Presuppositionless proxies of politeness: An (eventually) Optimality-Theoretic account

Abstract

Recent research has seen growing interest in the nature of social meaning, “the constellation of qualities and properties that linguistic forms convey about the social identity of language users” (Beltrama 2020, see also Eckert 2008). This dissertation concerns those linguistic forms termed polite pronouns. They convey one particular aspect of social identity: the social distance between language users, where a large social distance mandates the expression of politeness. Most existing work on polite pronouns has modeled polite meaning via dedicated grammatical mechanisms, such as a dedicated feature ([hon], Ackema & Neeleman 2016) or a dedicated dimension of meaning (McCready 2019). This dissertation shows that such dedicated mechanisms are superfluous, as existing grammatical mechanisms can be appropriated to describe and explain both the form and meaning of polite pronouns. Building on Wang (2023), I use a purpose-built typological sample of polite pronouns in >220 genetically and geographically diverse languages to establish the shapes of polite pronouns, showing that polite pronouns are obtained by asymmetrical recruitment of existing ϕ-featural values. Furthermore, the shapes of polite pronouns converge precisely with the shapes of semantic defaults, those ϕ-featural values underspecified in meaning, which must emerge when the context provides little or no information about number or person. To explain this convergence, I use the notion of negative politeness: respecting an interlocutor’s right to be unimpeded (Brown & Levinson 1987). I argue that semantic defaults and polite pronouns are morphologically identical because they are pragmatically identical: underspecification makes them well-suited to be avoidance mechanisms in the service of negative politeness. This captures the intuition that avoidance behaviors are a core component of expressing politeness. Specifically, polite pronouns enable speakers to avoid making presumptions about aspects of the interlocutor. Hence, presuppositionlessness, proxies, and politeness are what this dissertation is all about. I implement this intuition in Optimality-Theoretic terms, where polite grammars are defined by an outranking of Markedness (which mandates speakers to avoid presupposition-rich forms) over Faithfulness (which mandates speakers to be as informative as possible). The resulting system is restrictive but powerful, delivering a factorial typology which exactly mirrors the recruitment asymmetries. With no stipulations specific to polite pronouns, the account is furthermore able to make concrete predictions about politeness phenomena beyond the ϕ-featural domain. Happily, the resulting account is as lean as can be. Its main innovations are methodological and theoretical. Methodologically, the investigation is led by a large cross-linguistic sample, and the additional consideration of use conditions alongside morphological exponence. Theoretically, recruitment affords us with a fresh look at how the universal inventory of features is organized. This dissertation shows that recruitment is not only viable as a meta-grammatical operation, but even desirable for reasons of economy or computational efficiency.