Introduction

Quantitative sociolinguistics has been part of the language research landscape since the early 1960s, beginning with the work of William Labov in New York and Martha's Vineyard (Labov 1969; 1972a,b). In an early and often-cited study on the raising and centralization of vowels on Martha's Vineyard, Labov (1972b) found that centralization corresponded with certain age groups and with the speaker's orientation towards traditional life on the island. These findings represented one of the earliest uses of quantitative methods for arriving at conclusions about the structure and use of language. The results showed that linguistic variation is not random, that it is quantifiable, and that understanding variation is essential to understanding how language works. Numerous studies in the variationist tradition established by Labov have taught us a great deal about language structure and language change. This chapter focuses on what quantitative sociolinguistics has taught us about variation in sign languages and the implications of that knowledge for sociolinguistic theory and for linguistic theory more generally.

In contrast to the study of variation in spoken languages, the study of variation in sign languages is still in the relatively early stages. The first large-scale study of variation in American Sign Language (ASL) appeared only in 2001 (Lucas et al. 2001b). That study, based on data collected in the mid-1990s in seven areas of the United States, was following by similar studies in Australia and New Zealand (Schembri et al. 2009; McKee et al. 2011), the United Kingdom (Schembri et al. 2013), and Italy (Cardinaletti et al. 2011; Geraci et al. 2011; 2015), as well as by a study of Black ASL, the variety of ASL that developed in the segregated schools of the U.S. South before the civil rights era of the 1960s (McCaskill et al. 2011). Although we have ethnographically oriented studies of smaller signing communities, such as Green's (2014) work on Nepali Sign Language, as well as earlier work involving individuals or small groups of signers (see Patrick and Metzger 1996 for a review), large-scale surveys in several countries have provided the primary insights into the relationship between variation in sign languages and sociolinguistic (and linguistic) theory. We are fully aware of the advantages of recent trends in sociolinguistics that have combined close ethnographic observation with quantitative methods and sometimes focused on marginal members (e.g. Bucholtz 1999; Eckert 2000).

All human languages vary in both time and space as well as according to the linguistic environment in which a particular form is used. For example, the ASL sign DEAF has three main forms. It can be produced with a movement from ear to chin (the citation or dictionary form), from chin to ear or by contacting the check once (both non-citation forms). Even though the form of DEAF varies from signer to signer, and even within the signing of the same signer, the variation we observe is not random. Rather, signers’ choices among the three forms of DEAF are systematically constrained by a range of linguistic and social influences, or factors. For example, compared to signers in other parts of the United States, signers in Boston, Massachusetts use the citation form of DEAF more often. In contrast, signers in Kansas, Missouri, and Virginia tend to prefer non-citation forms. Indeed, a study of variation in the sign DEAF showed that signers in these states used non-citation forms of DEAF 85 percent of the time, more than twice the rate of signers in Boston (Bayley, Lucas, and Rose 2000: 92).

The region of the country where a signer lives is not the only influence on the choice of a form of DEAF. For example, although ASL signers in Boston generally used more citation forms of DEAF than signers in other areas of the United States, Boston signers aged 55 and older are far less likely to choose a non-citation form of DEAF than are younger signers. Bayley et al. (2000) reported that Boston signers aged 55 and older used the citation form of DEAF 76 percent of the time. In contrast, signers between the ages of 26 and 55 used the citation form 54 percent of the time, and signers between the ages of 14 and 26 used the citation form only 46 percent of the time. In addition, variation can be affected by linguistic factors. To continue with the example of DEAF, Lucas (1995) and Bayley et al. (2000) found that signers were very likely to use a non-citation form of DEAF when it was part of a compound, as in DEAF^CULTURE or DEAF^WORLD. However, when DEAF was a predicate adjective, as in PRO.1 DEAF (‘I am deaf’), signers were more likely to choose the citation form.

Introduction

Even casual observation reveals that language users sometimes have different ways of saying or signing the same thing. Variation may be realized at all different levels of a language. English, for example, contains numerous examples of variation in the lexicon. Some American speakers use the word couch, while others say sofa or davenport. American Sign Language (ASL) also exhibits many well-known examples of lexical variation. For example, a number of signs exist for the concepts BIRTHDAY, PICNIC or HALLOWEEN.

At the phonological level, variation exists in the individual segments that make up words or signs or in parts of those segments. For example, speakers of a wide range of English dialects sometimes delete the final consonant of words that end in consonant clusters such as test, round or past, the result being tes', roun' and pas' (Labov et al. 1968, Guy 1980). In ASL, phonological variation can be seen in all of the parts that make up signs – in the handshape, movement, location and palm orientation, and sometimes even in the nonmanual features that are part of sign production. The basic structure of these parts is discussed in other chapters of this volume.

Variation may also occur in the morphological and syntactic components of a language. For example, in African American Vernacular English (AAVE), the copula be is variably deleted, and the sentences He is my brother and He my brother both occur.

Introduction

The late 1960s and early 1970s witnessed the development of two subfields of linguistics: the quantitative study of linguistic variation pioneered by Labov (1966, 1969), Shuy (Shuy, Wolfram, and Riley 1968a), Wolfram (1969), and Fasold (1972), the focus of this volume, and the systematic investigation of second language acquisition (SLA), exemplified by studies such as Cazden, Cancino, Rosansky, and Schumann (1975) and Hakuta (1976). These two areas of research were motivated by a common concern to understand the underlying systems of language varieties, often socially stigmatized varieties in the case of quantitative sociolinguistics, and learner language in the case of SLA. Moreover, since its emergence as a distinct paradigm, variationism has been concerned with confronting the linguistic stereotypes of non-standard varieties by serious scientific study. Sankoff (1988a), for example, dates the development of variationism as a paradigm distinct from dialectology, ethnolinguistics, and traditional pidgin and creole studies from 1969, with the appearance of Labov's first major publication on the African American Vernacular English (AAVE) copula, rather than from 1963, the publication date of his study of Martha's Vineyard, or 1966, the publication date of his earlier work on New York City. The variationist paradigm was very quickly extended to speakers of other socially stigmatized language varieties, including American Sign Language (Woodward 1973), working-class British English (Trudgill 1974), Puerto Rican Spanish (Poplack 1980a), and Guyanese Creole (Rickford 1987), to name just a few.

Just as in variationist sociolinguistics, several early studies of second language acquisition focused on socially marginalized speakers, often working-class immigrants from the developing to the developed world.