On April 22, Earth Day, tens of thousands of people will assemble in Washington, DC, and several dozen satellite cities to march for science.
What does that mean, exactly, to march “for science”?
I have some thoughts on that — on how to make sense of the controversies around the march (as covered by our own Brian Resnick) and the larger question of science’s relationship to, well, Donald Trump.
First, though, a side note on the reality of the march. The fact is, the vast majority of people who attend it won’t care much about the somewhat abstruse issues I dig into in this post. Nor will they be familiar with the controversies among and about the organizers.
They will just show up because they’re angry at Trump and science is cool. And that’s pretty amazing, when you think about it — a testament to the incredible depth of civic energy and engagement that Trump has unleashed. People have awakened to a broad-based assault on US institutions and they want to do something, so they’re putting their bodies in the street, for women, for science, for climate, for transparency. It’s all to the good.
So this post isn’t really about the march itself. It’s about some of the thornier issues it has raised regarding science’s place in society and its relationship to politics.
I want to begin by distinguishing two different ways of thinking about science — or rather, two different things that “science” refers to. For the purposes of this post, I’m going to call them science-theory, or science-t, and science-practice, or science-p. (Hopefully that won’t get too annoying.)
They are very different, and keeping them distinct will help us figure out exactly what people are marching for and arguing about.
The idea of science, in the abstract
Science-t is the scientific method itself, the idealized essence of science. It claims nothing beyond what the empirical evidence supports.
That sounds simple, but in fact it is intensely difficult and profoundly unnatural. We humans were not shaped by evolution to identify truth. Why would we be? Membership in a tribe has far more survival value than objective accuracy.
We are built to seek out social connection and identity, a sense of rootedness. We crave an understanding of the world and our place in it. We don’t necessarily want or need truth for that, just good stories. So we instinctively round out our worldviews with stories, myths, inherited assumptions, and speculation.
It would take years of meditative, ascetic practice for an individual to fully overcome this — to truly and directly see only what is in front of her. We can’t all be monks, so instead we solve for the problem at the collective level. Science-t is a kind of social technology meant to counterbalance our individual propensity to error, through mutual fact-checking.
In order to restrict its practitioners to what can be supported with evidence, it imposes a strict set of rules and procedures.
Structured experiments produce observations. Scientists posit hypotheses to explain the observations. If those results can be replicated in other experiments, confidence in the hypotheses increases. Eventually hypotheses become theories, and mutually reinforcing networks of theories.
There is no such thing as empirical evidence sufficient to support certainty, so science-t offers none. In contains only probabilities and error bars, degrees of confidence on a continuum. Ultimately, everything is defeasible, subject to revision. Certainty is the domain of other human discourses, ideology and religion.
Similarly, there is no such thing as empirical evidence sufficient to adjudicate moral and prudential disputes. Science-t cannot bypass or short-circuit politics. The evidence can only tell us what is, not what should be.
Political debates should be informed by our best current understanding. They should not involve falsehoods. But even those are moral and prudential judgments, not derived from science-t.
Science-t can help us figure out which policies help the sick and vulnerable, but it cannot tell us whether, or how much, we ought to help them.
Similarly, science-t can tell us what effects air pollution and carbon emissions will have on public health, but it cannot tell us whether we ought to prevent that suffering, or how to weigh that suffering against other rights and obligations.
Similarly, science-t can tell us that vaccinations don’t cause autism, but it is silent on whether we ought to vaccinate children. That is a moral (and prudential) decision.
And finally, science-t might help us determine what effects diversity and representation have on institutions, but it cannot tell us whether subaltern populations should be better represented. That is a moral (and prudential) question.
Why “science-based policy” is a bit of a misnomer
Science-t can help us identify problems. It can help determine which policies will have which effects. It can help establish a common baseline of facts. But it cannot tell us whether the problem is worth solving, or how to weigh the costs and benefits of solving it. Ultimately, the choice of policies, though ideally informed by science-t, is based on moral and prudential considerations. All human decisions are.
Now, let me be clear: I personally favor reducing carbon emissions, vaccinating children, and making institutions more diverse and representative. To say that these are moral (and prudential) judgments, not scientific judgments, is not some kind of demotion. It doesn’t demean those judgments or make them purely subjective. Improving society’s moral and prudential decision-making is something everyone is always and already involved in; there’s no sense pretending science-t can do the job for us.
The point is, science-t is about empirical evidence and nothing beyond it.
The practice of science, in the real world
In the real world, the ideals of science-t are embodied by science-p, the actual institutions, norms, and people doing science. Science-p is different from science-t in two vitally important ways.
First, because it is composed of human beings, science-p never reaches the ideal of science-t. Work — lab research, surveys, field studies — is inevitably colored by personal history, professional and financial pressures, desire for recognition, and fear of failure. Replication is often absent or woefully inadequate; theories are often driven by hype and media attention; peer review is often weak. (Read Resnick on science’s replication crisis.)
It is a constant struggle for science-p to approximate the ideals of science-t. Which brings us to the second difference.
While science-t involves only first-order concerns about properly structured experiments, evidence, replication, etc., science-p, as a set of social institutions and norms, deals with a whole set of second-order considerations regarding its own institutional health. Science-p must be concerned not merely with doing good science-t, but with the conditions that make doing good science-t possible. (I made the same point about journalism in my post on tribal epistemology.)
The overarching issue facing science-p is how to get as close as possible to the ideal of science-t. That inevitably involves questions that are not themselves scientific. What’s the best way to fund science? How transparent should scientists be expected to be about funders, methods, and data? What steps should science-p take to diversify itself? To what extent should science-p engage in political lobbying? What role should science-p play in government rule-making? These are moral and prudential considerations about the health of institutions.
A certain sort of person (see: cognitive scientist and popular author Steven Pinker) thinks that the way to keep science-p healthy is for scientists and scientific institutions to remain silent on contested questions of values and policy. In this view, scientists, at least in their capacity as scientists, should confine themselves to that which can be supported by science-t, their only proper interest and authority. The way this is usually phrased is that they don’t want science “politicized.”
This seems a bit naive. Science-p has always been political, both outwardly, in the uses to which it is put, and inwardly, in the way it reflects society’s inequitable racial and gender power structures. Scientists Chanda Prescod-Weinstein, Sarah Tuttle, and Joseph Osmundson remind us of this in their bracing (if occasionally over-the-top) cri de coeur against the Trump administration.
More broadly, science-p is a set of institutions and norms amidst others, in a larger social ecosystem. It is part of the polity, so it is inescapably political.
The germ of value in Pinker’s argument is that scientists should not claim to speak on moral and prudential considerations as representatives of science-t. Again, science-t cannot speak on what is good or wise, only what is.
But there’s no reason scientists shouldn’t speak out on moral and prudential matters as representatives of science-p — that is, as representatives of the institutions that keep science-t alive and healthy.
The government agencies, research institutions, laboratories, and academies that do science-t must concern themselves with how good science-t can continue getting done. That might mean pushing for more funding (which Trump is cutting), more diversity, or more STEM education and outreach. It might mean protesting attempts by House Republicans to politicize the National Science Foundation grantmaking process. It might mean marching.
This distinction between science-t and science-p helps illuminate two of the controversies around the march.
Diversity is hugely valuable to science-p
First, much of the online fighting over the march has been about diversity and inclusion. Several early organizers have quit over the larger organizing committee’s ham-handed handling of these issues.
I left the @ScienceMarchDC organizing committee in March due to a toxic, dysfunctional environment and hostility to diversity and inclusion.
— Jacquelyn Gill (@JacquelynGill) April 16, 2017
In my experience — particularly my extensive experience as a white male — what motivates some scientists to resist these kinds of considerations is that they truly do have no place in science-t. At heart, science-t is nothing but a set of procedures meant to screen out positionality. The whole point is to determine what is true independent of any culturally specific perspective. In theory, a scientist’s background, race, socioeconomic status, etc. should not matter.
But we do not live in theory (science-t). We live in practice (science-p). We are all shaped by our backgrounds and the social infrastructure in which we operate. Any social monocrop (say, all white men) will only mass produce the blind spots and biases of its individual members. By bringing more perspectives to the table, diversity makes the social technology of science-t — mutual checking, correcting, amending, and building — more effective.
What’s more, closed systems cannot thrive. An institution that replicates society’s larger inequities also replicates its enormous waste of potential talent. Bringing in traditionally excluded groups brings in new energy and new ideas. (Meanwhile, Trump’s immigration policies are driving foreign science students away.)
What’s more, precisely because science-t is a cosmopolitan ideal, operating independently of parochial perspectives, it can welcome anyone. It’s a point well-made in an essay by particle physicist Yangyang Cheng, for whom science was a path from humble beginnings in China to the world’s most powerful particle accelerator, at Cornell University. “Despite the tainted history and current flaws of science,” she writes, “I, as a member of the underrepresented, still place my unwavering faith in its power as the great equalizer.”
In short, diversity and inclusion make science-p healthier and thus make for better science-t. They are not external, “political” considerations, they are, or should be, core concerns of anyone who supports science-t. That’s why they deserve a place at the heart of the march.
The purported danger of “politicizing” science-p
In a New York Times op-ed, geologist Robert Young warned that a march for science, “while well intentioned, will serve only to trivialize and politicize the science we care so much about, turn scientists into another group caught up in the culture wars and further drive the wedge between scientists and a certain segment of the American electorate.”
I hate to be the bearer of bad news, but that horse is already out of the barn. To paraphrase Leon Trotsky, you may not be interested in the culture wars, but the culture wars are interested in you. Scientists are already caught up; the wedge is already driven.
The primary reason, as I have explained at tedious length, is the decades-long campaign by right-wing media to convince its audience that America’s core institutions are irredeemably corrupted by leftists — government, academia, media, and science are the “four corners of deceit,” in Rush Limbaugh’s phrase.
Science, like media, has done a piss-poor job of defending itself against this onslaught. And now, with Trump ascendant, things have reached the level of crisis. Republicans are defunding research, politicizing grantmaking, diminishing science’s role in the agency rulemaking process, and flat out refusing to accept conclusions they don’t like.
That Republicans have been denying climate change for so long has made it seem familiar, but it remains incredible, a brazen rejection of scientific institutions and practices. They called society’s bluff and got away with it. And it turns out to have been a harbinger, a waystation on the road to what looks more and more like a wholesale rejection of empirics.
The conditions that make good science-t possible — robust, well-funded, independent institutions and a basic respect for accuracy — are under intense and immediate threat.
We should not pretend that science-t itself dictates an answer to this debate, or to any political struggle.
But we should also not pretend that science-p can remain silent. The institutions and people who do science, as well as all the people who value them, cannot remain neutral toward a threat to the conditions that make science possible.
The March for Science is (another) sign that those who believe in cosmopolitan values and democratic institutions are feeling some of the intensity that has, for too long, been the almost sole province of their tribalist opponents. Flaws of planning and communication aside, it cannot help but be a positive thing when thousands of people assemble to reaffirm that science is cool.