This is an extended excerpt from Charles Tilly, “Democracy Is a Lake,” Chapter 15 in George Reid Andrews and Herrick Chapman, eds., The Social Construction of Democracy, 1870-1990 (New York: New York University Press, 1995, pp. 365-366). The emphasis (bold format) is mine.
Timescale matters both theoretically and practically. Theoretically, democracy’s time could resemble that of an oilfield, a cultivated garden, or something in between. An oilfield, the specific product of millennial history, conforms to regularities strong enough that petrogeologists can spot likely untapped deposits or explain how an oil well works. Yet experts cannot produce a new oilfield at will wherever they want. The presence of oilfields depends on long, long conjunctions of circumstances that appear rarely in history, and are little amenable to human intervention.
Gardens are different. They will not flourish everywhere, but given adequate soil, sun, and precipitation, many different sorts of gardens grow in a variety of environments. As in the cases of oilfields, specialists in gardens are perfectly capable of explaining how they work, indeed of making contingent predictions about what will happen if X or Y happens first.
In both cases, experts know enough to intervene, within clear limits, to produce desired results with considerable probability. They know what aspects they cannot influence. They even know enough to identify a wide range of interventions which, however well meaning, are likely to fail. Differences between the cases do not concern the phenomenon’s regularity or intelligibility, but the nature and timescale of the regularities involved, which in turn determine the phenomenon’s susceptibility to deliberate promotion. We have no a priori warrant to think of democracy as resembling gardens more than oilfields. If oilfields offer the proper analogy, valid explanations of the presence or absence, waxing and waning, of democracy will combine very long histories with dense accounts of short-term dynamics. In that case, we might well conclude that Barrington Moore tells us how to analyze the foundations of democratization better than recent short-term planners.
Practically, the promotion of oilfield democracy will require the transformation of environments, indeed the creation of whole histories, over centuries or even millennia. Both planned ruling-class intervention and popular collective action will be irrelevant to the success or failure of democratic projects. The cultivation of garden-style democracy, on the other hand, can occur in a wide variety of environments with relative rapidity through many combinations of elite and popular action. If the garden analogy holds, the secret will be to find or create those environments that can support some kind of democracy, then adapt the design and cultivation to the capacities of each environment. Two essential points follow: (1) the validity of various theories and metatheories of democracy depends on the general character and timescale of the phenomenon, which remain highly contested; (2) the validity of theories and metatheories of democracy has profound practical implications. Both points are at issue in the effort to create a sound social history of democracy, constructed or otherwise.
Theorists of social change seem to be paying more attention to timescale right now than they have in a while, or at least they are being more explicit about it. A recent blog post by Brian Whitaker, Middle East editor for the UK’s The Guardian newspaper, thoughtfully considers how focusing our attention on the short-term consequences of the so-called Arab Spring might cause us to miss the long-term transformations that, he and many others believe, have helped to produce those upheavals and that give us reason for optimism about their eventual outcomes. Whitaker’s ideas about the sources of those long term-transformations echo a version of modernization theory postulated by Ronald Inglehart and Christian Welzel in their 2005 book, Modernization, Cultural Change, and Democracy: The Human Development Sequence. In spite of its teleological and normative baggage, modernization theory generates predictions that track broad trends over the past two centuries fairly well. Meanwhile, Francis Fukuyama self-consciously stretches the timescale even further in his latest work, The Origins of Political Order: From Prehuman Times to the French Revolution. In so doing, he follows a path recently traversed by Douglass North, John Wallis, and Barry Weingast in their 2010 book, Violence and Social Orders, which grew out of an NBER paper provocatively entitled “A Conceptual Framework for Interpreting Recorded Human History.”
Similar questions about cause, effect, and time are also crystallizing in the raging debate among development economists about the relative merits of randomized control trials (RCTs) and comparative economic histories as bases for evaluating the effectiveness of foreign-aid programs. I don’t know that literature well enough to point to specific sources on one side or the other of that debate, but Yale economist Chris Blattman recently weighed in on his blog with a tilt toward the long view, picking up on points made by Arvind Subramanian at the Center for Global Development in response to a recent column by The New York Times‘ Nicholas Kristof.