Measuring Populism from Political Text: (Thin) Ideology, Rhetoric, Discourse Quality, Civility and Linguistic Style

Abstract

The panel focuses on the measurement of the elusive concept of populism across sources. A variety of conceptual perspectives is encouraged. Moving from substance to structure, examples include definitions of populism as a (thin) ideology or a set of rhetorical elements stressing a perceived antagonism between the people and the elites, populism as indicated by low levels of discourse quality or civility or as reflected in a linguistic style marked by low lexical variety and simple syntax. Potential text sources include party manifestos, speeches as well as social and conventional media data. Typically, this requires some added efforts such as the creation of a dictionary, as for instance populist statements need to be identified and filtered from the bulk of party manifesto or Twitter texts. Methodologically, the panel is open to any approach from quantitative text analysis to qualitative assessments. Papers are invited which for instance: - Measure populist rhetoric, discourse quality or linguistic structure from a variety of sources. - Investigate which modeling approach fits which data source best. Is it for instance more appropriate to use classifiers on Twitter data rather than scaling models? - Show ways of dealing with the extreme scales of social media data and the resulting challenges (“big data”). - Explore options to bridge language barriers in the measurement of populism. - Apply measures of populism to substantial research questions, for example related to the variation of populism across parties or its consequences.