• Non ci sono risultati.

Touring across American English: A contrastive phonetic analysis of the New York City, Californian, and Southern Accents of American English Corpus (NCSAAE).

N/A
N/A
Protected

Academic year: 2021

Condividi "Touring across American English: A contrastive phonetic analysis of the New York City, Californian, and Southern Accents of American English Corpus (NCSAAE)."

Copied!
213
0
0

Testo completo

(1)

The word is a flame burning

in a dark glass.

(2)

Table of contents

Introduction ...1

CHAPTER 1 Background to the study ...9

1.1 Introduction ...9

1.2 Dialectology: preliminary remarks ...10

1.3 From dialect geography to sociolinguistics ...13

1.3.1 A Word Geography of the Eastern United States ...14

1.3.2 Innovative approaches to dialectology ...17

1.3.3 Generative dialectology ...20

1.3.4 Accents of English - an overview ...21

1.4 Social dialectology ...23

1.4.1 A few notions of sociolinguistic methodology ...26

1.4.2 The Martha’s Vineyard study ...30

1.4.3 The Social Stratification of English in New York City ...32

1.5 The Linguistic Atlas of North American English ...36

1.6 The three perceptual dialects of North America ...46

1.7 Mainstream perception of American English through time ...49

1.8 General American ...52

1.9 The development of English in North America ...61

1.9.1 Lexical attestations ...69

1.9.2 Phonetic evidence ...74

CHAPTER 2 Building a specialized corpus of telecinematic speech ...80

2.1 Introduction ...80

(3)

2.3 The language of films and television series ...82

2.4 The creation of the NCSAAE Corpus ...87

2.4.1 Data sources ...89

2.4.2 Editing the data ...102

2.5 Methodology: an integrated approach 105

2.6 Comparative analysis 107

CHAPTER 3 Analysis of accent features in the NCSAAE Corpus ...110

3.1 Introduction ...110

3.2 California English ...111

3.2.1 Lowering of short front vowels ...116

3.2.1.1 Theoretical background ...116

3.2.1.2 Lowering of short front vowels in the NCSAAE Corpus ...118

3.2.2 The low-back merger in California English ...124

3.2.2.1 Theoretical background ...124

3.2.2.2 The low-back merger in the NCSAAE Corpus ...126

3.2.3 Fronting of the mid-high back and high back vowels ...129

3.2.3.1 Theoretical background ...129

3.2.3.2 Fronting of the mid-high back and high back vowels in the NCSAAE Corpus ...131

3.2.4 Summary ...132

3.3 New York City English ...134

3.3.1 The short-a split system ...139

3.3.1.1 Theoretical background ...139

3.3.1.2 The short-a split system in the NCSAAE Corpus 142 ... 3.3.2 BOUGHT raising ...147

(4)

3.3.2.1 Theoretical background ...147

3.3.2.2 Raised BOUGHT in the NCSAAE Corpus ...150

3.3.3 TH/DH stopping ...153

3.3.4 Variable rhoticity ...155

3.3.5 Two accent peculiarities in the English of NYC ...156

3.4 English in the Southern United States ...159

3.4.1 The Southern Shift ...163

3.4.1.1 Theoretical background ...163

3.4.1.2 The Southern Shift in the NCSAAE Corpus ...167

3.4.2 The Back Upglide Chain Shift ...177

3.4.2.1 Theoretical background ...177

3.4.2.2 The Back Upglide Chain Shift in the NCSAAE Corpus ...180

3.4.3 Summary of findings ...182

Final remarks ...184

References ...194

(5)

List of figures

CHAPTER 1

Figure 1.1 Scale for (aeh) index ...28

Figure 1.2 Class stratification of (æh) ...29

Figure 1.3 Phonological environments favoring centralization on Martha’s Vineyard ...32

Figure 1.4 Class stratification of r ...35

Figure 1.5 Directions of movement in chain shifts along peripheral and non-peripheral tracks ...40

Figure 1.6 The low back merger in Indianapolis ...43

Figure 1.7 Split /æ/-/æh/ system of Rosanne V. ...45

Figure 1.8 Clustering solution for the three main perceptual dialects of North American English...47

Figure 1.9 Perceptual distances between American English dialects in a two-dimensional space ...48

CHAPTER 2 Figure 2.1 Corpus composition ...100

CHAPTER 3 Figure 3.1 Lowering of short front vowels in CE ...116

Figure 3.2 The Western vowel system of Ernest P. ...124

Figure 3.3 The California Vowel Shift ...133

Figure 3.4 Iconographic representation of NYCE raised /ɔ/ ...138

Figure 3.5 Split /æ/-/æh/ system of Nina B. ...140

Figure 3.6 Codas inside the box condition tense /æ/ in New York City 141 .... Figure 3.7 Long and ingliding vowels of a Telsur informant ...148

(6)

Figure 3.9 Iconographic representation of common mistakes

resulting from the Mary-marry-merry merger ...157 Figure 3.10 The Southern Shift ...164

(7)

List of maps

CHAPTER 1

Map 1.1 The merger of /o/ and /oh/ ...12

Map 1.2 Southeastern New England lexical isogloss ...16

Map 1.3 Location of the 69 informants on Martha’s Vineyard ...31

Map 1.4 An overall view of North American dialects ...38

Map 1.5 Extent of spacial diffusion of the Low Back Merger ...42

Map 1.6 Transitional low-back merger ...43

Map 1.7 The three major speech areas, as noted before the 1940s ...52

Map 1.8 The Mary-marry-merry merger ...56

Map 1.9 The horse-hoarse merger ...57

Map 1.10 The merger of /iw/ and /uw/ ...58

Map 1.11 Glide Cluster Reduction ...60

Map 1.12 Early English-speaking settlement areas in America ...64

Map 1.13 The Appalachian Region ...66

Map 1.14 The Old Northwest Territory ...67

Map 1.15 Geographical distribution of the term hasslet ...71

Map 1.16 Geographical distribution of the term sook ...74

CHAPTER 3 Map 3.1 An overall view of North American dialects ...112

Map 3.2 Differential fronting of /oʊ/ in North American English ...130

Map 3.3 New York City short-a system ...135

Map 3.4 Glide deletion of /aɪ/ ...160

Map 3.5 Core areas of the linguistic South ...161

Map 3.6 The second stage of the Southern Shift ...166

(8)

List of tables

CHAPTER 2

Table 2.1 Movies in the NCSAAE Corpus ...81

Table 2.2 Television series in the NCSAAE Corpus ...81

Table 2.3 Reality shows in the NCSAAE Corpus ...82

Table 2.4 Cast of Donnie Brasco ...92-93 Table 2.5 Cast of The Apostle ...93

Table 2.6 Cast of Gone in 60 Seconds ...94

Table 2.7 Cast of Law and Order ...95

Table 2.8 Cast of Memphis Beat ...95

Table 2.9 Cast of NCIS: Los Angeles ...96

Table 2.10 Cast of Jersey Shore ...97

Table 2.11 Cast of Duck Dynasty ...97-98 Table 2.12 Cast of Keeping up with the Kardashians...98-99 Table 2.13 List of lexical tokens subjected to phonetic evaluation ...105-106 Table 2.14 Subcomponents of the NCSAAE Corpus ...108

(9)

Introduction

English is destined to be in the next and succeeding centuries more generally the language of the world than Latin was in the last or French is in the present age. (John Adams, 1780)

In the fast-paced ever-changing reality of the twenty-first century, people are barraged with a boundless number of products that the world of the film industry continuously ‘churns out’, be they movies, television series, or the recently emerging trend of tawdry reality shows. Although the overwhelming popularity of recent blockbusters, such as the Harry Potter saga or the convoluted adventures of the aristocratic Crawley family in the box-office success Downton Abbey, has brought under the spotlight British English and, in particular, its sophisticated and smart-sounding accent, it is undeniable that American English has gained the attention of the general public since the inception of the film industry in the early 1900s. Even when films were silent, the American English of the captions aroused British anxiety; “it is a lingua franca, or a lingua californica” (Bailey, 2001: 495), bewailed English purists concerned about the growing predominance of American linguistic fashions in Britain and—we might add with the advantage of hindsight—across the globe. As improbable as Adam’s prediction in the above epigraph might have seemed at the time, it was accurate. Indeed, the influence of the United States has advanced the use of its language beyond that of any other language in human history.

Given the impact of the American variety of English internationally, the primary question that sparked off my interest in carrying out a study based on dialectal variation across the United States is whether the language to which people living abroad are exposed, mainly through the medium of television,

(10)

coincides with its usage on the part of native speakers. Is there such a thing as a default and exemplary state of American English? If so, what is its degree of ‘authenticity’ with respect to the manifold linguistic texture of North America? These are the nagging queries which prompted me to investigate the intricate and thorny subject of American dialectology; I hope that the present work will shed some light on the diversity and variation of U.S. English by examining the cultural framework and the most distinctive phonetic features of three dialectal regions of the country: California English, New York City English, and Southern English.

My first experience in the United States, a breathtaking ten-day tour along the Eastern Coast from Boston to Ocean City (MD), brought me into contact with an increasing quantity of variegated speech patterns the further away I travelled from my initial location in Massachusetts: I can still recall when, upon my arrival at Harvard University for a guided historical visit to its prestigious Yard, I happened to hear a local guard saying to our tour operator, with an inflection reminiscent of British dropped /r/, “Ya cahn’t pahk ye cah in Hahvahd yahd”; or, while standing in a queue in one of the endless flamboyant bars on 14th street, Manhattan, I was impressed by the peculiar and quaint accent of New Yorkers asking for a “cup of cawfee”. I was exposed at that 1 time to a tortuous ‘linguistic Babel’, made of oral sounds, which not only bewildered me but also kindled my interest to delve thoroughly into the field of U.S. dialects.

Prior to this, my knowledge of American English phonetics was limited to the vague notion of a relatively homogeneous and generic pronunciation to which I had grown accustomed through the media dimension. However, this latter idea, despite being quite a perfunctory perception, was not completely

The amateurish transcriptions reflect my initial impressions of direct contact with locally

1

(11)

detached from the mainstream view of American English which still dominates outside the borders of the country. As a concept, it corresponds to the layman’s perception (mine included) of an American accent once its local flavor has been removed, being the variety most acceptable on the television networks covering the whole of the United States. Hence, the term Network English, coined by Wells in 1982 but somewhat superseded nowadays by the less biased and authoritative Standard or General American English. In this regard, Metcalf (2000) observes, with his witty and entertaining flair, that the widespread use of the label General American (hereinafter GA) for a presumed most common form of American English essentially designates the colorless and unmarked speech of the Midwest:

Also of interest and importance is the residue, the common core, something that has been called “General American”: […] neutral speech, closest to that of parts of the North and much of the West, but without a sense of place. To say it has no accent isn’t fair. It has an accent, just one that doesn’t tell where you’re from. […] It’s the language also of national broadcasts and publications and thus is familiar to almost all Americans (p. X).

Metcalf’s implication that General American corresponds to the alleged ‘nonpartisan’ speech of the vast Western midsection triggered me to examine the legitimacy of such a straightforward and clearcut assumption. In accomplishing this task, I collected a sizable corpus of filmic material, differentiated into three subgenres (movies, television series and reality shows), which features the easily recognizable varieties of California English, New York City vernacular and Southern English. The disparate nature of the sources under scrutiny will also allow me to assess the potential impact of genre variation on the retention of well-defined aspects of regional pronunciation.

(12)

The reason why I decided to investigate the accent of California is that this region has a place of pride among dialects of the West; its language is amplified across the whole country thanks to the nerve center of American entertainment in the metropolitan area of Los Angeles. From the heights of Beverly Hills to the depths of San Fernando Valley, the “California word is broadcast to America and the world” (Metcalf, 2000: 134). Given its significance as the central hub of the telecommunications system, it goes without saying that California sets the tone for American media and Western American English. As stated above, the primary goal of the ensuing discussion is to observe the extent to which California English conforms to the often quoted phonetic structure of GA. Thus, in Chapter 1, I will provide some general information about the most salient features of the ‘all-encompassing’ variety of Standard American English. Formulaic and trite though it may be, a similar broad overview is unavoidable in order to question the stereotyped view of the Western accent as the perfect and exemplary state of canonical pronunciation. With the aid of corpus linguistic tools, I will retrieve symptomatic linguistic items in order to illustrate the relevant characteristics of California English pronunciation and, after describing them from a phonetic standpoint, I will proceed to ascertain their relation to the structure of GA sketched in Chapter 1. I expect to find that, although the California accent overlaps with GA in many respects, it nonetheless retains phonetic peculiarities which call into question the soundness of any attempt to set the norms of a default form of American English. Indeed, my ultimate objective is to corroborate Kretzschmar’s (2004) view of GA as a continuum of accents which is not modeled on the pronunciation of a specific geographic area:

Standard American English can best be characterized as what is left over after speakers suppress the regional and social features that have risen to salience and become noticeable. Decisions about which features are

(13)

perceived to be salient will be different in every region, even different for every speakers […]. When speakers travel outside of their native region, aspects of their pronunciation that are perfectly standard at home can be recognized by local speakers as being out of conformance with local StAmE preferences (pp. 262-263).

Since the object of the present work is not restricted to the analysis of a single regional parlance and its possible relation to the GA macrostructure, but it broaches the extensive area of American dialectology, I will also allocate ample space to an in-depth discussion of the best-known and most-discussed linguistic varieties of the United States: Southern English and New York City English. The criterion behind this latter choice might look haphazard at first. Not only do these varieties stand at opposite poles as regards phonetic habits but they also differ in the extent of their influence.

Despite being sizable in population, the dialect of the Big Apple is confined to the city limits and a few neighboring cities in New Jersey, an area of 530 square miles (Labov, Ash & Boberg, 2006). As we shall see, this geographic restriction appears to be associated with the negative prestige of the city vernacular, as documented in Labov’s (1966) ground-breaking sociolinguistic study The Social Stratification of English in New York City. On the contrary, the Southern linguistic region encompasses a much wider area ranging from North Carolina in the east to Texas in the west, and from the Ohio river in the north to the southern tip of Florida. Notwithstanding its extensive territory, the South appears to be undergoing a regional unification driven primarily by what is known as The Southern Shift in the literature devoted to the subject (Labov et al., 2006).

(14)

Inconsistent as it might seem, the rationale behind my decision to focus on the aforementioned varieties is borne out by the following considerations. In absolute terms, Southern English and New York City English show a much larger gap in their linguistic habits towards General American than does California English. It is this fact that gives some residual legitimacy to the older classification of American accents as Eastern, Southern, and General American (Wells, 1982). Moreover, among the numerous linguistic varieties 2 of the United States, Southern-States English and the New York City vernacular are the most widely recognized as regional dialects by the general public; a circumstance at odds with the mainstream observation that Americans pay very little attention to regional dialects and show little ability to recognize them, as argued by Labov (1966):

In Minnesota or Pittsburgh, the speech of lower class New Yorkers may be imitated by boys who think of this style as a symbol of the tough, hard life and defiance of authority. Indeed some of these sound features have entered into a folk mythology (p. 18).

A similar process of stigmatization concerns the Southern macroregion as well, where stereotypes are abundant. Most Northerners can produce some kind of imitation of a Southern accent and will do so when occasion demands (Labov et al., 2006).

As in the case of California English, a systematic review of the phonetic framework underlying Southern English and New York City English will be substantiated with evidence from my corpus so as to give a firm and practical basis to the theoretical approach of Chapter 1.

“Eastern” refers here to the non-rhotic accents of (i) Boston and Eastern New England, and

2

(15)

In accordance with what I said above about my personal approach to the subject of North American dialects, the following chapters move from a general premise to a more specific conclusion.

Chapter 1 reviews the fundamental nature of dialectological research and surveys the phonetic system of General American. In addition, the linguistic dimension will be integrated with a general ‘excursion’ in the history of the development of U.S. English, in an attempt to elucidate how the settlement patterns and the economic policy of the country have moulded the language of North America. Such a multilayered method will provide a more comprehensive and ‘tangible’ explanation as to why the various regional dialects came to be so remarkably different.

Chapter 2 discusses at length the design and collection of the corpus, which, as mentioned, can best be described as heterogeneric and diatopic. As far as the methodology is concerned, I intend to adopt an integrated approach that combines electronic text elaboration, which will prove useful in detecting specific lexical items, and a qualitative perusal of telecinematic material, an essential step for the selection of data to be subjected to phonetic investigation. Corpus analysis tools, moreover, will allow me to retrieve words in stressed initial and final positions where acoustic characteristics are ‘highlighted’ by the prosodic contours of the utterance. Thus, I will narrow down the procedure of selection of data to relevant inflectional environments.

Chapter 3 is devoted to the presentation of dialectally variable features characteristic of the three regional patterns on which the focus of my analysis is placed: California English, New York City English, and Southern English. A minute evaluation of the most exemplary items obtained in the preceding

(16)

section is intended to be used for corroborating the general framework of the U.S. dialects under analysis.

Before embarking on the ensuing discussion, a further clarification is required: the approach of the present work will be mostly corpus-based (Tognini-Bonelli 2001) since the items selected for analysis are aimed to validate or refine the ample literature of North American dialectology on which my study dwells.

(17)

CHAPTER 1

Background to the study

The language of individuals cannot be understood without knowledge of the community of which they are members. (William Labov, 1966)

1.1 Introduction

My aim in this chapter is to provide a full-circle profile of the discipline on which the study is based: dialectology. Therefore, in the next few sections, I will outline the gradual development of dialect research, from its tentative pioneering attempts in the latter half of the 19th century to the latest significant accomplishments of the early 2000s. The ensuing discussion, however, is not to be considered as a linear account of the various stages marking the growth of dialectology. Indeed, when reflecting on its historical background, I will introduce the textual sources which underpin the structure of the present paper. Each of them pertains to one of the different phases ‘punctuating’ the history of dialectological studies. A Word Geography of the Eastern United States, for example, is the first text to be presented since it belongs to an obsolete season of dialectology, entirely centered on an accurate scrutiny of regional lexical items to the detriment of phonetic variability.

Then, I will turn to the analysis of the ambiguous concept of General American, which, for far too long, has been considered the most common speechway of the United States. My view is that General American is actually a linguistic fallacy and it does not correspond to any given variety spoken in North America. Hopefully, when reviewing the principal U.S. dialects in Chapter 3, such a popular misconception will be definitely proved wrong.

(18)

Chapter 1 ends with a close-up look at the history of the development of English in North America, so as to draw the main strands of the discussion together under a diachronic viewpoint.

1.2 Dialectology: preliminary remarks

The inward-looking approach of traditional linguistics, hinged on the assumption that language can be modeled as an idealized system of abstract representations, resulted in a misconceived view of phonetic differences as unwanted attributes (Kenstowicz, 1994). As the epigraph at the beginning of this chapter epitomizes with eloquent clarity and in contrast to the conceptual framework just mentioned, dialectology systematically describes language variation as a consequence of regional and social differences.

Since language is so pervasive an aspect of human life, it is safe to assume that dissimilarities among groups of speakers have captured our interest from time immemorial, as pointed out by Chambers and Trudgill (1980):

One of the most venerable, and perhaps the most gory, instances of an observed dialect difference is recorded in the Old Testament, when the Gileadites were battling the Ephraimites along the Jordan. Some Gileadites had infiltrated the enemy lines and were posing as allies. A leader of the Ephraimites devised a way of detecting these impostors; he called a suspect in and asked him to name an ear of corn, which the Ephraimites called a shibboleth. According to the scriptural account, “he said sibboleth, for he could not frame to pronounce it right” (p. 15).

Archaic as it may be, the word shibboleth entered English and several other languages, as a technical term indicating a ‘diagnostic’ variable, in the latter

(19)

half of the 19th century. Around this time, observations of linguistic differences, intuitive and casual as they had hitherto been, were systematized into a new heuristic discipline: dialectology.

In the early stages, dialect researchers concentrated on the speech of the linguistically conservative, rural population. They favored, in fact, non-mobile, older, provincial males (hereinafter NORM) in the informant selection process. The rationale behind this narrow choice of subjects was deemed essential to guarantee the ‘authenticity’ of the speech patterns under investigation; urban communities involved too much mobility and flux (Chambers & Trudgill, 1980). Indeed, such a ‘lopsided’ principle informs the structure of an early and ambitious project, the Altas Linguistique de la France, which was used as the touchstone of many subsequent surveys. The research was begun in 1896 by Jules Gilliéron and the data collection process was carried out by his fieldworker, Edmond Edmont, who bicycled all around France, stopping in small villages, to interview aged speakers about a number of local vocabulary items. Once the interviews had been completed, the results were tabulated and plotted on maps, thus providing an atlas which showed how lexical items changed diatopically (Meyerhoff, 2006).

As we shall see, depending on the objective of the atlas, special emphasis can be placed either on lexical or phonetic variables. The maps obtained by sampling the ‘linguistic zone’ under scrutiny are often criss-crossed by meandering dividing lines, called isoglosses, which show how two or more regions differ with respect to some linguistic features. A well-known example from North America is the major isogloss symbolizing the so-called cot-caught merger (or low back merger), a sound change affecting the open and open-mid back vowels of three broad speech areas: Eastern New England, Western Pennsylvania and the Far West (Labov, 1994: 317). The visual impact of the

(20)

red jagged line in Map 1.1, cutting across the whole of the United States, promptly enables us to gauge the spatial diffusion of such a far-reaching phonological phenomenon.

Map 1.1 The merger of /o/ and /oh/ (http://aaronecay.com)

It should be specified that, although there is no other scientific method of determining boundaries between speech areas, an isogloss cuts arbitrarily through the ‘unknown’ territory which has not been studied by the fieldworker; here, a different, or opposing feature might well occur. As a matter of fact, Hans Kurath (1949), in the introduction to A Word Geography of the Eastern United States, comments on the benefits and drawbacks of drawing isoglosses as follows:

The reader must constantly bear in mind that an isogloss is a simplified statement of the geographic dissemination of a linguistic feature, and, on the other hand, dialect boundaries can be established only by means of isoglosses (p. V).

(21)

1.3 From dialect geography to sociolinguistics

A noteworthy step forward in the field of dialectology was achieved when the unbalanced selection of NORM informants, more and more perceived as a kind of ‘linguistic archeology’, was discarded. As early as 1939, the Linguistic Atlas of New England, designed by Kurath with the purpose of enlarging the social background of the interviewees, brought this obsolete methodology ‘up to speed’ on the current configuration of society. In fact, a mobile, younger, and urban structure, in sharp contrast to the one studied by Gilliéron, was sampled and investigated. In addition, the Linguistic Atlas of New England clearly demonstrated that the study of language could not be divorced from the analysis of non-linguistic variables such as age, social class, or education; these latter factors were the social parameters which guided the author through the selection of informants.

Nevertheless, the first phase of dialectology, of which Hans Kurath was probably the most influential and productive representative, avoided ‘coming to grips’ with any linguistic theorizing; it tended to treat linguistic forms in isolation rather than as parts of systems or structures. As Chamber and Trudgill (1980) poignantly observe:

The reaction of dialect geographers seems to have been a profound suspicion of linguistic theorizing under almost any guise […] The inevitable result was that dialectology and linguistics came to have less and less contact with one another. At its worst, there has been a kind of mindless friction between the two groups, with the dialectologists scorning linguists as ‘abstractionists’ who deal in ‘hocus-pocus’ rather than real language data, and the linguists dismissing dialectologists as ‘mere butterfly collectors’ who get so entangled in the bushes that they cannot see the trees (p. 17).

(22)

Hence, for the largest part of the nineteenth and twentieth centuries, dialect atlases were produced as works of reference without any immediate connection with the issues concerning theoretical or descriptive linguistics.

An additional characteristic of ‘bygone’ atlases was their lexically-based organization. As in the case of the Altas Linguistique de la France, once the items, expected to reveal dialect variation, had been pinpointed, the fieldworker proceeded to huddle them together into semantically related groups and to plot them on maps. A similar architectural design would help the reader visualize sharp regional divisions drawn according to local vocabulary. An example of such an outmoded yet fruitful approach, which fell into disuse after the advent of sociolinguistic studies in the early sixties (cf. section 1.4), is provided byA Word Geography of the Eastern United States.

1.3.1 A Word Geography of the Eastern United States

Kurath’s survey was published in 1949 as the culmination of a project begun in 1931 under the auspices of the American Council of Learned Society. The atlas investigation was concerned with the regional and local vocabulary of a geographically restricted area: the Eastern States. As Kurath (1949) reminds us in the preface to the first edition:

The collections of the Linguistic Atlas provide us with a systematic record of the currency of selected words and expressions in all the states on the Atlantic coast […]. The area treated in this book includes all of the original thirteen states of the Union and, in addition, the areas settled after the revolution prior to 1800, except Kentucky (p. V).

(23)

Despite its rather passé configuration, limited as it is to the Atlantic slope, A Word Geography achieved a real breakthrough in the field of dialectology in its attempt to go beyond the limits of the ‘isolationist policy’ of previous surveys. Here, lexical boundaries are not established ‘for their own sake’ but they are correlated with settlement and trade areas, thus giving the necessary foundation for historical interpretations. Albeit not entirely appropriate for an up-to-date review of American English lexical dialectology, the speech areas established 3 by Kurath (1949) on the basis of regional vocabulary will be used to corroborate the necessarily limited notions of social and political history given in section 1.9as preliminary tools for discussing the development of English in the United States.

As in the case of the Linguistic Atlas of New England, the prototypical traits of NORM informants give way to a more wide-ranging criterion of selection of subjects. Indeed, nearly every county in the Eastern States is represented by two speakers, one old-fashioned and unschooled, the other a member of the middle class (Kurath, 1949).

In many cases, a thorough record of various regional and local words, gathered from more than 1200 interviewees, is offered on maps delineating the major speech areas on the Atlantic Seaboard. Southeastern New England, for instance, is delimited by a curvilinear isogloss drawn on the basis of two regional words. As the author points out, this region has the character of a relic area: “from New London to Cape Cod and Nantucket, and northward almost to Boston, the local terms for storm and pet lamb are tempest and cade” (1949: 23). These latter words are not the only ones investigated in the survey as linguistically salient features of Southeastern New England: among them we

Language, in fact, evolves at a steady pace; a lexical feature, idiosyncratic of a specific

3

speech area, can well have become more widespread in use over the course of seven decades; Moreover, the items elicited during the field research of Kurath pertain to quite restricted semantic fields such as farming techniques, flora and fauna, or the weather.

(24)

find rare-ripe for spring onion, or quahog for a large round clam. Nevertheless, their diffusion and lexical significance in the region is such that they have been employed to set off its isogloss, as shown in Map 1.2.

Map 1.2 Southeastern New England lexical isogloss (Kurath 1949: 101)

In addition, the findings of the atlas survey allowed Kurath to replace the traditional North/South/General American division of dialects by compartmentalizing the eastern macroarea into a Northern, Midland, and Southern region. As the author remarks: 4

One fact of major importance seems to me to be fully established: There is an extensive Midland speech area that lies between the traditionally recognized “Northern” and “Southern” areas. This Midland area, which is linguistically distinct from the Northern and Southern areas and is in part set

As we shall see in section 1.6, a three-strata representation of North-American varieties is still

4

(25)

off by sharp boundaries, corresponds to the Pennsylvania settlement area (p. VI).

Kurath’s ‘recantation’ of this former linguistic doctrine, according to which the American English continuum could be broken down into three broad speech patterns, would turn out to be a crucial insight. Despite being restricted to the Eastern States and backed up only by lexical evidence, A Word Geography set the pace for a future confutation of the dominant outlook on North American variability.

1.3.2 Innovative approaches to dialectology

Until around 1960s, the problem of variation in language had been neglected in general linguistics. As argued by Petyt (1986):

Linguists were aware that heterogeneity did exist, but they treated it as if it were merely an uncomfortable but theoretically unimportant fact, which could be left to stylisticians, dialectologists, and other such ‘scavengers’, whose job was to tidy up the trivial matters on the periphery of linguistics proper (p. 46).

However, since the mid-1960s, another viewpoint has come to the foreground and variation in language, the primary concern of dialectology, has been subjected to rigorous examination by modern linguistic thinking. In fact, it was realized that the level of detail in many of the regional atlases offered a plethora of valuable findings to be utilized as ‘props’ to abstract conceptual frameworks. Viewed in this light, structural dialectology represents an attempt to apply some of the insights of linguistics to the ‘work’ of comparing language varieties.

(26)

The originator of the structural approach to dialectological studies is generally recognized in the figure of Uriel Weinreich, professor of Yiddish at Columbia University, who, in 1954, published an article titled, rather tentatively, Is Structural Dialectology Possible?. The objective of the scholar was “to iron out a few of the difficulties existing between the two very much disunited varieties of linguistics, structural and dialectological” (1986: 20) and to bring them closer together. Not only did Weinreich dispel the common fallacy that a given linguistic system operates as a ‘stand-alone’ entity, but he also reconciled the two areas of study by showing that a combined analysis of partially similar systems (i.e., dialects) would yield unexpected and illuminating results (Chambers & Trudgill, 1980) .

Existing dialectology already compared elements belonging to different varieties, but it did so by simply relying on ‘atomistic’ descriptions in which dialects were not conceived as systems, as observed by Weinreich (1986) in the aforementioned article:

A traditional dialectologist will have no scruples about listening to several dialect informants pronounce their equivalents of a certain word and proclaiming that these forms are ‘the same’ or ‘different’ (p. 24).

Structural dialectology, on the contrary, requires linguistic data be analyzed in terms of their intimate membership to the systems under scrutiny. Although Weinreich (1986) applied his theoretical framework to three Yiddish dialects, which will not be discussed here in detail, there should be no restriction on the feasibility of a structural analysis between two varieties of North American English. The concrete example below will help the reader gauge the benefits of this methodology.

(27)

If we compare from a structuralist standpoint New York City and Eastern New England accents (hereinafter NYCE and ENEE), we can observe that they agree in having the short-a vowel (i.e., /æ/) in their phonological inventories. 5 They do not agree, however, in terms of phonetic realization: there is a large group of words, including items such as Spanish, animal, or family, in which ENEE employs the tense vowel [æ̝]~[eə], whereas NYCE has the lax sound [æ]~[æː]. 6

Spanish Family • ENEE [æ̝]~[eə] [æ̝]~[eə] • NYCE [æ]~[æː] [æ]~[æː]

Traditional dialectology would have laid heavy emphasis on the features which differentiate these speech patterns, treating them as unconnected linguistic entities. Structural dialectology, on the contrary, by focusing on the systems of elements within the two varieties, shows how they both share the same phonological inventory, but differ in the allophonic realization of some of its components.

This being said, traditional and structural approaches to dialectology bear a fundamental similarity if viewed from the newfangled perspective of generative grammar, which will be the topic of the following section. As Petyt (1986) brilliantly observes:

The phoneme /æ/ is also called ‘short-a’. The short-a vowel class comprises lexical items

5

such as bat, sat, had, man, ham, lamb, batter, Spanish etc. (Labov et al., 2006: 13).

The feature [±tense] is a cover term for a complex of phonetic features: extended duration and

6

extreme articulatory position with an accompanying increase of articulatory effort (Labov et al., 2006: 16). In IPA, the diacritics [ ] and [ː] indicate a raised and longer realization respectively.

(28)

What traditional and structural dialectology have in common is the fact that they both concentrate mainly on the data of the different dialects […]. Hence their major task is to carry out surveys designed to collect linguistic material to be described. This is where the approach of generative dialectology is claimed to differ (pp. 36-37).

1.3.3 Generative dialectology

In the wake of the notable success achieved by Noam Chomsky’s (1957) Syntactic Structures and following up on his revolutionary claim that languages are more than a corpus of observable data, generative dialectology made good use of concepts and findings formulated by the father of modern linguistics. In fact, by focusing not so much on ‘empirical’ linguistic forms as on the grammars of dialects, such an innovative approach moved the spotlight from patent heuristic research to the investigation of rules underlying linguistic systems. Since the greatest number of dialect differences concern phonological variation, the ensuing discussion will be restricted to the area of generative phonology. As Chambers and Trudgill (1980) opine: 7

A generative approach to phonology consists of a two-level analysis which posits (a) underlying forms, which are the phonological forms in which lexical items are listed in the lexicon, and (b) phonological rules which convert these underlying forms into surface forms and thus, ultimately, into their actual pronunciation (p. 45).

Hence, adherents of this school of linguistics hold that the fundamental property of dialects rests on a common set of underlying forms; differences can

This limitation does not alter the fact that the same general principles would be applicable in

7

the case of grammatical differences between dialects (Petyt, 1986). However, it is beyond the scope of this study to take up a discussion of this theme.

(29)

be accounted for in terms of one or more rules which are ‘peculiar’ to them. A generative grammar-inspired approach largely underpins the structure of Wells’ (1982) Accents of English: Beyond the British Isles.

1.3.4 Accents of English - an overview

Accents of English is a three-volume text, accompanied by a cassette with recorded specimens (now converted to digital form), which aims to provide the reader with a wide-ranging examination of pronunciation patterns used by native speakers of English. The original contribution of Well’s study lies principally “in the description of certain neglected accents of the British Isles and the West Indies” (1982: 1). However, for the purpose of my paper, I will mainly rely on Volume 3, which comprises an analytical description of General American, NYC English and Southern English.

In the preface to the third volume, Wells (1982) prides himself on making appropriate use of all kinds of scholarly treatments, be they dialectological, structuralist, generative, or sociolinguistic. However, it seems unquestionable that generative phonology ‘reigns supreme’, as noted by Raven McDavid in his review to the books: “Well’s approach, through generative phonology, is for me an alien route, but generally clear; in fact, for the first time it seems to make sense”. 8

If we now return to the previous example concerning NYC and ENE different realizations of the phoneme /æ/ in the lexical items Spanish, animal, and family, which was viewed from the perspective of structuralism, and we adopt Wells’ (1982) stance, a concrete representation of how generative dialectology operates can thus be obtained. As mentioned above, a structural

http://eng.sagepub.com/.

(30)

standpoint merely expresses the systematic nature of partial differences and/or similarities of linguistic varieties (Chambers & Trudgill, 1980), whereas generative dialectology’s inductive reasoning overlooks empirical data—which at most can be its starting point—and attempts to specify a set of rules related to underlying forms. Hence, the phonetic difference between the two varieties under consideration can be subsumed by formulating the following rule:

[æ̝]~[eə] / N, M $ 9 NYC English: /æ/ →

[æ]~[æː] / $

ENE English: /æ/ → [æ̝]~[eə] / N, M

This means: in NYC English, the underlying segment /æ/ will be realized as a tense vowel (phonetically ranging from [æ̝] to [eə]) in the environment of a following nasal consonant occurring before a syllable boundary (i.e., in a closed syllable); on the contrary, in an open syllable, it will be realized as a lax vowel. Therefore, in the aforementioned lexical items Spa-nish and fa-mi-ly, the short-a sound, which occurs at the end of the syllable, has a lax realization. On the contrary, the vowel sound in man—a monosyllabic lexical item closed by a nasal consonant—is usually tense; the noun phrase a man will have to be phoneticized as [ə ˈmeən] (Labov, 1966). In ENE English, tensing of the underlying form /æ/ always takes place before a nasal consonant, regardless of the nature of the syllable in which it occurs (cf. the short-a split system and the nasal system in section 3.3.1.2).

A further illustration of a generative approach to phonology is given by Wells (1982) when treating the problems raised by vowels in the restricted

The symbol $ is employed to indicate a syllable boundary. For a complete list of

9

(31)

environment of a following /r/. The scholar, indeed, posits a linguistic situation whereby, phonetically speaking, [ʌ] plus /r/ is quite rare. With the exception of NYC and scattered parts of the east and south, what we should expect to hear in words such as furrow, hurry, courage, is an r-colored mid central sound [ɝ], the vowel of NURSE. At the level of deep structural relations, however, Wells 10 (1982) argues that:

[ɜ] can be regarded as a position allophone of /ʌ/ and [ɝ] as a realization of /ʌr/, from which it follows that there is no separate phoneme /ɜ/ (or /ɝ/) in this most usual type of General American […]. In view of the above argument, it is admittedly inconsistent to write NURSE words in General American with /ɜr/; nurse is to be phonemicized /nʌrs/ (pp. 480-481).

Describing Wells’ approach to North American English varieties solely in terms of generative dialectology would be reductionist. Nevertheless, it is evident that, in the aftermath of Chomsky’s major revolution in linguistics, the transformational interpretation of dialects predominates over other methodologies in Wells’ (1982) review of U.S. pronunciation patterns.

1.4 Social dialectology

In the mid-1960s, around the time that the interrelationship between dialectology and linguistics was growing stronger, some scholars began to recognize that the regional dimension of linguistic variation had been given exclusive attention. On this point, Chambers and Trudgill (1980) remark:

Throughout his book, Wells (1982) makes use of keywords, specifically chosen to refer

10

concisely to groups of words that generally share the same vowel in varieties of Southern British English and what is known as General American. The phonetic realization of the vowel may be rather different in Received Pronunciation and General American, but the set of words identified by the keyword generally will be the same.

(32)

This was gradually felt, by some workers, to be a drawback, since social variation in language is as common and important as regional variation. All dialects are both regional and social, since all speakers have a social background as well as a regional location (p. 54).

As previously mentioned, the criterion laid down by Kurath in the Linguistic Atlas of New England consisted in selecting socially different types of informants. This was an important preliminary step. However, before the issue of social variation would be given due consideration, dialectology still had a long way to go.

For years, in fact, dialectologists were not faced with the ‘challenge’ posed by variability in the linguistic behavior of individuals belonging to the same speech community. Many speakers of NYC English, for instance, occasionally pronounce pre-consonantal and word-final /r/ with oral constriction, while oftentimes they do not. Such a phenomenon was of no concern to traditional dialectology, grounded as it was on data usually obtained with one-word answers to individual questionnaire items (Chambers & Trudgill, 1980). Here, variation could hardly show up. Moreover, when observed, both interspeaker variability, and its necessary corollary intraspeaker variability, were downgraded to the level of free variation. By free, scholars meant that there was no clear pattern regulating such an ‘erratic’ linguistic behavior. As a consequence, variability was kept on the periphery of dialectology until the mid-60s, when another viewpoint gained ground.

By virtue of pioneering sociolinguistic studies, conducted by Dave Britain in England and by William Labov in North America, it was soon realized that variation in language could be constrained both by linguistic factors, as advocated by previous dialectological researchers, and by extra-linguistic

(33)

aspects congruent with the idea of regularity and structure, as observed by Petyt (1986):

There is ‘structured heterogeneity’, or patterned variation, within the individual to some extent, but more importantly, within the community. In fact, Labov (a pupil of Weinreich, who had emphasized the importance of structure in the study of dialects) suggested that true regularity of linguistic patterns was to be found not in the individual speaker, but in the speech community as a whole (pp. 46-47).

Having broadened the scope of dialectology to include the social dimension of language, traditional methods of analysis turned out to be ‘a house of cards’. In fact, long-established surveys, conducted within the guidelines of a questionnaire, were replaced by “tape-recorded stretches of quasi-conversational speech, usually obtained by the asking of questions designed to produce large amounts of talk” (Chambers & Trudgill, 1980: 58).

Providing a blow-by-blow account of what social dialectology is would take us far from the relatively restricted topic of this chapter. In fact, this discipline, as Meyerhoff (2006) wittily points out, “wears many caps” (p. 2). However, a discussion of Labov’s (1963) study of a sound change in progress on the island of Martha’s Vineyard, conducted with methods and principles essential to the field of sociolinguistics, will prove useful in illustrating its ‘prismatic’ perspective on language. As such, this survey offers an excellent entry point to one of the major contributions to sociolinguistic studies: The Social Stratification of English in New York City (Labov: 1966).

(34)

1.4.1 A few notions of sociolinguistic methodology

Before presenting Labov’s studies of social dialects, it is an absolute pre-requirement to lay down some key principle of sociolinguistics, as explained in the following paragraphs.

Variables and variants: the concept of variable refers to an abstract linguistic unit representing the source of variation. Similar to the relationship between phones and phonemes, variants, the actual realizations of a variable, can be constrained by social or non-linguistic factors (Meyerhoff, 2006). Conventionally, sociolinguistic variables are shown in rounded brackets. For instance, in the Martha Vineyard’s study, when investigating the pronunciation of the diphthong in words such as ice, time, Labov (1972a) talks about the (ay) variable, which is realized by different phonetic variants.

Markers and Indicators: variables which are behind the level of social awareness and consequently which have not acquired strong stereotypes, are classified as markers and indicators. Meyerhoff (2006) defines these concepts as follows:

Markers can be distinguished from indicators in that speakers show some subconscious awareness of them, and this is made evident in the fact that they consistently use more of one variant in formal styles of speech and more of another variant in informal styles of speech. Indicators, on the other hand, show no evidence that speakers are even subconsciously aware of them; hence, they consistently favor one variant over another regardless of who they are talking to or where (p. 23).

(35)

Stylistic Differentiation: in sociolinguistic studies, it is imperative that the

spectrum of registers be enlarged to encompass informal style. Indeed, it is at this level, least influenced by notions of correctness, that “linguistic tendencies and regularities are most clearly to be found and where many linguistic changes take place” (Chambers & Trudgill, 1980: 58). Therefore, sociolinguists have to develop methods to divert informants’ attention away form their speech. 11

When carrying out sociolinguistic surveys, the four basic parameters on which the focus of attention is usually placed are: social rank, ethnic background, gender, and age.

Social rank (i.e., social class) involves, in Labov’s (1966) words, the degree to

which an individual possesses “wealth, knowledge, power, and authority, relative to other members of society. In indices of class, these hierarchies are most commonly represented by income, education, and occupation” (p. 132).

Ethnic variability, on the contrary, is concerned with quantitative and

qualitative differences on the part of ethnic groups in their use of particular variables.

The gender of speakers is another social feature with which linguistic

differences have been found to correlate very closely. One of the most strikingly consistent findings to emerge from sociolinguistic studies is the fact that “women, on average, use more higher status variants than men do” (Chambers & Trudgill, 1980: 72).

We will see in the next two sections how these requirements are fulfilled.

(36)

Age is a paramount concern of sociolinguistic research. A simple method to

assess age-related variability consists in comparing the speech of older people with that of younger generations. If present, overt differences will be the result of diachronic linguistic changes. This latter technique is generally referred to as an investigation in apparent time, opposed to real-time studies which require fieldworkers to scrutinize a particular speech community at decade intervals.

As can easily be guessed, the aforesaid social constraints cannot predict with a 100 percent of certainty the outcome of a speaker’s linguistic behavior. Sociolinguistics is, in fact, relativistic in nature; its surveys will only tell us how likely we are to hear different forms, depending on contextual clues and the informants’ profiles. Hence, the benefits of calculating average values of variables in order to obtain index scores. An illustrative example is provided by Labov (1966) in The Social Stratification of English in New York City.

One of the variables examined in the NYC speech community is (æh), the sound of bad, or ask, which is ranged on a six-point phonetic continuum of vowel quality.

Figure 1.1 Scale for (aeh) index (Labov, 1966: 34)

The index score of the vowel is derived by multiplying by ten the average of the values assigned to all of its individual occurrences (Labov, 1966). Thus, (æh)-25 would be the index rating for a person who pronounced half of the

(37)

words in this group with (æh-3) and half with (æh-2). Once indices of this type have been computed, they can be refined by relating them to the above-mentioned sociolinguistic parameters. This procedure will allow the researcher to ascertain the extent to which a specific variable correlates with social class, ethnicity, style-shifting, and so on. Graphic mapping tools may be employed to show the consistent patterning of a variable with respect to some independent factor. In the case of (æh), Labov (1966) produced a graph displaying, on the vertical axis, the phonetic index, and, on the horizontal one, stylistic variation in three social classes, as illustrated in Figure 1.2. 12

Figure 1.2 Class stratification of (æh) (Labov, 1966: 142)

Figure 1.2 shows how variants associated with the (æh) variable are stratified by style. Regardless of social classes, there is a consistent tendency to use more open, front vowels as the formality of the talk increases until it reaches its apex in style D. The ensuing discussion of Labov’s major sociolinguistic studies 13 will give us ample opportunity to see such innovative procedures in full operation.

(0-2) lower class; (3-5) working class; (6-9) middle class.

12

In section 1.4.3, I will look a little more closely at the way in which the ‘Labovian’

13

(38)

1.4.2 The Martha’s Vineyard study

Martha’s Vineyard, a Massachusetts island located south of Cape Cod, is a beauty spot to which tourists swarm in summertime, causing local prices to soar high and forcing the year-round residents to ‘make ends meet’. In 1961, 14 Labov was one of those summer visitors, but his interest in linguistic ‘heterogeneity’ did not wane when on holiday. Having studied under the tutelage of the late Uriel Weinreich at Columbia University (NY), he soon embarked on an investigation of two specific variables, (ay) and (aw), which had struck him as markedly different from their counterparts in the neighboring areas of the mainland. His analysis (1972a) was already fully sociolinguistic in scope:

By studying the frequency and distribution of phonetic variants of (ay) and (aw) in the several regions, age levels, occupational and ethnic groups within the island, it will be possible to reconstruct the recent history of this sound change; by correlating the complex linguistic pattern with parallel differences in social structure, it will be possible to isolate the social factors which bear directly upon the linguistic process (p. 1).

On the Vineyard, the two variables under scrutiny were sometimes realized, at the level of both interspeaker and intraspeaker variation, with a raised centralized onset (i.e., [əi], [əʊ]). However, sometimes they normally had a lower, fronted starting point, as in the rest of the island’s mainland neighbors. In order to elucidate such a complex pattern of variation, Labov (1972a) gathered a wide range of tape-recorded stretches of speech which fell within a continuum from formal to casual style. As shown in Map 1.3, the sociolinguistic interviews were conducted in several areas of the island; each

As will be seen, the social background of year-rounders is an essential element of Labov’s

14

(39)

one was characterized by a different ethnic composition. In addition, the 69 informants came from various walks of life, and were of different age, so as to provide a fair cross-section of the local community. 15

Map 1.3 Location of the 69 informants on Martha’s Vineyard. Ethnic origin is indicated as follows:☐English, ■ Portuguese, ▴ Indians (Labov,1972a: 3).

As Meyerhoff (2006) reminds us, sociolinguistics cannot avoid taking into account purely linguistic constraints. Labov (1972a), in fact, started his analysis by assessing the influence of the consonant following the diphthongs under consideration; he tabulated five general articulatory dimensions. Complementing the phonological restrictions schematized in Figure 1.3, however, were social factors which seemed to affect the different variants of (ay) and (aw). A centralized, raised pronunciation appeared to be favored by people between the ages of 31 and 45, associated with the fishing industry in the more rural areas up-island (Labov, 1972a).

Some were employed in farming activities, some in the fishing industry, others in the tertiary

15

(40)

Figure 1.3 Phonological environments favoring centralization on Martha’s Vineyard (1972a: 20)

So far, Labov (1972a) had been measuring the influence of specific social parameters on the island’s societal networks. However, he did not confine his investigation to the hitherto sociolinguistic framework, but he set out to find whether his evidence could depend “on a larger factor which was logically prior to them” (1972a: 26). Hence, Vineyarders were rated on the basis of their attitudes to residing on an island where, as mentioned before, the cost of living had ballooned in recent years as a result of the ever-expanding holiday trade. What emerged from this further ‘psychological ranking’ of informants was an inverse relationship between centralization of the two diphthongs and negative reactions to being ‘stuck’ on the Vineyard. This combined analysis of linguistic aspects and social factors allowed for the conclusion that centralized variants represented an unconscious form of attachment to one’s roots on an island whose stability was threatened by “a long-range decline in the economy and the steady encroachment of the summer people” (Labov, 1972a: 37).

1.4.3 The Social Stratification of English in New York City

Traditionally regarded as the founder of the discipline of sociolinguistics, Labov’s (1966) research of the Lower East Side of Manhattan came to prominence as the model of numerous linguistically-oriented studies hinged on urban dialectology, in the years following its publication. More ambitious than

(41)

in his previous study, Labov examined a wide range of variables which receive a ‘special treatment’ in the NYC speech community, a dialect pocket on the eastern coast of the U.S. that is markedly different from neighboring areas. The reason for such a linguistic uniqueness is attributed to the stereotypical evaluation of NYCE as a “sink of negative prestige” (Labov, 1966: 338). Consequently, the City has failed to expand its own speech pattern into the surrounding hinterland. In Labov’s (1966) words:

Most of the important dialect boundaries of the eastern United States fall along lines which are natural troughs in the network of communication. […] But the New York City dialect area is an exception to this pattern, and a radical exception. The influence of New York City speech is confined to a narrow radius, hardly beyond the suburbs that form the ‘inner ring’ of the city (pp. 338-339).

Drawing upon the methodology first employed in the Martha’s Vineyard survey and honing many of its techniques, Labov (1966) set out to explain how social and stylistic variations impinge on the realization of a wide array of variables, in an attempt to reject what had been described as “a case of 16 massive free variation” (1966: viii), inaccessible to formal linguistic analysis. In fact, Labov (1966) proved that, when studied in the larger context of the speech community, a highly systematic structure of social and stylistic variation became fairly clear:

Native New Yorkers differ in their usage in terms of absolute values of the variables, but the shifts between contrasting styles follow the same pattern in almost every case (p. 6).

The five basic phonological variables under scrutiny are (r), (æh), (oh), (th), (dh).

(42)

We have already seen in Figure 1.2 how the variable (æh) reflected a consistent trend of social and stylistic stratification. A similar pattern can be observed in the case of the (r) variable. One of the most salient features, which sets NYC English apart from adjoining speech areas, is the issue of rhoticity. Despite being traditionally regarded as an r-less dialect, Labov’s (1966) analysis of the city vernacular sketched a somewhat varied image by revealing a patterned shift of variation. To achieve this goal, the scholar designed a series of sociolinguistic interviews, consisting of four parts arranged in increasing order of formality, so as to manipulate the amount of attention paid to the linguistic performance. As in the case of (æh), stylistic variation indices are 17 combined with socioeconomic parameters dividing the sampled population into three classes of different social rank. The first thing that catches our eyes in Figure 1.4 is the predominance of unconstricted realizations of the consonant sound in casual speech “where values for the use of (r) start at a very low point” (Labov, 1966: 143) and, then, steadily rise with increasing formality. Notwithstanding the regular pattern of stylistic stratification shared by the social ladder as a whole, it is equally evident, though, that the upper classes are differentiated from the lower ones by an unwaveringly higher rate of r-fulness in all of the five styles.

Style A: casual speech, usually obtained by recording stretches of conversation outside the

17

formal interview or by fostering an animated conversation about dangerous situations the speaker had been in.

Style B: careful speech (i.e., the type of speech which normally occurs when the subject is answering questions formally recognized as “part of the interview”).

Style C: reading style. The interviewee is asked to read a text constructed to contain the variables in as many linguistic environments as possible.

Style D: word list or reading of minimal pairs (i.e., pairs of words that have different meanings but only differ from each other in one sound).

(43)

Figure 1.4 Class stratification of r (Labov, 1966: 141)

A plausible explanation for the current status of post-vocalic /r/ in NYC lies in the pressure exerted by the introduction of the r-pronouncing norm, following World War II (Labov et al., 2006). After this watershed event, what was once the standard upper-class pronunciation began to be perceived as the ‘vulgar tongue of the masses’. Such a shift was due, as mentioned, to the mainstreaming of an r-full accent, which originated away from the British-influenced East Coast, but it was also caused by the loss of England’s imperial status. Widespread postwar migration from rhotic regions of the country partly contributed to the linguistic ‘role-reversal’ of r-dropping and r-pronouncing norms (Sȩn, 1979). Hence, NYC English should be regarded as a variably non-rhotic accent, which is basically r-less when people are speaking ‘the language of the street’, we might say in a rather prosaic manner.

To be fair, as early as 1950, Allan Hubbell, in The pronunciation of English in New York City, had noted a tendency for many New Yorkers to pronounce post-vocalic /r/ in reading more than in conversation. He had, nonetheless,

Riferimenti

Documenti correlati

The stimulation of the human skin can take place in different ways: principally by pressure, vibration or heat.. The information can be in the form of images, Braille script and

The only texts not presenting a pronoun as the only element in the post-verbal domain are the Kentish Homilies, which present no post-verbal pronouns, and the Lambeth Homilies,

We have provided combinatorial lower and upper bounds for the dimension, and we have shown that these bounds coincide if the dimensions of the underlying extended Tchebycheff

unum meum casamentum cum casa et edifficio super se, quod est positum in civitate Imole, in hora Taupadhe: ab uno latere tenet stratha publica, ab alio commune Imole, a tercio filia

In the case of a request for a 100 Gb/s lightpath between a single s–d pair, such that 6 equal cost shortest paths exist and just one is feasible, the average time to complete the

Università degli Studi di Modena e Reggio Emilia D IPARTIMENTO DI STUDI LINGUISTICI E CULTURALI.. C ORSO DI L AUREA M

Random samples of one hundred instances of whoops tagged as an interjection revealed one error in the British National Corpus (BNC) and no error at all in the Corpus of American