Lost in translation: Collecting and coding data on social relations from audio-visual recordings

Abstract

Some of the constitutive features of social relations fade from view when information naturally produced by sequential social interaction is translated into network ties. Building on core concepts and ideas developed within conversation analysis, in this paper we argue that this happens because the sequential, multimodal and embodied character of social relations can be fully understood only with reference to the sequential constrains that are generated by – and at the same time shape the micro-dynamics of social interaction. We suggest that the translation of social interaction into social networks precludes analysis of the multiple interfaces that sustain social relations (multimodality), and the material resources around which social relations are organized (embodiment). We highlight audio-visual recording as a data collection technology that facilitates storage, retrieval, and analysis of complex information on social relations that is typically absent from social network data. An illustrative video-supported case study based on the observation of social and task-related interaction among members of surgical teams provides the empirical context that supports and motivates our general reflection on network data collection strategies and technologies to study social interaction. The analysis highlights the need for social networks research to return to the study of social relations.