THR Web Features   /   September 2, 2015

In Defense of Scientific Asteroids

Jeffrey Guhin

Great Comet of 1577 (Georgium Jacobum von Datschitz). Via Wikimedia Commons.

People often ask me to define sociology, nearly always because I’ve just told them I’m a sociologist, but sometimes because the person asking me is also a sociologist, and none of us are actually that sure. I’ve got a few distinctions I keep at hand:

  1. Social workers help people. Sociologists just study them.
  2. Economists study how people make choices. Sociologists study how people aren’t able to make choices.
  3. Psychologists study individuals. Sociologists don’t believe individuals exist.
  4. Anthropologists never saw a category they couldn’t “problematize.” Sociologists never saw a category they couldn’t turn into a two-by-two table.
  5. Political scientists are positivists. Sociologists cross-dress as positivists.

None of these characterizations are true, in any of their directions. You can find a sociologist to disprove any of these distinctions, which, some might say, is part of the problem with sociology. We don’t have a clear sense of who we are.

In a recent article at The Smart Set, Michael Lind proposed an ideal university whose most important characteristic is its abolition of the social sciences. Sociology and all its siblings cannot escape the error of the “physics envy” that motivated their existence. Lind’s got a problem with social forces or anything bigger than the mess of individuals making their own individual decisions. It’s an ironic critique for someone who had just been criticizing rational choice theory (whose central conceit is exactly the same criticism), and it’s made more difficult by Lind acknowledging institutions, which is a category so capacious as to include everything from marriage to I.B.M. Yet whether Lind is right is less important than what the criticism represents: Something is rotten in the state of social science.

The article refers to similar debates in international relations, political science, and economics, all about the wrongheadedness of hyper-precise mathematical models and the absurdity of presuming a lab-like science of the social, one in which variables can be isolated as cleanly as they can be in physics. These debates are made more relevant by a recent paper showing how hard it is to reproduce some of psychology’s most important findings, and while the problems with psychology are not as big a deal as they might appear (it gets a bit wonky, but it has a lot do with P-values), they do raise the question: Is the science in social science worth defending?

Is There a Science in Social Science?

The short answer is yes, and the long answer is that it depends on how you define science. Physics, for example, is much broader than the equations mapping billiard balls that physics envy is trying to reach. Part of the problem comes from certain social scientists trying to shoehorn their methods into any one of many philosophies of science, but, as Susan Haack and others have shown, the only thing philosophers of science can really agree on is that science is basically just thinking carefully. Sure, it tends to be methodologically agnostic, but it isn’t always. It tries to make falsifiable arguments, except when it can’t. The same goes for control groups and lab conditions and appropriate models. Science is a lot more diverse than it appears. But perhaps the most important characteristic of science is that it builds upon itself. We know more about trees than we did ten years ago, and we will probably know more about them ten years from now that we do now. There have been some pretty important sociological and historical critiques of science, but, despite all the good points about the messiness of scientific politics and personalities, it really does seem to be the case that its knowledge is cumulative.

Can that be said of social science? Do we actually know more about society than Tocqueville did? It’s a stupid question in some sense, because in most ways the answer is obviously yes. We know more about voting patterns, about homelessness, about why people walk into schools and kill children. Are many of those studies wrong in bigger and smaller ways? Of course they are, but that’s also true for many studies in the physical sciences. Someone suggests an interesting argument, and then other people rip it apart and see what holds. That model is made more complicated by how journal publishing has been institutionalized (it’s easier to publish a new argument than a critique of an existing one) but the problem is not a damning one.

What is a damning problem is the stuff we still don’t know any more about than Tocqueville did. Social science is at its best when it’s scientific in its method but not in its ambition: We’re just not going to be able to solve the really big questions about the nature of human life and all societies in all times, at least not in the way that physics really does seem to be able to talk about the nature of the universe. All we have access to are bromides both cautious and banal: Humans tend to like sex, they tend to be social, and they tend to prefer more status to less. They like to accumulate resources and would rather buy them cheaply. These generalizations are wrong just enough to be imperfect, but they’re right enough to be heuristics for some of the most important theoretical approaches the social sciences have developed. And that’s where the more important envies start.

Economics Envy

While I can’t speak for other social scientists, I rarely encounter physics envy in sociologists. I sometimes explain sociology to people as being a middle space between political scientists who believe their studies of one survey can explain all human behavior and anthropologists who believe their twenty years with fifteen people can’t even explain those fifteen people. Sociologists are interested in explanations, but especially in the past few decades, we’re interested in doing so through mechanisms rather than laws: If anything, we’re envious of biology rather than physics. Like biology's explanations, ours are sometimes quite broad (a study of all vertebrae, or all rituals) but they tend to be much more focused: an explanation of a specific bacteria’s reproduction, or a specific social movement’s success. These explanations depend upon broad mechanisms, but the further you remove that mechanism from its original case, the harder it is to apply. The study of why the Civil Rights Movement worked might help explain something about the Progressive Movement, or the European revolutions of 1848. It couldn’t really explain Alexander the Great.

Sociologists are asteroids rather than planets. We’re linked more than anything else by a set of core texts (Marx, Weber, Durkheim, sometimes Simmel) and a set of normative concerns (stratification, urban blight, anomie) that we’ve managed to internalize without threatening the scientific objectivity of our work. Yet these are too many gravitational centers, and most sociologists bear only a family resemblance to each other. We can all identify people who call themselves sociologists who do work pretty similar to ours, but we also know other people who also call themselves sociologists with whom we have almost nothing in common, except that we study society. There’s a reason sociologists can get jobs in political science, anthropology, and any of the many “studies.” We study everything, and we study everything in just about every way.

We often decide this heterogeneity is a problem. There’s a recent movement in American sociology called critical realism, and one of its stated goals is to make American sociology more coherent. While the motives behind efforts to make pragmatism more central to mainstream sociology are more diverse, speakers at a recent conference on pragmatism at sociology’s major national meeting mentioned similar goals. Other movements—analytical sociology, the sociology of Pierre Bourdieu, and certain strands of New Left materialism, among others—have all identified the same problem at sociology’s coreless center. We’re asteroids. We’re too spread out. We need to get together, form a normal science, become a planet. Then we can be important and clear and consistent, attracting the moons of government attention and big grants and greater prestige within the university. Not like physics, but like our popular older sister, economics.

Part of the problem is constitutional: Sociologists are trained to be idiosyncratic and interesting. Nobody hires disciples, we’re told in grad school, and we’re warned not to work with an advisor who just wants us to reproduce his paradigm. (It’s almost always a he who wants us to reproduce his paradigm.) Lots of sociologists would like normal science, but none of them wants to be a normal scientist. Yet the bigger part of the problem is that the work is just too complicated. We study too many different things, and in too many different ways. Does that make us less important than economics? Maybe. But it also makes us more careful about generalizations and, I hope, less wrong about them too. I’d keep us in the university. But then, what do I know? I’m no physicist.