In search of a network theory of innovations: Relations, positions, and perspectives
结合卢曼的社会系统理论与拉图尔的翻译社会学，提出一个分析创新网络的多层框架，区分关系、位置与视角三个层次，用于研究创新周期。
As a complement to N elson and W inter's (1977) article titled “In S earch of a U seful T heory of I nnovation,” a sociological perspective on innovation networks can be elaborated using L uhmann's social systems theory, on the one hand, and L atour's “sociology of translations,” on the other. Because of a common focus on communication, these perspectives can be combined as a set of methodologies. L atour's sociology of translations specifies a mechanism for generating variation in relations (“associations”), whereas L uhmann's systems perspective enables the specification of (functionally different) selection environments such as markets, professional organizations, and political control. Selection environments can be considered as mechanisms of social coordination that can self‐organize—beyond the control of human agency—into regimes in terms of interacting codes of communication. Unlike relatively globalized regimes, technological trajectories are organized locally in “landscapes.” A resulting “duality of structure” ( G iddens, 1979) between the historical organization of trajectories and evolutionary self‐organization at the regime level can be expected to drive innovation cycles. Reflexive translations add a third layer of perspectives to (a) the relational analysis of observable links that shape trajectories and (b) the positional analysis of networks in terms of latent dimensions. These three operations can be studied in a single framework, but using different methodologies. L atour's first‐order associations can then be analytically distinguished from second‐order translations in terms of requiring other communicative competencies. The resulting operations remain infrareflexively nested, and can therefore be used for innovative reconstructions of previously constructed boundaries.