My undergraduate and graduate degrees are both in political science, but the chief work that helped me to understand the practice of politics is one of history: The Stakes of Power: 1845–1877, by Roy F. Nichols. Political science shares with sociology a bias toward presentism, describing political structures as they currently exist with no sense that they will ever change (at least not in the short run) or, for that matter, have ever changed. The more scientific practitioners of both disciplines claim to be, the more likely they are to become like economists, captives of “laws” of their own making, abstracting from real people making real decisions at a personal level in favor of masses of men who are (or at least seem to be) easier to quantify and to predict. That is why most political-science programs require students not to read Plato and Aristotle or even Machiavelli but to study statistics and to demonstrate their ability to design research projects to find hither-to unnoticed correlations in decades’ worth of polling data. Indeed, the fruit of such projects, over the last 20 years or so, has helped modern political polling occasionally to produce astonishingly accurate predictions, in a manner reminiscent of Hari Seldon’s psychohistory in Isaac Asimov’s famed Foundation series.
As a teenager interested in politics, mathematics, and physics, I was fascinated...