rality in a way that accounts for the patterned disorder that institutions routinely create.
To fix on questions of timing in institutional politics requires using history in ways that historians themselves seldom do. In our view, time is not the medium through which the story of this or that institution will unfold; it is itself the central problem. The discovery of patterns -- provisionally, even of orders -- is an indispensable step. This may entail uncovering several layers of institutional politics that compose a single moment, or reaching across historical periods to identify a repeated sequence or configuration. The idea is not to detach the analysis from time-bound descriptive minutiae, but to explicate the timing of such details, and to explain institutionally the shifts and developments that are constantly changing the political landscape. There are no "time lags" in this perspective, no institutional delays where older arrangements "catch up" with new; pieces held over from earlier patterns are part and parcel of the institutional composition and of the institutional construction of temporality itself.3
Institutions make history by the routine engagement of the tensions and contradictions among their various ordering principles and by their bending and reshaping of each other into patterns of change in time. Layers, not systems; dissonance, not fit; conjunctures, not regularities: These are the points of entry to a genuinely "new" institutionalism.
Colleagues who were kind enough to offer comments on this essay were Joyce Appleby, Paul DiMaggio, Morris Fiorina, Jeff Frieden, David Mayhew, David Plotke, Ian Shapiro, and Alex Wendt.