Concept Sustainability for Social Media and Politics Research: Reconstructing the Concept of Affordances through Science Mapping

The study of social media and democracy is rife with theoretical puzzles and methodological challenges. The resulting abundance of research questions and analytical designs have resulted in political scholars developing new concepts, refashioning old ones, and organizing concept typologies to guide empirical analyses. As the core building blocks of theory, clear and operationalizable concepts are fundamental to bridging knowledge across methodologies. This paper argues that without taking stock of how political scholars communicate concepts to each other, the legacy of our social media epoch may defined as an era where scholars produced empirical results that responded to the political transformations of the day, but ultimately fell short in developing concepts that provided longstanding theoretical or analytical value for future scholarly generations. Therefore, this study develops the notion of concept sustainability, which provides a diagnostic framework for assessing the current state of concept formation in social media research. To illustrate my arguments, I perform a concept reconstruction (Sartori, 1984) and computational science mapping (Chen, 2017) of the concept ‘affordances.’ Concept reconstruction is the structured extraction and ordering of characteristics derived from the definition of a particular word (here, ‘affordances’). It systematically identifies, organizes, and graphically represents the key properties of a concept as actually used in the scientific literature. Here, I am particularly interested in the practice of adding adjectives to affordances (e.g., perceived affordances or vernacular affordances). In hierarchical concept models (Sartori 1970; 1984), adding adjectives to a concept denotes a more specific sub-concept by adding definitional criteria. For example, bureaucratic autocracy carries a specific set of definitional criteria that makes it a sub-category of autocracy. Others argue that adding adjectives can reflect a radial concept structure (Collier & Mahon, 1993), where two sub-categories share common attributes with the central category but not with each other. For example, participatory democracy and deliberative democracy signify different properties of democracy, but they are not necessarily hierarchically ordered. My starting corpus is 998 published articles that mention “affordances,” “social media,” and “politics” in the title or abstract, suggesting these concepts are central to the paper. Using the doi’s of these papers, I then identify papers that cite them. This method, known as bibliometric coupling, allows for the uncovering of thematic clusters in research while presenting ‘the intellectual structure of a field’ (Donthu et al., 2021, p. 288). To focus explicitly on adjectives, I perform part-of-speech tagging on the full texts of the collected articles. This allows for the automated classification of ‘affordances-with-adjectives.’ From there, I manually annotate the definitional properties of each concept category and build a network map visualizing their relationships (to measure overlapping versus bounded properties). This provides a holistic mapping and conceptual analysis of the affordances concept and its sub-categories, specific to political research. This paper therefore provides a large-scale review of a key concept in social media and politics research. By assessing how affordances is defined and operationalized within political research, I aim to diagnose the utility and sustainability of the concept within and across various methodological approaches.