The civility that makes democratic politics possible can only be learned in the associational networks; the roughly equal and widely dispersed capabilities that sustain the networks have to be fostered by the democratic state.
If democratization is viewed as a long-term, complex, and partially open-ended process then our theatrical metaphor needs to be supplemented by a more extensive explanatory account. Before a democratic transition can begin there must be a political community receptive to democratic aspirations. After the regime change has taken place, the same community must respond to the new possibilities for political participation. The stability and overall direction of the process will depend on this larger social context. Several alternative strategies have been attempted to characterize the societal variables that may encourage democratization (or not). Causal connections have been sought between democratization and such elements as 'modernization', per capita income levels, the expansion of commercial society, the rise of a 'middle class', the emergence of organized labour, and more culturally specific variables such as Protestantism, ethnic homogeneity, and so on. However, if democratization is viewed as a partially normative process of social construction and persuasion, then such tight patterns of causal determination are improbable or, if found, are unlikely to remain stable. This chapter therefore considers an alternative type of explanation that directs attention to discursive and interpretative processes that are more compatible with the view of democratization developed in Chapter 1 .
Theorists of 'civil society' attempt to explain processes of democratization by reference to societal context, often touching on many of the same variables adopted by other research strategies, but their accounts are more normative and less determinist. This chapter first outlines the genealogy of such theories, and then settles for a