Political analysts have long used spatial representa tions as a substrate for behavioral models, especially for legislators’ voting choices. Most commonly, we think in terms of a single-dimension continuum of ide ology running from left to right. However, there are two weaknesses of such an approach for the study of elite political behavior. First, in a number of par liaments, the underlying space appears to be at least two dimensional; second, there are good reasons why, in many instances, political action cannot be under stood as a direct expression of preferences, but has to be understood in terms of a more complex strate gic orientation. While many political analysts work on generalizing core behavioral models, we here instead consider three different modes by which partisans can indicate their positions—the speeches they give, the votes they cast, and the way they interact with one another. We examine the Weimar Reichstag, a parlia ment in which there were not only two interpretable dimensions, but a shift in importance from one dimen sion to another. We show that this shift was presaged by interaction patterns, and that close attention to the interactions of parliamentarians may offer a new win dow into elite political behavior.
[more]
