Archive for August 6th, 2009
One of my favorite authors, Chris Hedges, is going to be giving a public talk at Deer Isle next week. Though it’s three hours away, I’m going to go there and hear and hopefully meet someone who has inspired my teaching as well as my current research.
Social science has gone through various phases. In the 19th century French sociologist Emile Durkheim posited society as an entity that could itself be the subject of study. Durkheim recognized that society was more than just the sum of individual choices and actions; social facts had to be understood at the societal level of analysis. Sociology paved the way for the rest of the social sciences by arguing for a scientific study of society and social systems.
Durkheim was not alone, of course. Karl Marx, studying political economy, recognized the connection between economic production and political outcomes, and attempted a scientific study of how this developed and progressed through history. Though Marx’s overall theory had many flaws, it brought a materialist bent to the analysis of politics. In 1903 political science established itself as a discipline with the founding of the American Political Science Association (APSA). In the 1920 the British started studying International Relations (IR) as a field separate from diplomatic history, and after World War II IR became a distinct subfield of political science, as people wanted to investigate international conflict and the international political economy.
In the early years, political philosophy and history were distinctly tied into the study of politics. Political Science was, in essence, an interpretive science, meaning that the practitioner used his (until the 70s the field was male dominated) expertise to analyze data and interpret results. By the 1960s that wasn’t good enough for a new generation of social scientists who wanted to speak with the authority of science, not just learned opinion.
With computers, statistics, and a new sense of purpose, they wanted to be truly scientific, develop theories to explain politics and social life, and understand the political world the same way a biologist might understand the botanical world. Unfortunately, a few problems were in the way.
The first was the problem of complexity. Except for areas like polling, election studies, and a few others which lended themselves very well to statistical analysis, it’s extremely difficult to control for the vast array of variables that influence politics. You can’t construct a time machine, go back to 1933 and kill Hitler, and then watch to see what difference removing that variable made. Controlled experiments are impossible in most cases, and thus it’s hard to know what variables are truly causal. The result is that models rely on statistical correlations that are extremely small in comparison to what would be required to claim causality in the natural sciences, and it becomes easy to ‘fudge the numbers’ by altering formula and weighting variables differently.
The second problem is perspective. Every researcher has a perspective, and brings that with them to their analysis. It might be an overt political bias, or simply a tendency to look at the world with the eyes of a western rational enlightenment thinker. In any event, perspective determines how problems are formulated, what issues are addressed in research, and how the data get interpreted. It becomes very easy for scholars to see what they want to see, and design studies to get the results they want to get. And though critiques and peer review, plus the self-conscious reflection of scholars who want to avoid this mistake, can mitigate this problem, it never completely goes away. Mix complexity and perspective, and the notion that social science is truly scientific becomes hard to hold.
The problem is deeper, however. Social scientists increasingly reject the original notion that their work can be value neutral. The idea was that an analyst could put his or her values aside and let the data and the methods give the results. However, thanks to the above problems, the choice of data and method, the framing of the research question, and the interpretation of the results are always influenced by the analyst’s perspective and values. One response to that is to be up front with ones’ biases and values, so that this is taken into account by critics and other analysts. That’s fine, but raises another question: is there a way to judge between different value systems?
Simply: social science as science runs into a problem when it tries to mesh science with philosophy: values and ethics are considered essential parts of the subject matter, but cannot themselves be scientifically investigated. Sure, one can investigate how ethical systems have been constructed, sociologists like Peter Berger can delve into trying to understand how religions function, but issues of meaning and value remain inherently relativized — ethical systems and values are interesting products of social interaction, nothing more.
Up through 2003 or so, I was slightly uneasy, but generally comfortable with that state of social science. Reality is a social construction, and we can examine how it was constructed, study the role of ideas in shaping it, and even be upfront in talking about how it might be transformed for the better, even if the “better” remains a matter of opinion. Then I read War is a Force That Gives Us Meaning by Chris Hedges, a book I stumbled upon while browsing at Barnes and Nobel. I started using that book in my classes, and reading other works by Hedges. Both his work, and the subject matter of classes I was involved in (co-teaching “Children and War,” as well as courses with people from the Arts department), I began to think there is a big hole in the way social scientists approach the political and social realm.
It’s not just that we cannot scientifically determine meaning, value, or ethics, leaving that up to philosophers or individual choice, but our methods and assumptions de-value and de-base such contemplations. We ignore and distrust the role of emotion and sentiment in judging and analyzing social reality. If as a class we are in tears talking about the experience of children in a war, the social scientist in me wants to come out and say “OK, it’s sad and all that, but let’s analyze the variables that caused this and how it might be mitigated — or if, perhaps, it was necessary given the larger picture.” The emotional impact of that human experience is an aside, something that is more like to mislead analysts since it distracts from rational thought.
That, however, is an untenable assertion. This simply replaces sentiment with abstraction (something I’ve already noted is dangerous). Abstract concepts and assumptions are treated as the “real” stuff of the social universe, human experience and sentiment is outside the realm of consideration. It’s subjective, immeasurable and irrational. Thus we have whole fields of political science studying war, but few even considering the impact on children. If one is talking about options in foreign policy and says “what about the impact on families and average people if we go to war,” that will be completely shouted down by an appeal to abstract theories, notions of national interest, and potential threats from our enemies.
To be sure, simply turning around the problem — putting sentiment and human experience first while ignoring abstract, rational thought — would be to simply replace one problem with another. The key is to find a way to balance the two, to be able to integrate concern about meaning, values and ethics into social science, even without making claims about what is ‘scientifically’ or ‘objectively’ right. Hedges work inspires me to think through that as I continue in my research this year. When I go to Deer Isle and hear him speak next week it’s not just the usual going to hear an author I admire or whose work interests me, it’ll be to listen to someone whose work has inspired me and altered my approach to my research.