Abstract :
To diversify the range of methods used in this symposium beyond statistics, I will begin with a case study, using my own training and experience in the field as evidence. I received my graduate training at the University of Rochester, perhaps the most specialized of the top-20 political science programs. We were required to take a course in scope and methods which focused on philosophy of science and a course in political philosophy; most students took a sequence of courses in statistical methods, and some took courses in formal theory. During my professional career, I have published in journals with different methodological inclinations according to the data presented by Andrew Bennett, Aharon Barth, and Ken Rutherford.1 Most, but not all, of my research has been technical, using a combination of game-theoretic models and statistics. Methodological specialization is the foundation of my career. I contend that specialization in graduate training and methodological focus in journals are necessary for diversity of methodology in the field. Specialization is a consequence of progress and diversity in the field. Imposing some form of universal training in all methodologies would either be so superficial that it could not create any depth of knowledge of any methodology or so demanding that graduate students would have little time left for substantive classes. Requiring journals to represent all methodologies with equal numbers of articles would not increase, and could decrease, the level of communication in the field. I recognize that the case of the American Political Science Review (APSR hereafter) is distinct and discuss it separately at the end of my comments. I use both formal models and statistics in my research, and my comments generally apply to both.
