Having had a foot in each camp for over 30 years, I am acutely aware of our discipline's customary division of the study of political parties between American and non-American subjects. The division remains most apparent in teaching programs despite increasing cross-national research efforts during the last few decades. I doubt that merger is entirely feasible. The division is deeply rooted in the general development of political science in the United States, and something like it is characteristic of other subjects as well as of parties. Legislatures, executives, and courts readily come to mind. Significantly, they are governmental institutions so linked to a country's constitutional and historical experience that a national context for their study seems plainly appropriate. Although parties are not governmental institutions in the same sense as are legislatures, executives, and courts, they have become more than merely private political associations. Most notably in the United States, they are plainly quasigovernmental in many respects. Perhaps this helps to explain why American political scientists have treated our parties, along with governing agencies, as American institutions while leaving parties in other nations for treatment under the rubric of comparative government and politics. Much can be said in behalf of that institutional tradition, but one must grant that it ties our work to geographic units and thus keeps many political scientists closer to historians, in at least one methodological sense, than to economists or sociologists. For better or worse, we thus appear less scientific, conceptually, than the ambitious title of our discipline suggests.