Treating Identity as a Variable

This paper outlines our initial thoughts on treating identity as a variable. It is part of a longer-term projectto develop conceptualizations of identity and, more importantly, to develop technologies for observingidentity and identity change that will have wide application in the social sciences. Heretofore the usualtechniques for analyzing identity have consisted of non-replicable discourse analysis or lengthy individualinterviews, at one extreme, or the use of large-N surveys at the other. Yet, much social science researchrelies on historical and contemporaneous texts. Specifically we hope to develop computer-aidedquantitative and qualitative methods for analyzing a large number of textual sources in order to determinethe content, intensity, and contestation of individual and collective identities at any particular point in timeand space. These methods will allow researchers to use identity in a more rigorous and replicable way as anindependent (and dependent) variable in a wide variety of research projects. They will also allow morerigorous testing among identity-based hypotheses—such as those drawing on social identity theory, roletheory, or cognitive theories—along with other variables in explaining behavior. Researchers may also beable to develop early warning indicators that might be used to track growing intensity of out-groupdifferentiation, a development which makes subjected groups more susceptible to identity-basedmobilization for conflict. Perhaps most important, scholars will, using these methods, be able to observemore systematically the contestation and construction of identity over time.