Anti-Climacus

JUST A QUICK THOUGHT: On the ongoing skills requirement debate: As per the (somewhere) below, one semester of stats not taught qua stats would be insufficient to train someone to understand statistics, but, frankly, even two semesters of it can only cover very little ground (probability theory, OLS, and a little bit of probit and logit). If you're a theorist, those requirements may be entirely foreign to what you want to do; if you're an Americanist/quantitativist (in comparative or IR), it probably won't be enough.

Similarly with theory, I think. One class, particularly not taught qua theory, would be almost entirely pointless. Two classes, if you're a theorist, won't be nearly enough, and for non-theorists, would only cover a little bit of ground. This leads me to two observations:

1. There's a tendency to place languages and methods skills in opposition to each other (they're both tools for doing a certain kind of work), but it seems like methods and theory are the more sensible opposites.

2. There are probably pretty good arguments for keeping both as requirements: a theory requirement definitely signals that theory is especially strong and valued in the department, in the same sort of way that a methods requirement would signal that sort of teaching being taken seriously. Both, for a person of a certain type of mind, are choiceworthy for their own sake, and are instrumentally valuable inasmuch as they make it possible for people to talk to each other across the potentially divisive positive-normative divide (I think that the obligation runs equally both ways there). Obviously the value of this one depends on what exactly one's graduate education is supposed to be doing, which has always seemed to me to be one of the great open questions in the discipline as a whole (though I could be wrong about that).