What is in PageRank? A Historical and Conceptual Investigation of a Recursive Status Index

Abstract

This paper proposes an analysis, based in a software studies mindset, of Google’s PageRank algorithm. It develops two lines of investigation: first, it situates this ‘evaluative metric’ in a larger genealogy of ideas, concepts, theories, and methods that developed, from the 1930s onwards, around the fields of sociometry, citation analysis, social exchange theory, and hypertext navigation. This backdrop is presented as a conceptual a priori, in the Foucauldian sense, from where PageRank becomes ‘sayable’. Second, by comparing the algorithm to a close cousin, Jon Kleinberg’s HITS, and by examining a particular parameter in the PageRank model, the paper shows that the concrete model does not follow teleologically from this historical a priori. Behind the particular empirical case lurks the larger epistemological and methodological question of how to study software. The paper advocates a multilayered approach that combines different types of methodological and conceptual resources.