ネットワーク論への再招待

――「特殊な方法」という神話を越えて――

抄録

  In social sciences, social network analysis has been viewed as a special method requiring highly complicated techniques mostly for quantitative data analysis based upon some mysterious ideology about social structure. Even after the abrupt emergence and popularization of so-called “the new science of networks” in a wider society including various natural sciences in the last decade, the myth of social network analysis as a special method apparently remains in sociology and cultural anthropology. This article attempts to disprove the myth by examining the usefulness and effectiveness of social network analysis or more broadly-defined social network approach, and to promote its application to studies on a wider range of social and cultural phenomena.

  It is especially emphasized that network approach is not only a collection of analytic techniques but also a theoretical perspective on societies and individuals as well as the relation between the two. The characteristics of the network perspective are explored in terms of three aspects. First, it focuses upon relations among individuals, not upon their attributes. Second, it focuses on networks, not on groups. Third, it explains relations in a relational context.

  To exemplify these points, I go back to the mid-twentieth century Manchester school of social anthropology and critically review pioneering social network studies by J. A. Barnes and E. Bott. Their followers’ later development in crystallizing the above three aspects of the network perspective is also discussed. Notorious weakness of network approach, too much emphasis on structure and too less emphasis on culture and human agency, is argued to be overcome by adopting concepts such as identities and (sub) cultures into network analytical studies, and by applying some network approaches to cultural studies which previously lacked structural perspective.

  In addition, I assert that more qualitative analyses should and can be incorporated into network analytical studies. In particular, mixed methods, combining both quantitative and qualitative data analyses, are full of promise to bring about more fruitful and productive outcome in many areas of social scientific research. I illustrate this issue with examples of network studies making the more or less use of mixed methods.