Variation and the indexical field1 - Stanford University

Variation and the indexical field1 - Stanford University

Journal of Sociolinguistics 12/4, 2008: 453–476 Variation and the indexical field1 Penelope Eckert Stanford University This paper argues for a focus ...

965KB Sizes 0 Downloads 3 Views

Recommend Documents

Parents and Stanford University - Giving to Stanford
means that even small gifts, which generally have few restrictions, are enormously valuable to Stanford. Giving to Stanf

stanford stanford - Stanford Alumni Association - Stanford University
Elizabeth Strauss Bing, PhD '62, and Ralph A. Bing*. Dorothy M. and William H. Bissell, '49, ..... Elizabeth Chandler Go

DETERMINATION, UNIFORMITY, AND - Stanford University
knowledge that, when added to the premises of the analogy, makes the conclusion follow soundly. It might be noticed that

The Localization and Regulation of - Stanford University
Sean B. Carroll, Robert A. Laymon: ..... The Sex combs reduced (SW) locus of the ANT-C is re- ... feet extra sex combs (

Grammaticalization, constructions and the - Stanford University
When we consider the topic of “language evolution: cognitive and cultural factors”, one of the first questions that

Poverty and the Great Recession - Stanford University
mid-1970s to the eve of the Great Recession in which ... Has the Great Recession contributed to a .... Great Recession's

lassicist - Stanford Classics - Stanford University
Julianne Groschupf ('74; MA. '75) married Paul Clouse 3 / 85. Bill Grueskin ('77) married Rosalind. Resnick 1987. Peter

Sweetness and Power - Stanford University
May 11, 2017 - followed by those of East Indians…Chinese,. Portuguese, and many other peoples. …Sugar brought a doze

Stanford Law Review - Stanford University
Apr 4, 2017 - (4) practitioners have begun implementing side agreements to the deal that have a commercial purpose along

Mood and Memory - Stanford University
recall of word lists, personal experiences recorded in a daily diary, and .... According to several indicators, the affe

Journal of Sociolinguistics 12/4, 2008: 453–476

Variation and the indexical field1 Penelope Eckert Stanford University This paper argues for a focus on the social meaning of variation, based in a study of stylistic practice. It is common in the study of variation to interpret variables as reflections of speakers’ membership in social categories. Others have argued more recently that variables are associated not with the categories themselves, but with stances and characteristics that constitute those categories. The paper reviews some variation studies that show that variables do not have static meanings, but rather general meanings that become more specific in the context of styles. Building on Michael Silverstein’s notion of indexical order, I argue that the meanings of variables are not precise or fixed but rather constitute a field of potential meanings – an indexical field, or constellation of ideologically related meanings, any one of which can be activated in the situated use of the variable. The field is fluid, and each new activation has the potential to change the field by building on ideological connections. Thus variation constitutes an indexical system that embeds ideology in language and that is in turn part and parcel of the construction of ideology.

KEYWORDS: Variation, style, indexical order, enregisterment, social meaning

INTRODUCTION The distinction between sociolinguistics and linguistic anthropology became an issue as the quantitative study of variation gained hegemony in sociolinguistics, subordinating the examination of the social to questions of linguistic theory and to the needs of regression analysis (see also Bucholtz and Hall this issue). By viewing the social as a fixed and external structure that is only reflected in linguistic variability, the study of sociolinguistic variation has remained safely within the bounds of linguistics as a cognitive science. I say this not to deny the interest and importance of the work that has been done in this vein, but to note that ultimately the variation (and the entire linguistic) enterprise must be integrated into a more comprehensive understanding of language as social practice. William Labov’s comment (2002: 283) that ‘[t]he great chain shifts sweeping across North America are more like ocean currents than local games’ suggests that the local indexical work that speakers do with variation is dwarfed in importance by the power of the internal workings of the great linguistic system.  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation  9600 Garsington Road, Oxford OX4 2DQ, UK, and 350 Main Street, Malden MA 02148, USA

454

ECKERT

To seek explanations for chain shifts in the day-to-day construction of meaning would certainly be futile and ridiculous. But to ignore what people do with the elements of these chain shifts to construct social meaning is to turn a blind eye to an aspect of human competence that is at least as mind-blowing as the ability to maintain distance between one’s vowels. As Kathryn Woolard (this issue) emphasizes, linguistic anthropology foregrounds the ongoing construction of meaning in human activity. In the study of sociolinguistic variation, on the other hand, meaning has been the stuff of casual speculation, but not part of the enterprise. In the following pages, I propose an approach to the study of social meaning in variation that builds upon linguistic-anthropological theories of indexicality, and most particularly Michael Silverstein’s (2003) notion of indexical order. I argue that the meanings of variables are not precise or fixed but rather constitute a field of potential meanings – an indexical field, or constellation of ideologically related meanings, any one of which can be activated in the situated use of the variable. The field is fluid, and each new activation has the potential to change the field by building on ideological connections. Thus variation constitutes an indexical system that embeds ideology in language and that is in turn part and parcel of the construction of ideology. This concept leaves us with a new (that is, an additional) enterprise of studying variation as an indexical system, taking meaning as a point of departure rather than the sound changes or structural issues that have generally governed what variables we study and how we study them.

THE FATE OF MEANING IN VARIATION STUDIES The early moments of the quantitative study of variation held promise for the analysis of social meaning. In his study of Martha’s Vineyard, Labov (1963) found correlations of centralized /ay/ with a range of social categories – fishermen, people living at the fishing end of the island, teenagers who planned to spend their adulthoods on the island. He interpreted these correlations as evidence of an association of the old island variant with local authenticity based in the English-descent island-based fishing community and its resistance to mainland incursion. This very local construction of meaning in variation, the recruiting of a vowel as part of a local ideological struggle, suggested that variation can be a resource for the construction of meaning and an integral part of social change. But this power of variation was lost in the large-scale survey studies of sound change in progress in the years that followed, as social meaning came to be confused with the demographic correlations that point to it. Social meaning remained as a subtext in community studies, but with no real place in the theory. Peter Trudgill (1972), for instance, called upon the perceived toughness of working-class men as a motive for middle-class men to adopt local working-class sound changes, accounting for the upward spread of change. But this account was vague about the nature of the connection between toughness, gender, and class, and did not open up an account of how meanings become associated with social categories  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

455

or with variables. Rather, it was absorbed into a view of the meaning of variables as consequences of the abstract demographic categories that structure survey research – socio-economic class, gender, and ethnicity. This view of variables began with the sociolinguistic focus on the spread of sound change. In this view, the socio-economic hierarchy is a social space through which change spreads, and speakers’ place in that space determines when they ‘receive’ the change. Speakers’ agency in the use of variables has been viewed as limited to making claims about their place in social space by either emphasizing or downplaying their category membership through the quantitative manipulation of markers. But clearly, women (and men) are not saying ‘I’m a woman’ when they use a ‘female-led’ change, nor are they saying ‘I’m not a woman’ when they do not. The generalization that women lead in a particular sound change is the outcome of a general statistical result by which, in the aggregate, women use advanced variants more than men (Eckert 1990). This generalization says nothing about the kinds of behaviors and ideologies that underlie these patterns, what kinds of meaning people attach to the conservative and innovative variant, who does and does not fit the pattern and why. It says nothing about language use and gender in everyday life, and it says nothing about why the same generalization applies to class stratification – that is, not only women, but working-class people, lead in sound change. Yet variationists continue to use labels such as ‘female-led changes’ as if such changes all had a direct relation to gender. Quantitative generalizations of the sort made in survey studies are important, but exploring the meaning of variation requires that we examine what lies beneath those generalizations. The very fact that the same variables may stratify regularly with multiple categories – e.g. gender, ethnicity, and class – indicates that their meanings are not directly related to these categories but to something that is related to all of them. In other words, variables index demographic categories not directly but indirectly (Silverstein 1985), through their association with qualities and stances that enter into the construction of categories.

VARIABLES, STYLE, AND SOCIAL MEANING Speaking in the social world involves a continual analysis and interpretation of categories, groups, types, and personae and of the differences in the ways they talk – in social cognition terms, a development of schemata (Piaget 1954). These emerge as we come to notice differences, to make distinctions, and to attribute meaning to them. Thus we construct a social landscape through the segmentation of the social terrain, and we construct a linguistic landscape through a segmentation of the linguistic practices in that terrain. The level of social practice that corresponds to distinctions in the terrain in which we study variation is style. In all areas of art, style is what characterizes schools, periods, and individuals. Style has a similar function in everyday language, picking out locations in the social landscape such as Valley girls, cholos,  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

456

ECKERT

cowboys, jocks, burnouts, Italian hoods. Variables occur only as components of styles, and interpreting variables requires an analysis of these components. The ability that a human being exercises from birth to acquire language no doubt persists through life as we strive to understand the social significance of linguistic form – as we analyze the nuances of the linguistic variability around us. This kind of style (what one might call persona style) is orthogonal to the formality continuum that is associated with style in traditional variation studies (e.g. Labov 1972). The focus on formality in these studies keeps the study of variation in the cognitive realm (see Eckert 2004) as it determines the amount of attention paid to speech, limiting stylistic agency to the manipulation of status in the socio-economic hierarchy. Styles associated with types in the social landscape bear an important relation to class, but not a direct one. They are the product of enregisterment (Agha 2003) and I might call them registers were it not for the common use of the term in sociolinguistics to refer to a static collocation of features associated with a specific setting or fixed social category. Asif Agha’s account (2005) of enregistered voices is quite precisely what I am talking about here, locating register in a continual process of production and reproduction. Sociolinguists generally think of styles as different ways of saying the same thing. In every field that studies style seriously, however, this is not so – style is not a surface manifestation, but originates in content. The view of style I present here precludes the separation of form from content, for the social is eminently about the content of people’s lives. Different ways of saying things are intended to signal different ways of being, which includes different potential things to say. I will return to the issue of content below. Persona style is the best level for approaching the meaning of variation, for it is at this level that we connect linguistic styles with other stylistic systems such as clothing and other commoditized signs and with the kinds of ideological constructions that speakers share and interpret and that thereby populate the social imagination. Ideology is at the center of stylistic practice: one way or another, every stylistic move is the result of an interpretation of the social world and of the meanings of elements within it, as well as a positioning of the stylizer with respect to that world. Whether the speaker is a teenage girl adapting a Valley girl feature to position herself as cooler than her interlocutors or a fisherman on Martha’s Vineyard (Labov 1963) centralizing the nucleus of /ay/ to position himself as an opponent to the incursion of the mainland economy on the island, stylistic moves are ideological (see also Woolard this issue). And while these styles and stylistic moves can be quite local, ultimately they connect the linguistic sign systematically to the political economy and more specifically to the demographic categories that both emerge from and constrain local practice (Bourdieu 1977) and that have been the preoccupation of variation studies. By stylistic practice, I mean both the interpretation and the production of styles, for the two take place constantly and iteratively. Stylistic practice is a process of bricolage (Hebdige 1984), in which individual resources (in this case, variables) can be interpreted and combined with other resources to construct a  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

457

more complex meaningful entity. This process begins when the stylistic agent perceives an individual or group style – perhaps the style will bring his or her attention to those who use it; perhaps the users will call attention to the style. But the noticing of the style and the noticing of the group or individual that uses it are mutually reinforcing, and the meaning of the style and its users are reciprocal. The style itself will be noticed in the form of features that the stylistic agent separates out for notice. Susan Gal and Judith Irvine (Irvine and Gal 2000; Irvine 2001) have provided an account of the semiotic processes by which categories of speakers and their linguistic varieties come to be perceived as distinct, as an ideological link is constructed between the linguistic and the social. These processes apply equally well to the construction of meaning for styles (Irvine 2001) and for individual variables. This process of selection is made against a background of previous experience of styles and features; a stylistic agent may be more attuned to particular kinds of differences as a function of past stylistic experience. (This point appears rather small in the present context, but social differences in stylistic perception and production are structured and fundamental to the role of political economy in stylistic practice.) Once the agent isolates and attributes significance to a feature, that feature becomes a resource that he or she can incorporate or not into his or her own style. The occurrence of that resource in a new style will change the meaning both of the resource and of the original style, hence changing the semiotic landscape. Material style, particularly clothing and other forms of adornment, provide important clues to the study of linguistic style. One example of bricolage in clothing style comes from a series of interviews that some of my students and I did in Palo Alto, California, in 1985. (This example is described in more detail in Eckert 2000.) We were interviewing students at Palo Alto High School about their styles, introducing ourselves as researchers simply interested in style and asking them to describe their own style. At the time, ‘new wave’ style was big on the Palo Alto adolescent scene, and a group of new wavers were the local counterculture. The new wavers dressed almost exclusively in black, with distinctive pegged pants (pants that narrowed at the ankles). In contrast, the mainstream ‘preppy’ group wore pastel colors and straight-leg designer blue jeans. Two girls who were members of the popular preppy group told us with great pride of a stylistic move they had made. These girls characterized themselves as school-oriented and quite conformist. Nonetheless, they admired the new wavers’ independence and sought to distance themselves just a little from the extreme conformity normally expected of preppies. They chose to adopt some small sign of independence by appropriating something of new wave style. Dismissing the dark eye makeup (which they took to be scary and slutty) and the wearing of black (which they took to be too rebellious, slutty and adult), they pegged their blue jeans. In other words, they segmented the new wave style into meaningful elements, most saliently cut and color, identified the cut of one’s jeans as indexing autonomy but not rebellion or sluttiness, and worked it into their otherwise preppy style, ultimately making a claim to being both preppy and independent.  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

458

ECKERT

While the social categories at work in these girls’ stylistic moves are local, they deal with fundamental issues related to gender and adolescence: innocence and independence. And the symbolic material the girls were working with is not local – there is a culture-wide association of pastels with innocence and black with adulthood, sophistication, and the ominous. At the same time, pegged pants had a more fluid association (indeed, the girls knew nothing about the heyday of pegged pants in the prewar era of zoot suits worn by African American and Mexican American working-class males, and the more conservatively pegged pants of the postwar beatniks), making pegging available for segmentation and (re)interpretation. It is also to be noted that the ratio of the width of the knee and the bottom of pants legs was already salient, as fashion had moved in the preceding years from bell bottoms to straight legs (Eckert 1980). One could say that in adopting the pegged look they were carrying this process one step further. Thus while these girls were engaging in stylistic moves that they viewed as strictly local to their school and social groups, the resources they were using were available and salient because they had been established at a much more generalized cultural level. These girls were able to articulate every detail of their interpretive process. And although linguistic style is rarely constructed in as intentional a fashion as clothing style, it is similarly a process of bricolage. The big question for the study of meaning in variation is how linguistic styles are constructed: what kinds of meanings can variables have, and how do they combine to yield the larger meanings of styles? I offer the following discussion, therefore, as a beginning in dealing with this problem.

Belten High adolescents In my study of white adolescent speech in Belten High in the Detroit suburbs (Eckert 1989, 2000), I found that the differential use of variables constituted distinct styles associated with different communities of practice: the school-oriented jocks and the urban-oriented and school-alienated burnouts. I considered six vocalic variables: the mid and low vowels involved in the Northern Cities Shift, and the raising and backing of the nucleus of /ay/ (so that fight sounds more like foyt). I also examined one syntactic variable: negative concord. As shown in Figure 1, the seven variables pattern quite differently across the categories of gender and style. I have marked only the leaders in the use of the advanced variant (black) and the ‘runners up’ (gray). The older components of this shift, which appeared to have stabilized across the suburban area, were used predominantly by girls while the newer changes, which were more advanced closer to the urban center, were used predominantly by burnouts. The burnouts, in other words, were leading in the use of urban variants, embedding a linguistic opposition between city and suburb within a community to support a local opposition between urban- and school-oriented kids.  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

459

Figure 1: Use of Detroit variables involved in the Northern Cities Shift (NCS) by gender and social category. Black = greatest use, gray = second greatest use But just as women are not making direct gender claims when they use femaleled changes, burnouts are not making direct urban claims when they use urban-led changes. The urban-suburban sociocultural opposition was salient to suburban kids because of what they associated with urban life and urban kids. Detroit, one of the most segregated urban areas in the United States, was discussed by white kids at Belten High as a scary place dominated by African Americans. The urban kids that they identified with were white kids who knew how to cope in the dangerous urban environment – kids they saw as autonomous, tough, and street-smart. Presumably in adopting urban forms, suburban kids were affiliating with those qualities, not claiming to be urban. This presumption is further supported by an additional division between two network clusters of burnout girls – the ‘regular’ burnout girls and the ‘burned-out’ burnout girls. The burned-out burnout girls were quite objectively wilder, more alienated, and more urban-oriented than the regular burnout girls; in fact, they explicitly took pride in this fact. The burned-out burnout girls led all other burnouts, male and female, in the use of all of the Northern Cities Shift urban variables as well as negative concord. Boys over all led in the use of negative concord in this school, as they do more generally in the population, but the burned-out burnout girls led all boys, including the burnout boys, in the use of negative concord. More than one academic who has heard these results has suggested that the burned-out burnout girls were trying to talk ‘like boys.’ But talking ‘like a boy’ is done at the level of style, not at the level of the individual variable, and these particular girls were into ‘femininity’ in many ways, particularly the sartorial, and also led the rest of the cohort in the use of the older ‘female-led’ changes. This embedding of a linguistic opposition between city and suburb within a community to distinguish urban or suburban orientation is a prime example of the kind of iteration that Judith Irvine and Susan Gal (2000) have called recursiveness. The burned-out burnouts’ even greater use of urban variants adds still another level of recursiveness.  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

460

ECKERT

Negative concord, of course, is a global resource in English in its stereotyped counter-standardness, and is associated everywhere with lack of education and alienation from legitimate institutions. The elements of the Northern Cities Shift are region-specific, but the spread of change outward from urban areas creates a structure in which the meanings of emerging changes have common elements across regions. An opposition between urban and suburban is common to metropolitan areas across the United States, and there is a commonality in the structure of meaning associated with urban variables across most U.S. cities. Inner cities tend to be poorer, more dangerous, and more ethnically diverse than the surrounding suburbs. Pittsburgh, New York, and Detroit urban personae no doubt share some qualities deriving from living in a poorer, more dangerous, and more diverse environment. But every city has its own character, and the specifics of associations with urban variables will depend on local characteristics of those cities. My next example, Qing Zhang’s (2005) work on variation in Beijing, provides an example of the local flavor of urban variables.

Beijing managers Zhang’s research has established a relation between particular Beijing Mandarin variables and urban personae that are associated in local ideology with Beijing as a specific urban site. Her study compared the speech of managers in foreignowned financial businesses with that of managers in state-owned businesses. Managers in the foreign financial sector constitute the emergent ‘yuppie’ culture. As highly paid workers in a global market, the yuppies represent an important social change in China, and are highly visible in their development of western styles of consumption and a generally cosmopolitan lifestyle. This cosmopolitanism characterizes their speech style as well, which contrasts quite starkly with the speech of managers in the more traditional state-owned businesses. Zhang notes, furthermore, that gender plays a different role in the two sectors. While the state sector has quite rigid policies about hiring men and women on an equal footing, the foreign businesses first hire women into front office jobs, on the basis not only of managerial qualifications but also of their linguistic decorative value: their knowledge of English and quite possibly their use of a more generally cosmopolitan Mandarin. Zhang found a significant difference in the use of two Beijing variables that have achieved the status of linguistic stereotype and are associated not simply with Beijing but with popular Beijing types. • The Smooth Operator variable. The rhotacization of syllable finals (giving vowels a retroflex quality) and the realization of retroflex initial obstruents as [ ] is popularly seen as giving Beijing Mandarin an ‘oily’ quality, which in turn is associated with an oily character: the ‘Beijing Smooth Operator,’ jing you-zi, recognized in the public imagination, including the literature of Beijing writers, as a male Beijing type. This is someone who is smooth and  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

461

Figure 2: Use of Beijing and international variables by managers in stateowned and foreign-owned businesses (based on Zhang 2005). Black = greatest use, gray = second greatest use streetwise, who can handle all kinds of situations, and who can talk his way out of situations, who has the ‘gift of gab.’ The salience of this type showed up in Zhang’s conversations with managers. Referring to rhotacization as ‘swallowing sound,’ one of the managers commented, ‘Beijingers are glib, talk fast, like to swallow sounds, hence appear to be smooth’ (Zhang 2005: 443). The connection between the sound of rhotacization and oiliness and between oiliness and a specific persona is a particularly striking example of iconization (Irvine and Gal 2000). Even more striking is the comment recorded by Zhang comparing Beijing speech to southern dialects, attributing rhotacization to a physical quality or ability: ‘Have you ever heard anybody saying the Cantonese have “oily accent, slippery tone”? That’s because their tongues can’t curl’ (Zhang 2008). As shown in Figure 2, the yuppie style downplays local Beijing features. But while male yuppies use the Smooth Operator feature less than the state managers, they use it considerably more than the female yuppies – indeed, the female yuppies use it very little. It appears that this feature does not fit the persona required of a woman in the foreign sector. • The Alley Saunterer variable. Another feature studied by Zhang, the interdental pronunciation of dental sibilants, is not just a ‘Beijing’ feature but appears to index another specific local type: the h´ut`ong chu`anzi, or ‘Alley Saunterer,’ a feckless character who hangs out in back streets waiting for something to happen. The managers in Zhang’s study pointed specifically to the Alley Saunterer when discussing this variable: ‘Mine is not the real Beijing Mandarin, because I’m not one of those Alley Saunterers . . . They speak the authentic Beijing speech. And they say “big cabbage” [ta paitsh ai] as [ta paitθ h ai]’ (Zhang 2005: 443). In an example of iconization similar to that discussed above regarding the Smooth Operator, another manager characterized Alley Saunterers as ‘big-tongued’ since they ‘bite their tongues when talking.’ While the Alley Saunterer image is not useful to the male or female cosmopolitan yuppie persona or to women in state-owned businesses, a little use of it works for state businessmen.  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

462

ECKERT

Beijing yuppies’ avoidance of these two Beijing variables builds on an association of the features with particular types and distances them from these types as part of constructing a new (yuppie) type. At the same time, this distancing process reinscribes the old types by creating a new space in the social map in opposition to them. Meanwhile, yuppies’ adoption of a non-Beijing feature, full tone, projects them out into transnational space. • Full tone. Unstressed syllables in Beijing Mandarin are sufficiently reduced that they lose their distinct tone, which assimilates to the preceding tone. Non-mainland dialects of Mandarin, however, retain the full tone in unstressed syllables. The yuppie use of full tone has been commented on in the media and is generally seen as evidence of a speaker’s being a Hong Kong-Taiwan wannabe, or of Chinese Mainlanders’ attempt to imitate the ‘Hong Kong-Taiwan accent.’ The yuppies – particularly the women – added an additional cosmopolitan flavor to their style with the occasional use of the full tone feature, which is associated with the global capitalist markets located outside the mainland – a resource that the state managers did not use at all. Of particular interest in this case is that since the use of the full tone is associated with the de-reduction of unstressed syllables, the result is a greater equality of syllable length in the stream of speech, giving the speech a more staccato sound – a crispness that one might contrast with the oily tone attributed to the typical Beijing speaker. Studies like Labov (1963), Eckert (2000), and Zhang (2005) clearly establish that variables that historically come to distinguish geographic dialects can take on interactional meanings based in local ideology. In all three cases, the meaning is based in ideologies about what the locality is about – what kinds of people live there and what activities, beliefs, and practices make it what it is. Local identity is never an association with a generic locale but with a particular construction of that locale as distinct from some other. Local identity claims are about what it means to be from ‘here’ as opposed to some identified ‘there.’ The Martha’s Vineyard fishermen, in appropriating the centralized variant of (ay), were not simply claiming to be Vineyarders but were making a claim about what a Vineyarder is. This claim immediately raises the potential for the distinction to be activated inside the community as well, as there will be differences in the community that are construed in direct relation to differences across communities. The most salient phonological feature that distinguishes Vineyard from mainland speech came to mark opposition to mainland incursion on the island. Furthermore, the fact that the English-descent fishing community was leading in this particular linguistic opposition – not the Portuguese nor the Native Americans – suggests that the English claim to authenticity on behalf of the island was inseparable from their claim to greater authenticity within the island.

 C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

463

INDEXICAL ORDER AND INDEXICAL FIELDS As the above examples suggest, ‘acts of identity’ (Le Page and Tabouret-Keller 1985) are not primarily a matter of claiming membership in this or that group or category as opposed to another, but smaller acts that involve perceptions of individuals or categories that fall under the radar of large sociolinguistic surveys. This is not to say that these acts are independent of the larger social order; on the contrary, they are systematically related to the macrosociologist’s categories and embedded in the practices that produce and reproduce them. It is in the links between the individual and the macrosociological category that we must seek the social practices in which people fashion their ways of speaking, moving their styles this way or that as they move their personae through situations from moment to moment, from day to day, and through the life course (see also Heller this issue). In this process, they do not simply use social meaning – they produce and reproduce it. Michael Silverstein’s (2003) concept of indexical order provides a crucial perspective on this process. It gives a foothold on the relation between the macrosociological facts and linguistic practice by providing a theoretical account of the role of construal in context in the process of indexical change. In his discussion of the indexical value of variation, Silverstein engages with John Gumperz’s (1968) distinction between dialectal and superposed variability and with Labov’s (1971) distinction between indicators, markers, and stereotypes. In Labov’s terms, indicators are dialectal variables that distinguish social or geographic categories but have attracted no notice and do not figure in variation across the formality continuum. Markers and stereotypes are variables that have attracted sufficient attention to emerge within those categories in stylistic variation. The difference between markers and stereotypes lies in the level of consciousness: stereotypes are subject to metapragmatic discussion, while markers are not. An indicator in variation is what Silverstein terms a first-order index. A first-order index simply indexes membership in a population – it designates people as Martha’s Vineyarders, Beijingers, Detroiters. In the case of Labov’s (1966) New York City study, which Silverstein uses as his example, the populations may be social class strata. But the social evaluation of a population is always available to become associated with the index and to be internalized in speakers’ own dialectal variability to index specific elements of character.2 At that point, the linguistic form becomes a marker, a second-order index, which figures stylistically as speakers position themselves with respect to the elements of character selected out for internal use. The difference between the notion of marker as used in variation studies and the index of Silverstein’s treatment is in the ideological embedding of the process by which the link between form and meaning is made and remade. Participation in discourse involves a continual interpretation of forms in context, an in-the-moment assigning of indexical values to linguistic forms. A form with an indexical value, what Silverstein calls an nth order usage, is always available for reinterpretation – for the acquisition of an n + 1st value. Once established, this new value is available  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

464

ECKERT

for further construal, and so on. While the terms first order, second order, etc. may imply a linearity,3 this is clearly not Silverstein’s intention. On the contrary, the reconstruals are ‘always already immanent’ (2003: 194) precisely because they take place within a fluid and ever-changing ideological field. The emergence of an n + 1st indexical value is the result of an ideological move, a sidestepping within an ideological field. In order to understand the meaning of variation in practice, we need to begin with this ideological field, as the continual reconstrual of the indexical value of a variable creates, in the end, an indexical field.4 An indexical field is a constellation of meanings that are ideologically linked. As such, it is inseparable from the ideological field and can be seen as an embodiment of ideology in linguistic form. I emphasize here that this field is not a static structure, but at every moment a representation of a continuous process of reinterpretation. The traditional view of a variable as having a fixed meaning is based in a static, non-dialectical, view of language. In this view, a variable is taken to ‘mean’ the same regardless of the context in which it is used, and while we know, for example, that variables may change their meanings over time, the mechanism for this process is not well understood. This dilemma is reminiscent of the one that Weinreich, Labov, and Herzog (1968) sought to solve by arguing that the social structuring of variability provides orderly heterogeneity in the constant process of change. The notion of the indexical field is an argument for a similar orderly heterogeneity in the ever-changing indexical value of variables. Variables have indexical fields rather than fixed meanings because speakers use variables not simply to reflect or reassert their particular pre-ordained place on the social map but to make ideological moves. The use of a variable is not simply an invocation of a pre-existing indexical value but an indexical claim which may either invoke a pre-existing value or stake a claim to a new value. As noted above, the Martha’s Vineyard fishermen, in appropriating the centralized variant of /ay/, were not simply claiming to be Vineyarders but were making a claim about what a Vineyarder is. I would argue that as disagreements about the future of the island became more prominent in daily life, the terms of those disagreements entered into the local ideological field, available to be pointed to with the use of a linguistic variable already associated with Vineyarders.5 The new construal of meaning for the re-centralization of /ay/ was no doubt not just a product of who used it, but when and how they used it. The use of the centralized variant may have been a way of asserting one’s local authority and/or loyalty in a range of utterances that gave specificity to the source of authority and the object of loyalty: an argument about the fate of the island; recounting heroic or nostalgic stories about the fishing life; taking a negative stance towards the tourist industry or towards mainland or mainland-oriented individual; etc. It is not just the meaning of phonological variants that change in discourse – lexical change does as well. A word’s denotation can absorb connotations through association with aspects of the context in which it is used and most certainly, stances. Andrew Wong (2005) has traced the development of tongzhi ‘comrade’ in Hong Kong in spoken and written discourse, from its appropriation  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

465

by the local gay community as a positive term, through its pejoration at the hands of the media through the selective use of the term in news stories. The pejoration of many English words referring to females is a perfect example of the systematic absorption of ideology into the lexicon. Sally McConnell-Ginet (1989) has argued that changes such as the pejoration of hussy, originally meaning ‘housewife,’ happened over years of situated use, in which: (1) the term was used repeatedly in negative utterances about specific women or categories of women; and (2) the utterances of those who said such negative things were registered disproportionately. It is worthwhile at this point to return briefly to the apparent toughness of working-class men. I believe that toughness is an important social meaning – one that is a component of many styles, and that is embraced situationally, and by women as well as men. It is reasonable to guess that toughness is a common part of the indexical field for some variables originating in the working class and for urban variables. But it is important to consider the ideological work behind this connection in order to look beyond it. Working-class men are quite diverse, and while they fall everywhere on the toughness scale, toughness nonetheless remains central to the working-class stereotype. There are a variety of other qualities that one might say are characteristic of many working-class men including, for instance, egalitarianism. The process of erasure (Irvine and Gal 2000) elevates toughness by downplaying working-class men who are not tough, as well as downplaying other qualities such as egalitarianism. The fact that toughness is separated out is an ideological fact that is part of the very central construction of an opposition between working-class (physical) and middle-class (technical) masculinities (Connell 1995). In other words, the indexical values of variables are part and parcel of the ideological work of society and vice-versa – and it is for this reason that the survey studies of variation have found over and over that variables correlate with the fundamental, ideologically laden experiences of class and gender. I would argue, therefore, that no trivial meanings come to be associated with variables but that any meanings that are associated with variables will be based in highly salient ideological issues.

An example: (ING) Kathryn Campbell-Kibler’s (2007a) experimental work on (ING) provides us with a starting point for thinking about how indexical fields are configured. She has shown that listeners develop an impression of a speaker based on general speech style and the content of the utterance, and interpret the particular use of (ING) on the basis of that impression.6 Using advanced matched guise techniques with college students as subjects, Campbell-Kibler demonstrates that (these) hearers associate the velar variant with education, intelligence, and articulateness. Central to this perception is a view of the velar form as a full form and therefore effortful and of the apical form as a reduced form, hence a sign of lack of effort.7 One might then presume that this lack of effort can be further construed as a  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

466

ECKERT

Figure 3: Indexical field of (ING) (based on Campbell-Kibler 2007a, 2007b). Black = meanings for the velar variant, gray = meanings for the apical variant result of laziness, not caring, or even rebellion, and by extension, impoliteness. In further experiments with a broader socioeconomic range, Campbell-Kibler (2007b) found that hearers additionally interpret the apical form as casual or relaxed. Based on these potential meanings for (ING), one could imagine an indexical field for the velar form as shown in Figure 3. The adjectives in black represent potential meanings for the velar variant, while those in gray represent potential meanings for the apical variant. One cannot assume that the pairs work in lockstep. Among other things, while a negative evaluation of a speaker using the apical variant might be that the speaker is inarticulate or lazy, a favorable evaluation might be that he or she is unpretentious or easygoing. Similarly, the speaker using the velar variant can be heard as simply articulate or as pretentious. Which of the meanings in the indexical field the hearer will associate with a given occurrence will depend on both the perspective of the hearer8 and the style in which it is embedded – which includes not only the rest of the linguistic form of the utterance but the content of the utterance as well. Campbell-Kibler found that hearers judged the nature of the social move being made by the speaker based on presupposed indexicality. If their impression was that the speaker was relatively uneducated or from the South, they expected to hear the apical variant; if their impression was that the speaker was educated and to some extent from the North, they expected to hear the velar variant. The use of the ‘wrong’ variant, then, led them to interpret the speaker as pretentious, condescending, insincere, and so on. Since the same variable will be used to make ideological moves by different people, in different situations, and to different purposes, its meaning in practice  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

467

will not be uniform across the population. This is a point that Barbara Johnstone and Scott Kiesling (2008) have argued, and it cannot be overemphasized. Variability across communities is not limited to linguistic form but is present in the understanding of what that form means and ultimately in the ideologies that underlie language use (cf. Sidnell this issue). Beatriz Lavandera (1978) argued decades ago that once we leave the level of phonology, that is, once we leave linguistic units that have no meaning of their own, we can no longer claim that class differentiation is simply differentiation of form and not of content. Suzanne Romaine (1984) has pointed out that the correlation of syntactic variation with social categories cannot proceed before taking account of pragmatic meanings. These meanings, in turn, may be related to social differences. People at different places in the political economy see the world differently, do different things, have different preoccupations, and say different things. I would argue that this principle applies to phonological variation as well.9 While the entire population might agree on first-order indexicality – who uses what variant – the evaluation of that differentiation can differ across the population. It is reasonable to suggest that the social stratification of variables may in fact reflect socially-stratified ideological orientations. One might consider that socio-economic differences in the use of (ING) potentially involve highly ideological choices, with the use of the velar form indexing one’s association with institutions of legitimacy and the power they represent, and the use of the apical form indexing opposition to such institutions and the power that they represent. It is also reasonable to speculate that the class stratification of this form can reflect class-based ideologies about formality. Presumably the almost categorical use of the apical form by Lord Peter Wimsey in the murder mysteries of Dorothy Sayers indexes the effortlessness that comes of entitlement. And it is the embedding of this apical form in a broader upper-class style that allows it to take on this meaning. If I am correct in my assumption that class differences involve ideological differences about formality and displays of education, then one might expect working-class speakers to have the more positive evaluations of this form and middle-class speakers to have the more negative ones.

Another example: /t/ release in American English Most variables that have been studied socially have been either sound changes in progress or a relatively small set of stable linguistic stereotypes. But recent explorations of hyperarticulation in the form of /t/ release where one would normally not expect release in American English have found correlations with an oddly unconnected set of social categories: nerd girls, Orthodox Jewish boys, and gay men. In a study of the speech of high school students in northern California, Mary Bucholtz (2001, in preparation) encountered a group of girls who fashioned themselves as ‘nerds’ – a particular kind of intellectual identity stereotypically limited to boys. One might characterize these girls as intellectual mavericks, since they distanced themselves from teachers, whom they viewed  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

468

ECKERT

as mere consumers and purveyors of standard intellectual fare. These girls developed a distinctive linguistic style through a variety of means, among them the use of /t/ release. /t/ release has also been associated with Jewishness. In a study of an orthodox Jewish school in northern California, Sarah Benor (2001) found that boys produced significantly more released /t/s than girls, and that the leaders in /t/ release among boys were those who had been formally studying Talmud in a Yeshiva. Furthermore, anecdotal evidence suggested that these boys were particularly likely to release /t/ when making a strong point in an intellectual argument. Finally, released /t/ is part of a common stereotype of gay speech. Rob Podesva, Sarah Roberts, and Kathryn Campbell-Kibler (2002) studied the speech of two lawyers in a debate about the exclusion of gays in the Boys Scouts, one a ‘straight’ lawyer representing the Boy Scouts and the other a ‘gay’ lawyer representing the Lambda League. (While the sexual orientation of the lawyers was not specified in the debate, each lawyer can be seen as officially representing a group defined by sexual orientation.) In this debate, the Lambda League lawyer released significantly more /t/s than the Boy Scout lawyer. What makes this variant ‘nerdy,’ ‘Orthodox,’ or ‘gay’ is its participation in three quite different styles. So what is the underlying potential of this variable? The use of /t/ release in the nerd-girl style and in the Yeshiva-boy style no doubt builds primarily on the social significance of clear speech, which in turn is associated with a school-teachery standard. (There is also the potential of Yiddish as an additional source for Orthodox Jewish /t/ release.) The contrast between the flapped intervocalic /t/ of the United States and the released /t/ of British English further evokes stereotypes of the British as cultured, refined, and articulate, and Americans as anti-intellectual and loutish. Indeed, /t/ release is a common resource for Americans imitating British English, and loutish types in the United States are commonly portrayed as using hypoarticulated /t/. Needless to say, this leads us into a broader view of language ideology – the association of hyperarticulation with care and hypoarticulation with laziness. The combination of the North American-based association with clear speech and the international association with British speech opens up another set of possible meanings having to do with refinement and elegance. And the intersection of refinement and care opens up the terrain of politeness. The potential of clear speech is further clarified in Rob Podesva’s work (2004, 2007) on gay speech. The very notion of gay speech has been elusive, and Podesva has pursued the study of stylistic variation in the speech of several gay professionals in situations where gayness had different salience. Particularly relevant here is his study of a doctor, Heath, in several settings, including at work in the clinic and at a barbecue with friends. In the clinical setting, Heath needed to present himself as a competent, educated, and articulate professional. But at the barbecue, he was with good friends with whom he was in the habit of presenting himself as a gay ‘bitchy diva’ – as meticulous, style-conscious, critical, and flamboyant. In keeping with the doctor image as an educated and precise professional, Heath released significantly more occurrences of /t/ in the clinic  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

469

than at the barbecue. But while there were more occurrences of /t/ release in Heath’s speech in the clinic, those occurrences of /t/ that he did release at the barbecue had significantly longer and stronger bursts than those in the clinic. In other words, his /t/ release at the barbecue was a parody of his /t/ release in the clinic, yielding what Podesva (2008) calls a ‘prissy’ effect in keeping with his diva persona. An additional aspect of stop release is its potential to express emphasis, which is related, but not identical, to clarity. Stop release commonly emerges to index exasperation and even anger – and one might consider that Heath’s bitchydiva persona combines the prissiness of the teacher’s pet and the expression of some kind of generalized attitude of exasperation. Figure 4 is an approximation of an indexical field for /t/ release. Needless to say, while Figure 3 is based on experimental evidence of hearers’ interpretations, Figure 4 is based on interpretations of correlations in speech and hence is more speculative. The point, however, is to propose both a way of looking at the indexical value of variation and a goal for further analyses. For purposes of discussion, Figure 4 differs from Figure 3. While I included both variants of (ING) in Figure 3, Figure 4 does not include alternatives to /t/ release, although I will discuss alternatives below. In Figure 4, I have distinguished between what might be considered permanent qualities (in black) and stances (in gray). I do this not to distinguish between two distinct categories of meanings, but to emphasize the fluidity of such categories and the relation between the two in practice. While anger or cynicism may be momentary and situated stances, people who are viewed as habitually taking such stances may become ‘angry’ or ‘cynical’ people through stance accretion (Rauniomaa 2003, cited in Bucholtz and Hall 2005: 596). By ‘becoming an angry or cynical person,’ I mean that one may come to be socially

Figure 4: Indexical field of /t/ release. Boxes = social types, black = permanent qualities, gray = stances  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

470

ECKERT

positioned as angry or cynical – that anger and cynicism become part of one’s identity (in the sense of Bucholtz and Hall 2005) or one’s habitual persona. This possibility of stance accretion is central to the fluidity of indexicality, as a mechanism for the elaboration of the indexical field. Figure 4 also includes, in boxes, the social types, or enregistered voices, that live at the less fluid end of the field, anchoring the process of interpretation. The indexical field of /t/ release does not exist in a vacuum but is part of a vast system. While I have not included alternatives to /t/ release in Figure 4, an understanding of these alternatives is crucial to understanding this feature and ultimately they belong in this picture. In all his work, Podesva emphasizes that phonetic details are crucial to social meaning in variation, and his attention to the range of variability in stop production leads to an even broader view of the indexical value of /t/ release. Podesva did not limit his study to this form alone but included the full range of articulation of /t/, from hyperarticulation to deletion. In his study of Heath’s variation, Podesva not surprisingly finds significantly more /t/ deletion in his casual speech at the barbecue than in the clinic. It is the combination of deletion with the exaggerated bursts that begins to construct the diva style. The salience of hyperarticulation goes beyond clarity and into broader ideologies associated with this trait. The prissy meaning of Heath’s strong bursts is no doubt associated with the ‘goody-goody’ image associated with purist teachers who insist on clear standard speech, and with their teachers’ pets. At the other extreme of the legitimacy hierarchy, the deletion of /t/ is a stigmatized feature of African American Vernacular English (AAVE). The relation between power and hyperarticulation cannot be underestimated, because it provides the basis for the depreciation – and perhaps the maintenance – of AAVE in particular as actively counter-cultural. Some of the main stereotypes of AAVE (cluster reduction, absence of plural and third singular –s) involve absence or hypoarticulation of consonants, throwing the variety directly into opposition with the hyperarticulation of the hyperstandard (Bucholtz 2001). Andrea Kortenhoven (personal communication) recalls a humorous pronunciation of anyway [εntiweyz] developed in her African American childhood friendship group, parodying the hyperarticulation of teacher talk.) This attention to the continuum of articulation level thus can be seen as part of a broader national ideology that links hyperarticulation to clarity and clarity to education and power. I might be tempted, then, to revise the indexical field for (ING), inasmuch as in English language ideology, the velar variant is commonly seen as clearer and as hyperarticulation (in some circles), while the apical variant is seen as hypoarticulation (in some circles). Certainly (ING) and /t/ release co-occur stylistically to a great extent, but this has not to my knowledge been studied quantitatively as yet. Whether the two overlap enough that they contribute the same thing to styles is an empirical question. So far I have examined a very limited kind of variable – hyperarticulation – which is not the kind of variable that is the subject of most variation study. It remains to be seen what kinds of indexical fields other kinds of variables have.  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

471

No doubt the longer a variable is around the more nuanced meanings it can take on (Haeri 1997), so the fields of sound changes in progress may be less well defined than those of stable variables. On the other hand, they may leave more room for local interpretation. But even stable variables may have a broad potential for social meaning. While Labov’s study of (DH) (commonly referred to as /dh/ stopping) in New York City focused on class differentiation, it is a common variable across immigrant communities and has been found to be ethnically salient among Germans (Rose 2006), Cajuns (Dubois and Horvath 1998a, 1998b), Poles (Edwards and Krakow 1985), and Latinos (MendozaDenton 2007). This variable no doubt began in all of these communities as a substratal form, providing a quite pure first-order index. This situation then made the variable available to index ideologies associated with those ethnic groups. Mary Rose’s (2006) study of a farming community in Wisconsin found that this variable was overtly associated with the German farming community, although German ancestry no longer correlates with its use. Rather, it appears to have come to be linked to lack of education on the one hand and with the value of hard work associated with small farming on the other. Meanwhile, in Louisiana Cajun English (Dubois and Horvath 1998a), this variable has come to be associated with the prestigious Cajun renaissance. Students in my sociolinguistics class years ago studied this variable in Chicago’s Little Italy and found a correlation between /th/ stopping and integration into the Italian community. In my own experience in suburban New Jersey, where the working class was predominantly Italian, /th/ and /dh/ stopping were highly salient, highly conscious variables that we associated overtly with certain personae in our school. These personae were boys we saw as Italian, tough in an endearing way, and not terribly bright, and people regularly used /th/ and /dh/ stopping in performances of each of these characteristics separately. The use of /th/ and /dh/ stopping usually involved emphasis of some sort, and to this day my own speech features /dh/ and sometimes /th/ stopping for emphasis. Of course, this retrospective analysis is highly suspect, but it raises, among other things, the prospect that the indexical field of /th/ stopping in my community overlapped in interesting ways with the field of /t/ release. First of all, /t/ release could be an exaggeration of /th/ stopping – one could say that /th/ stopping feeds /t/ release. But it does not have the same indexical field, for while as a fortition it is certainly emphatic, as a hypoarticulation, it is unlikely to have the indexical values associated with clarity. Thus, /th/ stopping and /t/ release are likely to co-occur stylistically only in cases in which /th/ stopping is quite explicitly indexing annoyance. Once again, this is pure speculation, but with the intention of suggesting directions to follow.

CONCLUSIONS Labov’s introduction of class into the study of language was a landmark of immense importance. But his view of the class hierarchy, and of the relation of standard and vernacular language to that hierarchy, is only the beginning of a  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

472

ECKERT

theory of the social value of variation. The social is not just a set of constraints on variation – it is not simply a set of categories that determine what variants a speaker will use – it is a meaning-making enterprise. And while one’s place in the political economy has an important constraining effect on how one makes meaning, and on the kinds of meanings one engages with, this place cannot be defined in terms of a simple model. A theory of variation ultimately must deal with meaning, and not only does a view of meaning in variation as predetermined and static seriously undershoot human capacity, it cannot even account in any principled way for the changes in correlations that have been observed over the lifetime of a sound change (e.g. Labov 2001: chapter 9). Ultimately, all change unfolds in the course of day-to-day exchange, and that exchange involves constant local reinterpretation and repositioning. Ultimately, it is in this action that we can get at the meaning-making that gives life to variation. While the larger patterns of variation can profitably be seen in terms of a static social landscape, this is only a distant reflection of what is happening moment to moment on the ground. The study of style and of the meaning of variation will raise many new questions. I have focused in this paper on the indexical potential of a few variables, and claimed that variables combine to constitute styles. I have left undiscussed the question of the process of bricolage and the structure of style: Are there constructions? Is stylistic meaning compositional? And implicit in this discussion is the need to examine a far greater range of variables than is commonly done in the field. A study of style and social meaning will focus us on variables that are not apparently changes in progress, or that currently have no particular structural interest. Sadly, these variables will not have been included in large community studies, so we will not benefit from the kind of macrosociological information about them that we do for the more commonlystudied variables. This study will also draw our attention to more abstract aspects of variables such as fortition-lenition, hypo-hyperarticulation, or the differential stylistic potential of different phonetic classes. And it will draw us into details of prosody and voice quality. This may also lead us to think about ‘edgier’ issues of emotional expression in language, following F´onagy (e.g. 1971). What I have presented here is just an idea, and an expectation that others will carry the project forward.

NOTES 1. The idea of the indexical field emerged in my many discussions with Miyako Inoue, particularly as we prepared and taught a course at the Linguistic Society of America Summer Institute at Stanford University in 2007. I am deeply indebted to Miyako for immense intellectual fun and inspiration. I am grateful also to Michael Silverstein not only for providing the theoretical basis for this paper, but for remarkably painstaking, gracious, and helpful comments on the first draft.  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

473

2. Of course, the very fact of distinction of social groups entails evaluation, and by its ideological nature, linguistic practice entails an evaluation of linguistic difference. Thus whether ideologies about groups lead to the perception of their linguistic differences or whether a perception of a linguistic difference calls for a perception of the group makes little difference: the two are indistinguishable. 3. Johnstone, Andrus and Danielson (2006) equate indexical order with a progression from indicator (first order index) to marker (second order index) to stereotype (third order index). Since the only difference between a marker and a stereotype is the level of consciousness, this distinction is orthogonal to changes in indexical value. 4. I am a little hesitant to use the term indexical field for this configuration because of its mismatch with the term semantic field, which refers to a field of words of similar semantic content rather than to polysemy. Indeed, it might be better to use the term indexical field to refer to the various variables that might have related indexical value. But at the moment it is not at all clear that this would be a profitable way to look at relations among variables since the social world is interlocked in a way that the world of the lexicon is not. (The indexical field also has no clear relation to Bourdieu’s notion of field.) 5. It is notable that the social correlations for (ay) and (aw) are dissimilar – the Native American community at Gay Head, which trailed in the centralization of (ay), led in the centralization of (aw). I have suggested elsewhere (Eckert 2000: 23–24) that these two variables were associated with possibly competing claims to local authenticity. 6. Describing it this way implies a temporal relation between noting the style and interpreting the variable (ING), which clearly is not intended. The actual cognitive processes behind this interpretation remain to be understood. 7. While the apical variant is not in itself initially a reduction of the velar variant, it is subject to reduction while the velar variant is not. 8. Campbell-Kibler (2007b) has even found that hearers are more likely to give negative evaluations of speakers’ (ING) production when they are feeling grouchy. 9. I understand that Michael Silverstein made this point in his Collitz Lecture in 1989. I wish I’d been there to hear it.

REFERENCES Agha, Asif. 2003. The social life of cultural value. Language and Communication 23: 231–273. Agha, Asif. 2005. Voice, footing, enregisterment. Journal of Linguistic Anthropology 15: 38–59. Benor, Sarah. 2001. The learned /t/: Phonological variation in Orthodox Jewish English. In Tara Sanchez and Daniel Ezra Johnson (eds.) Penn Working Papers in Linguistics: Selected Papers from NWAV 2000. Philadelphia, Pennsylvania: University of Pennsylvania Department of Linguistics. 1–16. Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge, U.K.: Cambridge University Press. Bucholtz, Mary. 2001. The whiteness of nerds: Superstandard English and racial markedness. Journal of Linguistic Anthropology 11: 84–100. Bucholtz, Mary. In preparation. Styling: Language and White Youth Cultures. Cambridge, U.K.: Cambridge University Press.

 C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

474

ECKERT

Bucholtz, Mary and Kira Hall. 2005. Identity and interaction: A sociocultural linguistic approach. Discourse Studies 7: 585–614. Campbell-Kibler, Kathryn. 2007a. Accent, (ING) and the social logic of listener perceptions. American Speech 82: 32–64. Campbell-Kibler, Kathryn. 2007b. What did you think she’d say?: Expectations and sociolinguistic perception. Paper presented at the annual conference on New Ways of Analyzing Variation. Philadelphia, Pennsylvania, October. Connell, Robert. 1995. Masculinities. Berkeley, California: University of California Press. Dubois, Sylvie and Barbara Horvath. 1998a. From accent to marker in Cajun English: A study of dialect formation in progress. English World-Wide 19: 161–188. Dubois, Sylvie and Barbara Horvath. 1998b. Let’s tink about dat: Interdental fricatives in Cajun English. Language Variation and Change 10: 245–261. Eckert, Penelope. 1980. Clothing and geography in a suburban high school. In Conrad Phillip Kottak (ed.) Researching American Culture. Ann Arbor, Michigan: University of Michigan Press. 139–145. Eckert, Penelope. 1990. The whole woman: Sex and gender differences in variation. Language Variation and Change 1: 245–267. Eckert, Penelope. 2000. Linguistic Variation as Social Practice. Oxford, U.K.: Blackwell. Eckert, Penelope. 2004. Elephants in the room. Journal of Sociolinguistics 7: 392–397. Edwards, Walter and Cheryl Krakow. 1985. Polish-American English in Hamtramck: A sociolinguistic study. Paper delivered at the annual Conference on New Ways of Analyzing Variation in English. Georgetown University, Washington, D.C., October. F´onagy, Ivan. 1971. The functions of vocal style. In Seymour Chatman (ed.) Literary Style: A Symposium. London and New York: Oxford University Press. 159–174. Gumperz, John. 1968. The speech community. In David L. Sills (ed.) International Encyclopedia of the Social Sciences. London: McMillan. 381–386. Haeri, Niloofar. 1997. The Sociolinguistic Market of Cairo: Gender, Class and Education. London: Kegan Paul International. Hebdige, Dick. 1984. Subculture: The Meaning of Style. New York: Methuen. Irvine, Judith. 2001. Style as distinctiveness: The culture and ideology of linguistic differentiation. In Penelope Eckert and John Rickford (eds.) Style and Sociolinguistic Variation. Cambridge, U.K.: Cambridge University Press. 21–43. Irvine, Judith T. and Susan Gal. 2000. Language ideology and linguistic differentiation. In Paul V. Kroskrity (ed.) Regimes of Language: Ideologies, Polities, and Identities. Santa Fe, New Mexico: School of American Research Press. 35–83. Johnstone, Barbara, Jennifer Andrus and Andrew E. Danielson. 2006. Mobility, indexicality, and the enregisterment of ‘Pittsburghese’. Journal of English Linguistics 34: 77–104. Johnstone, Barbara and Scott F. Kiesling. 2008. Indexicality and experience: Exploring the meanings of /aw/-monophthongization in Pittsburgh. Journal of Sociolinguistics 12: 5–33. Labov, William. 1963. The social motivation of a sound change. Word 18: 1–42. Labov, William. 1966. The Social Stratification of English in New York City. Washington, D.C.: Center for Applied Linguistics. Labov, William. 1971. The study of language in its social context. In Joshua A. Fishman (ed.) Advances in the Sociology of Language, Vol. 1. The Hague, The Netherlands: Mouton. 152–216. Labov, William. 1972. The isolation of contextual styles. In Sociolinguistic Patterns. Philadelphia, Pennsylvania: University of Pennsylvania Press. 70–109.  C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation 

VARIATION AND THE INDEXICAL FIELD

475

Labov, William. 2001. Principles of Linguistic Change: Social Factors. Oxford, U.K.: Blackwell. Labov, William. 2002. Review of Penelope Eckert, Linguistic Variation as Social Practice. Language in Society 31: 277–284. Lavandera, Beatriz R. 1978. Where does the sociolinguistic variable stop? Language in Society 7: 171–182. Le Page, R. B. and Andr´ee Tabouret-Keller. 1985. Acts of Identity: Creole-Based Approaches to Language and Ethnicity. Cambridge, U.K.: Cambridge University Press. Mcconnell-Ginet, Sally. 1989. The sexual (re)production of meaning: A discoursebased theory. In Francine W. Frank and Paula A. Treichler (eds.) Language, Gender and Professional Writing: Theoretical Approaches and Guidelines for Nonsexist Usage. New York: MLA. 35–50. Mendoza-Denton, Norma. 2007. Home Girls. Oxford, U.K.: Blackwell. Piaget, Jean. 1954. The Construction of Reality in the Child. New York: Basic Books. Podesva, Robert. 2004. On constructing social meaning with stop release bursts. Paper presented at Sociolinguistics Symposium 15. Newcastle upon Tyne, April. Podesva, Robert. 2007. Phonation type as a stylistic variable: The use of falsetto in constructing a persona. Journal of Sociolinguistics 11: 478–504. Podesva, Robert. In press. Three sources of stylistic meaning. In Terra Edwards, Kate Shaw, Sarah Wagner and Eiko Yasui (eds.) Proceedings of SALSA XV. Texas Linguistic Forum 51. Podesva, Robert J., Sarah J. Roberts and Kathryn Campbell-Kibler. 2002. Sharing resources and indexing meanings in the production of gay styles. In Kathryn Campbell-Kibler, Robert J. Podesva, Sarah J. Roberts and Andrew Wong (eds.) Language and Sexuality: Contesting Meaning in Theory and Practice. Stanford, California: CSLI Press. 175–190. Rauniomaa, Mirka. 2003. Stance accretion. Paper presented at the Language, Interaction, and Social Organization Research Focus Group, University of California, Santa Barbara, February. Romaine, Suzanne. 1984. On the problem of syntactic varation and pragmatic meaning in sociolinguistic theory. Folia Linguistica 18: 409–439. Rose, Mary. 2006. Language, place and identity in later life. Unpublished PhD dissertation, Stanford University, Department of Linguistics. Silverstein, Michael. 1985. Language and the culture of gender: At the intersection of structure, usage and ideology. In Elizabeth Mertz and Richard Parmentier (eds.) Semiotic Mediation: Sociocultural and Psychological Perspectives. New York: Academic Press. 219–259. Silverstein, Michael. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23: 193–229. Trudgill, Peter. 1972. Sex, covert prestige and linguistic change in the urban British English of Norwich. Language in Society 1: 179–195. Weinreich, Uriel, William Labov and Marvin Herzog. 1968. Empirical foundations for a theory of language change. In Winfred Lehmann and Yakov Malkiel (eds.) Directions for Historical Linguistics. Austin, Texas: University of Texas Press. 95– 188. Wong, Andrew. 2005. The re-appropriation of Tongzhi. Language in Society 34: 763– 793. Zhang, Qing. 2005. A Chinese yuppie in Beijing: Phonological variation and the construction of a new professional identity. Language in Society 34: 431–466.  C The author 2008

C Blackwell Publishing Ltd. 2008 Journal compilation 

476

ECKERT

Zhang, Qing. 2008. Rhotacization and the ‘Beijing Smooth Operator’: The meaning of a sociolinguistic variable. Journal of Sociolinguistics 12: 210–222.

Address correspondence to: Penelope Eckert Department of Linguistics Stanford University Stanford, CA 94305-2150 U.S.A. [email protected]

 C The author 2008 C Blackwell Publishing Ltd. 2008 Journal compilation