Meaning & Measurement Mini-Conference

David Smilde— University of Georgia dsmilde@uga.edu My main sociological interest is in the way marginalized groups use culture to gain agency over aspects of their social world. In my dissertation I looked at the way some men in Caracas Venezuela adopt the meaning system of Pentecostal Christianity as a way to gain control over certain aspects of their lives (substance abuse, involvement in crime and violence, relationship problems). I have worked with concepts of narratives and network analysis in this study. In a separate project, two Venezuelan colleagues and I carried-out ethnographic research on street protest in Caracas. In this project we work against stereotypes of developing world protest as reactive "bread riots" and demonstrate the rationality of the culture work these actors carry-out through protest. The data from these projects was collected from 1996 to 1999. I have just returned from seven weeks in Caracas in which I carried-out follow-up data collection for both projects. My interest in the M3C is two fold. First, in both of the studies described above I found that intersubjective concepts of culture (such as frame, narrative, discourse, code, context) were more compatible with both the data and the idea of agency. Nevertheless, when "culture" is not located in people's heads, it's hard to know where or even what it is. One way of arguing for the existence of any phenomenon is to measure it. Thus I support any attempts to measure meaning and look forward to discussing the dilemmas of this task. I also have an abiding interested in improving the reliability and generalizability of qualitative methods in two ways. First, by using software for qualitative data analysis, sample sizes can be dramatically increased to the point that statistical significance can be reached. The two projects described above are based on quantities of qualitative data that would be unanalyzable without the help of software (in the former, 84 life histories, in the latter, 50 protest events with approximately 20 in situ qualitative interviews each). For each of the projects I/we used a German software package called Atlas.ti to analyze the data. Second, I think that quantitative techniques can greatly improve analytic penetration of qualitative data. In the article on networks in Pentecostal conversion that I am currently working on, I use Charles Ragin's qualitative comparative analysis based on Boolean algebra. Both of these ways of improving qualitative methods depend on, in some way or another, measuring, or at least isolating and defining, however momentary, particular constellations of meaning. (more information on my work than you would ever want is available on my web page: www.arches.uga.edu/~dsmilde)