What is a “Democratic Experiment”?

Abstract

This paper examines a key concept of asking what is a experiment? Starting with Donald Campbell's argument in favor of an experimenting society, I critically examine whether randomized controlled experiments should be the gold standard for experimentalism. I next suggest that a more compatible conception of might be found in recent work on design experiments that iteratively improve design through real world application. Avoiding verificationism, Dewey's treatment of as a basis for modem authority adopts a more general interpretation of as a form of cooperative inquiry. Finally, I examine the meaning and application of in the area of adaptive management, an approach to natural resource management. This analysis finds many of the same tensions in practice as they are described in theory, particularly between experimental and collaborative forms of adaptive management, pointing to the challenges of making experiments democratic.1. IntroductionPragmatist philosophy holds in high esteem. Experimentation is a leitmotif that captures, in a single stroke, the Pragmatist commitment to fallibilism, to active learning through discovery and experience, and to the iterative growth of knowledge. It is not surprising then that recent work in the Pragmatist-inspired social sciences places this leitmotif at the center of their analysis. Under the banner of experimentalism, social scientists have identified and explored the possibilities of new forms of political practice. But as a richly evocative concept, also carries with it the intellectual baggage associated with the concept experiment. In this paper, I will examine this intellectual baggage, taking the idea of and experimentalism seriously. To do this requires that we examine a number of issues. First, in what sense is the language of experiment and experimentation, derived from the laboratory, useful and appropriate when applied to the social world? Second, what is the meaning of when we apply the adjective democratic to it? How can a technique designed to isolate the effect of an independent variable on a dependent variable be said to be democratic? If we are to take the concept of seriously, I will argue, we must directly confront these issues.I think it is fair to say that the literature on has taken the idea of more in its connotative than in its denotative sense. In the classic article on by Dorf and Sabel (1998), has a general connotation implied by the key governance processes outlined in their paper - benchmarking, simultaneous engineering, and learning by monitoring. Although never explicitly defined, we can infer from these processes that refers broadly to a process of learning from the experience of others. Similarly, in a more recent article applying the concept of to EU governance, Sabel and Zeitlin (2008) identify four key features of experimentalist governance: (1) framework goals and measures established through joint deliberation between member states and the EU; (2) lower-level units with the autonomy to address these framework goals and measures as they see fit; (3) regular reporting by lower-level units on their progress in meeting framework goals and measures; and (4) periodic revision of framework goals and measures. Again, the authors do not focus in a literal way on the meaning of but we can infer from their description that refers to the iterative efforts of lower-level units to devise strategies for meeting framework goals, followed by information pooling about the results of those efforts.It may be something akin to a category error to literally scrutinize a concept like experimentation if it is only being used in a connotative way. …