Voicing Politics and the Debility of Political Science

Voicing Politics claims that political attitudes vary with the language that people speak. Experiments with random assignment of bilinguals to answer survey questionnaires presented in either Estonian or Russian show that their answers do change when the language of the questionnaire is switched. First, relative to gendered Russian, genderless Estonian elicits more support for women’s rights. Second, relative to futured Russian, futureless Estonian elicits more willingness to invest now for future gain. Third, relative to dominant Estonian, minority Russian elicits more awareness of the “most nationalist” party in Estonian politics. But the first claim is inconsistent with the authors’ own evidence. The second claim is inconsistent with both Estonian and Russians’ joint use of linguistic aspect to express futurity; neither is “futureless.” The third claim is invalidated by Russian’s lack of any translation for English “nationalist” that does not signal “anti-Russian,” to which Russian speakers will be more sensitive regardless of dominant or minority status. The experimental asymmetries reported by the authors are attributable, not to differences between the Estonian and Russian languages, but to unspoken context, on which the use of any language for communication must rely but which varies from one language to the next and for which survey research cannot control. Since no practical survey questionnaire can control what context respondents choose to activate in interpreting a question and deciding how to respond, rather than identifying “beliefs” or “attitudes,” any survey research reveals unspecifiable variation in context. Since much of what purports to be known about politics has been inferred from inherently unreliable surveys, this implication of the errors in Voicing Politics is debilitating enough. But since the institutions that political science attempts to explain are uniformly consequences of language use, the uncritical endorsement of Voicing Politics by the discipline’s most prestigious American academic press, its board, editor and referees, prominent endorsers and multiple reviewers is evidence that the debility revealed by this study’s errors afflicts a broad sweep of the discipline extending far beyond survey research alone. Keywords: Survey Research; Experiment; Cross-national; Linguistics; Eastern Europe; Discourse