Research

Broadly speaking, my research is concerned with the social meaning of linguistic variation. Recent approaches to variation have demonstrated that variation patterns do not merely correlate with membership in predefined social categories; they encode the meanings from which social categories are derived. Rather than focusing solely on demographic categories such as female or working-class, or even locally salient social identities like nerds or jocks, the social meaning approach additionally considers stances and characteristics, like ‘tough’ and ‘prissy’ which may be more relevant in their interactional contexts than identity categories per se. This strand of research began with William Labov’s groundbreaking work on Martha’s Vineyard in the early 1960s, but has only recently witnessed a resurgence of interest among variationists. In an effort to better understand the social and linguistic dimensions of the social meaning approach, my work seeks both to further develop theories of indexicality and to explore different methodologies for tapping into the social meaning of variation. To this end, I have drawn on a range of data and data collection methods, each offering different insights into what linguistic variants mean. Read on for brief descriptions of some of my current projects.

One of my primary areas of interest lies in the social significance of variation in voice quality. I have examined the meaning of non-modal phonation in a number of speaking populations, including gay professionals across a range of speaking situations, a Japanese pop star over the course of her career, prominent U.S. politicians in the context of political speeches, and most recently, in the speech of Washington, D.C. residents. Much of this work examines and unpacks the indexical associations between non-modal phonation types and gender/sexuality. In spite of strong conventional associations between particular voice qualities and gender, social meanings are arbitrary and can be subverted. My most recent work in Washington, D.C. (conducted with the assistance of Sinae Lee, Georgetown University) shows that creaky voice, which due to its characteristic low fundamental frequency has been iconically associated with masculinity in the literature, predominates in the speech of women; falsetto predominates in the speech of African American women in our data, but is relatively rare in the speech of white women. Our analysis argues that voice quality indexes gender, but that crucially it does so in non-iconic and culturally-specific ways.

California Vowels and Gay Identity

A large strand of my research deals with the role of language in the construction of sexuality. I have emphasized that sexuality is not directly indexed by the use of linguistic features, but rather emerges from the stances enacted through the use of linguistic features. While in previous work, I have examined the potential of consonantal and prosodic variables and voice quality to index sexuality, this work turns its attention to vowels (specifically, the California Vowel Shift, or CVS). Vowels present an interesting case, due to their strong associations with regional identity and accent. An analysis of one gay man’s vowels across three speaking situations reveals significantly more fronted variants of BOOT and BOAT, more raised variants of BAN, and more backed variants of BAT when speaking with friends than in other situations. The speakers’s use of these advanced variants of the CVS correlates with non-heteronormative prosodic patterns, such as frequent and acoustically radical falsetto and extreme fundamental frequency contours on falling declaratives. Based on these patterns and an analysis of the discursive contexts in which they are produced, I suggest that the speaker under analysis is constructing a ‘partier’ persona. It is argued that the social meanings associated with the CVS (e.g. ‘laid back,’ ‘fun’) derive from stereotypes about California character types (e.g. ‘surfer,’ ‘valley girl’) that led to the enregisterment of the variety. These meanings can be recruited in constructing particular kinds of gay identity. The analysis crucially leaves room for regional accent features to index meanings that may be divorced of their geographical roots.

Language and Communication in Washington, D.C.

I am co-directing (with Natalie Schilling, Georgetown University) a large-scale project on language and communication in the Washington, D.C. metropolitan area. Through this project, we investigate how D.C. residents use variation and discourse to showcase and shape identities in their neighborhoods, communities, and the wider city and suburban area. Our primary data source is a corpus of over one hundred sociolinguistic interviews conducted by linguistics students and faculty since 2006. In collaboration with students in my sociolinguistic variation and sociophonetics classes, I have begun examining a number of phonological variables, including (-t/d) deletion, (ing), and the pin-pen merger. I have been particularly interested in how patterns of variation relate to locally significant discourses of race and place (e.g. gentrification, the urban-suburban divide). Projects conducted by other Georgetown sociolinguistics have examined the construction of place in narrative and a number of other sociolinguistic variables, including (ay) monophthongization, post-vocalic r-lessness, and variable (of) in exceptional degree modification (e.g. “It’s not that big (of) a deal”).

Variation in the Speech of U.S. Politicians

Over the six five years, I have been interested in the use and perception of phonological variation in political discourse. In my collaborations with scholars at Georgetown University (Jermay Jamsu, Patrick Callier, Jessica Heitman) and Stanford University (Lauren Hall-Lew, Jason Brenier, Rebecca Starr, Stacy Lewis), I have approached the intersection of variation and politics from a number of angles, using the methods of quantitative variation analysis, discourse analysis, and speech perception. Included among the topics investigated thus far are how released variants of word-final /t/ are interpreted depending on who utters them and how social meanings change depending on phonological context; how particular variants of phonological and phonetic variables are distributed over discourse to make texts cohere in the genre of political speeches; and the varied ways comedians exploit phonological features in parodies of prominent politicians.

Phonetic Detail in Sociolinguistic Variation

In my dissertation, Phonetic Detail in Sociolinguistic Variation, I argued that if sociolinguists intend to explore the social meaning of phonological variables, they must consider their phonetic properties. Outside of vowels, variationists typically focus solely on variation across phonological categories, often failing to recognize the significance of the considerable phonetic variation within categories. By examining phonetic and phonological variation at three levels of phonological analysis – the segment (variation in the realization of word-final coronal stops), prosody (variation in declarative contour shape and phonetic implementation), and voice quality (variation between falsetto, modal, and creaky phonation) – I showed that phonetic details at all three levels encode social meanings that are inaccessible through a solely categorical analysis. For example, two of the speakers studied used falling (H*L%) declaratives equally frequently across speaking situations, but the phonetic properties (e.g., fundamental frequency range and slope) characterizing falling declaratives differed significantly across contexts, with more acoustically extreme falls used in some social situations in which speakers indexed an ‘animated’ meaning. The methodological approach I took in my dissertation, in which I conducted long-term participant observation of three gay professionals, enabled me to locate social meaning in social practice, cross-situational variation, and interactional/discourse context. This micro-level ethnographic approach is one fruitful methodology for studying the social meaning of variation.

The Indeterminacy of Social Meaning

I have been working with Elaine Chun (University of South Carolina) to develop a theoretical framework for conceptualizing the widespread indeterminacy characterizing the social meanings of variables. We have suggested that the ambiguity and underspecification of sociolinguistic meaning may be partly resolved along three contextual dimensions. First, forms do not occur in isolation, but within the linguistic context of other variables, each contributing its own meaning to a style. For example, a medical school student may make use of frequent stop releases to index a ‘competent’ meaning while at the same time using rising declaratives to signal a ‘non-threatening’ stance toward his patient. The meaning of one feature is vivified by the meaning of the other, and together they compose a caring-doctor persona. Second, sociocultural contexts inspire and constrain linguistic interpretation. For example, while stop releases uttered by a gay medical school student can index a ‘competent’ or ‘prissy’ meaning, the situation of a medical examination would foreground the ‘competent’ meaning and render the ‘prissy’ meaning highly unlikely. Finally, speakers negotiate meanings in interaction. In conversations among Korean American youths, for example, rising intonation can be interpreted as constituting a preppy ‘white girl’ style or one mocking the same style. The ambiguity is resolved after the rising intonation is assigned a new meaning, and crucially, such assignments of meaning are achieved collaboratively. Finally, although we argue that indeterminate social meanings can be partly disambiguated, we suggest also that speakers can lively peacefully or even exploit indeterminacy to index multiple social meanings simultaneously. For example, the Asian American comedian, Margaret Cho’s Mock Asian style is both racist and subversive, and both meanings must be noted in order to comprehend her social commentary. Indeed, indeterminacy is a crucial component of the art and humor of language.

Most recently, we have turned our attention to the multiplicity of meanings that linguistic forms can index. Specifically, we have examined the social meaning of non-modal phonation in constructed dialogue, where non-modal phonation can simultaneously signal a shift in footing (i.e. ‘These are not my words’) or construct the current speaker’s or constructed speaker’s identity (taken to encompass interactionally grounded stances, more enduring qualities, and personas). We have drawn on quantitative and qualitative methodologies to make two main points regarding the indexical potential of non-modal phonation. First, correlations between a linguistic feature and multiple social objects (like stances, qualities, and social types) make these objects available as potential indexical values, giving shape to the indexical field. Second, interactional negotiations between participants give further shape to an indexical field by either (a) ironing out meanings, rendering some irrelevant and narrowing the indexical field; or (b) maintaining indeterminacy or even manufacturing new indexical meanings, thereby broadening or shifting the focus in an indexical field.

Style, Indexicality, and the Social Meaning of Tag Questions

Emma Moore (University of Sheffield) and I have worked to elucidate the connection between stance-based social meanings tied to interactional contexts and much more general, enduring social identities. Moore and I have combined our specializations in morphosyntactic and phonological variation, respectively, to examine the use of tag questions in the speech of adolescent girls in a working class community in northwestern England. While most analyses of tag questions take a purely discourse analytic approach, we additionally examined tag questions from a quantitative variationist perspective. Coding for a number of discourse, morphosyntactic, and phonological factors, we found that although all four groups of girls studied – Townies, Populars, Geeks, and Eden Village girls – used tag questions with a similar ‘conducive’ pragmatic function, the linguistic composition of the tags differs greatly across groups. For example, the prototypical Popular tag question was characterized by a feminine 3rd person subject, typically referring to someone located outside of the Popular group, and the interlocutor usually explicitly agrees with the proposition expressed in the declarative preceding the tag. In other words, Popular tags are about other girls – more specifically, they describe, evaluate, and position Populars individually and collectively relative to other people in their school. Popular tag questions are used to build walls, negotiate borders, and regulate and monitor norms. In sum, the many linguistic components of tag questions index a variety of social meanings which combine in unique ways to construct identities through stance accretion. In future work, we plan to examine intonational influences on the social meaning of tag questions.

Non-Pronominal Self-Reference in Japanese

Sakiko Kajino (Georgetown University) and I have investigated the social meaning of non-pronominal self-reference (NPSR, or the use of one’s first name to refer to oneself) in Japanese. To explore the feature’s range of social meanings, we examined patterns of NPSR use in child language, Japanese anime, women’s fashion magazines, a Japanese pop star’s speech over an eight-year span, and in the speech of three women across a range of social and professional situations. We found that NPSR originates in child language and as a result carries connotations of youthfulness, and the association between NPSR and youthfulness is evident in and reproduced by the media. For example, immature anime characters categorically use NPSR, while their mature counterparts use only traditionally feminine first-person pronouns. Similarly, a Japanese pop star’s use of the form declines over time as she presents a more mature and serious self. At the same time, NPSR is often interpreted as heterosexually feminine. One of the participants in the study, for example, discusses having received unwanted romantic attention from a man who intimates that her NPSR use ‘suits [her] well.’ The meaning of NPSR is thus currently under contestation; NPSR users traverse a sociolinguistic terrain composed of youthfulness, femininity, and their interrelations. While most studies on the linguistic practices of Japanese women locate femininity in relation to masculinity, this study reveals the importance of an orthogonal dimension of identity – youthfulness, and in particular its fetishization – in constructing femininity.