Science is Awesome, Science is Broken
Why we need radically different institutions for science, and what achievements they must preserve.
Life expectancy is on the rise, with new high-tech diagnostics tools and medication. Our technology has reached “magic” level, in air, space, computation and telecommunication.We can grow lots of food with little human effort, and we can kill lots of people with little human effort. We have a pretty good understanding of many systems and processes, from the formation of galaxies and stars to the biochemical machinery in our cells. Much of this we would not have without science, the methodical empirical testing of our beliefs. Moreover, much of this we would not have without industry and publicly supported, institutional science.
Institutional science can be divided into three parts. The first part is institutions, with addresses, budgets, staff, and aims: school and university science departments, university research labs, government and private research institutes, government and philanthropic funding agencies, scientific publishers, and science museums. The second part is communication methods: conferences, journals, archives, citations, textbooks, databases, websites, and patents. The last part is shared tools: theories, paradigms, models, experiment protocols, machines, code, reagents, jargon, scepticism, rules of argument, ways of assessing skills and expertise, accepted behaviour, public engagement.
Not all is well in the land of science, and the ills are numerous and varied. Private companies finance the publication of positive research on their products, via ghost authors behind academic fronts. Leading researchers spend more time applying for funding than doing research or training their lab members. Young academics are rushed to publish incomplete research or pushed to manipulate their data, in order to get the publication record required to stay in a science career. It is hard to publish unorthodox or interdisciplinary research, and even harder to get funding for it. Too much of the research supported by the public, including via tax breaks to private R&D, cannot be optimally used by that public. Public science is not accessible because it is protected by paywalls and patents, or because it is written in inaccessible language, or because its reliability cannot be estimated due to lack of replication and lay expertise. Finally, public funds are spent on unnecessary machines in a prestige arms race.
The institutions of science don’t have to look like they do now. They were all created, often by a relatively small set of individuals, at some point in the history of science, which means anytime between the 17th century and the present. Their initial design was made to fit very specific social and political constraints, and they have since evolved in the semi-chaotic development of large bureaucratic organisations, under diverse and dynamic pressures from within and without. While this should indicate some remarkable resilience among these institutions, especially the older ones, it can also suggest that we might be able to do a lot better, with some new tools and some open discussion.
To start off the discussion, here are some thoughts. Publicly funded science should ultimately serve the public. Specifically, it should serve it in the three following ways. First, public science should eventually provide affordable and reliable information for the shaping of public policy and for the running of public services. Second, it should address areas of research that are of ultimate commercial interest, where expected payoffs are too far off or too disconnected to be of interest to private firms. Finally, it should provide answers to questions that are of interest to individual citizens, or provide citizens with the information and tools to explore these questions themselves.
Research is inherently an unpredictable endeavour, and should be treated as such. While we may want research to address certain questions, we cannot be sure in advance how to come up with the answers. After all, science is not magic. It is simply a very broad set of semi-reliable tools for coming up with answers, and a loose set of instincts for coming up with more of these tools. If we learn to accept this we can bring about a significant lowering of unnecessary pressure and some healthy reduction in expectations. We will then be able to start eliminating the voodoo practices whose sole purpose is to hide the unpredictability of research. Such practices include mile-long research proposals, and high-profile multi-year projects to discover the origin of everything. Perhaps we can start distributing science funds more freely and widely, and perhaps we should do it by lottery to guard against all sorts of biases and corruption.
We should worry less about what gets researched, and worry a lot more about what happens with the results of research. Where is all that information going, and how do we make sure it is put to good use? Perhaps some of the outputs of research should be recognised as valuable or dangerous to public well-being, and should be published at public expense, or protectively-patented, or placed under restricted access. This would be particularly relevant to research that affects underprivileged groups, such as health research, or research into social inequalities.
We should make the results of research more accessible as well. This should be approached from two sides. From science to citizens, potential science popularisers (animators, bloggers, creative persons of all flavours) should be encouraged and allowed to make a living from turning high-quality but jargon-heavy papers into accessible, reliable, consumable formats. From citizens to science, self-education, amateur experimentation, and locally coordinated curiosity spaces can help us make use of the information that is already there, and to formulate better questions for the next generation of scientists. Internet and web video can be brought into every corner of science, especially amateur science. It will generate lots of skilled persons rather quickly, and it would also be jolly good fun.
While we push these new horizons, and do away with musty old practices, we should be careful not to lose the gems of science. In science, more than anywhere else, disagreements are resolved by data, not by rhetoric or popular vote. Science is not a popularity contest, and should never become one (so no crowdfunding for science please). We have vast tomes of knowledge and experience inherited from centuries of finding things out, more or less imperfectly. To throw them away because they are not in the right format would be unfortunate and irresponsible. Finally, the people staffing and engaging with the institutions we have now may be old fashioned at times, but they know way more than we do about living with science and living to do science, so we should listen to them. But we should remember to raise an eyebrow when they say everything is hunky dory, or that it would all be perfect if we throw some more money their way.
