Tool-Driven Revolutions in the Social Sciences

There are two kinds of scientific revolutions, those driven by new tools and those driven by new concepts. Thomas Kuhn in his famous book, The Structure of Scientific Revolutions, talked almost exclusively about concepts and hardly at all about tools. His idea of a scientific revolution is based on a single example, the revolution in theoretical physics that occurred in the 1920s with the advent of quantum mechanics. This was a prime example of a concept-driven revolution. Kuhn’s book was so brilliantly written that it became an instant classic. It misled a whole generation of students and historians of science into believing that all scientific revolutions are concept-driven. The concept-driven revolutions are the ones that attract the most attention and have the greatest impact on the public awareness of science, but in fact they are comparatively rare. In the last 500 years, in addition to the quantum-mechanical revolution that Kuhn took as his model, we have had six major concept-driven revolutions, associated with the names of Copernicus, Newton, Darwin, Maxwell, Freud, and Einstein. During the same period there have been about twenty tool-driven revolutions, not so impressive to the general public but of equal importance to the progress of science. Two prime examples of tool-driven revolutions are the Galilean revolution resulting from the use of the telescope in astronomy, and the Crick-Watson revolution resulting from the use of X-ray diffraction to determine the structure of big molecules in biology. The effect of a concept-driven revolution is to explain old things in new ways. The effect of a tool-driven revolution is to discover new things that have to be explained…

The above quote is from Freeman Dyson’s book Imagined Worlds (49–50), where he goes after Thomas Kuhn for having too abstract a notion of how scientific advancement occurs. The idea that tool-driven revolutions “discover new things that have to be explained” might be apt in the harder sciences, but applying that thinking to the social sciences might be a bit dangerous.

Political methodologists, who are keen to tell anyone who’ll listen that they have the journal with the highest impact factor, are the people in political science who develop new tools to analyze politics. They’ve created a bevy of new standard errors, new estimators for different data types, and new question designs meant to better tap some latent dimension (how should we measure racial animosity, for instance?). Some of these things have been very useful — the addition of hazard models to political science has been a tremendous boon to those studying how policies diffuse across states or countries. What makes a useful method useful is that it was created with an eye toward how to better answer a question that already existed.

At the same time, plenty of tools get built by political methodologists who put the cart before the horse a bit — creating methods in the hope that others will find a use for them, or bolting on a “question” mainly to demonstrate the method’s value, without the question itself being especially important. In other words, some political methodologists create a new tool without fully having a “new thing that has to be explained.” In my view, this needlessly complicates political science, creates redundant projects that all examine the same question with newer methods, and distracts from discovering genuinely new things to explain.

I think richer political science can be had if people study politics first and then seek the best available tool second, to help understand the politics around them. I’d bet that for most questions, the tools already in existence will suffice.