800-1000 essay sociolinguistic
/r/ and the construction of place identity on New York City's Lower East Side † (Becker,2009)
Abstract
This paper argues that a group of white residents on the Lower East Side of Manhattan use a New York City English (NYCE) feature – non‐rhoticity in the syllable coda – in the construction of a place identity, one aspect of identity tied to localness and authenticity. A quantitative analysis confirms that the change in progress towards rhoticity in NYCE ( Labov 1966 ) continues to advance slowly, so that non‐rhoticity remains a resource for New Yorkers, imbued with local social meaning. Ethnographic observation of the Lower East Side reveals conflict among residents, which motivates one group to highlight their place identity by using non‐rhoticity. These Lower East Siders utilize micro‐variation of /r/ in stretches of interview talk, increasing non‐rhoticity when discussing neighborhood topics. Results support a social practice approach to stylistic and sociolinguistic variation, where Lower East Siders use /r/ in constructing a place identity in order to present themselves as authentic neighborhood residents.
What is going on in the Lower East Side, or anywhere else for that matter, is simply an example of what is going on in all human life – a struggle over the very basis of human experience – the need for place as a bedrock of human meaning and social relations.
(Cresswell 2004: 32)
1. WHERE ARE YOU FROM?
In sociolinguistics, places normally serve a typological function. Studies of speech communities rely on geographic placement as an objective division that bounds and unites a group of speakers, and allows us to categorize and analyze the social variation and social meanings within that group ( Johnstone 2004 ). Places circumscribe a speech community, often offering little more than a location for our investigations into social categories and the linguistic practices that correlate with them. As Eckert (2004) notes, descriptions of social variation within class‐based, ethnic, or gendered categories are often disconnected from the places where they occur. We rarely investigate the linkages between speakers and places, knowing little about how speakers negotiate the places we use to define them.
This paper presents sociolinguistic data from speakers who reside on the Lower East Side of Manhattan in New York City. Sociolinguists know this place well; it is the site of Labov's (1966) study establishing the field of variationist sociolinguistics. In The Social Stratification of English in New York City, he describes how New York City English (NYCE) features correlate with the stratification of speakers according to various social characteristics like socio‐economic status, age, and gender, observing both stable variation and linguistic change in progress in this regional variety. The location of the study on the Lower East Side allowed Labov to identify long‐term residents of European ancestry who spoke NYCE, advancing the hypothesis that speakers of this type in an urban community produce regional features and share norms about their usage. Yet the study could conceivably have been conducted in another neighborhood with a similar set of speakers, or in another city altogether. The geography was the location for an investigation into social stratification that would be generalized to the modern urban experience.
Of course, Labov (1966) is unique to New York City because the linguistic variables under analysis come from the repertoire of NYCE. One implicit consequence of place in sociolinguistics is that it often defines the linguistic variables we choose to investigate, particularly in sociophonetics. The recent interest in social practice ( Bucholtz 1999 ; Coupland 2007 ; Eckert 1999, 2003b ; Eckert and McConnell‐Ginet 1992 ; Mendoza‐Denton 2008 ; Podesva 2007 ; Schilling‐Estes 2004 ; Zhang 2005, 2008 ), where speakers actively create and change social meaning through talk, recognizes that social agents do not normally have total linguistic freedom, but select features that are both available and meaningful to them. Features that become available for social practice are often constrained by regionality; these can be local fashion trends, local vocabulary, or local phonology. However, we often analyze these place‐constrained linguistic resources for their contributions to other aspects of identity, whether gender, class, race/ethnicity, or age ( Bell 2001 ; Bucholtz 1999 ; Eckert 1999 ; Fought 1999 ; Le Page and Tabouret‐Keller 1985 ; Schilling‐Estes 2004 ).
The present study identifies a regional dialect feature of NYCE, non‐rhoticity in the syllable coda, and describes the ways that this feature links speakers to their local place. I follow Johnstone (2004) , Eckert (2004) , Modan (2007) , and others in a call for a sociolinguistics of place that complicates the linkages between speakers and the places they are from. The following treatment of NYCE on the Lower East Side brings place and its meanings to the forefront of analysis, operationalizing place as an aspect of identity that can be constructed through social practice.
1.1 Constructing place
My conception of place as an aspect of identity builds on work in cultural and humanistic geography that distinguishes between space and place ( Agnew 1993 ; Cresswell 2004 ). Space refers to spatial organization and location, while place involves the attachment of meaning to space – for instance, by turning a newly inhabited house into a home by moving furniture, adding personal knick‐knacks, and decorating. Cultural geographers understand place as socially constructed –‘in general places are never complete, finished or bounded but are always becoming – in process’ ( Agnew 1993 : 253). If place is itself a social construct, and not just the setting for other identity‐making events, then place has the potential to be part of an individual's identity ( Feld and Basso 1996 ; Mills 1993 ). Myers says:
In sociolinguistics and discourse studies, as in geography, researchers are moving from the assumption that place defines identity, to studies of the ways participants may make place relevant to their identities in situated interactions. (2006: 325)
A growing number of sociolinguistic studies are investigating place and place identity, among them Modan's (2007) study of the Mt. Pleasant neighborhood of Washington, D.C. and the work of Johnstone and her colleagues in Pittsburgh ( Johnstone 2004 ; Johnstone 2009 ; Johnstone, Bhasin and Wittkofski 2002 ; Johnstone and Kiesling 2008 ). While most previous work on place is largely ethnographic and discursive ( Basso 1996 ; Feld and Basso 1996 ; Johnstone 1990, 2009 ; Modan 2007 ; Myers 2006 ; Remlinger 2009 ), Johnstone and her colleagues have utilized quantitative data on a local phonological feature –/aw/ monophthongization – to support the indexical link between that feature and a local Pittsburgh identity. Similarly, Zhang (2008) identifies the ‘imbued social meaning’ of rhotacization in Beijing Mandarin Chinese as indexical of localness. In fact, sociolinguistics has a rich history of linking regional dialect features to localness ( Modan 2007 ), starting with Labov's (1963) analysis of /ay/ and /aw/ centralization on Martha's Vineyard. Those Vineyarders with positive local identities resisted a larger change in progress that lowered the nucleus of /ay/ and /aw/, and increased their rates of this local feature as a form of resistance to social changes on the island.
These are only a few examples of many where speakers use regional features to indicate alignment with local identity. Although studies can often be more concerned with other aspects of a multivalent identity ( Mendoza‐Denton 2002 ), or socially stratified variation within a regional speech community, the speakers themselves in these studies assert a place identity in utilizing regional features that link them to the places they are from. In effect, speakers are engaging in what Bucholtz calls authentication (2003), a process of socially constructing their own ideologies of authenticity. While authenticity is often a researcher‐imposed category privileging the most vernacular data ( Eckert 2003a ), authentication recognizes the practices speakers use to construct the most ‘real’ group in a given place. This acknowledgment that speakers agenticly authenticate their local identities through the use of regional features has been enriched by the introduction of the concept of enregisterment ( Agha 2003 ), a process by which a linguistic repertoire becomes a socially recognized register or dialect. Recent work on enregisterment has identified everything from t‐shirts ( Johnstone 2009 ) to bumper stickers ( Remlinger 2009 ) as items that contribute to the enregisterment of a dialect, linking features to values that are often related to a local identity; or in other words, linking ‘language, people, and place’ ( Remlinger 2009 : 118). I will continue that work here with evidence that acknowledges the work speakers do on the Lower East Side to assert a place identity using non‐rhoticity, a feature imbued with a local social meaning.
1.2 Levels of place
The ensuing analysis of non‐rhoticity on the Lower East Side draws from multiple levels of sociolinguistic evidence to argue that a group of speakers asserts a place identity in utilizing this local phonological feature.
Section 2 presents ethnographically informed notes on the sociodemographic community context of the Lower East Side to argue that rapid social change and community conflict threaten a local group's sense of place, motivating these speakers to assert a place identity in situated talk.
Section 3 focuses on non‐rhoticity and its continued salience as a marker of NYCE. Evidence from popular culture and from Lower East Siders themselves is presented to show that non‐rhoticity remains indexical of a New York identity. In addition, because forty years have passed since Labov described a change from above for non‐rhoticity in New York City, section 4 presents a summary of a variationist analysis of the current status of rhoticity for a group of Lower‐East‐Side speakers. This analysis confirms that non‐rhoticity, while it has decreased, continues to be a robust feature of NYCE, imbued with local meaning and available for social practice and place identity construction.
In section 5 I turn to micro‐variation to describe how an individual speaker's variation in the use of /r/ constructs a place identity. Micro‐variation or practice‐based variation focuses on micro‐shifts in the rate of a linguistic variable as talk unfolds in time ( Mendoza‐Denton 2002 ; Schilling‐Estes 2004 ). We find that the Lower East Siders shift rates of /r/ according to topic, becoming more non‐rhotic when discussing the neighborhood than when discussing non‐neighborhood topics. These place‐related shifts align with a Speaker Design model of style ( Coupland 2007 ; Schilling‐Estes 2002 ) which acknowledges micro‐variation as a resource for speakers in the active construction of identities; in this case, a place identity.
2. COMMUNITY CONTEXT
We have seen in previous work that speakers in communities in the midst of social change are particularly likely to highlight a place identity, as was the case in Martha's Vineyard ( Labov 1963 ) or in the Mount Pleasant neighborhood of Washington, D.C. ( Modan 2007 ). In Pittsburgh, Johnstone notes the importance of identifying the ‘linguistic correlates of the rapid social and economic changes of the last few decades and their effects on people's senses of self and place’ ( Johnstone, Bhasin and Wittofski 2002 : 148). Places in flux, then, motivate place identities to emerge at the forefront of identification practices.
2.1 Contested turf
The Lower East Side has always been an area characterized by transience, beginning in its earliest days as a port of entry for the immigrant poor, with rapid turnover ( Abu‐Lughod 1994 ; Labov 1966 ; Mele 2000 ). In the last forty years, continued changes to the sociodemographic landscape have provoked conflict among residents, which can be characterized in two ways: external conflict resulting from macro‐economic forces of disinvestment 2 and gentrification; and internal conflict between community groups.
The 1970s and 1980s saw a period of disinvestment on the Lower East Side on the part of government agencies and real estate developers, resulting in extreme urban blight, poverty, increased crime, and the displacement of local residents ( Mele 2000 ). But the tide of investment turned quickly, and by the late 1980s came forces of gentrification, mainly due to the neighborhood's location near trendy Greenwich Village ( Abu‐Lughod 1994 ). Local residents now found themselves at risk of another kind of displacement due to the rising cost of living ( Mele 2000 ). Gentrification has only increased in force in recent decades, and is a visible presence on the Lower East Side with upscale bars, restaurants, boutiques, and condominiums dotting the landscape.
While disinvestment and gentrification are quite different economic forces, each serves to place local residents in competition over dwindling resources. Crucially, socio‐economic divisions at the macro‐social level are often perceived and perpetuated as a racialized conflict by community members, who see other ethnic groups as opposition. What was until the latter half of the twentieth century a working‐class amalgamation of various white‐ethnic groups (pre‐Civil War Irish and German; post‐1880 Italian, Russian, Polish, Ukrainian, and Eastern European Jewish) is currently home to a diverse group of ethnic minorities, most notably a large Puerto Rican population who arrived in large numbers post‐World War II, but also African Americans, Asians, and other Latinos. At the same time, artists and counterculturalists, who began entering the neighborhood in the 1950s and 1960s, combine with the gentrifying middle class to produce a large white population, which remains distinct from the descendents of earlier white‐ethnic immigrant groups.
The diversity of the population can be seen in the various terms available for the neighborhood. If naming is one way to lay claim to a place and give it meaning ( Cresswell 2004 ; Feld and Basso 1996 ), then we can see signs of conflict manifested in the many terms for the area (pictured in Figure 1 ), historically considered the northern tier of the Lower‐East‐Side neighborhood ( Abu‐Lughod 1994 ; Mele 2000 ). For Puerto Ricans and other minority groups the neighborhood is Loisaida, a term coined in 1974 by the poets Chino Garcia and Bimbo Rivas ( Mele 2000 ). For the gentrifying urbanites and artists, it is the East Village, a term introduced by the real estate industry to index the status of neighboring Greenwich Village ( Mele 2000 ; Smith 1992 ). This term is also used widely by other New Yorkers to refer to the neighborhood ( Abu‐Lughod 1994 ). For the white‐ethnic residents who are the vestiges of the once vibrant European immigrant community, the area where they live remains the Lower East Side.
Figure 1
Open in figure viewer PowerPoint
A Map of Manhattan, with the Lower East Side shaded. Note that the area under study here is the northern portion.
It is within this context that a small group of white ethnics live out their own struggles over community membership and choose to assert place as a valued aspect of their identity. I call them Lower East Siders because they use this name for their neighborhood, although other groups certainly claim to be authentic Lower East Siders. The speakers described here do have long‐term neighborhood roots – one criterion they use to authenticate themselves amidst a complex community landscape where competition arises over who are the ‘true’ Lower East Siders.
2.2 The Lower East Siders
Because meanings are fluid and change from group to group, looking locally at small communities ensures that we better understand speakers’ own interpretations of local meanings ( Eckert 1995 ; Eckert and McConnell‐Ginet 1992 ). In this spirit, I set out to identify white‐ethnic residents who were born, raised, and still live on the same block where I live in the neighborhood. I was interested in talking with these residents as they are a demographically dwindling group, and I sensed that those who had remained on the block amidst great change were staunchly attached to the neighborhood. I spent time at a community center on the block, where I first met a few of the interviewees, who then introduced me to their siblings and children. Because I chose to focus on such a small locale, I was only able to identify and interview seven residents who met my criteria, despite various tactics of recruitment. The sample of seven speakers (shown in Table 1 ), while small, does allow the ethnographic ‘looking‐locally’ which was my initial goal, although it presents problems for quantitative analysis and generalizability.
Table 1. Demographics of the Lower East Siders
|
|
In light of the community context above, some features of this group are salient. First is their age – of the seven Lower East Siders, five are over the age of seventy. This is representative of the population of white ethnics on the block and nearby. Many of the children of Michael, Lucille, Mae, Ann, and Maggie's generation were raised in the 70s and 80s during the disinvestment period, and have since moved to more desirable locations. Lindsey and Tobias are exceptions to this general rule; there are few native white ethnics below the age of fifty in the area (I was unable to locate any more on our block). Their dwindling numbers may influence another characteristic of the group, that of their self‐professed identity as ‘white’ despite diverse European heritages. They think of themselves more in terms of authenticity than in terms of ethnicity – they are ‘true’ Lower East Siders. Despite earlier ethnic differentiation on the Lower East Side, these differences have been erased ( Irvine 2001 ) so that white ethnics are opposed to newer, non‐white neighborhood groups – Latino, African American, Asian. The white ethnics are united through a shared neighborhood history in which other residents cannot participate. This longevity forms the basis for their claims of authenticity; as Eckert says, ‘authenticity implies stasis – the “real” peasant is just like the peasant that came before’ ( Eckert 2003a : 393). Here Mae recounts seeing an African American woman interviewed during the neighborhood riots against gentrification in the early 1990s:
· Mae: I remember they had this woman on television, interviewing her, African American woman, and she says ‘yes, and these Caucasians who came in here.’[referring to the gentrifiers] I thought ‘oh my lord, we Caucasians were here thousands of years before you were.’ But that's how, so many of them had this attitude, that, this neighborhood belonged to them, and the way they spoke you think that they had lived here all their lives. And, you know, obviously they didn’t.
In authenticating themselves as Lower East Siders, these speakers oppose themselves to newer minority residents. Yet minorities are not the only new residents. Despite similar ethnic traits, the white urbanites flooding the neighborhood as a result of gentrification are also viewed as inauthentic, both because of a lack of time depth and because they appear uninterested in community involvement. Here Tobias expresses frustration at the newer residents in his building:
· Tobias: How can you live, I mean, if I were to suddenly move somewhere, into a new neighborhood, if I didn't know my neighbors, I’d feel isolated. See that's the sad part about the whole deal, these people are so into what they’re doing, they don't take the time to be human with anyone else, or even care about anyone else. Like even my neighbors are pretty much kind of, they’ll say hello briefly, but it's not like they say, ‘hey, come on in for a cup of tea,’ or just sit down and shoot the breeze – no!
Having watched newer groups settle the neighborhood, these white‐ethnic Lower East Siders’ sense of place, their sense of their own status as authentic Lower East Siders, is threatened. As such, they distance themselves from other groups, minorities and gentrifiers alike, a position that leaves them increasingly isolated.
The Lower East Siders’ perception of macro‐economic forces like gentrification is also complicated. As seen in Tobias’ quote, these residents resent newer urbanites encroaching on their ‘urban village’ ( Abu‐Lughod 1994 ). Yet they all live without exception in rent‐stabilized housing, which allows them to also see the benefits of gentrification that increase their quality of life without the risk of displacement. This is a rare stance among neighborhood residents, who often protest neighborhood change with signs such as ‘Gentrification = Class War!’ ( Smith 1992 ). Mae, on the other hand, feels threatened by those who want to stop gentrification:
· Mae: We were the evil ones, especially too when they would talk about the gentrification? They’d rather live in a slum than have them fix up the houses and stuff? That I don't understand, you know?
· Kara: So what do you think about the gentrification?
· Mae: Well, to a degree I, I think you have to give, affordable housing, there's no question about that. I don't think anybody, I can't afford you know, five thousand dollars a month rent or whatever. I just can't do that. But you can't just keep slums, you have to improve things. That's the way I feel, and yes, have it affordable. Or at least, most of it affordable.
A final note on Lower East Siders concerns socio‐economic class. Although a speaker like Michael is a pastor with a graduate degree, while speakers like Tobias and Ann have an elementary education and do not work, I found the assignment of a class index ( Guy 1988 ) difficult and subjective given the small sample. Day to day, the speakers live fairly similar lifestyles that would be considered lower‐middle class. This lack of stratification, as well as the lack of age stratification, is problematic for the statistical results of the influence of social‐category membership on non‐rhoticity in section 4.
The Lower East Siders create a space within their community landscape that isolates them from other residents; they are distinct from minorities, from those poorer than them, from those wealthier than them. In a sense, they are threatened from all sides by neighboring groups they see as ‘other.’ As a result, they turn to a ‘distributed memory’ ( Mendoza‐Denton 2008 : 180) of happier times on the Lower East Side in order to construct a group place identity. As Mele says, ‘through nostalgic memories and everyday practices, older ethnics view themselves as the caretakers of the quickly vanishing old Lower East Side, to which they remain staunchly attached’ (2000: 135). Amidst a community changing economically and socially, the Lower East Siders cling to a remembered place, and it is that Lower East Side that they wish to present to me in our conversations. Within this context I turn to the variation of the linguistic resource /r/, which the Lower East Siders utilize in order to construct a place identity.
3. THE SOCIAL MEANING OF NON‐RHOTICITY IN NEW YORK CITY
Labov's fieldwork on the Lower East Side occurred in the post‐World War II period, a turning point for NYCE due to a general shift in the prestige of Britain in the United States as well as large‐scale population movement during and immediately after the war. This was accompanied by a shift in the prestige of features inherited from or associated with British English, including non‐rhoticity ( Labov 2006 : 225). The pronunciation of /r/ in NYCE at the time showed fine stratification across social classes and sharp shifts in formal styles. This behavior indicated a change from above in apparent time ( Labov 1966 : chapter 9), suggesting New Yorkers were embarking on a change in progress from non‐rhotic to rhotic speech. Further evidence of the change came from the severe linguistic insecurity New Yorkers had about their own speech ( Labov 1966, 1972 ), with the middle class hypercorrecting /r/ in formal styles. There was also external stigma, still present today – NYCE is consistently ranked as one of the ‘worst’ U.S. varieties by speakers of American English ( Niedzielski and Preston 2003 ).
This might suggest that today we would expect to find the change towards rhoticity well advanced or even complete in NYCE. In fact, Labov, Ash, and Boberg note that rhoticity in NYCE appears to have advanced slowly, at a rate of 1.5 percent per year (2006: 47), and is a feature confined to more formal speech. This is quite different from the advancement of rhoticity in other historically non‐rhotic locales in the U.S., like the South, which saw a rapid change to rhoticity in all domains in younger speakers ( Feagin 1990 ; Labov, Ash and Boberg 2006 ). Little work has followed the progress of the change in /r/ in New York City ( Fowler 1986 ), but it has corroborated this slower rate of advancement. The data here also provide evidence that, while rhoticity continues to increase, it does so slowly. The result is that non‐rhoticity is still available as a resource for NYCE speakers.
3.1 Non‐rhoticity as a socially imbued New York variable
Section 4 will show that non‐rhoticity is still used by this sample of Lower East Siders. However, observations of production alone cannot establish that non‐rhoticity is imbued with social meaning, particularly a local meaning as I argue here. Further evidence is needed, especially given the fact that non‐rhoticity is not a feature used only in New York City. It was once used in cities along the eastern seaboard with ties to London ( Labov, Ash and Boberg 2006 ) and is still used today robustly in Eastern New England as well as New York City ( Irwin and Nagy 2007 ; Labov, Ash and Boberg 2006 : 48). This does not mean, however, that NYCE speakers cannot and do not index their local dialect using a feature that may have a different social meaning in another locale ( Eckert 1995 ).
There is much evidence that both New Yorkers and non‐New Yorkers alike do identify non‐rhoticity as a salient feature of NYCE, one that (in combination with other NYCE features or even alone) can index a New York persona ( Eckert 2003b ). The Lower East Siders in Labov's (1966) study certainly viewed non‐rhoticity as salient in New York City, as shown through measures of linguistic insecurity and meta‐linguistic commentary ( Labov 1966 ), while stereotypes of NYCE at the time were full of non‐rhotic shibboleths (like the classic phrase ‘toity toid and toid’ for thirty third and third). The question remains if non‐rhoticity continues to index ‘New York‐ness’ today.
While an entire article could outline the combination of historical, cultural, ideological, and linguistic practices that together allow for the enregisterment ( Agha 2003 ; Johnstone 2009 ; Remlinger 2009 ) of a dialect like NYCE, a quick look at popular culture and the media shows that non‐rhoticity continues to participate in the indexing of the dialect. One example linking non‐rhoticity to New York City in popular circulating discourse is the orthographic variability in the spelling of the city's name, found online, on bumper stickers, in tourist materials, and a variety of other media. A Google search, often a fruitful site for current popular discourse ( Hill 2005 ), found over 800,000 hits for the phrase ‘new yawker;’ over 100,000 for ‘new yawk,’ and over 10,000 for ‘new yawkah,’ which appropriately utilizes both non‐rhotic environments. Entries on the popular website Urban Dictionary, where individuals can supply their own popular definitions for terms ( http://www.urbandictionary.com ) show that non‐rhoticity is often central to definitions of New Yorkers. One lists as part of the definition of ‘New Yorkah:’‘This breed of American absolutely refuses to pronounce an “er” sound if it appears at the end of a word’ (and this is the only linguistic characteristic mentioned).
These are just a few examples which indicate that non‐rhoticity popularly retains a social meaning that indexes New York City. Importantly, the Lower East Siders also appear to have this socially‐imbued meaning for non‐rhoticity. Although I did not explicitly elicit meta‐linguistic commentary on /r/, one speaker brings it up spontaneously during the word list reading when comparing source and sauce:
· Maggie: Source
Maggie's meta‐linguistic commentary provides evidence that the Lower East Siders continue to view non‐rhoticity as a salient marker of a New York identity, although there is no other discursive data from the seven interviews, nor any indication of whether these speakers still view non‐rhoticity as a source of linguistic insecurity ( Labov 1966, 1972 ). My own experience from continued fieldwork on the Lower East Side is that non‐rhoticity continues to be used to index a New York City identity by white‐ethnic speakers, while at the same time speakers who use the variable are less insecure about expressing their ‘New York‐ness.’ This anecdotal impression is in line with what we see here from the Lower East Siders, who not only continue to use non‐rhoticity, but use it to assert a positive place identity at the local level.
4. QUANTITATIVE ANALYSIS
A brief summary of a multivariate analysis of variation in non‐rhoticity is presented here to provide current quantitative evidence that non‐rhoticity is still involved in a change in progress in NYCE, albeit a slow one.
The data for /r/ come from sociolinguistic interviews with the seven Lower East Siders, which include interview speech, a reading passage, and a word list. The quantitative analysis presented here utilizes data from all three speech styles (with the exception of the topic factor group, which uses only interview speech data, and the emphatic stress factor group, which excludes word list data). Each interview lasted from one to one‐and‐a‐half hours and was conducted in the interviewee's home on our block. The total corpus across seven speakers consists of 3,000 words with syllable‐coda /r/, which were perceptually coded for rhoticity/non‐rhoticity (hereafter referred to as [r‐1] and [r‐0] respectively). Dubious cases were coded but excluded in this analysis. In order to verify the reliability of coding, two additional listeners coded 193 random words (six percent of the corpus). The three coders show an overlap in perceptual coding of 90 percent. I produced an [r] index for each speaker by dividing the total [r‐1] by the total number of instances in which /r/ could vary.
The variable [r] can be produced as non‐rhotic whenever it is in syllable coda position. This excludes non‐rhoticity in words like tyrant or lottery, where /r/ is in the syllable onset, as well as in words like hiring, where /r/ is reanalyzed as a syllable onset following the addition of a suffix. However, it includes word‐final intervocalic /r/ or linking /r/, as in phrases like matter of fact, where speakers can in fact be non‐rhotic (
Irwin and Nagy 2007
). Both lexical and functsion words are included in the corpus, although function words are limited to a restricted set of words and coded only when they are not reduced: here, there, where, were, our and her. Finally, the stressed nucleus schwa [
Completely non‐rhotic as late as 1890 ( Babbitt 1896 ), NYCE is slowly becoming rhotic as part of a general rhotacization of all American English dialects that were non‐rhotic ( Bonfiglio 2002 ; Feagin 1990 ; Labov 1966, 2006 ; Labov, Ash and Boberg 2006 ; Parslow 1971 ; Schonweitz 2001 ). In 1966, Labov found rates of [r‐1] as low as zero for working‐class speakers in casual speech. Today, the Lower East Siders are, on average, 36 percent rhotic. The following discussion highlights some salient factors, both internal and external, that constrain r‐pronunciation for the Lower East Siders. Table 2 shows the factor groups that are statistically significantly predictors of the realization of /r/. 4
Table 2. Statistically significant factor groups for [r‐1], from Goldvarb X (Sankoff, Tagliamonte and Smith 2005)4
|
Factor Group |
Factor |
Percent [r‐1] |
N |
Factor weight |
|
Preceding vowel* |
[i] |
42 |
278 |
.44 |
|
|
[e] |
40 |
260 |
.41 |
|
|
[ |
47 |
480 |
.61 |
|
|
[o] |
42 |
592 |
.49 |
|
Following context |
consonant |
33 |
2269 |
.44 |
|
|
pause |
30 |
405 |
.41 |
|
|
vowel |
78 |
326 |
.91 |
|
Syllabic stress |
unstressed |
29 |
1318 |
.40 |
|
|
stressed |
44 |
1682 |
.58 |
|
Emphatic stress |
no |
36 |
2670 |
.49 |
|
|
yes |
42 |
206 |
.59 |
|
Topic |
neighbourhood |
33 |
1385 |
.47 |
|
|
other |
39 |
916 |
.55 |
|
Speech style |
interview |
35 |
2304 |
.46 |
|
|
reading |
41 |
572 |
.58 |
|
|
word list |
67 |
124 |
.78 |
|
Age |
older (70+) |
29 |
2054 |
.39 |
|
|
younger (45–55) |
57 |
946 |
.72 |
|
Class |
6–8 |
34 |
1462 |
.47 |
|
|
2–4 |
42 |
1538 |
.53 |
· *Only tokens with /r/ in stressed position were included in the preceding vowel category. In addition, /r/ preceded by /u/ was removed from analysis because of a very small N (50).
4.1 Phonological factors
One salient phonological factor that constrains r‐pronunciation is stress. When /r/ is in stressed position, [r‐1] is favored, and within that stressed syllable, a preceding vowel further constrains the realization of /r/. In addition, unstressed syllabic environments
5
provide a window into change towards rhoticity, as previous studies of /r/ have described this environment as the most likely to disfavor [r‐1] (
Feagin 1990
;
Irwin and Nagy 2007
;
Miller 1998
). In her work on rhoticity among Southern Whites, Feagin argues that [
Another finding in line with previous research is that the use of emphatic stress (‘Not the bar, the car!’) favors [r‐1], as it did in Labov's department store study ( Labov 1966, 1972 ), which showed higher rates of [r‐1] in the repeated, emphatic phrase fourth floor. Finally, as noted above, although previous studies have excluded contexts where word‐final /r/ is followed by a vowel across a word boundary (another apartment), arguing that [r‐1] is categorical in this context, here [r‐0] is realized in 22 percent of words followed by a vowel. Irwin and Nagy (2007) also find instances of non‐rhoticity in the context of a following vowel in their corpus of Bostonians.
4.2 Intra‐speaker factor groups
The results for speech style (which distinguishes interview, reading, and word list speech) show stratification according to formality ( Labov 1966 ). Interview speech disfavors [r‐1], with a factor weight of .46, while both reading (factor weight .58) and word‐list speech (factor weight .78) favor [r‐1]. In addition, the topic factor group, which splits interview speech into neighborhood versus non‐neighborhood topics, finds that non‐neighborhood topics favor [r‐1] (factor weight .55) and neighborhood topics disfavor it (factor weight .47). This finding will be addressed in detail in section 5 as evidence of social practice and the construction of a place identity.
4.3 Inter‐speaker factor groups
Despite a small sample, the data from The Social Stratification of English in New York City ( Labov 1966 ) provide a point of comparison in real time for the change in r‐pronunciation in NYCE over the last forty years. Labov does not report an overall rate of [r‐1] for his speakers, which we could compare here to an overall rate of 36 percent. Instead, he focuses his analysis on the stratification of speakers across class boundaries, a salient categorization that motivates him to suggest that rhoticity is a change from above (1966). With only seven speakers in the present study, we cannot examine any fine stratifications by class. 6 Instead, age emerges as a strong predictor of r‐pronunciation. Older speakers (70 years and older) disfavor [r‐1] with a factor weight of .39, but use more [r‐1] than the oldest age group reported in 1966 (55+), which supports change. Younger speakers (here two speakers, one aged 45 and one aged 55) are even more distinct from their 1966 equivalents with an average [r‐1] of 55 percent, favoring [r‐1] with a factor weight of .72. In contrast, Labov's lower‐middle‐class speakers aged 40+ in 1966 used just under 21 percent [r‐1] in careful speech. The speakers here have more than doubled the rate of [r‐1] of their 1966 age equivalents.
Another way to compare these data with those from (1966) is in real time: the five Lower East Siders who are aged 70+ would have been in their 30s at the time of Labov's study. In 1966, lower‐middle‐class speakers aged 30–39 used 15 percent [r‐1] in careful speech ( Labov 2006 : 223). This same demographic now uses 26 percent [r‐1] in interview speech. This may indicate nothing more than change in speech across the lifespan of adults ( Sankoff and Blondeau 2007 ) but it is in the direction of an increase in [r‐1]. Labov acknowledges this as a possibility, saying:
We would expect that those who show the most linguistic mobility in their stylistic shifts would be most likely to show such an overall shift as they advance in years. The lower‐middle‐class speakers would be pre‐eminent in this respect. ( Labov 2006 : 202)
The two younger speakers here, Lindsey and Tobias, would have been children in 1966, so a direct comparison cannot be made.
Again, the small sample size of the current data calls for hesitation in generalizing conclusions about a change in progress with respect to Labov (1966) . Still, the results suggest that NYCE speakers continue to increase their rhoticity. At the same time, this change is in no way complete. The younger speakers Lindsey and Tobias use [r‐1] just more than half of the time, while the older speakers use it only a third of the time. Non‐rhoticity still remains a feature of NYCE, a resource available for social practice and one that may be used to indicate a place identity due to its salience as a local sociolinguistic marker. In the next section we will see one way that the Lower East Siders use non‐rhoticity in situated interaction to construct a place identity.
5. MICRO‐VARIATION AND THE CONSTRUCTION OF PLACE IDENTITY
A quantitative analysis of /r/ shows that non‐rhoticity is still a robust feature of NYCE, and section 3.1 confirms that it is indexically linked to a local New York identity. Here, I argue that non‐rhoticity operates on a more local level for the present group, indexing a Lower‐East‐Side place identity. In section 2 we saw that a tumultuous community context threatens the status of the Lower East Siders and their ‘authentic’ neighborhood. The relationship between non‐rhoticity as an NYCE feature and non‐rhoticity as used by the Lower East Siders is a recursive one, where speakers take an opposition that exists at one level, in this case New York vs. non‐New York, and project it onto another level, in this case Lower East Side vs. non‐Lower East Side. Non‐rhoticity, used in the wider community to distinguish NYCE speakers from others, is now used recursively in the neighborhood to distinguish community members ( Gal and Irvine 2000 ; Irvine 2001 ). The result is the assertion of a Lower‐East‐Side place identity manifested through individual speaker's style‐shifts in r‐pronunciation.
To demonstrate the recursive use of non‐rhoticity at the local Lower‐East‐Side level, I return to the quantitative results that showed a significant effect of neighborhood vs. non‐neighborhood topics on the realization of /r/ (see Table 2 ). I use topic as a way to identify discourse that is concerned with place, with the hypothesis that speakers will use more [r‐0] when discussing neighborhood topics as they construct a place identity. Similarly, Schilling‐Estes (2004) , Rickford and McNair‐Knox (1994) , and Eckert (1995) analyze topic shifts in the sociolinguistic interview, identifying the ‘key cultural content’ ( Eckert 1995 : 58) that evokes the usage of a socially‐meaningful variant. A large portion of my sociolinguistic interviews centers on neighborhood talk, as I presented this as my main interest in conducting interviews with my neighbors. I facilitated neighborhood talk by asking specific questions; speakers also initiated many topic shifts towards neighborhood topics. Following the interviews, I coded the speech into general topics (i.e. childhood memories, family relations, opinions of gentrification, etc.), listening broadly for changes in topics (so that some topics, like ‘childhood memories,’ might last for as long as ten minutes), using my own questions and shifts from the interviewees as cues to change of topic. I split these topics into two categories: ‘neighborhood’ and ‘non‐neighborhood’. This process was repeated by a third party to control for the possibility of my own bias; our coding showed an overlap of 96 percent. Examples of neighborhood and non‐neighborhood topics are given below.
Examples of topics considered neighborhood talk (58 percent of corpus):
· •
Housing talk (discussions of housing availability, rent, landlords, etc.)
· •
Local food (what restaurants in the neighborhood do you like? where do you order from?)
· •
Changes in the neighborhood over time (drugs, gentrification, tenement life, etc.)
· •
Narratives that are situated on the block (e.g. a neighbor who is robbed in her apartment)
· •
Memories of the block (games on the street, sleeping on fire‐escapes, entertainment, etc.)
Examples of topics considered ‘non‐neighborhood’ talk (42 percent of corpus):
· •
Food (general discussion of food – what kind of food do you cook? do you like?)
· •
Spirituality
· •
Career
· •
Schooling (Catholic vs. private vs. public education) 7
· •
Asides (on the phone or to others present)
· •
Tangents not related to local topics
To illustrate the topic coding, here is a short example from Lindsey, who discusses a topic that is important to her – change in the neighborhood. I coded this excerpt and the discussion surrounding it as a neighborhood topic, but using the Decision Tree ( Labov 2001 ) it was coded as careful speech (see section 5.1 for a discussion of these two approaches to interview speech).
Example 1
|
Lindsey: |
It really has changed, and, I just never[r‐0] thought it would get that high. But it has. And I’m forty‐five[r‐0], so I didn't really anticipate things getting so expensive. |
|
Kara: |
Do you find, being in the neighborhood hard, I don't know groceries, and things like that? |
|
Lindsey: |
Well we only have, really have like one super[r‐1]market[r‐0]. |
|
Kara: |
Where do you guys go now when you go out, I don't know restaurants or, [other places to go. |
|
Lindsey: |
[We or[r‐1]der[r‐0] in. (laughs) Well we or[r‐0]der[r‐0] from Odessa's and that's a landmark[r‐0], that's been here for years. |
Lindsey's overall rate of [r‐1] is 60 percent, but in this excerpt, her rate of [r‐1] is 22 percent.
As we saw in Table 2 , when the Lower East Siders speak about topics related to their neighborhood, they disfavor [r‐1], with a factor weight of .47. When speaking about other topics, they favor [r‐1] with a factor weight of .55. One way to see the Lower East Siders constructing place identity through neighborhood talk is to track levels of [r‐1] within one sociolinguistic interview. Figure 2 shows the [r‐1] levels for Michael across topics, which are bounded into neighborhood and non‐neighborhood topics.
Figure 2
Open in figure viewer PowerPoint
Michael's levels of [r‐1] across topics
Michael varies his rate of [r‐1] so that it is consistently lower when he discusses neighborhood topics than when he discusses other topics like his career, family relations, or spirituality. In fact, his highest rate of [r‐1] for a neighborhood topic is still lower than his lowest rate of [r‐1] for a non‐neighborhood topic. Topic is statistically significant as a predictor of his use of [r‐1]. The same is true for a younger speaker like Tobias (see Figure 3 ). His overall rate of [r‐1] is 55 percent, yet we see him drop down to a low of 39 percent [r‐1] when discussing his housing troubles on the block.
Figure 3
Open in figure viewer PowerPoint
Tobias’ levels of [r‐1] across topics
5.1 Style‐shifting and place identity
Micro‐shifts in r‐pronunciation accomplish what proponents of social practice would call the creation of a persona ( Eckert 2003b ; Podesva 2007 ); in this case, the persona authenticates the speakers as ‘true’ Lower East Siders by projecting a place identity. The construction of identity categories through social practice, in particular through micro‐variation in situated interaction, aligns with the Speaker Design model of style, where ‘stylistic variation is viewed not as a reactive phenomenon but as a resource in the active creation, presentation, and recreation of speaker identity’ ( Schilling‐Estes 2002 : 388). Here, I argue that non‐rhoticity has a socially‐imbued local meaning ( Zhang 2008 ), and speakers actively draw on that meaning when they shift.
This is not to say that other mechanisms for style‐shift are not present in the data; I believe none of the motivations for style‐shift are mutually exclusive. 8 We saw that the Lower East Siders increase [r‐1] when switching to more formal styles like reading passages and word lists (section 4.2), so certainly the Attention to Speech model ( Labov 1966, 2001 ) is operating. However, within interview speech, the casual/careful distinction laid out in Labov (2001) does not significantly predict r‐pronunciation. I tested this by coding my data according to Labov's (2001) Decision Tree. The casual/careful and neighborhood/non‐neighborhood categories are fairly orthogonal, meaning that codes for /r/ that agree as neighborhood/casual and non‐neighborhood/careful account for only 57 percent of tokens, so that there is not considerable overlap between the two groups. I ran two separate but otherwise identical GOLDVARB runs using the casual/careful distinction in one, and neighborhood/non‐neighborhood in the other. The run using the Decision Tree coding did not select the casual/careful distinction as significant, while the run using neighborhood/non‐neighborhood topics selected this group as a significant predictor of r‐pronunciation. This provides further evidence that there is an aspect of agency in the Lower East Siders’ shifts; more than relaxing into vernacular speech, they are actively engaged in meaning‐making ( Bell 2001 ; Eckert 2001 ).
The Attention to Speech model of style takes as its point of entry the linguistic variable, seeking to understand why it shifts. In Speaker Design, the point of entry is the speaker, with a more ethnographic approach to style that allows the speakers to tell us what is meaningful by watching what they do through micro‐variation. The style work of the Lower East Siders is a form of social practice where speakers increase their non‐rhoticity to assert a place identity that indexes ‘Lower East Side.’ We see this in the stretch of Lindsey's talk in Example 1 above, concerned not with a greater New York identity, but focused on extremely local topics related to neighborhood change. By taking a feature that indexes New York City on the whole and applying it recursively in their local context, the Lower East Siders distance themselves from other neighborhood residents who do not use non‐rhoticity as a resource indexing localness, whether Latinos, Asians, or white urbanites. 9 As described in section 2, conflict in the Lower East Side over place identity emerges as opposition between both class and ethnic groups. Yet neither of these categories alone is enough to account for these speakers’ identification practices; instead, speakers react to multiple oppositions in order to retain an identity that is decidedly local. As a salient marker of NYCE, non‐rhoticity allows speakers to authenticate their claims that the Lower East Side belongs to those with roots in the neighborhood; i.e. those who use NYCE speech. By manipulating /r/, these speakers assert their own local meaning of the neighborhood as Lower East Side, not Loisaida, not East Village. They resist what they see as a loss of the ‘authentic’ Lower East Side they once knew.
6. CONCLUSIONS
I have argued here that individual speakers use non‐rhoticity to construct a place identity through micro‐variation in the use of that feature in a local context. I have also argued that the motivation for these agentic style‐shifts lies in part in the social landscape of the Lower East Side and conflict over place. By increasing non‐rhoticity, the Lower East Siders index a social meaning that authenticates them as ‘true’ residents of the Lower East Side ( Coupland 2003 ). The use of non‐rhoticity or any other NYCE feature as a resource for local authentication practices is certainly not a random choice; these are the very resources that speakers can utilize to construct a place identity and contribute to the enregisterment of their local dialect ( Agha 2003 ).
Authentication practices and place identity construction that arise from speakers’ desires to project an authentic local persona may have implications for language change. If the most authentic speakers in a place are those with the longest history ( Eckert 2003a ), they may be those that maintain local linguistic features that are linked to historicity ( Coupland 2003 ). The micro‐variation of /r/ in section 5 shows that the Lower East Siders increase non‐rhoticity to authenticate their status as Lower East Siders. One question worth asking is if these micro‐practices will bleed into macro‐level rates of variation as speakers like the Lower East Siders attempt to resist what they see as social threats to their place identity. To maintain an authentic claim on the Lower East Side or even a New York identity, speakers might maintain or increase their non‐rhoticity. This proposal is certainly not borne out in the current data, and would contradict the implications of work on NYCE that suggest an advancement of the change in progress for /r/. Yet we have observed here that this change in progress is incredibly slow, quite different from other areas of the U.S., like the South, where a change to rhoticity occurred quickly ( Labov, Ash and Boberg 2006 ). Perhaps speakers like the Lower East Siders have contributed to this slow advancement due to a desire to assert their place identities. Johnstone, Bhasin and Wittofski state:
A number of recent studies have suggested that, at least to a certain extent, the leveling forces of increased dialect contact … which encourage people to sound more like people elsewhere via linguistic accommodation, may be counteracted by attempts to cling to local identity by preserving at least one or two features that sound local. ( Johnstone, Bhasin and Wittkofski 2002 : 160)
There is some evidence of this in Johnstone's work, where working‐class Pittsburghers maintain /aw/ monophthongization ( Johnstone, Bhasin and Wittkofski 2002 ). On Martha's Vineyard, young locals with positive orientation towards the island used the highest rates of centralized /ay/ and /aw/, despite an expected trend towards lowering of the nucleus ( Labov 1963 ); this was in effect a reversal of the trend in the New England dialect community towards lowering. Speakers still centralize /aw/ and /ay/ today in the face of continued threats from tourism on the island ( Blake and Josey 2003 ; Pope, Meyerhoff and Ladd 2007 ).
In New York City, non‐rhoticity persists as a socially imbued marker of localness – one that speakers access in their attempts to assert local authenticity. Continued study of non‐rhoticity across time in NYCE will allow us to watch the progression of a locally‐meaningful variable which individual speakers utilize to make place identity salient.
Labov’s big new York study: