Myth of Objective Science Busted
Students and citizens are taught a very distorted view of what science is and how it actually works.
Basic science is not an unbiased knowledge generator. Daniel Sarewitz pulls no punches in Nature this week. “Kill the myth of the miracle machine,” he shouts in his column that stabs one of science’s most treasured sacred cows: the value of “basic science.” The very, very few cases where undirected investigation has actually produced some worthwhile findings do not justify calls for non-targeted political funding, he argues, nor do the leftist attacks on funding cuts for basic research justify labels of a “war on science.” Science is not some kind of “miracle machine” where you turn a crank of scientific method and out pops knowledge. “Exceptional science is produced not by a miracle machine, but by institutions that tie scientific curiosity to problem solving,” he says. In fact, promoting the myth of the miracle machine can actually backfire.
Vast improvements in the scientific system could be had if science agencies strengthened the ties that link research agendas to societal needs, and counteract the perverse incentives that commit scientists to careers measured by publications and grant dollars rather than the creation of socially valuable knowledge.
Impact factor has been a counter-productive measure. Speaking of killing old myths, Nature is also glad to read an obituary for the dubious measure of scientific value called “impact factor.” What was supposed to provide a “bibliometric” measure of scientific value actually did the opposite. “It should never have been used and has done great damage to science,” complains Richard J. Roberts. “Let us bury it once and for all.”
The impact factor is often used, improperly, to provide a mathematical measure of a scientist’s productivity, on the basis of where they published their results. It has proved popular with bureaucrats, and even with many researchers, because it seems to offer an easy way to determine the value of a scientist’s output for someone who is either unable or too lazy to read that scientist’s papers and judge their true worth (see et al. Nature 544, 411–412; 2017).
Science is not supposed to be a money prize. The Editors of Nature worry about perverse incentives at work in China, where the government rewards scientists too quickly with grants and bonuses for what they consider successful research. “Don’t pay prizes for published science,” they argue. “For one thing, it creates a culture in which scientists look at their research as a means to make quick cash.” It also “rewards science that is not yet proven.” Like impact factor, metrics for what constitute successful research are often “greatly overblown.”
Scientists are not above data manipulation. We asked last month (6/12/17), “If science is superior, why does it need fixing?” More evidence that scientists are like other fallible humans led Nature to complain about the problem of “image doctoring” in scientific papers – a problem that has mushroomed with the rise of digital manipulation tools like Photoshop. Publishers and editors do not always catch the digital trickery, and algorithms to detect image doctoring are not good enough yet. “By both human and technological means, research organizations, researchers and journals need to do more to counter the image-manipulation challenge.” But wait; weren’t we all taught the myth of the unbiased scientist seeking only truth for its own sake?
Models do not always catch important details. Mathematical models, frequently used in science, try to simplify reality by focusing on pertinent details. But which details are pertinent? Researchers decided to check a popular “quarter vehicle” model used by auto manufacturers to gauge ride dynamics. They added in other factors omitted by the model and compared the results. They found that omitted details do make a big difference. Their paper in PLoS One says,
The results clearly indicated that these details do have effect on simulated vehicle response, but to various extents. In particular, road input detail and suspension damping detail have the most significance and are worth being added to quarter vehicle model, as the inclusion of these details changed the response quite fundamentally. Overall, when it comes to lumped-mass vehicle modeling, it is reasonable to say that model accuracy depends not just on the number of degrees of freedom employed, but also on the contributions from various modeling details.
What other models in science, for the sake of convenience and simplicity, are failing to consider significant details that could fundamentally change the conclusions? The more complex the problem, such as with global climate, the more it seems the simplicity is the enemy of accuracy – especially when conclusions are not readily testable as they were in this case.
Language can manipulate rather than enlighten. We shared examples recently (7/02/17) of leftist bias in science. Sometimes leftist researchers are open about their manipulation. Phys.org reports on psychologists who found that saying “climate change” instead of “global warming” reduces the “partisan gap by 30 percent in U.S.” This is a clear attempt to nudge voters rather than educate them (6/11/17). We can also add to the list another manipulative article posted by Phys.org that claims, “How bills to replace Obamacare would especially harm women.” Conservative researchers could easily argue the exact opposite conclusion from well-grounded data; so why don’t they get the microphone of Phys.org or The Conversation?
Bad definitions of science prevent scientists from finding truth. The editors of Nature tried to be nice to Catholics on May 15, only to be shouted down by a reader, who repeated the myth of scientism in the June 22 issue of Nature. According to Frank W. Nicholas, the editors forgot to be naturalistic enough:
Your Editorial suggests that Pope Francis’s meeting with patients and researchers is evidence of “a new openness [of religion] towards science”, in the spirit of his 2015 encyclical Laudato si’ (Nature 545, 265–266; 2017). This is tempered by your view that the encyclical nevertheless illustrates “a chasm between religion and science that cannot be bridged”.
In my view, the encyclical’s most fruitful comment on science and religion is that they have “distinctive approaches to understanding reality” (paragraph 62; see go.nature.com/2swk22m). The essence of this distinctiveness is that the modern scientific approach never invokes God as an explanation for any phenomenon. This restatement of ‘methodological naturalism’ is not science being anti-God: it is science being science. All scientists adhere to this approach, including scientists who believe in God. In the religious approach, by contrast, God is at the heart of phenomena.
It follows that the fundamental distinction between science and religion has nothing to do with the question of whether or not God exists.
These insights can inform the debate around what should and should not be taught in science classes on, for example, evolution. In shedding light on the nature of the “chasm” between science and religion, these insights can also inform the new openness to which you refer.
It seems lost on Nicholas that methodological naturalism of this sort is guaranteed to come to wrong conclusions if God does exist and was involved. For instance, if God did create life, all the efforts and funds to find a natural origin are doomed to failure. If evolution is false, then all the published papers about natural selection creating man from molecules are also false. By excluding intelligent causes, would Nicholas insist on a natural explanation for Stonehenge? Would he insist on unguided natural causes as the only tools to explain his own righteous indignation? If so, his arguments would implode.
Nicholas bought into the NOMA myth of Stephen J. Gould without apparently being aware of its weaknesses. His letter illustrates the unchallenged assumption of a particular philosophy in scientific institutions – methodological naturalism – which, as intelligent design advocates have frequently argued, becomes indistinguishable in practice from philosophical naturalism (for some of the debate, search on ‘methodological naturalism’ at Evolution News). Causation is a long-standing debate where the extreme positions obfuscate the productive middle ground. Surely no theistic scientist is going to attribute the precipitation of chemicals in a flask to the direct intervention of God. But neither should an atheistic science rule out convincing evidence for intelligent causes just to maintain his materialist philosophy. That could guarantee a false conclusion.
The list above reports only some of the most recent debunkings of the myth of scientism from secular sources themselves. Big Science is a political force that once in awhile discovers interesting facts about nature, just like Big Education is a political force that once in awhile teaches something valuable to a student, or like Big Labor is a political force that once in awhile helps a worker. The real contributions usually come not from the top, but from the individuals who, through their own integrity and moral character, decide to help their fellow man.