Political science spent a century arguing over how scientific it should be
In 2000 a group of political scientists launched what they called the Perestroika Movement, a revolt against the mathematicisation of their own field. They wanted room for many methods and research that mattered to people outside universities. It was the latest swing in a long argument over what studying politics scientifically should mean.
The discipline split away from history, philosophy and law only in the late nineteenth century, riding the rise of empiricism and positivism as universities professionalised and created chairs with political science in the title. Even then many still saw it as a branch of history. The American Political Science Association, founded in 1903, now serves more than 11,000 members in over 100 countries. After a UNESCO push in the late 1940s, an International Political Science Association appeared in 1949, with national bodies in France the same year, Britain in 1950 and West Germany in 1951.
The first big shift came with the behavioural revolution of the 1950s and 1960s, which moved attention from institutions and legal texts to the measurable conduct of voters and groups, in work by Robert Dahl, Philip Converse, Paul Lazarsfeld and Bernard Berelson. By the late 1960s William Riker and his colleagues at the University of Rochester were importing game theory and economic models to study Congress and voting. Scholars nonetheless concede that progress toward a systematic theory of politics has been modest and uneven.
The difficulty is partly built in. Unlike minerals, chemical elements or stars, the subjects of political science make conscious choices and can be only partly observed, and controlled experiments are hard to reproduce, although experimental methods are increasingly common. Researchers therefore mix field experiments, surveys, case studies, process tracing, ethnography, interviews, agent-based models and equation-based models of opinion dynamics.
Some of the more intriguing work looks for warning signs. One proposed statistical indicator of an approaching crisis is a simultaneous rise in variance and correlation across large groups. Evolutionary psychologists add a longer view, arguing that human political instincts evolved for small ancestral groups rather than modern mass states, which may help explain systematic biases in how people think about politics today.
Source: Political science
