Note: this was an essay I wrote as homework for one of my experimental physics classes (in Institute X, they bend over backwards to drill scientific integrity into students).
Constraint: react to On Being a Scientist: A Guide to Responsible Conduct in Research, by various authors.
First and foremost, the job of a scientist is to uncover the truth. His measure is the clarity and novelty of his thought. If he is exacting and thorough, lining every table, checking every decimal point, but in the end writes only echoes of his teachers, then he is no greater for it. If he becomes head of his department by a curious doggedness but in doing so neglects to pursue new ideas, then he is no closer to the business of science than pedestrians on the street. A scientist should keep his identity small: in doing so he avoids the rituals and inauthentic behaviour to which science is no less immune.
The Atomic Age thrust science in the public’s view. No longer were scientists seen alone in their labs, concocting various mechanisms to demonstrate in cramped classrooms. We started writing letters to governments, affecting foreign and economic policies. We began dominating intellectual circles for better or for worse: in the 1920s it was customary to apply “relativity” to the world’s moral dilemmas [insert citation here]. This narrative is not new, but herein lies an important distinction: all science is a set of tools. If these tools are used for nefarious goals, if they are used for the betterment of humanity, they are still truthful knowledge. The responsibility lies in the person who finds their use. It is because of this that we must distinguish the scientist from the human: we hold ourselves culpable for consequences of our research because we are human, not because we sought dangerous knowledge that must not be sought.
To this end, I shall call ‘scientist’ that who seeks knowledge for knowledge’s sake and ‘science-practitioner’ the scientist with human faults and foibles. The science-practitioner in today’s world faces two main challenges: to collect citations, and to secure funding. Immediately this brings us to various conflicts of interest that may hamper the progress of scientific research. The first one is the maligned focus on credit. We are extremely protective of our ideas. It is as if the conception of an idea in one’s mind (particularly if one has reason to claim priority) leaves us with the same psychological imprint as finding a coin in the street. This is evident in the well-known rivalry of Isaac Newton and Gottfried Leibniz on the discovery of calculus, where what could have been a fruitful correspondence turned into a 20-year-long state-backed superiority contest no better than the red-team, blue-team bickering often seen in sports or politics. The scientist, in his pursuit of knowledge, must learn to avoid this pitfall lest this primal need for social status overwhelms his path to his original goal.
This issue of credit is such an important point that it warrants a longer discussion. Aside from priority, scientists also bicker about authorship. A great importance is placed on the relative order of authors in a paper. First authorship implies a bulk of the intellectual contribution to the paper; a project is after all never equally divided. More issues arise when advisors and researchers with higher administrative positions demand inclusion in the list of authors of a study, regardless of their intellectual contribution to it. Thus, a scientist is never alone. In his pursuit of knowledge, he is haunted by the spectre of his heroes, he must walk briskly keep pace with his highly competitive peers, he must let students trail behind him to carry on when he falls by the wayside, and he must carry the ever-increasing burden of administrative responsibilities on his back.
A more subtle issue on authorship arises when different parts of a paper are written by different authors. A broth is spoiled by many cooks, so why should a paper be any different? The coherence of such a multi-authored paper is rarely apparent, and problems arise when one of the authors is involved in an academic scandal. The case of Jan Hendrik Schön, an experimental physicist who was found to have fabricated semiconductor data, is now a classic warning to budding scientists. The larger question, however, is the culpability of his co-authors. Are they as guilty as the fraudulent physicist by having signed off the paper for publication? What about the people who peer-reviewed his papers? How apt is it to fault them for Schön’s career to have lasted as long as it did?
A naive attempt to cut the Gordian knot would lead one to express culpability for all those involved. “Everyone gets the whip” is a common sentiment for those who would like to get on with their lives after scandals such as these. However, a brief pause would lead one to conclude that this undermines the foundation of trust that the scientific community puts on its members. The pure scientist demands empiricism in everything, but practicality compels the science-practitioner to leave the replication to others in their respective fields and focus on extracting the pieces of information relevant to his problem. To say that we must check and double-check our co-authors for fraudulent behaviour undermines this web of trust and wastes precious hours that could have gone to research.
Of course, it is lunacy to suggest that checks and balances be done away with. That web of trust only works if it can be trusted (and this is not a trivial tautology). What I am advocating rather, is for the community to spend some of its research-hours into building automated verification systems. The advance of machine learning has increased steadily in previous years. Everywhere, we are experiencing the fruits of breakthroughs in recognition systems, from self-driving cars to a cleaner inbox. There is no physical law that prohibits spam classifiers to be aimed at scientific papers instead. Perhaps, it would even be possible to automatically rate the credibility of a researcher and possible conflicts of interest. Science may be methodical, but it need not be manual.
This brings me to my next point. Science, as a profession that defines itself in how it prods and tests the borders of human knowledge, is strangely resistant to alternative systems. There is a huge pressure to converge on professional standards detailing and constraining the various aspects of one’s work. Strange, that we extol empiricism in everything but our own practice, that we so easily peer outside the public window of discourse yet fail to see more efficient research processes. Science may pride itself on being methodical, but it need not be slow.
Consider how a random civilisation would develop its scientific community. Would you imagine that it would start with lone individuals speaking out against the mores of society, like our Grecian philosophers of old? Would you imagine mathematics intertwined with the practicalities of business and war and the economy, until its parts are distilled one by one? A Galileo perhaps, waging a public battle against old institutions. Then guilds and universities. Then the Industrial Revolution. If so, then one must work to broaden one’s horizons. It is a fallacy to suppose that societies will converge to our own, an implicit assumption of the superiority of one’s culture. If not, then you understand: there is a vast number of paths our science could have taken to get to this point. Therefore, there is also a vast number of scientific processes that could have gained foothold by the arrival of global communication. Our science is the conglomeration of different ways of expressing empiricism, some more affected by extenuating socioeconomic goals than others. It is therefore a curious and frightening prospect that our science was grown, not designed.
The second aspect of a science-practitioner’s career centers on funding. Money is the prime mover. It directs the actions of humans much more so that it is polite to admit. Its main use to the science-practitioner is in the procurement of devices necessary to conduct research. Computer systems, laboratory equipment, technicians and operating crew for heavy apparatuses: the list goes on and on. Money buys tools for the toolmaker, and as tools are said to multiply forces, so do they expand the range of phenomena within reach. A scientist without tools is left to use only his mind, and the mind can only carry so much.
What compels science-practitioners to spend an inordinate amount of hours writing grant proposals? The production of tools for toolmakers is an economy unto itself. There is a huge variety of laboratory equipment accessible to research institutions, if they have the money. Always there is a drive to purchase better and better equipment, and this is not entirely unreasonable: all the eyes in the world could never have guessed the existence of microbes without having seen one for itself. As the phenomena we investigate get more and more exotic, so must our sensory capabilities expand.
By virtue of interacting with the economy, however, this procurement process gains its own incentives. A pure scientist will whittle away everything he has to spend on furthering his research. A science-practitioner, by having to exist in reality, must treat his money as a resource and strategically place his bets on lines of study that to him would prove most fruitful. Immediately, this wrests control from the scientist to pursue his own research directions and gives it to the ebbs and flows of the economy.
There is no problem with this picture: usually we do not have perfect knowledge of what we must know (if we did, it would not be called research). This foible more than makes up for the efficiency we would otherwise gain from giving complete control to the scientist.