Our account of this science will be adequate if it achieves such clarity as the subject matter allows.
Previous chapters have presented a succession of definitions, metaphors, models, normative constructs, and theoretical traditions that can help organize our thinking about contemporary processes of democratization. If theory-building means only the construction of tightly defined covering law type predictive theories, then this exercise has generated almost nothing that can be 'tested' against the evidence. If looser probabilistic claims are counted as theory then some provisional judgements become possible, but their predictive power remains quite low. After all, although comparative democratization studies were already well advanced by the end of the 1980s, almost all political scientists were caught by surprise when the great majority of communist regimes underwent mostly peaceful and almost simultaneous transitions to what may at least provisionally be labelled 'democracy' in the 1990s. Some eminent scholars had even explicitly predicted that this would not come about. 1 Looking ahead from the standpoint of 2001, existing scholarship provides only limited guidance on the parallel question of whether or not democratization is predictable in the Islamic world. This is not to decry rigorous predictive theorizing in those domains where it is feasible, but only to note that most of our subject-matter is not of that kind.
Nevertheless we have generated a considerable range of categories, concepts, and hypotheses that are intended to assist thinking