Conclusions and Difficulties
My contention is that this scheme is not just analytically simple, but is empirically true: that these types of government, studied historically (including contemporary history), reveal these broad characteristics; and also that these characteristics, if the questions they answer separately are put together, are in some way systematically related — as best expressed by the three main concepts. One should reach the same results whichever way one works. This is, of course, too neat by far. The characteristics are themselves generalisations of a considerable degree of abstraction, like any generalisations, so liable to particular exceptions and peculiar and famous hybrids at any given time and place. But the argument comes down to this (if advanced in full): to convince the historian that generalisation is sensible and that any generalisation involves simplification or abstraction; and to convince the political scientist that these apparent systematic relationships can be expressed thus simply — without any more elaborate conceptual framework or any special vocabulary of induced (i.e. quite unknown to the actors in political events) processes.