Biology as Ideology: dangers of reductionism and science as a job


I was recently recommended Lewontin’s Biology as Ideology, which added to a growing ferment I’ve been experiencing. There are three themes I saw in this book that I’ve been thinking about:

First, science as legitimation. Lewontin draws a comparison between Western organized religion in the Middle Ages and science in the modern age. Both are groups of people taking actions while also presenting themselves as something greater than or beyond individuals. Both present themselves as methods for identifying eternal, nonpolitical truths. Both require the intermediating presence of priests/scientists between those arcane truths and the laity/public. Both are ostensibly apolitical (e.g., “spiritual”) but also are important for political (e.g., “temporal”) affairs: kings ruled by the divine right of God (the pope crowned the Holy Roman Emperor), and modern states and affairs are built, to some degree, on the legitimacy offered by a scientific, or pseudo-scientific, approach.

This latter point is the one that requires convincing and explanation. Lewontin’s first example is scientific racism (which is treated excellently in Stephen Jay Gould’s Mismeasure of Man). Scientific racism was not the ravings of some fringe group; it was the position (what we might call “opinion” but what they called “scientific fact”) of leading anthropologists, doctors, psychologists, and politicians.

Predictive policing is my example. The dream has tremendous appeal: you use “big data” to figure out where police should patrol and how to intervene with individuals who are at the greatest risk to commit or be victim of a crime. The catch, summarized by William Isaac in a 2016 Science article, is that predictive policing algorithms “[are] not predicting the future. What they’re actually predicting is where the next recorded police observations are going to occur.” In other words, those algorithms are “scientific”, but that doesn’t mean that their output is not biased nor does it mean that their application is not racist in intent or outcome.

If that was confusing, SMBC has a comic about it.

Second, reductionism, or, confusion of cause and agent. Lewontin’s example is tuberculosis. In one sense, tuberculosis is caused by Mycobacterium tuberculosis. More reductionist-ly, TB is a result of the body’s response to the behavior of the bacterium. But saying that M. tuberculosis causes TB is like saying that bullets (or, more reductionist-ly, the hemorrhaging that the path of the bullet through the body cases) are what kill people. We feel very comfortable saying that bullets don’t kill people. It should at least be the guns, and it should probably be the person operating the gun. I think you can go even further, to say that the cause of a gunshot death is partly attributable to the ease of acquiring a gun as well as to the things that caused the person to decide to use that gun to kill someone.

Lewontin’s point is that a scientific treatment of M. tuberculosis legitimizes the reductionist approach to the disease: TB is caused by a bacterium, so we need a drug that kills the bacterium to cure it. To make drugs, we need molecular biology, so we need to do a lot of study about the genetics and biology of M. tuberculosis. In a way, that’s as crazy as saying that we need to fully understand the metallurgy of bullets in order to prevent gun deaths.

Focusing on molecular biology as a cure means that the cure, if produced, will be a drug that an individual company owns and distributes. Is that a good thing? It also means that the problem, which is one of human health, is translated into a scientific, molecular biology problem. TB will be cured only when our understanding of the molecular biology of M. tuberculosis is complete. Is that a good thing?

I’m not saying that antibiotics are useless, but I am saying that most of the reduction in deaths from infectious disease in the developed world are due to improved nutrition, infrastructure, and healthcare in general. (OK, vaccines were really important too.) I’ve started to think about it this way: the developing world mostly has antibiotics but not generally strong healthcare and infrastructure. I think it would better for it to be the other way around.

Third, science as enterprise. Weber famously proposed that one of the reasons people continue to support the status quo is habit. We mostly don’t re-evaluate the world, our role in it, and our moral system everyday. We mostly did what we did yesterday.

I think this is part of the danger of science as enterprise. Being an academic scientist is a career. You pay the bills by getting grants. You get grants by having done work that makes the granting agency think you deserve another grant from them. You can mostly do work in things that you know how to do. If you have NIH grants to study the molecular biology of M. tuberculosis, then you are probably going to keep doing that.

I encourage scientists (and anyone who interacts with scientists) to ask: What is the important problem? How do you know it’s important? If the answer is a dollar sign (e.g., disease X costs Y dollars a year), then ask, why does the disease cost that? One way to say that cancer is more important than malaria is to compute the lost income from the two diseases. I encourage the reverse thinking: malaria is a disease that disproportionately affects the less wealthy people in the world. How do you measure the positive moral quality of working on malaria against the dollars of cancer?

I end with an anecdote. A friend of mine was at a dinner, seated next to a scientist. It turned out that that scientist was a PhD student whose project was, roughly speaking, to make a potato that doesn’t sprout. My friend asked, “If potatos don’t sprout, does that mean the farmer will need to buy the ‘potato seed’ from the ag company every season?” In other words, wasn’t this a project about Monsanto-ing the potato? The scientist was miffed and turned away.

Addendum— I was later reminded of Huxley’s Brave New World, where the World Controller Mustapha Mond gives some exposition near the end of the book about the role of science in the dystopian, highly-controlled community. He says that “[e]very discovery in pure science is potentially subversive” and then goes on to talk about how his own past:

I was a pretty good physicist in my time. Too good—good enough to realize that all our science is just a cookery book, with an orthodox theory of cooking that nobody’s allowed to question, and a list of recipes that mustn’t be added to except by special permission from the head cook.

This possibility, that a biologist can simply cook from a book of orthodoxy, is what scares me, especially because the book of orthodoxy need not be obviously a book.