Guerrilla Science

Observing various scientific “controversies” over the past few years, I’ve seen a pattern repeated again and again between the scientific mainstream and dissenters. Whether it’s about global warming, a link between vaccines and autism, the safety of GMO crops, or any other issue, the conversation looks pretty much like this.

Scientists: “We have developed the following hypothesis due to a preponderance of evidence, and the scientific consensus says we should enact the following policies.”
Dissenters: “Well, that’s just your theory. What about this evidence which argues something entirely different?”
Scientists:  “That evidence is methodologically and theoretically flawed, and we have dismissed it for scientific reasons.”
Dissenters: “No, you’re dismissing it because you’ve been bought off by Big Pharma/Monsanto/Al Gore!”
Scientists: “Well, you’re anti-science, and you aren’t responsible enough to participate in this debate. Come back when you’re willing to accept the truth.”

And then the two camps go their separate ways: The dissenters to fringe websites where they catalog the corruption of mainstream science and develop their alternative bodies of evidence, and the scientists to letters pages of scientific journals, where they write pleads for better science education and communication, so that science can drive out all the frauds out of the public sphere and we can have rational policies again, like we did in the good old days.

The other thing that I’ve observed while following these controversies is that mainstream science is losing. The American political system has become entrenched around the truth or falsity of anthropogenic global warming, despite an overwhelming scientific consensus that it is happening and it is a problem. Fewer people completely vaccinate their kids each year, even though the original Wakefield study has been totally discredited and disease rates are on the rise. And fears of GMO have become a permanent part of European politics, and a rising force in America and China.

This leads to one of two conclusions: either despite all the calls by respected scientists for more communication and education efforts by the scientific community have been falling short and should be increased, or the conventional framing of the problem is essentially wrong and misleading. I believe it is the latter; that the arena of public scientific debate has changed in recent decades, that the dissenters are “guerrilla scientists” who like guerrilla fighters use asymmetric strategies to avoid the superior strength of their foe, and that to win, mainstream science must find an equally adaptable counter-strategy.

To explain this idea, I’m going to need to talk about science and guerrilla warfare. Please bear with me.

As a PhD student in science and technology studies, one of the biggest questions that we face is “What is science?” There are lots of good definitions: facts about the natural world, systemic knowledge, a method for generating said facts and knowledge and ensuring their reliability using experiment and observation, but all of these definitions conceal the process of how science is made; how specific claims become true facts or false hypothesizes. To understand that process, we need to go inside science, inside scientific writing, and inside the lab. Bruno Latour has developed some of the most powerful lenses on the actual practice of science in his books Laboratory Life and Science in Action.

For Latour, science is a rhetoric; a way of convincing other people to believe your claims. The form of the modern scientific paper has been careful developed to be as convincing as possible. A successful scientific paper integrates itself in a network of previous scholarship, explains how it will extend the results of previous scholars, presents a method that can be duplicated by others, shows results (typically in graphical form), and then discusses those results. I’m going to use as my example a paper by the other author of this blog, “A model for the origin and properties of flicker-induced geometric phospenes”, but any other paper would work just as well.

The paper begins by summarizing the previous work in the field, starting in 1819 and moving through the 1970s and into the present day. The introduction establishes the paper vis-à-vis previous work on the visual system, and a question about the origins of flicker in either the retina or the visual cortex. The method section describes using the Wilson-Cowan equation for modeling flicker in a simulated neural network, how to implement that equation in a computer program, the images that are produced by the model, and finally a discussion of how those images might relate to what happens in the brain and what we can perceive when we close our eyes and press on our eyeballs (or use ze goggles).

At every turn, the paper preemptively parries those who would try to doubt it. “You think that this problem is unimportant. Here are people who have worked on it before me.” “You doubt my math? Write your own program and check my results.” “You disagree with my choice of the Wilson-Cowan equation? Here are 1040 papers that also use it. Do you disagree with all of them?” The paper is structured and linked such that to disagree with it either requires opposing a much larger and more authoritative body of scholarship than “things that Mike Rule, Matt Stoffregen, and Bard Ermentrout say”, or going into their lab, checking that all their machines work, that the graphs in the paper are actually reproducible, and essentially duplicating their effort and expertise.

This post-modernist view of science can be disconcerting at first; what about objective physical reality? What about the search for truth? Has science just become another kind of blind faith, based on appeal to past authority?  No. What Latour tells us is that scientists do not know a priori what is ‘real’ and ‘true’. Those words are only applied to hypothesis after an intense process of purification and examination by the community of scientists that rules out every other possible explanation.

The picture of science that Latour develops is an interlocking network of claims about the natural world, linked into a mutually reinforcing pattern with more accepted claims at the center and weaker claims at the fringes. Science as a totality is like a fractal star fort, defensible from any angle.  But this picture is partial and passive.

The life of science is in active disputes, for example, “What is the structure of DNA circa 1953?” Disputes are opposing versions of reality, and they are only settled by the destruction and absorption of alternative facts and theories into the final ‘science truth’. Actor-Network Theory describe this process as one of enrollment wherein scientists enlist facts, instruments, and people as allies in their cause, with the aim of building the strongest rhetorical network.

Looking at this, it struck me that the contest is much like a battle, with the scientist deploying his or her enrolled facts like a general committing his soldiers. There are many points of congruence between these models: critical questions and strategic points, the ability to generate new results and supply, predictive power and firepower, but they key point is that in science as in war there are rules.

Science has its own kind of Geneva Convention. Not an explicit treaty, but social norms that describe how science should be done, and how the contest should be decided. I’m not going to provide a complete set of norms, but a some of the more important ones might be: results must be reproducible; judge the idea and not the person; cite your sources; do not present others’ work as your own; remain open-minded; accept the accolades of your peers humbly. A scientist whole held onto a clearly discredited theory would not be respected, and it is considered bad form to pillage a rival’s lab and enslave their postdocs.

In action, the difference between mainstream scientists and dissenters is that dissenters don’t play by the rules. Dissenters care more about their personal commitments than the structure of science as a whole. They do not cede the field gracefully if their facts are overturned. And they accept a wider variety of evidence as a basis for their claims, including social, moral, economic, and political factors. They may not be good scientists, their theories are frequently shoddy and mystical, but it is important to recognize that they are engaged in essentially the same kind of work as mainstream scientists: making cause-and-effect claims about the natural world. Calling dissenters ‘anti-science’ implies that scientists should ignore and belittle them as unworthy of serious critique. Calling them guerrillas suggests a very different approach for understanding their aims and engaging with them.

Guerrilla warfare is political warfare. In conventional war, the goal is defeat of the enemy through decisive battle, and strategy is the art of staging the decisive battle on favorable terms. The aim of guerrilla warfare is to demonstrate the political illegitimacy of the people in charge while building popular support for revolution. Strategy is focused on swift strikes to demonstrate the ineffectiveness of the governance and provoke reprisals against the people, the preservation of the guerrilla’s own forces, and the use of time to wear out the enemy’s will to resist. While conventional military firepower is still important in guerrilla war, it is of secondary importance compared to psychological and political factors. In guerrilla warfare, the winner is the side that everybody believes has won; not the side that maintains control of the battlefield afterwards.

Guerrilla warfare is a complicated subject, and no two conflicts are alike, but some common patterns can be drawn. Lt Col John Boyd developed a theory of warfare based on learning systems, and he noticed that as a force slides towards defeat, it becomes isolated and insular, it stops taking in information from the outside world, and is eventually confined to irrelevance. Information and morality are central; as the American military learned in Vietnam in the 1960s and in Iraq in 2003 and 2004, firepower is useless if targets cannot be located, and support can only be gained through demonstrating moral strength and sensitivity. To beat guerrillas, the government must demonstrate its superiority through active policies that improve the lot of the people while avoiding internal corruption. Successful counter-insurgency strategies, such as the Iraq Surge implemented by General Petraeus, aims to isolate guerrillas, to draw wavering fighters back into the government’s camp, and to find and kill the most hardcore commanders who could not be converted.

Combining these two theories, Latour’s Actor-Network Theory and Boyd’s OODA Loop, the shape of the problem and its solution begin to emerge. Scientific guerrillas exist because scientific expertise is a key buttress of democratic decision-making. 21st century American culture is such that a policy must appeal both to the will of the people and an external reality as informed by expert, i.e. scientific, opinion. But science, to put it bluntly, is hard. It requires a long and grueling apprenticeship and then access to expensive and specialized laboratory equipment. And worse from a political perspective, science is not democratic. No matter how many geeks wished that the OPERA neutrinos were truly faster-than-light, that result stubbornly remains an experimental error. It’s far easier to don the guise of expertise when it’s needed to support a policy position than it is to genuinely discover the truth according to the strict rules of science.

In this context, saying that the dissenters need to play by mainstream standards of evidence is like saying that we just need Al Qaeda to put on uniforms, gather around Tora Bora, and have that decisive battle we’ve been waiting for. It’s a fantasy, because it involves convincing guerrillas who are winning to fight a conventional battle that they will surely lose. Science education, science funding, and more public understanding of science are equivalent to sending in more troops, more weapons, more airstrikes. It can stabilize the situation, but it is unlikely to actually defeat the guerrillas.

I worry that science is becoming isolated in a Boydian sense. Scientific papers only cite other scientific papers; most scientists work and live in enclaves around major research universities. There are extremely good reasons for this, from a conventional perspective it generates stronger science, but it has also made science more brittle, less relevant, and less politically legitimate.

Like it or not, scientists have become embroiled in a wide variety of guerrilla disputes on major issues, and I’ve not seen a robust strategy for countering the guerrillas. I love and respect science; it’s the best tool for understanding and improving the world that we have, but it is under attack in ways that most people can’t even see, and is not effectively defending itself. Guerillas can be beaten, but it will require an active strategy of integrity, candor, and two-way communication. The stakes could not be higher. As Henry Kissinger said on the Vietnam War in 1969, “The guerilla wins if he does not lose; the conventional army loses if it does not win.”