ANTA Actor-Network Text Analyzer

Having its roots in the laboratories studies movement, actor-network theory has always had in ethnography its privileged research method. Still, at least in the words of its founders, ANT has always longed for a more quantitative grasp of its objects. Following social actors along their alliances and oppositions, Bruno Latour, Michel Callon, John Law and many others soon touched the limits of qualitative investigation, as scientific and technical networks revealed too vast and heterogeneous to be explored with interviews and observatory participation. At the same time, though, standard quantitative methods seemed unfit for ANT purposes. Devised to average and aggregate, pools and statistics proved useless to follow individual actors along specific chains of action. Social network analysis seemed to offer a more promising alternative and early ANT scholars recognized an elective affinity with this research method. More precisely, it was scientometrics that drew most of ANT founders’ attention for its capacity to explore and represent heterogeneous networks of articles, scientists, keywords, institutions…

Scientometrics, however, remained confined to scientific literature and was unfit to analyze other, less formalized, types of discourse. In the effort to extend the limits of scientometrics, ANT scholars conceived a variety of methods and developed several pieces of software (Leximappe, Réseau-Lu, Prospero…). Until recently, however, all the attempts to devise an integrated methodology for actor-network text analysis were frustrated by the scarcity of text to be analyzed. A part from scientific literature and media discourses it was difficult to find large amount of digitized text to investigate. In the last few years, this bottleneck has been spectacularly removed by the advent of electronic media and of digital traceability. The deluge of digitized texts made available online by all sort of actors (institutions, individuals, associations, media, activists, scientists…) calls for new tools of analysis at the same time more user-friendly and more powerful. ANTA or Actor-Network Analyzer is one of such tools. It has been developed at Sciences Po médialab to offer social researchers a simple text-analysis toolkit attuned with the theoretical tenets of actor-network theory. Striving to make actor-network theory compatible with modern text-analysis, we learned much about both. In this workshop we’ll explore the basic functions of ANTA.