Voicing Politics, by Efrén Pérez and Margit Tavits, offers readers something remarkable in two respects—one intended and one presumably unintended. The authors bring their respective bodies of expertise in political psychology (Pérez), comparative politics (Tavits), and the politics of racial, ethnic, and gender identities (both) to shed light on the narrow but profound matter of how the language one speaks may shape or constrain one’s worldview and one’s opinions on politics and policy. It is, on the whole, an impressive work that pushes the authors’ research agendas forward substantially. Although Voicing Politics deserves to be read by those with an interest in its core topics—language and politics, comparative public opinion, and identity politics—it also merits a far wider audience as an exemplar of rigorous, yet accessible, social-scientific writing. Pérez and Tavits offer a master class in the iterative process of multimethod research and how to write a book that weaves together a set of individual studies into a coherent narrative. Dissertation advisers take note: this book makes an excellent gift for graduating doctoral students about to be lectured by publishers on the difference between a dissertation and a book manuscript!
Pérez and Tavits grapple deftly with a topic that could be viewed as having niche appeal but that deserves appreciation by a broad audience. Their introduction and first chapter assume no special expertise beyond perhaps a social or behavioral science orientation. Recognizing that most readers will have little background in linguistics or cognitive psychology, they motivate interest through an exciting general-interest science writing style while also managing to catch us up to speed as scholars. Nonlinguists—even those who speak multiple languages—may not appreciate just how widely languages vary in grammatical structures and prominent features such as gendering and time/tense. As an early example, the authors note the tendency of English speakers to use a transitive construction for accidental actions (“Jeremy spilled the coffee”), whereas Spanish speakers more often use a passive sentence construction, even omitting the agent (“The coffee got spilled”) One can imagine the potential implications for politics, where the interplay of individual agency, relative power, and institutional constraints forms the basis for much of what we understand or wish to understand.
The theoretical foundations of the book, laid out in the first chapters, derive from the marriage of the linguistic and political behavior literatures, particularly Dan Slobin’s “thinking for speaking” (“Thinking for Speaking,” Proceedings of the Thirteenth Annual Meeting of the Berkeley Linguistics Society, 1987)—whereby one’s language shapes the mental associations on which one relies not simply to describe but also to attend to and reason about one’s environment—and John Zaller’s “belief sampling” (The Nature and Origins of Mass Opinion, 1992), which holds that temporarily salient “considerations” that happen to be top of mind for low-information members of the public guide their responses to public opinion surveys. The resulting language-opinion hypothesis at the heart of the book articulates a natural, yet profound, implication of these two theories: the language in which one communicates should condition the relative accessibility of certain considerations over others and manifest in detectable cross-language differences in expressed political opinions. This hypothesis drives the book’s investigations fruitfully, with empirical results of earlier studies raising new questions subsequently examined through further research.
Although the implications of the language-opinion hypothesis are limitless, some are especially consequential for political psychology, behavior, and the institutional arrangements that emerge over time from these micro-foundations. The second and third chapters of Voicing Politics focus on variability in the genderedness of languages. Does speaking a genderless language prime one to place less importance on human gender differences than speaking a more rigidly gendered language? To evaluate this hypothesis, the authors begin with two experiments in Estonia, in which bilingual speakers of Russian and Estonian are randomly assigned to be surveyed in one of these two languages. The first experiment is conducted on 1,200 participants, with a follow-up replication and extension conducted on 262 respondents. They find small but statistically distinguishable differences in expressed support for certain gender stereotypes, with those speaking Russian (gendered) slightly more likely than those speaking Estonian (genderless) to lend credibility, for example, to the notion that women are more emotional and men more rational. The authors view this as a likely mechanism through which language may affect respondent judgments about policies promoting gender equality and the suitability of women to hold public positions traditionally reserved for men. Their follow-up study involved just 262 respondents, but the results mostly hold up (detectable nonzero effects).
The internal validity of the first two gender studies is complemented by a successful attempt to establish external validity through a large observational study based on a trusted worldwide survey of 170,000 respondents across 90 countries. Next, the authors refine their investigation, inquiring whether the discovered language effects are not simply grammatical but also lexical; that is, do the words themselves affect opinions? This study is more subtle but has greater policy implications than the first because it relates to practical linguistic changes that we may choose to embrace. Although it focuses on the introduction of gender-neutral pronouns in Swede, it will resonate with English speakers as well: When we consciously use gender-neutral pronouns such as the individual “they” or professional labels such as “representative” instead of congressman and congresswoman, might this actually shift perception and public opinion on matters of gender equity and LGBTQ rights, or is it simply a symbolic gesture to please the like-minded? To gather evidence, the authors again proceed through a sequence of studies, including a small pilot study, a larger replication study, and an online survey to rule out social desirability effects. Over and over throughout the book, Pérez and Tavits demonstrate an incredible commitment to subjecting results to various checks, poking and prodding their findings for possible holes. Their perseverance should stand as a model for our students and to all of us who may be too easily satisfied by one suggestive study.
In later chapters, the human stakes remain high. The authors examine the possible impact of speaking futured versus futureless languages, asking whether the degree of separation between now and later might have an impact on concern about long-term threats such as climate change. Their broad language-opinion hypothesis leads the authors to expect that those forced to conceptually separate today from tomorrow in communicating through a futured language will perceive future threats as more temporally distant, whereas those speaking a futureless language should be less likely to experience a salient distinction between present and future, with crucial implications for their opinions on proposed policies. Two experimental studies and a large-scale survey build more support for their claims. Next, they take on ethnic divisions and the impact of speaking a majority versus minority tongue within a country, finding evidence that speaking marginalized languages may sensitize one to the distinct experiences of ethnic minorities and to the divisions that disadvantage them. Pérez and Tavits push this line of inquiry further in their final substantive chapter, asking whether majority-language speakers enjoy a “language premium” when expressing political opinions. Carrying out experiments on bilingual Spanish and English speakers in the United States, they examine whether such speakers can more quickly retrieve public affairs knowledge from memory when speaking English, the majority tongue, because they are more likely to have consumed such information in English. The experimental manipulation no longer works through the mechanism of grammar or lexicon but rather through the relative social status of the language as a whole. More specifically, bilingual speakers in the United States build more mental associations regarding public policy and politics through English-language sources. If the linguistic tags that help us retrieve ideas from memory are encoded in English, it should be easier to access and articulate them when speaking English. In two experiments and a large-scale survey, they find evidence supporting a modest but meaningful language premium.
The book concludes with a review of the substantial progress the authors have accomplished on their research problem and the implications for real-world politics and policy. They point the way forward for other intrepid souls who wish to delve into related questions, offering a generous appendix with many details related to research design and statistical analysis.
