance, problems which are gravely exacerbated by the constant acceleration of academic practice. Nobody has time anymore to follow complex trains of thought. An argument either fits your thinking habits, what you see as the zeitgeist of your field, or it is preemptively trashed upon review. In the U.S., for example, an empirical study has found that those biomedical grant applications are favoured that continue the work of previously successful projects. More of the same, instead of exploration where it is most promising. And so the monoculture becomes more monotonous yet.

FROM AN INDUSTRIAL TO AN ECOLOGICAL MODEL OF RESEARCH PRACTICE

How can we escape this vicious circle? It is not going to be easy. First, those that are profiting most from the current system are extremely complacent and powerful. They can show, through their quantitative metrics, that academic science is more productive than ever. The loss of originality (and the suffering of the victims of this system) is hard to measure, and therefore no major issue. What cannot be measured does not exist. In addition, the current flurry of technological innovations (mostly in the area of information technology) give us the impression that we have the world and our lives more under control than ever. All of this supports the impression that science is fully performing its societal function. 

But appearances can be deceptive. Indeed, we do not need more facts to tackle the existential problems of humanity. What we do need is deeper insight, more wisdom, and just like originality, these cannot be measured. There are cracks appearing in the facade of modern science, which suggest we must change our attitude. I've already mentioned the Human Genome Project, which cost a lot of money, but did not deliver the expected profusion of cures (or any deeper insight into human nature). Even less convincing is the performance of the Human Brain Project so far, which promised us a simulation of the entire human prefrontal cortex, for a mere billion euros. Not much happened, but this is not surprising, because it was never clear what kind of insights we would gain from such a simulation anyway. These are signs that the technology-enamoured and -fixated system we've created is about to hit a wall.

Since the main problem of academic science is an increasing intellectual monoculture, it is tempting to use ecology as a model and inspiration for a potential reform. As mentioned at the outset, the current model of academic research is indoctrinated by free-market ideology. It is an industrial system. We want control over the world we live in. We want measurable and efficient production. We foster this through competition. As in the examples of industrial agriculture and economic markets, the shadow side of this cult of productivity is risk-aversion and the potential of a ruinous race to the bottom.

What we need is an ecological reform of academic research! Pretty literally. We need to shift from a paradigm of control to a paradigm of participation. Young researchers should be taken seriously, properly supported, and encouraged to take risk and responsibility. What we want is not maximal production, but maximal depth, sustainability, and reproducibility of scientific results. We want societal relevance based on deep insight rather than technological miracle cures. We need an open and collaborative research system that values the diversity of perspectives and approaches in science. We need a focus on innovation. In brief, we need more lasting quality rather than short-term quantity. Our scientific problems, therefore, mirror those in society at large pretty exactly.

STEPS TOWARDS AN ECOLOGICAL RESEARCH ECOSYSTEM

How is this supposed to work in practice? I assume that I am mostly addressing practicing researchers here. This is why I focus on propositions that can be implemented without major changes in national or international research policy. Let me classify them into four general topics: 

• Good research practice: as a researcher, it is easy to complain about the system and how it makes everybody suffer. What is often lost is the obvious fact that we are the system! Many inconveniences could be avoided by simple changes in our own behaviour, in the context of our own local research community. Much would already be achieved if we could refrain from engaging in cheap ad hominem attacks and, in turn, not take every criticism personally. A more constructive discussion culture is key. This is especially important for (anonymous) peer-review. It is virtually impossible these days to get past referees with a daring or speculative idea. A bit more trust and appetite for risk would not harm anybody. And remember this when reviewing: you are supposed to review the quality of the manuscript or project as proposed by the applicant, not replace it with your (undoubtedly better and more sophisticated) idea of what should be done. Positive feedback is much too rare in academic circles. All of this could be easily changed by a slight tweak of your own habits. Encourage and support collaboration over competition locally, in your own research group, your own department or institute. The same applies to the implementation of rules for transparency and quality control. Sharing is caring. The format of seminars, talks, workshops, conferences, and retreats should be focused on an exchange of ideas (not the usual parade of show-offs), and needs to be properly moderated to achieve this. And why not take young researchers more seriously, for instance, by including them in relevant discussions and decisions? Most of them are intrinsically motivated, responsible, and not at all stupid. It is amazing what kind of difference a tolerant and inspiring environment can make. I have experienced this many times first hand on my journey through numerous research institutions and cultures.

 

• Incentives: the core problem of academic science today lies in its one-sided and perverse incentives. Productivity is all that counts. Here also, much can be achieved at the local level. Researchers should not only be evaluated based on the funding they bring in and the publications they push out. Teaching, outreach, and services to the local community must be officially acknowledged. Risk-taking and interdisciplinarity should be fostered and supported, not only locally but at a more general level. More researcher-focussed long-term funding should be available to talented young researchers, to support early independence and to minimise bureaucratic overhead. The misguided standard notion of accountability should be dropped entirely: it makes no sense to insist that any project follow its work plan exactly, month for month, as long as work progresses in an interesting direction. Less accountancy, more creativity! What should be evaluated at an interview is the potential of a candidate, not their past productivity (which puts risk-taking explorers at a disadvantage). In general, a steady flow of research publications is no indicator for the quality of a researcher's ideas.  Things like obsessive-compulsiveness or a 60-hour work week are warning signs, not criteria for excellence. It is time that group leaders stop to enforce or even encourage such behaviour. Researchers who are mentally balanced deliver more interesting and solid results in the long run.

 

• Education: we urgently need a radical reform in the way we train young investigators. Today's higher education is still too much focused on rote learning and the regurgitation of facts at exams. This does not make anyone a creative researcher. Although solid foundational knowledge is essential, it is only a means to an end. And that end is the formulation of interesting research questions and the perseverance to pursue them. The best researchers excel through original approaches and thinking, skills which are best leant through applied practice and independent study. Unfortunately, flipped-classroom teaching techniques put high demands on workload and personnel. A hybrid approach with online lectures and on-site discussion/mentoring groups can render it quite practical in most settings. What we need are mature graduates, independent explorers who are driven by a fascination with the unknown (and who can live with the uncertainty that comes with it), not high-bred lab technicians who meticulously and efficiently do the bidding of some tyrannical group leader. A flipped-classroom education, by the way, also profits those students who decide not to go on in science. Instead of useless facts, they get transferrable skills for their onward journey.

• Philosophy, Science, Society: finally, I think it's extremely important that researchers start to seriously question themselves again. More reflection is required, not just in terms of one's personal motivation, but also concerning academic research in general and its broader role in society. Why are we here? Why are we doing this? What is the purpose of basic research? How do we generate knowledge? What is scientific knowledge anyway? (A notoriously difficult question to answer.) And in what relation does it stand to wisdom? (Even harder.) These are discussions we must have, over and over again. Philosophical questions belong back in any scientific curriculum. Society needs science more than ever. But not the industrial kind. As the famous sociobiologist E. O. Wilson aptly put it: "We are drowning in information, while starving for wisdom." In this sense, we do not need more highly qualified specialists, but more daring thinkers and explorers, who tackle the existential questions of humanity. More humanity in the sciences, instead of more quantification in the humanities.  Wouldn't that be nice?