Science is Rotting from the Top

Science leadership has forgotten … science

The texts for today are from Dr. Marcia McNutt, editor-in-chief of Science from 2013–2016 and current president of the National Academy of Sciences. First, we have an interview with the Columbia University Center on Global Energy Policy from 2018:

Interviewer: I want to touch on a topic that’s another of your priorities here at the Academy, and that’s diversity. What are you trying to do and why?

McNutt: Well … diversity is a topic that is very close to my heart ever since I came here to the Academy … Scientists in America are far more white and male overall than is the majority of the American population … I would like to have prominent representatives of the diversity of American scientists that we can use to encourage a more diverse science workforce, because if young people can look to those in the Academy and say, “My God, look at these people, they were elected to the National Academy of sciences, they are the top people in American science, and that person looks just like me.” Then that gives them this hope that they, too, can succeed in that field, and there just aren’t enough role models for these young people who are from diverse backgrounds … So that’s my goal …

Interviewer: How would you like to be remembered when your term as president of the Academy is done?

McNutt: Well, I would certainly like people to look at the makeup of the National Academy when my term is done and say she changed the face of the Academy [emphases added].

McNutt wants to fix the racial and gender mix of the Academy. Her opinions are not novel. Similar views are expressed almost daily by institutions ranging from the U.S. Securities and Exchange Commission to universities and colleges, the military, the federal government, the UK Royal Society, and the current editor of Science magazine. McNutt is just echoing the same DEI (Diversity, Equity, and Inclusion) message as almost every other leader in America, from politicians and executives to media figures. What’s wrong with that, you may ask?

What’s wrong is that McNutt is a scientific leader. While ignoring real problems of contemporary science (more on those in a moment), she promotes views that actually undermine science. To see how, let’s look at some of her comments a bit more closely:

Scientists in America are far more white and male overall then is the majority of the American population.

So what? Why should we “encourage a more diverse science workforce …”? Nurses are mostly female, engineers mostly male. Why is a disparity in science of the slightest concern? There are three possibilities: either McNutt believes 1) that females (and blacks) are somehow being unfairly excluded from science, 2) that females and nonwhites have a unique contribution to make—or 3) she is an anti-white-male racist.

Are men preferred over better-qualified women scientists? There is almost no evidence for this, and some evidence to the contrary. Perhaps females are either better at science or can add something that males cannot? This isn’t true either. Women are not uniquely better at science. Indeed, a case can be made that males are better at ‘hard’ science than females—a suggestion that drew an emotional reaction from female scientists when soon-to-be-ex-president of Harvard Larry Summers presented some relevant data a few years ago. There are, of course, individual exceptions, excellent scientists of every color and gender. Unfortunately, Dr. McNutt, like her fellow DEI fans, studiously avoids talking about individuals. Theirs is a collectivist philosophy—only groups matter, and, perhaps conveniently, group disparities will always exist, so their task is never-ending.

“Diversity enhances creativity” is the conventional wisdom. But, in fact, there is no evidence that scientific breakthroughs are more likely to come from non-white groups. There are a few papers purporting to show more “impactful” (citation-count) research from mixed-gender research teams—and much more research is now done by teams than individual researchers. Of course, some research (e.g., nuclear physics) requires a team. But the great advances even in that field have usually been made by individuals, or at most a very small group. Moreover, the increase in team research has not led to a comparable increase in scientific breakthroughs. There is no reason why teams, diverse or not, are the best way to advance science, and no evidence that diversity fosters real creativity.

[Related: “Science Should Leave the University”]

If there is little or no actual discrimination, what is the need for ‘diversity’? Why should society urge women to prefer, say, engineering or physics to nursing or public relations? Only because diversity is now in fashion (or is it anti-male bias?). An oft-cited study shows that, in gender-neutral countries like Sweden, the gender disparity in STEM subjects is greater than it is in more patriarchal countries, suggesting that (a) women and men do not have identical interests and abilities, and (b) they may not favor the same occupations. There is no reason to expect, or encourage, population-proportional representation of every group in every branch of science. There is no reason to pressure women (or men or people of color, for that matter) into professions they would not otherwise choose simply because of their sex or race. It isn’t even fair, because favoring one group necessarily disfavors others who may be better suited to the work.

[E]ncourage a more diverse science workforce … just aren’t enough role models …

Another justification for diversity is the role-model hypothesis: role models are important, and the role model must look like the student. In our present era, when race is so salient and colorblindness is on the back foot, it is natural that a black student will notice the dearth of black scientists and be excited when she sees one. But in other eras and at other times, the race of the admired scientist was more or less irrelevant. For a talented young scientist, it’s the ideas, not the complexion, of the teacher that matter. The role-model hypothesis has little scientific basis.

Censorship and Consensus

Marcia McNutt recently clarified her views on science in our second text, a 2023 article with Michael Crow, president of Arizona State University. First, consensus is king:

[P]ublic understanding of climate policy has been stymied by disinformation that has called into question the scientific consensus about the nature of the threat …

In fact, there is no real consensus about climate change. Many able scientists dissent. The volume of publications is heavily skewed in favor of the received view by a massive asymmetry of funding: pro-anthropogenic-warming studies find it much easier to get support than critics of the theory. In any case, major advances in science usually come from “calling into question” an existing consensus, not from following it. Consensus may be relevant for policymakers (they’re not scientists; what else can they go on?), but not for scientists.

In a particularly pernicious passage, McNutt and Crow explicitly advocate the censorship of science—not of truly dangerous excursions, like “gain-of-function” research with infectious viruses, but of science that fails “to promote democratic ideals.” McNutt and Crow comment:

In Science in a Democratic Society, philosopher Philip Kitcher reminds us that “science should be shaped to promote democratic ideals.” To produce outcomes that advance the public good, scientists must also assess the moral bases of their pursuits.

The view that science should be guided not just by empirical facts but also by conformity to extra-scientific values has been around for a long time. In the 1930s, for example, famed polymath J. D. Bernal in Britain urged scientists to get involved in politics and recognize, even welcome, the intrusion of—in his case, Marxist—politics into their work. It didn’t do too much harm in this instance—X-ray crystallography is pretty immune to political influence. But political influence was fatal for millions of Russians when the genetical errors of Trofim Lysenko were made dogma. Yet, in 2020, Scientific American, the premier popular science journal, could proudly proclaim: “Yes, science is political,” urging scientists to get politically involved.

Political, nonscientific values are especially dangerous to the human sciences. Philosopher Philip Kitcher, cited by McNutt and Crow, has a history of subordinating the major value of science, truth, to political considerations. In a 1997 article, he justified the censorship of scientific results that might cause societal harm. The eminent scientific journal Nature has explicitly endorsed this practice.

Suppressing uncomfortable findings can be disastrous. Here is an imaginary example. Assume there is a group disparity in wealth, say, between group A and group B: A is wealthier than B. The obvious scientific question is: What’s the difference between the two groups? One commentator, X, focuses on the treatment they receive from society. Group B, he points out, is not allowed to work, has limited freedom to travel, and is often under the control of others. In other words, it is discriminated against. It’s society’s fault. Another commentator, Y, disagrees. Maybe Group B commits more crimes than Group A; perhaps it is less well educated. This is unfair, says commentator X; you are blaming the victim. Well, says Y, group B are children after all; that’s the main reason they are poorer than Group A, who are adults.

[Related: “In Defense of Merit: Is it too late?”]

This seems like a silly argument. Of course, you can’t compare adults and children without recognizing the intrinsic, endogenous differences between the two groups. Yet Kitcher has used that same argument to dismiss the study of endogenous differences between racial groups as one possible cause of differences in societal outcomes. The proposition has been taken up by one group of economists who openly reject anything other than societal influences—racial discrimination—as a cause of racial wealth disparities.

Mainstream science, following Kitcher’s advice, has suppressed research on endogenous causes, differences in the interests and abilities of blacks and whites. Well, it looks like bad science, but is it a big deal, really? Here’s the problem: wealth disparities exist, and something must be causing them. If endogenous causes are ruled out, all that is left is the unmeasurable, ineradicable systemic racism, a sinister social force for which white people must be responsible. In other words, suppressing research on one set of causes for a societal problem doesn’t lead either science or the public to admit that they don’t know. Instead, human nature being what it is, some other imaginary cause, one that serves political rather than epistemic ends, will come to the fore. Suppressing truth is just as dangerous as promoting lies. Indeed, the first will often lead to the second.

The failure of the scientific elite to recognize the dispassionate epistemology essential to real science has led to a justified distrust of science by the population at large. Suppressing research on a true cause will often have much more dangerous effects than frank and open inquiry into a politically sensitive topic. Yet, Dr. McNutt, by claiming that “science should be shaped to promote democratic ideals,” is advocating this dangerous position. Since science is just the search for verifiable truth, McNutt is, in effect, saying that “truth should be shaped …” Does anyone think that is a great idea?

What Now?

The quest for DEI is endless: numerical equity will never be achieved (nor should it be). It is also detrimental to science. There are many examples, but here is one that is as dangerous as it is ludicrous: so-called citation justice. The University of Maryland library defines it this way:

Citation Justice is the act of citing authors based on identify [sic] to uplift marginalized voices with the knowledge that citation is used as a form of power in a patriarchal society based on white supremacy.

A research paper cites other articles that have contributed, directly or indirectly, to the reported work. A key article may be accidentally omitted, or an irrelevant one included, for bad reasons (e.g., you like the researcher or wish to impress). Both of these are errors. To these occasional mistakes, DEI advocates, following their crypto-Marxist philosophy that everything is about power, wish to add another: including citations just because of the race or gender (identity) of their authors. Citation justice absolutely corrupts science. A ‘just’ reference list represents not the field in question but politics.

This grotesquerie arises from a critical flaw in the existing science system: evaluating scientists (for grants, promotion, etc.) not on the quality of their work (tough to measure), but on grants received, number of publications, and citation count (easy). McNutt has quite properly criticized citation counts. One can only hope that she will do the same for citation justice, although, based on her other comments, we can’t be sure.

If she were not so preoccupied with social justice, Dr. McNutt might be able to spend some time on the real problems of science. There are many. Most dramatic is the replication crisis, which is still going on almost twenty years after it was first identified. At least fifty percent of results in social and biomedical science turn out not to be replicable. The main reason is a misapplication of statistical methods so as to get positive, publishable results, which are necessary for career advancement because of the citation/publication/grants-awarded evaluative criteria that are used. Another error is the widespread application to, say, individual psychology of data obtained from groups (we are all subject to ‘confirmation bias’; well, no we’re not, but you wouldn’t know it from most reports). Fraud is also skyrocketing; papers with fake data, imaginary clinical trials, and the like grow in number every year. Correlations (easily obtained by appropriate statistical tests) are almost always presented as causes; just look for the phrase “linked to.”

The ethics of science have obviously deteriorated in recent decades. Scientists are more willing to follow trends, fake data, misuse statistics, suppress criticism, etc. Why? Are there too many scientists relative to the number of solvable problems and the distribution of human talent? Have career incentives become destructive? Is science funding too concentrated?

Problems like this should be at the top of Dr. McNutt’s mind. But social justice anxieties are widespread, and conformity to them is comfortable and well rewarded. Better worry about the number of white men dominating science. Meanwhile, ruinous policies are advocated, and the real problems of science are ignored.

Image: Cmichel67, Wikimedia Commons, Creative Commons Attribution-Share Alike 4.0 International license