Since Robert Dahl's seminal writings on democracy more than two decades ago, interest in the topic has emerged again, especially among scholars analyzing democratic transitions. Great strides have been made in revealing the uncertain nature of these transitions (O'Donnell et al. 1986; Malloy and Seligson 1987; Diamond, Linz, and Lipset 1989; Hakim and Lowenthal 1991; O'Donnell 1994), in methodologically analyzing them as contested and “crafted” rather than spontaneous (Di Palma 1990), and in documenting the class and social forces that make democratic outcomes more likely (Rueschemeyer, Stephens, and Stephens 1992; see also Moore 1966). Despite these advances, there has been little change in our theoretical understanding of democracy. As Bruce Cumings has perceptively noted, recent studies of democratic transition have “given way to atheoretical and idiosyncratic explanations of more or less successful democratic ‘openings’” in which little time is spent elaborating “the decision rule for saying this person is hard-line or soft-line, that system is ‘liberalized autocracy’ instead of ‘limited democracy,’” or for defining democracy itself. If scholars do bring theory into their writings “through the back door of the obscure but telling footnote,” he observes, “rather than advancing their own conception of democracy, [they] uniformly define democracy by reference to Robert Dahl's Polyarchy, a classic pluralist account of the North American system” (Cumings 1989:15–17).