On the Sociometry of Search Engines: A Historical Review of Methods

This contribution traces the changing history of a method fundamental to all current search engines - measuring the authority of a website through its links, and it discusses the perspectives on society underlying this. I start out with references provided by search engine designers on the historical roots of the technologies they use. Then I follow multiple traces on a search to see how these tech-nologies determine authority. I examine concepts of social relationships and how they are transformed into socio-technical forms of communication that today seem natural. I will refer to three related forms of handling social relations: 1) Ever since the 1930s, a "sociometric revolution" has been advocating group-psychological interventions and the visualization of how they are embedded in social structures as a means of self-realization. To this end, mathematical methods were developed that were designed to assign authority directly based on group behavior, without drawing on any prior scheme of classification. 2) Present-day bibliometric methods for developing and surveying the sciences continue this tradition of thought. External references were eliminated in favor of an allegedly politically independent science, and self-referential models aggregating objective authority were developed - e.g. the Impact Factor and the Science Citation Index - both of which soon became bench-marks exercising a strong influence on their objects of measurement. As auto-mation and data archiving progressed, such measuring methods turned into behavioral instructions. 3) Search engines incorporate and expose a particular relationship between social relations and authority, although clear qualitative points of reference are missing. As mediators (and notorious data collectors) search engines produce and represent authority themselves and render this self-referentiality visible. However, this strategy is not only found in search engines, it is part of a more general tendency of social optimization.