Panel Studies of Variation and Change

Suzanne Evans Wagner, Isabelle Buchstaller, Suzanne Evans Wagner, Isabelle Buchstaller

• Publisher: Routledge
• Year: 2017
• eBook ISBN: 9781317446392
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

The relationship between the individual and the community is at the core of sociolinguistic theorizing. To date, most longitudinal research has been conducted on the basis of trend studies, such as replications of cross-sectional studies, or comparisons between present-day cross-sectional data and 'legacy' data. While the past few years have seen an increasing interest in panel research, much of this work has been published in a variety of formats and languages and is thus not easily accessible. This edited volume brings together the major researchers in the field of panel research, highlighting connections and convergences across and between chapters, methods and findings with the aim of initiating a dialogue about best practices and ways forward in sociolinguistic panel studies. By providing, for the first time, a platform for key research on panel data in one coherent edition, this volume aims to shape the agenda in this increasingly vibrant field of research.

1
Introduction

Using Panel Data in theSociolinguistic Study of Variation and Change

Isabelle Buchstaller and Suzanne Evans Wagner

1. Panel Studies

In this volume, we aim to present the unique benefits and challenges that arise out of longitudinal sociolinguistic studies of individuals, with the explicit goal of informing future studies. Most models of linguistic behavior have relied on the assumption that the language of individual speakers remains stable from late adolescence or early adulthood onwards (the critical period hypothesis, Lenneberg 1967) and thus assume that many fundamental aspects of individuals’ mental grammar are locked into place. In variationist sociolinguistics, the critical period is crucial to the “use of the present to explain the past” heuristic (Labov 1975: 226): If it is true that speakers retain their early adult grammars, then age-related variation in the community can be interpreted as a diagnostic for diachronic change (Wagner 2012a, b; Prichard and Tamminga 2012; MacKenzie forthcoming). Yet from the earliest days of the field, it has been understood that this assumption would need to be refined (Labov 2006[1966]: 200–202). Speakers self-evidently do change their linguistic habits as they age. The questions for sociolinguists have been: Which aspects of the sociolinguistic repertoire can change in post-adolescence? If an identified feature happens to be involved in a community-wide change in progress, what are the implications for the accuracy of our models of language change? More specifically,

• If individual speakers participate in ongoing changes, models of language change that assume speaker-internal stability underestimate the rate of community-wide change.

(Sankoff and Blondeau 2007)

• If speakers withdraw from ongoing changes, models of language change that rely on a strict interpretation of the apparent time hypothesis overestimate the rate of change.

(Buchstaller 2015)

The relationship between the individual and the community is thus at the very core of sociolinguistic theorizing. Until quite recently, however, longitudinal research has been conducted almost exclusively on the basis of trend studies, such as replications of cross-sectional sociolinguistic studies (e.g., Trudgill 1988; Pope, Meyerhoff, and Ladd 2007) or comparisons between present-day cross-sectional data and ‘legacy’ data (e.g., Maclagan and Gordon 2004; Buchstaller 2014). Consequently, the degree to which the grammar of individual speakers changes across the lifespan could not—or could only very indirectly—be investigated. This volume advances inquiry into individual linguistic lability by highlighting a complementary area of longitudinal research: Panel studies of language variation and change.

Panel studies, in contrast to trend studies, rely on a longitudinal intra-speaker design. They collect repeated recordings from the same speaker(s) as they age, providing us with the most direct evidence for peoples’ linguistic habits across their lifespan. ‘Panel studies’ is the term that has been most frequently employed in variationist sociolinguistics to date, and we use it as an umbrella term for any study in which individual speakers are recorded at multiple time points. Elsewhere in this volume, however, authors have elected to use the term ‘panel surveys,’ in line with broader social science norms, or to distinguish specific types of panel research, such as cohort studies (wherein speakers from a specific age cohort are tracked longitudinally) and case studies (wherein a single speaker is tracked).

Although panel studies have been part of the sociolinguistic landscape for decades (e.g., Prince 1987; Rickford and McNair-Knox 1994), there has been increasing interest in tracking speakers across the lifespan, as evidenced by the growing number of presentations based on panel data at international conferences on variation and change such as New Ways of Analyzing Variation (NWAV), UK Language Variation and Change, the International Conference on Language Variation in Europe (ICLaVE) and the Sociolinguistics Symposium. The greater availability of older digital audio and video recordings from academic repositories, public media and other sources has provided opportunities for even undergraduate students to embark on panel study projects. There has been a particular upswell of panel studies since the publication of Sankoff and Blondeau’s much-cited 2007 paper in Language, which found that although the majority of speakers persisted in using either the older trilled apical [r] or innovative posterior [R] in Montreal French, a non-trivial number of individuals increased their use of the innovative form as they aged. By focusing squarely on a change in progress in the community, Sankoff and Blondeau underscored that it is not enough to determine if and how individuals are sociolinguistically malleable in post-adolescence. Rather, it is the interaction between individual malleability and community-wide language change that deserves the field’s attention.

Indeed, to date, panel studies have amassed an increasing body of evidence that changes in linguistic habits occur in two broad patterns of instability across the adult lifespan during community-wide language change:

• 1. lifespan change, when individuals adapt their language use towards the direction of the community-wide trend;

(Sankoff and Blondeau 2007;Harrington and Stevens 2014 inter alia)

• 2. retrograde movement against the community-wide trend, when post-adolescent speakers move away from ongoing change.

(Wagner and Sankoff 2011; Buchstaller 2016)

What is less well understood are the factors conditioning intra-speaker (in)stability across the lifespan. While it is relatively uncontroversial that cognitive maturation alone cannot fully explain age-related effects in language use (Birdsong 1999; Loewen and Reinders 2011), the sheer number of determinants, including cultural, historical and individual factors, on shaping a person’s dispositions, including their speech patterns, makes the creation of a comprehensive model of intra-speaker instability daunting to say the least (McCrae and Costa 2008). What is needed, therefore, are large-scale studies that examine language use across the entirety of maturational watersheds of human life, including the effect of full-time employment, professional promotions, career changes, mentoring colleagues and all the “life experiences that give age meaning” (Eckert 1997: 167). Such a coherent empirical and theoretical knowledge base will allow us to examine the impact of lifespan specific effects and maturation on the linguistic malleability of the post-adolescent speaker (Gerstenberg and Voeste 2015). Another large unknown is the longitudinal linguistic effect of parenting (Arnett 1997, 2014), both regarding the trajectory of the parents as well as in terms of providing the input into children’s language acquisition and thus the future progression of the change.

Finally, as panel research continues to advance our understanding of the ways in which individual speakers’ internal grammars can and do change across their lives, as well as the conditions which drive intra-speaker malleability, we hope that findings such as the ones represented in this collection can contribute to bridging the schism “between the many claims about linguistic” change (Hruschka et al. 2009: 466). Linguists have long argued that individual speech patterns are more durable dispositions than other habitual forms of human behavior, such as voting patterns and lifestyle choices (Labov 1972). Indeed, formal schools of linguistics have tended to dismiss the notion that changes in internal grammar can occur post-critical age (Sobin 1997; Lightfoot 1999; Anderson 2016) and relegate the locus of language change uniquely to child language acquisition. Observed variability in the adult speaker is thus interpreted as situation-specific performance effects with no repercussions on the stability of speakers’ internal grammar (Chomsky 1964; Kiparski 1995). Consequently, the two schools of linguistics sometimes give very different theoretical weight to recent findings from panel research.

Only a comprehensive view of the “determinants and the limits of linguistic malleability” (Bowie and Yaeger-Dror 2016: 608) will allow us to formulate a coherent theoretical model that captures the relationship between speaker-internal grammar and the community grammar. This is important because findings from panel research have ramifications beyond the language sciences. Age-related changes in adult grammars can provide key insights into the development of cognitive performance since language as a human form of communication constitutes an abstract, highly systematic cognitive structure (Hruschka et al. 2009). Therefore, sociolinguistic research on the malleability of adult speakers’ linguistic habits can generate essential new information on behavioral dispositions across the lifespan of the individual that are less susceptible to experimenter effects, conscious suppression and short-term changes than other areas of human behavior. Consequently, the analysis of linguistic malleability across the lifespan can give important insights into one of the most disputed issues across the developmental sciences: The rate and the limits of neuroplasticity (Chang 2014; Barry 2010).

Until now, there has been no single publishing venue for sociolinguistic panel studies. By bringing together key research in one volume, we hope to open up a more holistic perspective on the development of linguistic habits across the lifespan of the individual and thus to accelerate our understanding of how language change in the individual intersects with language change in the community. The contributions featured in this collection are grouped loosely into four subsections, outlined in 1.1 to 1.4 below.

1.1 Methodological Conundrums in Building, Sharing and Analyzing Panel Corpora

Whereas ‘big data’ is increasingly common in the humanities and the social sciences, relatively few large-scale follow-up studies have been conducted in sociolinguistics. To date, research on change across the lifespan tends to be conducted on the basis of almost antithetical data-sources: Small samples characterized by detailed interpretation of local social meaning and stylistic performance (Sankoff 2004; Sankoff & Blondeau 2013; Rickford and Price 2013 inter alia), or large samples which consider larger time spans and focus on broad sociodemographic categories (Sankoff and Blondeau 2013; Gregersen, Maegaard, and Pharao 2014). A critical discussion of the methodological issues involved in collecting and analyzing panel data is thus both timely and useful.

The present collection of chapters provides a platform for panel researchers to share their experiences with longitudinal data sets. While many of the challenges are unique to the individual research project, we hope that the connections and convergences across and between chapters, methods and findings will initiate a dialogue about best practices in sociolinguistic panel studies. The first two chapters in our volume in particular discuss concerns related to collecting, analyzing and sharing large-scale panel corpora. Sankoff’s contribution in Chapter 2 discusses the creation of the first ever sociolinguistic panel corpus, which grew out of the Montreal French project (Sankoff and Sankoff 1973) and which has been used for dozens of research projects on intraspeaker and community wide change. Cieri and Yager-Dror explore in Chapter 3 the advantages and disadvantages of existing longitudinal recordings from independent sources, including cross-discipline reuse of recordings as well as raw “found” data. In Chapter 4, Van Hofwegen and Wolfram describe a collaboration with researchers in child development that facilitated their access to—and eventual expansion of—a rich source of longitudinal linguistic data: The Frank Porter Graham corpus.

By discussing the methodological challenges and the benefits involved in constructing and using these panel data sets, we hope these chapters can serve as a guide and an inspiration for researchers in the field of sociolinguistics, as well as in related disciplines, such as linguistic anthropology, second language acquisition, phonetics and historical linguistics with an interest in conducting panel studies.

1.2 Key Life Stage Events Across the Lifespan

The human lifespan offers a dynamic context for research on the mechanisms of language change. At the time of writing, however, sociolinguistic investigation is not yet in the position to establish a comprehensive picture of linguistic malleability across the lifespan as a whole (Buchstaller 2006; Cheshire 2006; Wagner 2012a). This is due, partly, to the fact that our evidence base is severely biased towards the younger age ranges, especially focusing around the adolescent peak (Tagliamonte and D’Arcy 2009). The life stage of “emerging adulthood” (Arnett 1997, 2014) has only recently begun to be investigated from a linguistic perspective, and there is no consensus as to the relevance of this life stage on speakers’ trajectories (Bigham 2012; Rickford and Price 2013; Buchstaller 2015). Very little data exist on the earlier stages (most coming out of research on the Frank Porter Graham corpus mentioned above), and we know even less about the middle and later stages of life in sociolinguistics (Gerstenberg and Voeste 2015; Eckert 1997), but also in developmental research (Levinson 1986; Wahl and Kruse 2005; Terracciano, McCrae, and Costa 2010). The contributions in section two of this volume are devoted to exploring the methodological, conceptual and practical issues involved in conducting panel research with different age groups.

The section contains two chapters in which researchers discuss the issues they encountered when exploring individual age brackets in the early and late stages of the lifespan. Kohn and Farrington discuss in Chapter 5 the important issue of physical changes to vocal tract morphology, which strongly affect the acoustic signal across time points in (early) childhood. In Chapter 6, Reubold and Harrington explore the ways in which phonetic change can be disentangled from the influences of increasing age (especially into old age) when measuring the fundamental frequency and the vocal tract resonances.

While giving linguists interested in panel research access to cutting-edge technologies, the chapters in this section suggest that some of the methods and the conceptual frameworks we employ to analyze our data need to be tailored to the different stages of life. The overall aim of this section is thus to build a “comprehensive, systematic [inventory of the methods and concepts involved in studying language change across the] … life course [so that] the parts will become less isolated and each part will enrich the others” (Levinson 1986: 4).

1.3 Stylistic Determinants of Linguistic Malleability

The creation of speaker identities is fundamentally sensitive to age-specific stages across the lifespan (Rickford and Price 2013). At the same time, panel research needs to be able to capture the fact that speakers’ interactional prese...

Table of contents

• Cover
• Title
• Copyright
• Dedication
• Contents
• List of Figures
• List of Tables
• Acknowledgments
• 1 Introduction: Using Panel Data in the Sociolinguistic Study of Variation and Change
• PART I Methodological Conundrums in Building, Sharing and Analyzing Panel Corpora
• PART II Key Life Stage Events Across the Lifespan
• PART III Stylistic Determinants of Linguistic Malleability
• PART IV Interdisciplinary Approaches
• Contributors
• Index