Table of contents : Table of contents List of contributors List of figures and tables 1. Introduction I. Enregisterment 2. Northern English and enregisterment 3. Dickens and northern English: stereotyping and ‘authenticity’ reconsidered 4. The linguistic landscape of north-east England 5. Lenition and T-to-R are differently salient: the representation of competing realisations of /t/ in Liverpool English dialect literature II. Phonology 6. External and internal factors in a levelling process: Prevocalic (r) in Carlisle English 7. Scouse NURSE and northern happy: vowel change in Liverpool English III. Syntax and discourse features 8. Are Scottish national identities reflected in the syntax of Scottish newspapers? 9. Final but in northern Englishes IV. Sociolinguistics 10. Education, class and vernacular awareness on Tyneside 11. Changing domains of dialect use: A real-time study of Shetland schoolchildren V. Language and corpus 12. New perspectives on Scottish Standard English: Introducing the Scottish component of the International Corpus of English Index
Like this paper and download? You can publish your own PDF file online for free in a few minutes!Sign Up
File loading please wait...
Citation preview
Sylvie Hancil and Joan C. Beal Perspectives on Northern Englishes
Topics in English Linguistics
Editors Elizabeth Closs Traugott Bernd Kortmann
Volume 96
Perspectives on Northern Englishes
Edited by Sylvie Hancil and Joan C. Beal
ISBN 978-3-11-044865-8 e-ISBN (PDF ) 978-3-11-045090-3 e-ISBN (EPUB) 978-3-11-044874-0 ISSN 1434-3452 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de. 6 2017 Walter de Gruyter GmbH, Berlin/Boston Cover image: Brian Stablyk/Photographer’s Choice RF/Getty Images Typesetting: RoyalStandard, Hong Kong Printing and binding: CPI books GmbH, Leck ♾ Printed on acid-free paper Printed in Germany www.degruyter.com
Table of contents Contributors vii List of figures and tables
ix
1
Sylvie Hancil and Joan C. Beal Introduction 1
I
Enregisterment
2
Joan C. Beal Northern English and enregisterment
3
4
5
17
Katie Wales Dickens and northern English. Stereotyping and “authenticity” re-considered 41 Michael Pearce The linguistic landscape of north-east England
61
Patrick Honeybone, Kevin Watson & Sarah van Eyndhoven Lenition and T-t-R are differently salient: the representation of competing realisations of /t/ in Liverpool English dialect literature 83
II Phonology 6
Sandra Jansen External and internal factors in a levelling process. Prevocalic (r) in Carlisle English 111
7
Marten Juskan Scouse NURSE and northern happY: vowel change in Liverpool English 135
III Syntax and discourse features 8
Sanna Hillberg Are Scottish national identities reflected in the syntax of Scottish newspapers? 169
vi
9
Table of contents
Sylvie Hancil Final but in northern Englishes
191
IV Sociolinguistics Marie Jensen 10 Education, class and vernacular awareness on Tyneside
11
215
Mercedes Durham Changing domains of dialect use. A real-time study of Shetland schoolchildren 245
V Language and corpus Ole Schützler, Ulrike Gut & Robert Fuchs 12 New perspectives on Scottish Standard English. Introducing the Scottish component of the International Corpus of English 273 Index
303
List of contributors Joan C. Beal University of Sheffield Les Coudrais, 22150 Plouguenast, France [email protected]
Sylvie Hancil University of Rouen Normandie 4, Parc de la Varenne, 76130 Mont Saint Aignan France [email protected]
Mercedes Durham Cardiff University ENCAP/Cardiff University John Percival Building/ Colum Drive Cardiff CF 10 3EU Wales durhamm@cardiff.ac.uk
Sarah van Eyndhoven University of Canterbury University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand [email protected] Robert Fuchs Hong Kong Baptist University Department of English Language and Literature 224 Waterloo Road Kowloon Tong Hong Kong SAR [email protected] Ulrike Gut University of Münster English Department Johanisstrasse 12-20 48143 Münster Germany [email protected]
Sandra Jansen Universität Paderborn Warburger Str. 100 33098 Paderborn Germany [email protected] Marie Møller Jensen Aalborg University Department of Culture and Global Studies Kroghstræde 3 9220 Aalborg Øst Denmark [email protected] Marten Juskan University of Freiburg Englisches Seminar Rempartstrasse 15 D-79098 Freiburg Germany [email protected] Patrick Honeybone University of Edinburgh Linguistics and English Language Dugald Stewart Building 3, Charles Street Edinburgh EH8 9AD, UK [email protected]
viii
List of contributors
Michael Pearce University of Sunderland Reg Vardy Centre, St Peter’s Campus, Sunderland SR6 0DD, UK. [email protected]
Katie Wales University of Nottingham 2, The Orchards, Great Shelford, Cambs. CB22 5AB, UK [email protected]
Ole Schützler University of Bamberg English Department An der Universität 9 96045 Bamberg Germany [email protected]
Kevin Watson University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand [email protected]
Repertoire of nineteenth-century “Geordie” features 33 Repertoire of nineteenth-century Yorkshire features 33 Repertoire of modern Yorkshire dialect features 34 The LL corpus 67 North East features represented in the LL corpus 70 74 Features of the LL corpus attested in dictionaries and corpora Realisation of utterance-final /d/ (in environment. V_##) in 16 adolescent speakers 90 Realisation of (th) in four male and four female speakers from OLIVE’s Archive 91 subcorpus Vowel plots from 4 female speakers and 4 male speakers from OLIVE’s Archive subcorpus 92 93 CHLDL respellings. (d), (th) and NURSE / SQUARE Realisation of utterance-final /t/ (in environment. V_##) in 16 adolescent speakers 96 Frequency of occurrence of T-to-R (in environment. V__#V) in four male and four female speakers from OLIVE’s Archive subcorpus. 98 Spelling of /t/ in the LE corpus of CHLDL in three phonological environments 100 Spelling of /t/ in Liverpool CHLDL split by words which exhibit T-to-R in OLIVE’s Archive subcorpus and words which do not, in three phonological environments 101 Location of Carlisle 115 Participant sample in the study 116 Variants of prevocalic (r) in % 118 118 Distribution of taps in % according to age group, sex and style Rbrul results for (r) in prevocalic position. Application value. tap (oral history data are not included) 119 Rbrul results for (r) in VrV position. Application value. tap (Oral History data are 121 not included) Distribution of (r) in VrV position 122 Distribution of (r) in Vr#V position 122 Distribution of (r) in V#rV position 123 Distribution of (r) in CrV position 123 Distribution of (r) in C#rV position 124 Rbrul results for (r) in V#rV position. Application value. tap (Oral History data are not included) 124 Rbrul results for (r) in CrV position. Application value. tap (Oral History data are not included) 125 number of speakers by age, gender, and social class 138 happY. mixed linear effects regression models (extracts) 140 happY by age 141 happY (F1) by gender and age 142
happY (F2) by age and class 143 happY- FLEECE . mean vowel position by age 144 NURSE . mixed linear effects regression models (extracts) 145 NURSE by age 146 NURSE by gender and age 147 NURSE by age and class 148 NURSE (F1). t-tests of age by social class 149 NURSE (F2). t-tests of age by social class 150 NURSE -SQUARE . Pillai scores by age and style (groups) 150 NURSE -SQUARE . vowel space by style (old speakers) 151 NURSE -SQUARE . vowel space by style (middle-aged speakers 153 NURSE -SQUARE . vowel space by style (young speakers) 154 awareness of NURSE by age 156 Screenshot from http.//www.ayecan.com/scottish_census_2011.html 175 Frequency of relativizers which, who, that, zero, whom and whose in SSE news, absolute numbers normalized per 10,000 words 177 The frequency of restrictive relative clauses in descending order (adverbial relativizers excluded). 178 Frequency of can, cannot/can’t, may, ought, shall, should, want, will, ‘ll and won’t in SSE news, absolute numbers normalized per 10,000 words 181 Number of occurrences of sentence-final in the NECTE corpus and the Spoken Scottish Corpus 197 Syntactic category of sentence-final but in the two corpora 202 Distribution of the semantic values in relation to the subjectivity scale 203 Markers of agreement in the NECTE and the Scottish corpora 205 First person pronouns 228 Overview of vernacular variables in Tyneside questionnaire study with examples 229 Filler sentences 230 Questionnaire counterbalancing scheme 231 Overview of participants 232 Significant correlation between frequency judgment scores and affiliation 233 index Significant correlations between language use scores and affiliation index 234 Comparison between self-identified class membership and education 235 Distribution of self-identified class membership and education 235 Distribution of participants in the four participant groups in terms of age and gender 236 Task 1 (perceived frequency), descriptive statistics across the four groups 236 Task 2 part 1 (attested own language use), descriptive statistics across the four groups 237 Difference between task 1 (perceived frequency) and task 2 part 1 (attested own language use), descriptive statistics across the four groups 237 Task 2 part 2 (language awareness score), descriptive statistics across the four groups 238
List of figures and tables
xi
Task 3 (affiliation score), descriptive statistics across the four groups 238 Map of the United Kingdom with the Shetland Islands highlighted 247 Distribution across schools of the 2010 questionnaires 251 254 Distribution of children's origins in 1983 and 2010 Percentage of children’s origins by school (2010) 254 Percentage of answers to first question broken down by year and origin 256 Shetland heritage children’s rates of use according to level of acquaintance 258 Figure 11.5 Shetland born children’s rates of use according to level of 260 acquaintance Figure 11.6 Non-Shetland children’s rates of use according to level of acquaintance 260 Figure 11.7 Shetland heritage children’s rates of use according to level of formality 261 Figure 11.8 Shetland born children’s rates of use according to level of formality 262 Figure 11.9 Non-Shetland children’s rates of use according to the level of formality 263 Table 11.3 Summary of key findings in terms of different domains of use 264 Figure 11.10 Students’ reports on whether there are situations where it is not appropriate to use the dialect 265 Figure 11.11 Specific examples of when it would not be acceptable to use the dialect 266 Figure 11.12 Student responses to whether there are situations in which English is not acceptable 267 Figure 12.1 Deficits and gaps in research on SSE 279 Figure 12.2 Traditional model of the SSE-Scots continuum of usage (McArthur 1979:59) 281 Figure 12.3 The elastic boundaries of Scottish Standard English 282 Figure 12.4 The elasticity of individual traditional features of the Scottish Standard English accent 285 Figure 12.5 Annotation of the written part of ICE-Scotland using Pacx 291 Figure 12.6 An example of the ELAN transcription of a spoken file in ICE-Scotland 292 Figure 12.7 An example of the phonemic annotation of a spoken file in ICE-Scotland, extract from the phrase “members of the Scottish Parliament” 294 Figure 12.8 Search results for the word Scots 296 Table 10.14 Figure 11.1 Table 11.1 Figure 11.2 Table 11.2 Figure 11.3 Figure 11.4
Sylvie Hancil and Joan C. Beal
1 Introduction Abstract: Northern English has been the object of much attention linguistically over the last thirty years but scholars have had a tendency to focus on the phonology of the dialects and varieties encountered. The purpose of the present volume is to complement and enrich the existing studies by providing readers with a kaleidoscopic perspective, allowing for a holistic interpretation and understanding of northern English. It is based on a selection of papers delivered at the International Conference on Northern British English on May 12–13, 2014, Rouen. It includes studies not only on phonology but also on semantics, syntax and sociolinguistics from a synchronic and diachronic point of view, with a special emphasis on the process of enregisterment. The varieties covered include Scottish Standard English, Shetland as well as varieties from the North of England.
1 Background The ‘North-South Divide’ is a phrase that has been used with reference to Britain in a wide variety of fields including human geography (Baker and Billinge 2004, Dorling 2010), public health (CLES 2014), economics (Mason and Harrison 1991) and cultural studies (Morley 2013). In terms of dialectology, sociolinguistics and typology, the North of England, and the North of Britain more widely (including Scotland and Northern Ireland) have been widely researched in recent years. There has been a major study by Wales (2006) and chapters by Beal (1993, 2008a/b) on northern English; volumes on specific varieties within the North of Britain (Millar 2007, Corrigan 2010, Beal et al 2012) and a perceptual study of the “North-South divide” in English dialects (Montgomery 2006). The vitality of research in this area is reflected in the existence of a biennial Northern Englishes Workshop, which had its seventh meeting in Edinburgh in 2016. The workshop at which many of the chapters in this volume were presented also testifies to a growing awareness of the importance of northern Englishes amongst anglicists in France. Whilst research on northern Englishes presented at conferences such as the Northern Englishes Workshops has tended to be dominated by variationist Sylvie Hancil, University of Rouen Normandie Joan C. Beal, University of Sheffield DOI 10.1515/9783110450903-001
2
Sylvie Hancil and Joan C. Beal
studies and by sociophonetics in particular, this volume presents a wider range of perspectives with an overarching emphasis on data-based studies. In this respect, it also differs from the recent volume edited by Hickey (2015), which provides more extensive geographical coverage of dialects of the North of England, but largely from a sociophonetic and/ or variationist perspective. Linguistic diversity in Britain was first extensively discussed in Higden of Chester’s Latin history Polychronicon (Wales 2006: 64), which was then elaborated by John of Trevisa (1385) who remarks how the Northumbrian “ys so scharp, slyttyng, and frotyng, and unsschape” that “we southern men (nos australes) can hardly understand it” (cited Sisam 1967: 150). It is not until the sixteenth and the seventeenth centuries that scholars offered a fuller description of the prevailing dialectal situation. Carew (1595) discussed the binary opposition in the dialects in Britain in these terms: “Wee have Northern and Southerne [. . .] which differ from each other, not only in the terminacions, but also in the words, terms and phrases” (cited Görlach 1990: 243). Under the entry ‘dialect’ in his dictionary, John Bullokar (1616: 65–6) claims that “the dialect, or manner of speech, in the north is different from that in the south”. Wales argues that there has been “both a metropolitan bias, and a southern one: what I shall term metrocentrism and austrocentrism respectively” (2006: 2). There are essentially three reasons for this: first, the emergence of the standard written English out of London in the late fifteenth century; second, London’s influence on ‘correct’ pronunciation; and third, the enregisterment of a variety based on southern English as Received Pronunciation. Therefore, northern English has mainly been defined in relation to, if not against southern standards, offering a negative view of the variety examined. The perception of the North as being “alien and barbaric” (Wales 2006: 65) dates back to the medieval period, as underlined by Pollard (1997: 39): “This frightening north was a cultural construct, a state of mind”. Wakelin (1972: 35) quotes the use of northern dialect by Chaucer for the characterization of two Cambridge students John and Aleyn as being “comic because they speak a regional, non-Standard, dialect”, which anticipates the view of superiority and inferiority in varieties of English starting to emerge in the mid sixteenth century, with London English being particularly associated with the prestige of courtly English and the other types of English with lower social status, i.e. “provincial boorishness and country bumpkins” (Wales 2006: 77). John Hart wrote “of the farre West, or north Countryes, which use differing terms from those of the Court, and London, where the flower of the English tongue is used” (cited Blank 1996: 106 in Wales 2006: 75). These derogatory remarks contributed to the movement among grammarians in favour of “the Proper, or London language” (Hugh Jones 1724) to the disparagement of regional dialect.
Introduction
3
Nevertheless, the perception of northern dialect does not boil down to a condescending attitude in the literature of the period. Boorishness can be synonymous with plain speaking and honesty without the affectation of courtly speech and northern speech can be praised for its “antiquity, purity and also rusticity” (Wales 2006: 80), as in the so-called ‘Scottish Chaucerian’ poets, from John Barbour in the fourteenth century and James I of Scotland (1394–1437), through to Robert Henryson and William Dunbar (d. 1516). Katie Wales’s study in this volume pays homage to the way Dickens re-created and created northern English in his works. The late nineteenth century provided crucial developments in education, which contributed to help Northerners climb the social ladder, especially from the 1930s onwards, when the industries of weaving, ship-building and mining started to decline. School-educated Northerners have had to redefine their sociolinguistic defining criteria, along with being confronted with the requirement to meet the expected norms of the Received Pronunciation instilled by dialectologist Ellis (1890) and associated with social advancement and mobility. This feeling of “being in social limbo” (Wales 2006: 143) is well illustrated by two studies in this volume (Marie Jensen; Mercedes Durham). The linguistic variation is largely defined in phonological terms. One of the first systematic studies of English dialects was Ellis’s On early English Pronunciation Part V, The existing phonology of English Dialects (1889). Ellis’s isoglosses were based on the following phonological criteria: the pronunciation of words like some; the pronunciation of r; the pronunciation of the definite article; and the pronunciation of words like house. The first, second and fourth criteria are under discussion in the phonological part of this volume (Sandra Jansen; Marten Juskan). It is Trudgill’s The Dialects of England (1990, 1999) which offers the most often cited classification of dialects, with the northern dialect situated above a line drawn from the Humber to the coast to the north of Lancaster, which is a line slightly to the north of Ellis’s. Over the last ten years, linguists have provided other linguistic variants to set up slightly different frontiers between the North and the South, thereby contributing to the fuzziness of the North-South divide (see Beal, this volume). There has been an effort over the last thirty years to reveal and underline the characteristics and specificities of the North, relying on almost 1,500 years of history to revive its periods of cultural and literary prestige. The salience of these features is underscored in the part on enregisterment in this volume (Joan Beal; Katie Wales; Michael Pearce, and Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven).
4
Sylvie Hancil and Joan C. Beal
In northern English grammar, there still remain salient characteristic markers such as third person feminine pronoun ho(o), second person pronoun singular thou, reflexives, deictic referents, definite article reduction and negation, which partake of the richness and complexity of sociolinguistic usage in spoken language. There are nevertheless some other areas which remain little or not explored in corpora. Syntax is here illustrated by one study (Sanna Hillberg). Discourse features are exemplified by a corpus-based analysis (Sylvie Hancil), which examines particles that “are in some cases difficult to assign to a word class” (Wales 2006: 190) but remain important-to-grasp aspects of northern colloquial speech. To conclude, the relationship between language and corpus is examined in an article by Ole Schützler, Ulrike Gut and Robert Fuchs on Scottish Standard English.
2 Structure of the volume The first part of the volume opens up with four chapters which deal with the ways in which linguistic features become associated with the North and northern characters. Joan Beal discusses the sociohistorical processes whereby northern varieties of English became recognised as distinct from those of other parts of the country. The theoretical framework is that of indexicality and enregisterment, the latter term defined by Asif Agha as a set of “processes through which a linguistic repertoire becomes differentiable within a language as a socially recognisable register of forms” (2003: 231). Although histories of English note that differences between ‘northern’ (e.g. Northumbrian) and ‘southern’ (e.g. West Saxon) varieties are apparent from the earliest records of Old English, there is no evidence that speakers and writers were aware of these distinctions. Beal’s account therefore begins in the Middle English period, when, as Katie Wales (2006: 33, 61) points out, authors such as Trevisa (1385, translating Higden (1327), comment disparagingly on the language of Northumbrians. Linguistic commentary in the Middle and Early Modern English period establishes northern English as the ‘other’, in contrast with the emerging standard of the court and London, but neither recognises the northern origin of certain ‘standard’ features, nor singles out specific features of northern English for condemnation. However, Beal demonstrates that literary representations of northern speech were beginning to establish a ‘repertoire’ of northern features along with a stereotyped persona of the Northerner.
Introduction
5
As Johnstone et al. (2006) point out, dialect contact is a prerequisite of enregisterment, so the greater social and geographic mobility of the population in the Late Modern period led to increased awareness of northern English, and of distinct dialects within the North. Evidence for this awareness, and the social values associated with northern speech, becomes much richer in this period, so Beal discusses on the one hand, how normative texts such as pronouncing dictionaries stigmatised northern features, and on the other hand how the representation of northern dialect features in literature and popular entertainment disseminated both positive and negative stereotypes of northern English and Northerners. Katie Wales re-considers stereotyping in the works of Charles Dickens. It is now over 40 years since Stanley Gerson produced the only detailed analysis of sound and spelling in the works of Charles Dickens. In this chapter Wales takes a fresh look at Dickens’ handling of northern British dialect speech (including morphology, syntax and lexis) in the particular context of an approach to literary dialect and linguistic stereotyping that has been so far under-used. So she focusses on social cognition and schema theory, which is concentrated on people’s mental biases and attitudes. Looking at Dickens’ characters Wales shows the beginning of the process of literary formation in the mind of a novelist. She unpicks the creative process by analysing mainly the portrayal of John Browdie, the North Yorkshire corn-factor, in Nicholas Nickleby (1838–9); and of Stephen Blackpool, the Lancashire cotton-mill worker, in Hard Times (1854). She argues that characters like Browdie and Blackpool, far from being either one-dimensional or a hotch-potch of random features, are constructed intertextually though schemas and social ideologies, mental images of common beliefs about the under-classes, and about Northerners in general; and also through cultural practices, including other literary texts. The result for each novel is a different set of schemas, and a different linguistic repertoire. Dickens’ idio-regiolects have their own rules, their own ‘gaps’. Importantly, however, Dickens is not so much ‘representing’ northern speech as creating it, so traditional questions of ‘authenticity’ or ‘accuracy’ are not relevant. There are further implications even for the creative processes of those writers who represent the dialect speech of an area with which they are familiarly associated. Common cultural images can still be influential on the process of ‘representation’. Michael Pearce’s chapter is a first foray into an under-researched aspect of the region’s local linguistic ecology. Research on language attitudes and perceptual dialectology has shown that North-East English is one of the most widely recognized and positively evaluated varieties in Britain. There is also a rich tradition of dialect writing associated with the region, and a long history of both ‘folk’ and scholarly attention to local forms of language – for example, it
6
Sylvie Hancil and Joan C. Beal
is the only part of England to have a major corpus devoted to it – the Diachronic Electronic Corpus of Tyneside English. However, one element is missing from the otherwise well-charted dialectological terrain: an account of localized forms in the ‘linguistic landscape’ (that is, the public display and representation of written language on road signs, advertisements, house names, vehicles, and so on). Pearce’s chapter describes and contextualizes a corpus of signs compiled in 2014–15, showing how they draw extensively on a set of features which previous studies have revealed to be enregistered as part of north-east dialect. The implications of the findings are discussed and the – perhaps surprisingly – infrequent appearance of such forms in the linguistic landscape is addressed. Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven’s chapter deals with the non-standard spelling practices found in a corpus of Contemporary Humorous Localised Dialect Literature (CHLDL) from Liverpool. Dialect features are represented to different extents in the Contemporary Humorous Localised Dialect Literature (CHLDL) that is commonly published for northern (and other) varieties of English (Honeybone and Watson 2013). In this article, Honeybone and Watson consider a corpus of CHLDL for Liverpool English and show that the notion of differential phonological salience (Trudgill 1986, McMahon 2000, Kerswill and Williams 2002) can account for some of these differences. The authors consider a number of Liverpool English consonantal dialect features which are represented to some degree in the spelling conventions found in the corpus. Although these features are all robustly characteristic of the dialect’s phonology (Watson 2007), some of them are represented very frequently in spelling (for example: DH–stopping, in which a lenis dental stop is used in place of the lenis fricative found in most other varieties of English), while others are represented hardly at all (for example: Liverpool lenition, in which underlining stops such as /t, k, d/ are realized as affricates or fricatives). After a general introduction to the issues, and a consideration of the features just mentioned, Honeybone and Watson focus on the representation and status of ‘T – to – R’, which is common across dialects of English in the North (Wells 1982). In T – to – R, an underlying /t/ in a small set of lexical items (such as but, not, get, that) is realized, when followed by a vowel, using the same surface form as the underlying rhotic. T – to – R is frequently represented in CHLDL (in forms such as gerrup ‘get up’ and norra ‘not a’). A precise quantitative report on the occurrence of T – to – R is presented and it is compared to other consonantal features such as lenition. Honeybone, Watson and van Eyndhoven argue that, while T – to – R and lenition have a number of things in common, they differ in terms of their phonological salience and this accounts for the different extents to which they are written in Liverpool CHLDL texts.
Introduction
7
The second part of the volume focusses on the phonological aspect of northern English and its varieties, paying attention to the concept of levelling. Sandra Jansen deals with changes in /r/ in Carlisle English. Diffusion as a geolinguistic process has been discussed in detail in Britain in recent years as features such as t-glottalling, th-fronting and r-labiodentalisation have spread across the country. However, in particular, language internal factors in the levelling process of local features have been overlooked in many cases. The aim of this chapter is to investigate real time changes of /r/ in Carlisle English during the twentieth century and in particular pay attention to the internal and external constraints in the levelling of [ɾ] and postvocalic /r/ in this variety. The status of /r/ is highly variable in Carlisle English. Hughes et al. (2012: 124) describe the use of /r/ of an 80 year-old female speaker from this city. She varies between [ɾ] and [ɹ] in pre-vocalic position and between [ɹ], [ɹ̥] and completely r-less pronunciation in post-vocalic position. This very short and limited analysis already gives some indications of the varying nature of /r/ in Carlisle English. The analysis is based on sociolinguistic interviews conducted between 2007 and 2010. To add the real time view to this data, oral history recordings conducted in the 1980s by the local museum have also been included in the analysis. The speakers in these recordings were born around the turn of the twentieth century and were mainly factory workers, i.e. speakers who probably represent the Carlisle dialect in its most vernacular form. Marten Juskan’s chapter investigates change in the NURSE and happY vowels in Liverpool English (Scouse) across three generations of speakers and discusses if and how the results might be connected to questions of salience, local identity, and Liverpool’s changing fortunes in the twentieth and the 21st century. Based on a sample of 20 sociolinguistic interviews, this study finds that younger speakers use more local variants of the NURSE-SQUARE merger, a highly salient variable (Honeybone and Watson 2013, Watson and Clark 2013), than their parents’ or grandparents’ generation. Realisations of happY, on the other hand, become laxer, which is a change away from the (tense) traditional local norm, and towards the majority of the other varieties spoken in northern England (Trudgill 1999). These changes in production are linked up with qualitative data from the interviews, which indicate that younger Liverpudlians not only readily express pride in their city and its accent, but that they also feel a strong connection to the north of England more generally. Phonetic change in the two vowels under scrutiny is interpreted as being governed by a combination of salience and questions of identity: younger speakers use Scouse variants of the socially salient NURSE vowel to express their ‘primary’ identity as Liverpudlians, and laxer realisations of less-salient happY to also associate themselves with other towns
8
Sylvie Hancil and Joan C. Beal
and cities in the North – a strategy which allows them to simultaneously express both their local, and their regional identity linguistically. The third part includes studies on syntax and discourse features. Sanna Hillberg discusses Scottish Standard English (SSE), an under-researched variety of British English, and the possible reflections of Scottish identities in its syntax. In her databases she uses the Corpus of Scottish English Online Press News and some spoken parts of the Scots Corpus of Texts and Speech. She discusses the use of relative clauses and modal verbs in written and spoken varieties in the so-called Educated Scottish Standard English (Aitken 1984). These are syntactic features that have been reported to be “massively different” in Scots in comparison to Standard English (Herrmann 2003; Tagliamonte et al. 2005; Miller 2008: 304). Scots is the dialect or language of Lowlands of Scotland, which is not standardised according to the rules of Standard English, and whose status has been debated for decades. Thus far, there has been very little scholarly interest in SSE and in the influence of Scots on this. Scottish identity with respect to the use of Scots lexical items in SSE has been discussed by Douglas (2009), but the syntactic features have not been investigated thus far. The Scotland’s Census 2011, published in 2013, show that the majority of the population in Scotland, approximately 80 per cent, identify themselves at some level of ‘Scottishness’ and only eight per cent as ‘British only’. However the percentages of those who are competent in the indigenous Scottish languages, i.e. Scots and Scottish Gaelic, are remarkably lower, while nearly all are competent in English. The findings of her PhD research on SSE relativisation strategies indicate that some features of Scots relativisation infiltrate written use. Scots influence is evident, for example, in the frequency of relativiser that in restrictive relative clauses as well in prepositional complementation. The use of the features that are strongly attested in spoken Scots, may be a reflection of Scots identity in educated writing, but also the use of predominantly StE features in news may reflect the fact that English, at least in educated writing, today forms the Scottish linguistic identity. Sylvie Hancil examines the semantics and syntax of final but in Northern English. It is a fact that final adverbials in English are traditionally associated with VP-oriented adjuncts, especially manner, time and space adverbials, whereas the presence of clause-oriented adverbials is said to be rare and even problematic in such a position. But the examination of recent spoken English corpora has shown that they are attested in various dialects of English, such as American English and Australian English (Mulder and Thompson 2008, Mulder, Thompson and Williams 2009). The present study complements existing studies and focusses on northern British English (Tyneside English and Scottish English), as illustrated in the Newcastle Electronic Corpus of Tyneside English (NECTE) and the spoken
Introduction
9
part of the Scottish Corpus of Texts and Speech. Various discourse functions of the adverbial can be identified. Besides, the use of sentence-final adverbials by the speaker is a way to manage recipient turns. The turn-taking system is also examined in relation to the study of the CTRP (Complex Transition Relevance Place) theory (Ford and Thompson 1996). Moreover, the use of sentence-final constructions by the speaker contributes to maintain the preference organization of conversations. The fourth part of the volume relies on a sociolinguistic approach to best describe how Northerners describe their identity. Marie Jensen re-considers the interaction between social class and language awareness in Newcastle. Investigations of the impact of socio-economic class membership on language use are numerous within the field of sociolinguistics. After decades of research, certain patterns seem have been established when it comes to the use of standard and non-standard forms by members of different socio-economic classes (Ash 2013, Chambers 2003). But how can socio-economic class best be described? As observed by Ash (2013) there seems to be little consensus among sociolinguistics as to the best ways to operationalise this variable. The consensus only diminishes when looking to sociology and the intricate parameters used within this field. However, one might wonder if speakers are solely products of their social environments? Does a speaker’s own definition of him/herself play a larger role than a supposedly objective classification based on income, education, etc.? A questionnaire study carried out in Newcastle upon Tyne in the summer of 2012 forms the foundation for an investigation into Tyneside speakers’ awareness of vernacular morphosyntactic forms. Statistical analyses showed interesting patterns with regard to speakers’ definition of own social class and their level of education and the different groups’ performance on different questionnaire tasks. In short, there was only a very weak correlation between participants’ own definition of social class and that based on their level of education. Furthermore, the group of participants who identified as middle class but were not highly educated performed significantly differently on the tasks compared to the rest of the participants. These results raise questions not only about methodology (e.g. how to operationalize class and the usefulness of this distinction in investigations of language use in the north of England, c.f. Wales (2000) and Lancaster (2005)) but also about speaker agency. The results suggest that perhaps taking into consideration how speakers define themselves in terms of different parameters (e.g. social class) might be a useful way to investigate the socio-cognitive factors which influence language use and perception. Mercedes Durham shows real-time change in dialect use in Shetland. Even as regional dialect distinctions in the United Kingdom are reported to be levelling, it is clear that, as a whole, dialect use is generally viewed to be more acceptable
10
Sylvie Hancil and Joan C. Beal
now than it was in the past. In what ways can these seemingly opposite tendencies be related? And how does this shift away from the dialect take place in areas which were traditionally diglossic or bidialectal? This chapter considers the degree to which shifts in dialect use are correlated with shifts in domains of use by examining over 850 attitudinal questionnaires completed by Shetland schoolchildren in 1983 and in 2010. Because research has demonstrated that there has been a widespread shift away from the dialect by the youngest generations, particularly in the main town of Lerwick (Smith and Durham 2011, Tait 2001, van Leyden 2004), the longitudinal perspective is particularly valuable. In the nearly thirty years between 1983 and 2010, the situation on the islands has changed, linguistically, but also socially: there has been an influx of newcomers to the islands following the discovery of oil in the North Sea in the late 70s. By comparing the data from 1983 to that in 2010 it is possible to examine whether the dialect is seen as being used less, but also whether the range of situations where it is thought to be used have changed. Among the questions students were asked, several focus on situations when it would and would not be appropriate to use the dialect. While the rate of children reporting that there are circumstances where it is not proper for a Shetlander to use Standard English has gone down from 1983 to 2010 (40% to 18% for the children born in Shetland to Shetland parents), the rate of children reporting that there are no circumstances where the use of the Shetland dialect is unacceptable has gone up (from 14% to 35% for the Shetland origin children). Although the rates are less dramatic, the direction of these shifts is the same in the children who are not originally from the Islands. This underlines that even as the dialect is used less, it is seen as acceptable in a wider range of places. It could mean that the dialect itself is less marked and already ‘levelled’ to some extent, but not necessarily: within these questions, students also give examples of when it would be particularly appropriate or inappropriate to use the dialect. Although not all of them responded to this part of the question, the answers provide insight into shift over time and into the perceived acceptability of the dialect in a range of domains. By examining these responses more closely this chapter establishes whether the domains where dialect use is acceptable have extended or whether they are merely perceived to have done so. The fifth part comprises a study on the relationship between language and corpus. In their chapter, Ole Schützler, Ulrike Gut and Robert Fuchs argue that Scottish Standard English (SSE) is often regarded as a standard Scottish English accent combined with a standard grammar shared with the rest of Britain. In consequence, SSE is underexplored (except, perhaps, for its phonology) and has virtually no place in the research field of World Englishes. They argue that this is because linguists studying grammatical variation concentrate on Scots,
Introduction
11
rather than SSE. In the second part of the chapter, the Scottish component of the International Corpus of English (ICE-Scotland) is presented as a tool that can remedy this situation. Its text categories cover domains of usage ranging from the standard pole to the interface of standard and non-standard, and their exploration will contribute to a better definition of SSE and its boundaries on all linguistic levels, at the same time enabling comparisons with other standard varieties of English.
References Agha, Asif. 2003. ‘The social life of cultural value’ Language and Communication 23: 231–273. Aitken, A.J. 1984. Scottish accents and dialects. In Peter Trudgill (ed.), Language in the British Isles. Cambridge: Cambridge University Press, 94–114. Ash, Sharon. 2013. Social Class. In Jack K Chambers and Natalie Schilling-Estes (eds.), Handbook of language variation and change, 350–367. Oxford: Blackwell Publishers. Baker, Alan R. H. and Mark Billinge (eds.) 2004. Geographies of England: The North-South divide, material and imagined. Cambridge: Cambridge University Press. Beal, Joan C. 1993. The Grammar of Tyneside and Northumbrian English. In James Milroy and Lesley Milroy (eds.), Real English: the grammar of English Dialects in the British Isles, 187–242. London: Longman. Beal, Joan C. 2008a. English dialects in the North of England: phonology. In Bernd Kortmann and Clive Upton (eds.), Varieties of English: the British Isles, 122–144. Berlin: Mouton de Gruyter. Beal, Joan C. 2008b. English dialects in the North of England: morphology and syntax. In: Bernd Kortmann and Clive Upton (eds.), Varieties of English: The British Isles, 122–144. Berlin: Mouton de Gruyter. Bullokar, John. 1616. An English expositor. London: John Legatt. Chambers, Jack K. 2003. Sociolinguistic theory. 2nd ed. Oxford: Blackwell. CLES (Centre for Local Economic Strategies). 2014. Due North: report on the inquiry on health equity for the North. Liverpool: University of Liverpool and Centre for Local Economic Strategies. Corrigan, Karen P. 2010. Irish English, volume 1 – Northern Ireland. Edinburgh: Edinburgh University Press. Dorling, Danny. 2010. Persistent north-south divides. In Neil M. Coe and Andrew Jones (eds.), The economic geography of the UK, 12–28. London: Sage. Douglas, Fiona. 2009. Scottish newspapers, language and identity. Edinburgh: Edinburgh University Press. Ellis, Alexander J. 1890. English dialects: their sounds and homes. London: English Dialect Society/Kegan Paul, Trench, Trübner and Co. Ltd. Ford, Cecilia and Sandra A. Thompson. 1996. Interactional units in conversation: syntactic, intonational, and pragmatic resources for the management of turns. In Elinor Ochs, Emanuel A. Scheglof and Sandra A. Thompson (eds.), Interaction and grammar, 134–184. Cambridge: Cambridge University Press.
12
Sylvie Hancil and Joan C. Beal
Görlach, Manfred. 1990. Introduction to Early Modern English. Cambridge: Cambridge University Press. Herrmann, Tanja. 2003. Relative clauses in dialects of English: a typological approach. Freiburg: University of Freiburg PhD Thesis. Available at http://www.freidok.uni-freiburg.de/volltexte/ 830/, accessed Feb 2, 2012. Hickey, Raymond (ed.) 2015. Researching Northern Englishes. Amsterdam: Benjamins. Honeybone, Patrick and Kevin Watson. 2013. Salience and the sociolinguistics of Scouse spelling: exploring the phonology of the contemporary humorous localised dialect literature of Liverpool. English World-Wide 34(3): 305–340. Jones, Hugh. (1724) 1967. Accidence of the English tongue. Menson: Scolar Press Reprint no. 22. Johnstone, Barbara, Jennifer Andrus and Andrew E. Danielson. 2006. Mobility, indexicality, and the enregisterment of “Pittsburghese”. Journal of English Linguistics 34: 2, 77–104. Lancaster, Bill. 2005. Newcastle – capital of what? In Robert Colls and Bill Lancaster (eds.), Geordies: roots of regionalism, 53–69. Edinburgh: Edinburgh University Press. Mason, Colin M. and Richard T. Harrison. 1991. Venture capital, the equity gap and the “northsouth divide” in the United Kingdom. In Mitford B. Green (ed.), Venture capital: international comparisons, 202–248. Abingdon and New York: Routledge. Millar, Robert McColl. 2007. Northern and Insular Scots. Edinburgh: Edinburgh University Press. Miller, J. 2008. ‘Scottish English: morphology and syntax’. In Bernd Kortmann and Clive Upton (eds.), Varieties of English. 1, The British Isles, 299–327. Berlin: Mouton de Gruyter. Morley, Paul. 2013. The North and (almost) everything in it. London: Bloomsbury. Montgomery, Chris. 2006. Northern English dialects: a perceptual approach. Sheffield: University of Sheffield PhD thesis. Mulder, Jean and Sandra A. Thompson. 2008. The grammaticalization of final but in English conversation. In Ritva Laury (ed.), Crosslinguistic studies of clause combining: the multifunctionality of conjunctions. 179–204. Amsterdam: John Benjamins. Mulder, Jean, Thompson Sandra A. and Cara Penry Williams. 2009. Final but in Australian English conversation. In Peters, Pam, Peter Collins and Adam Smith (eds.), Comparative studies in Australian and New Zealand English: grammar and beyond, 337–358. Amsterdam: Benjamins. Pollard, A. J. 1997. The characteristics of the fifteenth-century North. In John C. Appleby and Paul Dalton (eds.), Government, religion and society in Northern England 1000–1700, 131–143. Stroud: Sutton Publishing. Sisam, Kenneth (ed.). 1967. Fourteenth-century verse and prose. Oxford: Clarendon Press. Smith, Jennifer and Mercedes.Durham. 2011. A tipping point in dialect obsolescence? Change across the generations in Lerwick, Shetland. Journal of Sociolinguistics 15, 197–225. Tagliamonte, Sali, Jennifer Smith and Helen Lawrence. 2005. No taming the vernacular! Insights from the relatives in northern Britain. Language Variation and Change, 17: 75–112. Tait, John. 2001. Whit is Shetlandic? Lallans 58, 7–16. Trudgill, Peter. 1990. The dialects of England. Oxford: Blackwell. Trudgill, Peter. 1999. The dialects of England. Second Edition. Oxford: Blackwell. Upton, Clive. 1995. Mixing and fudging in Midland and Southern dialects of England: the cup and foot vowels. In Windsor Lewis, Jack (ed.). Studies in general and English phonetics, 385–394. London: Routledge. van Leyden, Klaske. 2004. Prosodic characteristics of Orkney and Shetland dialects. An experimental approach. Leiden: Leiden University PhD Dissertation (LOT Dissertation Series 92, Utrecht: LOT.)
Introduction
13
Wakelin, Martyn F. 1972. English dialects: an introduction. London: Athlone Press. Wales, Katie. 2000. North and South: an English linguistic divide? English Today 16(1). 4–15. Wales, Katie. 2006. Northern English: a social and cultural history. Cambridge: Cambridge University Press. Watson, Kevin and Lynn Clark. 2013. How salient is the NURSE~SQUARE merger? English Language and Linguistics 17: 297–323. Wells, John C. 1986. Accents of English 2: The British Isles. Cambridge: Cambridge University Press.
I Enregisterment
Joan C. Beal
2 Northern English and enregisterment Abstract: In the introduction to this collection, Hancil and Beal discuss the cultural and linguistic meanings of terms and concepts such as the North, the North-South divide and northern English(es) in the context of the British Isles. This chapter presents, within the theoretical framework of indexicality and enregisterment, an account of the historical evolution of these meanings. Enregisterment is defined by Agha as a set of “processes through which a linguistic repertoire becomes differentiable within a language as a socially recognisable register of forms” (2003: 231). A historical account of enregisterment concentrates on metalinguistic comments on language at various points in time in order to trace the development of “socially recognisable” varieties such as northern Englishes. Johnstone identifies the key research question in studies of enregisterment as “How do particular words, ways of pronouncing words, grammatical patterns, and patterns of intonation come to point to particular identities and activities?” (forthcoming) In the context of northern Englishes, this entails tracing the development of concepts of northern-ness and of the ways in which linguistic features of northern varieties became associated with these concepts. This chapter begins with an account of how ‘the North’ has been defined culturally, historically and socially as well as linguistically. It then goes on to outline the framework of indexicality and enregisterment in order to set the theoretical context for the historical account of the enregisterment of northern English.
1 The North In her historical account of the North-South divide, Jewell notes that this division is “literally as old as the hills” (1994: 6, 28). She is referring here to the geology of Britain, whereby the North and West are divided from the South and East by the Highland Line separating the northern uplands from the southern lowlands, so that even from prehistoric times there would have been economic and cultural differences between these areas. There is, of course, no evidence from prehistoric times of attitudes towards the North, as such evidence depends on written records. The Romans divided conquered Britain into two provinces: Britannia superior in the South, and Britannia inferior in the North. The northern Joan C. Beal, University of Sheffield DOI 10.1515/9783110450903-002
18
Joan C. Beal
boundary of the Roman Empire was Hadrian’s Wall, beyond which was Britannia barbara. These names are deceptively evocative to the modern reader, but at the time superior indicated that the South was closer to Rome and barbara that the region beyond the wall was not Roman. However, Wales suggests that the Roman terms “do also appear to have their evaluative connotations” (2000: 8). The Anglo-Saxon kingdom of Northumbria, stretching from the Humber to the Forth, was for a time the most culturally rich area, such that Alfred of Wessex, bemoaning the decline in literacy in the country, noted that there were very few “this side [i.e. south] of the Humber” who could understand church services in English or translate Latin to English, and not many “beyond” (Onions [ed.] 1959: 4). The contrast implies that the North was viewed as more cultured and literate than the South. However, Alfred’s revival of learning was centred on Winchester, and the greater influence of Viking raiders and later settlers in the North, consolidated by the establishment of the Danelaw in 886, led to the North becoming associated with a non-literate, pagan and, from a southern point of view, alien culture. Although the Normans, as the etymology of their name (“north men”) suggests, were themselves of Viking descent, the conquest of 1066 further consolidated the South, and London in particular, as the centre of power in England. For centuries after the Norman Conquest, the North was to be a thorn in the side of London-based monarchs. Jewell tells us that “the North gave William some trouble” (1994: 35), resulting in several campaigns against northern rebels, culminating in the harrying of the North in 1069–70. Later northern rebellions included the Pilgrimage of Grace in 1536 and the Rising of the North in 1569, both of which were met with harsh reprisals from Henry VIII and Elizabeth I respectively. These rebellions, along with the skirmishing between families on either side of the border with Scotland, gave the North and Northerners a reputation for lawlessness and truculence. Jewell tells us that the southern author William of Malmesbury, writing in the twelfth century, coloured his historical account of Northumbrian challenges to the southern rulers of the tenth and eleventh century with derogatory terms. Malmesbury described the Northumbrians as “a people ever ripe for rebellion” and “a ferocious race of people” (Jewell 1994: 38). By Tudor times, the lawlessness of the Northumbrians had entered the popular culture of the time via the Border ballads, songs about the warring families of the English-Scottish border probably dating back to at least the fifteenth century, and orally transmitted, but distributed as broadsides from the late sixteenth century and later published in collections such as those of Percy (1765), Scott (1802) and Child (1882–1898). Lamont quotes as evidence for the popularity of these songs in sixteenth-century London Sir Philip Sydney (1580):
Northern English and enregisterment
19
“Certainly, I must confess my own barbarousness, I never heard the old song of Percy and Douglas that I found not my heart moved more than with a trumpet; and yet it is sung but by some blind crowder [=fiddler] with no rougher voice than rude style” (cited in Lamont 1997: 61). The “old song of Percy and Douglas” would be either The Battle of Otterburn (Child 161) or Chevy Chase (Child 162)1. It is interesting to note Sydney’s use of the term barbarousness in confessing to admire a song set in the Britannia barbara of Roman times. By the sixteenth century, terms such as barbarous were increasingly used to refer to places distant from the “civilized” capital, and the far North was such a place. Eyewitnesses also testified to the barbarousness of Northumberland. Watson cites Sir Robert Bowes, Elizabeth I’s ambassador at the Scottish court, describing “the country of North Tynedale” as “plenished with wild and misdemeanoured people” (Watson 1974: 119). In the eighteenth century, the Industrial Revolution was, to a certain extent, changing the image of the North from that of a barbarous wasteland to a place of relative prosperity. Defoe described Newcastle as “a large and exceeding populous town” and the Tyne as “a deep and noble river” with “a Bridge, which consists of seven Arches, as large, at least, as those of London Bridge” (1748: 220–1, italics in original). However, even at this early stage in the Industrial Revolution, Defoe noted the drawbacks of pollution: “the Situation of the Town being on the Declivity of two high Hills [. . .] and the Buildings being very close and old, render it incommodious, to which the Smoke of the Coals contributes not a little” (1748: 223). This association of the North with industrial grime was to continue throughout the nineteenth and twentieth centuries as the prosperity of the North waxed and waned. Wales notes that “many of the current stereotypes of the north of England derive from the industrial revolution and the huge expansion of industry and growth of the Midland and northern towns.” (2000: 5–6) Sayings such as “it’s grim up north” and “where there’s muck there’s brass [= money]” testify to these stereotypes. Where there is industry, there are industrial workers, and the association of the North with the industrial working class persists, even in today’s postindustrial age. As Wales points out “for many people even today ‘northern’ and ‘working class’ are synonyms” (1999: 2–3). The North is also associated politically with the party which has its origins in working-class activism: Labour. As Montgomery points out, the Labour Party was founded in Bradford and by the time of its success after the end of the second World War “was an institution which was synonymous with the north” (2006: 13). Although the Labour Party downplayed these associations in the late twentieth and early twenty-first 1 It is customary in traditional music scholarship to refer to the Border ballads by their numbers in Child’s collection.
20
Joan C. Beal
centuries, the geographical distribution of results from the 2010 general election shows a very clear pattern of Labour-held constituencies in England being located in the industrial heartlands of the North-west, North-east and West Yorkshire.2 From the first recorded distinctions between Britannia superior in the South and Britannia inferior in the North of Roman Britain, a set of dichotomous stereotypes has evolved, contrasting the two regions in ways that define the North as cold, harsh, uncivilized, poor, working-class and socialist, and the South as warm, soft, civilized, rich, middle-class and Conservative. These associations have some basis in (historical) fact, but have been built up over the centuries via discourses and images which promote and propagate these contrasts. The process whereby these stereotypical associations are constructed is enregisterment. In the next section, I explain this process and its application to linguistic attitudes.
2 Indexicality and Enregisterment Recent research in the field of language and ideology has shed light on the relationship between linguistic features and their social connotations. This languageideological approach was first introduced by Silverstein (1976), developed by Agha (2003) and further interpreted by Lesley Milroy (2000, 2004), Johnstone and her co-authors (Johnstone et al. 2006), amongst others. Silverstein posits successive orders of indexicality whereby linguistic forms are indexed, or associated with, social categories. At the nth order of indexicality, there is a correlation between a particular linguistic form and a social category, observable by an outsider such as a linguist, but not noticed by speakers themselves. Silverstein (2003: 2005) states that such observations are “scientific” and that the indexicality is “presupposing”, i.e. it depends on a pre-existing set of values. At the n + 1-order the feature has been given “an ethno-metapragmatically driven native interpretation” (2003: 212). In other words, speakers come to rationalise and justify the link between the linguistic form and the social category in terms of native ideologies such as correctness of speech. At the (n + 1) + 1 order of indexicality, forms which have been indexed at the n + 1-order become associated with another ideological schema. In Silverstein’s model, each increase in order is a meta-level interpretation of the one below. This model has been further interpreted by Lesley Milroy (2000) and by Johnstone et al. (2006) in terms of three
2 See the map at http://news.bbc.co.uk/2/shared/election2010/results/region/48.stm This pattern was maintained in the 2015 general election.
Northern English and enregisterment
21
orders of indexicality. The first order is described by Johnstone et al. as “potential indexicality”: a correlation between a linguistic form and a social feature exists “because linguists have noted it” (2006: 83). The second order occurs “when people begin to use first-order correlations to do social work” (2006: 83), whereas the third order involves explicit metalinguistic comment about the correlation: the link between a linguistic feature or features and a certain social category is by this stage the subject of overt comment. This may take the form of direct metalinguistic commentary, or it may involve the use of such features in literature and other media. At first glance, these orders of indexicality seem to parallel the stages in linguistic change posited by Labov (1972: 178–180): indicators, markers and stereotypes3. However, Labov’s definition of stereotype implies that the linguistic features concerned are negatively valued, become increasingly divorced from linguistic reality, and may eventually disappear, whereas the schemas of indexicality proposed by Silverstein and by Johnstone et al. have no such implications. The concept of enregisterment, first introduced by Agha (2003), sheds further light on this process. Agha points out the evaluative connotations of the word accent. The folk-term ‘accent’ does not name a sound pattern alone, but a sound pattern linked to a framework of social identities. The social identity is recognised, indexically, as the identity of the speaker who produces the utterance in the instance, and described, metalinguistically, through the use of identifying labels (2003: 232–3).
Whereas the orders of indexicality outlined above refer to the association of linguistic forms with non-linguistic characteristics and the interpretation of these within ideological frameworks, enregisterment involves the establishment of a register, a whole repertoire of linguistic forms associated with “characterological figures or social personae” (Agha 2003: 243). Agha (2003) uses as a case study the enregisterment of Received Pronunciation as an elite register associated with the characterological figure of the [British] Public School Man in the nineteenth and early twentieth centuries. Agha argues that “the dissemination or spread of a register depends on the circulation of messages typifying speech” (2003: 243, emphasis in original). Each time such a message is transmitted and received, the process of enregisterment is taken further. These messages may take the form of face-to-face communication, but can also have the more enduring form of printed text. In establishing historical processes of enregisterment, as Agha did in his study of RP, we are, of course, dependent on the existence of written records. In the following sections, I examine the earliest evidence for the enregisterment of northern English as a variety or 3 See Johnstone and Kiesling (2008: 8–9) for a comparison of Silverstein’s and Labov’s schemas.
22
Joan C. Beal
set of varieties associated with the stereotypical characteristics of northernness outlined in section 2.
3 Potential indexicality: the beginnings of northern English We saw in section 2 that the first evidence of an evaluative distinction between the North and South of England came with the division of Roman Britain into Britannia superior and Britannia inferior. However, this had no bearing on the evaluation of language, not least because English as a language did not yet exist. The non-Roman inhabitants of Britain would have spoken Celtic languages, and the elite would have acquired Latin. Meanwhile, the ancestors of those who would bring what with hindsight we call English to these shores were speaking Germanic languages in continental Europe. What we do know is that the language brought to England after the departure of the Romans was not monolithic, since the settlers came from various tribes associated with different continental territories. These settlers were not literate, so the first records of what we now call Old English or Anglo-Saxon come from the period following the introduction of Christianity. At this stage, dialectal differences between texts from the North and the South are already evident, as can be seen from a comparison of the Northumbrian and West Saxon versions of Caedmon’s Hymn below. Northumbrian version Nu scylun hergan hefaenricaes uard
West Saxon version Nu sculon herigean heofonrices weard
metudas maecti end his modgidanc
metodes meahte and hid modgeþanc
uerc uuldurfadur sue he uundra gihuaes
weorc wuldor fæde swa he wundra gehwæs
eci dryctin or astelidae
ece drihten, or onstealde.
he aerist scop aelda barnum
He ærest sceop orðan bearnum
heben til hrofe haleg scepen.
heofen to hrofe, halig scyppend;
tha middungeard moncynnaes uard
þa middangeard moncynnes weard,
eci dryctin aefter tiadae
ece drihten, æfter teode
firum foldu frea allmectig
firum foldan, frea ælmihtig4
4 A rough translation of Caedmon’s hymn into modern English: Now [we] must honour the guardian of heaven, the might of the architect and his purpose, the work of the father of glory as he, the eternal lord, established the beginning of wonders. He, the holy creator, first created heaven as a roof for the children of men. Then the guardian of mankind, the eternal lord, the lord almighty, afterwards appointed the middle earth, the lands, for men.
Northern English and enregisterment
23
Caedmon’s Hymn is the earliest extant poem in English. It was probably composed in the second half of the seventh century, but the first written version we have is the Northumbrian version above, from Bede’s Historia Ecclesiastica Gentis Anglorum. The West Saxon version is taken from the ninth-century translation of Bede’s work. There is, therefore, a diachronic dimension to the comparison of the two versions, but sources of Old English dialects are so scarce that these two texts have often been used to demonstrate the differences between northern and southern dialects of Old English. No knowledge of Old English is needed to see that there are regular contrasts between these two versions, such as we would expect to find when comparing dialects of the same language. In the first line, the words hergan and uard in the Northumbrian version correspond to herigean and weard in the West Saxon. Where Northumbrian has a monophthong spelled , West Saxon has a diphthong , and the same correspondence can be found in the fifth line, where barnum in Northumbrian corresponds to bearnum in West Saxon. A similar monophthong/ diphthong contrast can be noted between Northumbrian heben and uerc and West Saxon heofen and weorc. These observations are made by modern readers across a considerable distance in time from the composition of the original manuscripts, and so are what Silverstein would describe as “scientific” or “nth order” (2003: 205). As a twenty-first century scholar, I am noting the differences between these two texts and interpreting them according to my presuppositions about the diatopic differences between dialects. There is no evidence that these distinctions were noticed, let alone assigned n + 1 or second-order indexicality by Anglo-Saxons (though presumably scribes from other areas could have seen the Northumbrian gloss of Caedmon’s hymn and perhaps remarked on its strangeness without having any concept of a Northumbrian dialect as such5). Whether the West Saxon of Alfred’s reign had the status of a standard variety is a matter of debate (see Gneuss 1972, Lenker 2000 and Gretsch 2006), but, since there is no contemporary metalinguistic commentary on either the status of West Saxon or the dialectal differences between this and other varieties of Old English, any indexicality attached to the forms discussed above remains potential. Indeed, as Hogg points out, although the linguistic similarities between certain texts from this period suggest the existence of focussed varieties, the idea of a standard variety “would be a mere anachronism in the Old English context” (2002: 22). As we shall see in the following sections, enregisterment, depending as it does on the transmission of messages about language, can only happen when there is contact between speakers (or writers and readers) with knowledge of different language varieties and where those varieties have been 5 I am grateful to Christine Wallis for information on the Old English Bede and its transmission.
24
Joan C. Beal
assigned indexical values. From the Old English period we have very little evidence of the former and none of the latter.
4 Second or n + 1 order indexicality: Middle English to Early Modern English When outlining the characteristics of second order indexicality, Johnstone et al. note that, at this stage “regional forms become available for social work” (2006: 82). As speakers become aware of the social connotations of linguistic forms, they begin to style-shift, avoiding in formal situations forms that are indexed as incorrect or inferior. Following the Norman Conquest, all varieties of English became indexed as socially inferior to French, because William placed his followers, speakers of Norman French, in all positions of power and prestige. Such metalinguistic commentary as we have from the eleventh to the thirteenth century focuses on the prestige of French. To do social work in this period meant switching from English to French rather than modifying your native variety of English. The late thirteenth-century chronicler Robert of Gloucester provides an example of such metalinguistic commentary: Thus came, lo, England into Normandy’s hand: and the Normans then knew how to speak only their own language, and spoke French as they did at home, and also had their children taught (it), so that noblemen of this land, that come of their stock, all keep to the same speech that they received from them; for unless a man knows French, people make little account of him. But low men keep to English, and to their own language still. (Modern English translation from Barber et al. 2010: 146)
Ranulf Higden’s Polychronicon, written in Latin in the early fourteenth century, and translated into English by John of Trevisa between 1385 and 1387, provides further metalinguistic commentary on the indexicality of French. He states that the English Language is “corrupted” because it has been influenced by Norse and French, then goes on as follows. This corruption of the mother-tongue is because of two things. One is because children in school, contrary to the usage and customs of all other nations, are compelled to abandon their own language, and to construe their lessons and their tasks in French, and have since the Normans first came to England. Moreover, gentlemen’s children are taught to speak French from the time that they are rocked in their cradle and are able to speak and play with a child’s trinket; and rustic men want to make themselves like gentlemen, and strive with great industry to speak French, in order to be more highly thought of. (Modern English translation from Barber et al. 2010: 152–153)
Northern English and enregisterment
25
The last two lines of this extract provide clear evidence that French was indeed being used to do social work – “in order to be more highly thought of”. Whilst this may well have been the case when Higden wrote the original Latin version of this text, Trevisa felt compelled to add a note of his own to the translation: This custom was much in use before the first plague [that is, the Black Death of 1349], and since then has somewhat changed. For John Cornwall, a licensed teacher of grammar, changed the teaching in grammar school and the construing from French into English; and Richard Pencrich learnt that method of teaching from him, and other men from Pencrich, so that now, in the year of Our Lord 1385, in the ninth year of King Richard II, in all the grammar schools of England children are abandoning French, and are construing and learning in English. (Modern English translation from Barber et al. 2010: 153)
Trevisa’s intervention marks an important turning-point: once “gentleman’s children” are no longer speaking French, the association between this language and gentility would diminish.6 It is no coincidence that Trevisa’s text also provides one of the earliest examples of metalinguistic commentary on northern English. Higden’s original text notes that there are differences between northern and southern varieties of English, and that the language of the North, especially at York, is hard for southerners to understand. We could interpret this as a scientific observation with first-order indexicality: Higden has noticed that there are dialectal differences and correlated them with geographical regions. However, Trevisa goes on to augment this comment with evaluative descriptors. “All the language of the Northumbrians, and especially at York, is so sharp, slitting and unshaped, that we Southern men may that language unnethe (=‘hardly’) understand” (Trevisa 1385, trans. Caxton 1480). “Sharp, slitting and unshaped” are alliterative, onomatopoeic and highly suggestive of the stereotyped view of the North as a harsh and uncouth place outlined in section 2 of this chapter. There is no comparable metalinguistic comment from this period about southern English being incomprehensible or sounding harsh to Northerners, but a much-quoted passage from the Second Shepherd’s Play of the Wakefield Mystery Cycle (c. 1430) involves a character disguising himself as an official of the court by assuming a “sothren tothe” (southern tooth). In other words, this character (albeit fictional) is using his awareness of dialectal differences in English to do social work. A century earlier, he would have switched to French to achieve the same end. We can also see in the transmission of the comment about “the language of the Northumbrians” from Higden to Trevisa to Caxton (and their readers) a good example of the communication of metalinguistic messages which according to
6 It would not disappear altogether, though: see Beal (2012a) for an account of attitudes to French in eighteenth-century Britain.
26
Joan C. Beal
Agha (2003) constitutes the process of enregisterment. Caxton himself provides us with further metalinguistic comments about the diversity of dialects in England. In the preface to his translation of Virgil’s Aeneid Eneydos (1490), Caxton tells a story about English merchants visiting a house in Kent to ask for food: And one of them named Sheffelde, a mercer, came into an howse and axed for mete; and specially he axyd after eggys: And the good wyf answerde that she coude not speke no Frenshe. And the marchaunt was angry for he also coude speke no Frenshe but wolde have hadde egges and she understode hym not. And thenne at laste a nother sayd that he wolde have eyren: then the good wyf sayd that she understod hym wel. Loo what sholde a man in thyse dayes now wryte, egges or eyren? Certaynly it is harde to playse every man by cause of dyversite and chaunge of langage.
Caxton does not specify here which dialect the mercer is speaking, only that it is different from that of the “good wyf” in Kent. The fact that he shares his name with a town in Yorkshire and that the form of the word for eggs used by the merchant is one that has the originally Norse form with rather than might suggest that he was from the North7, but Caxton gives no explicit information about the merchant’s provenance, nor any detrimental comment on his dialect. What concerns him here is that English is variable and, as a printer, he needs to use a variety that can be understood by all. Ironically, the form egg was already on its way to becoming the standard form of the word by 1490: the Oxford English Dictionary has the forms with occurring up to the end of the fifteenth century, with the forms as the norm from the sixteenth century onwards (www.oed.com). Likewise, Caxton, whilst complaining about the printer’s dilemma in choosing the most comprehensible form of English, was writing in an embryonic standard form. By the sixteenth century, a standard written form of English had become accepted and was the norm for all printed texts. Although, as the case of eggs demonstrates, certain linguistic forms occurring in this standard variety had their ultimate origin in northern dialects, the standard was indexed as being associated with London as the seat of government. Comments about northern English in this period single it out (along with the dialect of the far South-west) as maximally divergent from the London-based norm. During the sixteenth century, English began to replace Latin as the language of education and of learned and religious discourse. A great deal of metalinguistic commentary is devoted to the question of what style / variety of English is appropriate to fulfil the higher functions formerly carried out in Latin. Thomas Wilson makes the following statement about the importance of provenance: 7 The British Library’s ‘English Timeline’ makes a statement to this effect http://www.bl.uk/ learning/timeline/item126611.html
Northern English and enregisterment
27
The Realme declares the nature of the people. So that some Countrey bringeth more honor with it, then an other doth. [. . .] As it is much better to bee borne in Paris, then in Picardie: in London then in Lincolne. For that both the ayre is better, the people more ciuill, and the wealth much greater, and the men for the most part more wise. ([1560] ed. Mair 1909, 12–13)
Here, Wilson explicitly puts forward the view that Londoners are inherently more “civil” than those born in the provinces, from which it could be inferred that the further from London a person is raised, the less civil (in the sixteenth-century sense of “urban” or “sophisticated”) that person must be. It is no coincidence that Lincoln is in the North as well as alliterating with London. Elsewhere, Wilson writes about “evill voices”, but the only regional variety explicitly mentioned is that of the North: “This man barkes out his English Northern-like, with I say, and thou lad” ([1560] ed. Mair 1909, 219). Here, Wilson singles out specific features as northern and implies an aesthetic evaluation in the phrase “barkes out”– northern speech sounds as rough as a dog’s bark. We are reminded here of Trevisa’s “sharp, slitting and unshaped”: Wilson’s comments index the North and northern English in a way that is consistent with Wales’s assertion that “the North was ‘constructed’ from the medieval period onwards as alien and barbaric” (2006: 65). One of the most frequently cited metalinguistic comments of the sixteenth century is that taken from Puttenham’s Arte of English Poesie (1589). Puttenham is here discussing what kind of English the poet should use. After warning the reader against using the language of the lower classes, Puttenham turns to geographical distinctions: Neither shall he take the termes of Northern-men, such as they use in dayly talke, whether they be noblemen or gentlemen, or of their best clarkes all is a matter: nor in effect any speach used beyond the river of Trent, though no man can deny but that theirs is the purer English Saxon at this day, yet it is not so Courtly nor so currant as our Southerne English.
Puttenham’s attitude to northern English is not wholly negative. His remark about the purity of northern English highlights an alternative indexicality in which the dialects of regions remote from the capital are viewed as more ancient and authentic. Blank (1996: 100) cites Gill in Logonomia Anglica (1616) as stating that poets “use the Northern dialect quite frequently for the purpose of rhythm and attractiveness” because it is “the most delightful, the most ancient, the purest, and approximates most closely to the speech of our ancestors”. Blank notes that several Renaissance authors, most notably Spenser, made use of northernisms especially for the purpose of archaism, but that “the literary use of northern words and archaisms remained controversial throughout the period”
28
Joan C. Beal
(1996: 100). As Puttenham’s remarks suggest, the corollary of northern English’s archaic purity is that it is neither “courtly”, that is, associated with the fashionable manners of the royal court, nor “current”, meaning “up to date”. So, between the thirteenth and the seventeenth centuries, as first French and then Latin begin to lose their status as elite languages for official and learned discourse, varieties of English become indexed according to prevailing ideologies of civility, courtliness, purity and authenticity. Northern English is indexed on the one hand in accordance with the long-standing view of the North as uncivil, harsh and rough, and on the other hand as ancient, pure and authentic. In either case, it is contrasted with the emerging standard variety, geographically centred on London, so that the North-South divide is applied to language. These metalinguistic comments provide evidence for what Silverstein would term n + 1th order indexicality: observations about the differences between dialects have been interpreted in the light of ideologies concerning the nature of places and their inhabitants.
5 A new order of indexicality: eighteenth-century prescriptivism According to Silverstein, what happens at the n + 1 + 1 order of indexicality is that forms which have been indexed at the n + 1-order become associated with another ideological schema. In the course of the eighteenth century, specific features of northern English are singled out for metalinguistic commentary within the ideological framework of prescriptivism. From this point of view, the variety of English used by educated Londoners and the elite more generally was now indexed not only as the most courtly and civil but as the only correct variety and therefore a model against which all other varieties were judged incorrect. This point is made by Sheridan. Almost every county in England has its peculiar dialect. [. . .] One must have preference, this is the court dialect, as the court is the source of fashions of all kinds. All other dialects are sure marks, either of a provincial, rustic, pedantic or mechanical education, and therefore have some degree of disgrace attached to them. (1761: 29–30)
Although Sheridan’s statement, like those of the sixteenth-century authors cited in the previous section, singles out the dialect of the court as the most fashionable, his remarks about other dialects clearly point to these indexing social class. Anything other than the elite variety marks out the speaker’s
Northern English and enregisterment
29
upbringing (“education”) and brings “disgrace”. Holmberg remarks that “it is in the eighteenth century that the snob value of a good pronunciation began to be recognised” (1964: 20). Social and geographical mobility at the start of the Industrial Revolution led to heightened awareness of speech differences, and the rising middle class sought out guides to correct speech to help them avoid the stigma of provincial dialect. From the second half of the eighteenth century onwards, elocutionists such as Kenrick (1773), Sheridan (1780) and Walker (1791) provided models of correct speech in the form of pronouncing dictionaries. Their comments on northern (and other) dialects of English act as warnings to their readers, but also play their part in the enregisterment of these features. One of the most enduring and salient features of northern English pronunciation today is the lack of what Wells (1982:196) terms the FOOT- STRUT split: south of a line crossing England approximately at Birmingham, pairs of words such as put and putt; could and cud have different vowels, but north of that line they are homophonous. In other words, northern English lacks the phoneme /ʌ/. This feature was not subject to comment in the sixteenth and seventeenth centuries, because the southern phoneme /ʌ/ is an innovation first evidenced in the late seventeenth century. By the eighteenth century, the contrast between northern and southern pronunciations of words such as trunk was already a matter of comment. Earlier in the century Kirkby, himself a Northerner, makes what we might term a scientific observation on the subject, remarking that the “seventh vowel” in his notation, that found in the words skull, gun, supper, figure, nature “is scarce known to the Inhabitants of the North, who always use the short sound of the eighth vowel instead of it” (1746: 7). Kirkby’s “eighth vowel” is described as long in too, woo, food and short in good, stood, foot so we can interpret the latter as /ʊ/. Kirkby’s remark includes no metalinguistic information beyond identifying the lack of the “seventh vowel” as northern. Later in the century, this northern characteristic is clearly indexed as incorrect. Kenrick, referring to the same vowel as Kirkby’s seventh, notes “that the people of Ireland, Yorkshire and many other provincials mistake its use” (1773: 36) and Walker writes: If the short sound of the letter u in trunk, sunk etc., differ from the sound of that letter in the northern parts of England, where they sound it like the u in bull, and nearly as if the words were written troonk, soonk, etc., it necessarily follows that every word where that letter occurs must by these provincials be mispronounced. (1791: xiii)
Here, Kenrick and Walker are not merely making the scientific observation that northern speakers pronounce these words differently. They are also indexing the
30
Joan C. Beal
pronunciation as provincial (a keyword in eighteenth-century prescriptive statements) and incorrect. In using semi-phonetic spelling to indicate the northern pronunciation, Walker also facilitates the transmission of messages about this pronunciation by providing a means of representing it. To this day, spellings such as are easily found8. This shift in the indexicality of northern features in the second half of the eighteenth century corresponds to what Jones terms “a sea-change in the way linguistic usage is perceived to relate to criteria such as social status and place of geographic origin (the two often vitally interconnected)” (2006: 117). Another example of such a shift with respect to a specific feature of northern pronunciation can be seen if we compare the following comments on what is now known as the Northumbrian burr: a uvular pronunciation of /r/ now recessive and only found in older and/ or rural speakers in the very far North-east of England.9 Defoe ends his account of Northumberland with one of the few linguistic comments in his Tour: I must not quit Northumberland without taking notice, that the Natives of this Country, of the antient original Race or Families, are distinguished by a Shibboleth upon their Tongues in pronouncing the Letter R, which they cannot utter without a hollow Jarring in the Throat, by which they are as plainly known, as a foreigner is in pronouncing the Th: this they call the Northumberland R, or Wharle; and the Natives value themselves upon that Imperfection, because, forsooth, it shews the Antiquity of their Blood. (1748: 232–3)
Defoe’s remarks here are not derogatory: to him, the “wharle” is a curiosity. It is worth noting, however, that, according to Defoe, the Northumbrians themselves are aware that this pronunciation is local, and indexed with the antiquity and authenticity attributed to northern English by some sixteenth-century witnesses. By the second half of the eighteenth century, the northern pronunciation has become indexed as incorrect. Kenrick and Sheridan both describe it as a speech defect. Kenrick notes that “in the northern parts of England, particularly in and about Newcastle, we find the r deprived of its tremulating sound, and very awkwardly pronounced somewhat like a w or oa” ( 1773: 31). Sheridan writes: The letter R is very indistinctly pronounced by many; nay in several of the Northern counties of England, there are scarce any of the inhabitants, who can pronounce it at all. Yet it would be strange to suppose, that all those people, should be so unfortunately distinguished, from the rest of the natives of this island, as to be born with any peculiar defect in their organs; when this matter is so plainly to be accounted for, upon this principle of imitation and habit (1762: 24).
8 See Beal (2012b) for a fuller discussion of the enregisterment of this feature. 9 See Påhlsson (1972) for a full account of this feature.
Northern English and enregisterment
31
It is no coincidence that Sheridan’s comments come from his Course of Lectures on Elocution. He is advocating a very interventionist approach whereby elocutionists such as himself teach others how to avoid what he elsewhere terms the “disgrace” of provincial pronunciation. This “imitation and habit” is a specific kind of social work, something which Johnstone et al. (2006) see as one of the defining characteristics of second-order indexicality. The next section will consider the enregisterment of northern dialects as repertoires of features represented as stable varieties.
6 Enregisterment: northern English “characters” and their repertoires In Agha’s framework, enregisterment involves “characterological figures stereotypically linked to speech repertoires (and associated signs) by a population of users” (2005: 45). Evidence of enregisterment therefore involves identifiable sets of linguistic features associated with specific personae. Enregisterment presupposes indexicality at at least the n + 1 or second order, since it is only at this stage that linguistic forms are imbued with meaning according to prevailing ideologies. It is at the third or n + 1 + 1 order that enregisterment is unequivocally evidenced, for, as Johnstone et al. state, at this stage speakers “use regional forms [. . .] to perform local identity” (2006: 83). In a historical context, evidence for such performance of identity comes mainly from literary representations of these “characterological figures”. Although there are a few earlier examples of northern characters being marked out by their language, such as the students in Chaucer’s Reeve’s Tale, evidence for a literary repertoire of northern features becomes more abundant from the second half of the sixteenth century. Ruano-Garcia’s study of northern English lexis from a corpus of Early Modern English texts singles out William Bullein’s Dialogue Against the Fever Pestilence (1564) as the text in which “northern traits were first used with specific purpose” (2010: 55–56) in this period. I have also discussed elsewhere (Beal 2016) how Bullein drew upon resources such as the Border ballads to depict the Northumbrian beggar in his Dialogue as outlandish. Stewart (2011) analyses a corpus of sixteenth- and seventeenth- century literary texts which include representations of northern and/or Scottish English. She identifies a repertoire of features used by these Early Modern dramatists to mark out characters as northern/Scottish, and notes that “all representations of dialect in these plays always index ‘foreignness’ or the ‘other’ in the dialectspeaking character” and that they also index a variety of ideas, concepts, and
32
Joan C. Beal
attitudes about the North and Scotland, and of Northerners and Scots (2011: 363). The features which regularly occur in the speech of northern English characters include the representation of the definite article as or
(henceforth Definite Article Reduction), the verbal suffix –en and the representation of as a rounded vowel before nasals. Other features, such as the word bairn for “child” occurred in representations of both northern and English and Scottish speech, whilst a third set of features, including ken for “know” and kirk for “church” were exclusively Scottish (see Stewart 2011: 403 for a full list). Stewart establishes this repertoire with hindsight, but at the time when these texts were written, read, performed and watched, each occurrence of a northern feature in the speech of a character displaying stereotypical northern characteristics would form a link in the chain of communication that constitutes enregisterment. Stewart identifies recurring characterological figures, such as the pure, innocent northern lass or the belligerent Scot, who are signalled by their use of features which likewise recur whenever such figures appear. Stewart’s evidence indicates that, by the end of the seventeenth century, northern English was enregistered as a recognisable repertoire of features indexed in accordance with the prevalent ideology in which the North was at once ancient and pure, harsh and barbaric. Although, as we have seen in section 6, northern features were overwhelmingly indexed as incorrect in the comments of eighteenth-century prescriptivists, dialect literature from the late eighteenth century onwards presented a more positive enregisterment of northern dialects. Joyce argues that the effect of the Industrial Revolution in northern towns and cities was to create urban, working class communities on which dialect literature “conferred citizenship”: Dialect ‘spoke to ‘working folk’ of all occupations and geographical locations, conferring on them citizenship in the nationalities of ‘Lanky’, Yorkshire ‘Tyke’ or northeast ‘Geordie’. [. . .] Dialect literature created its own symbolic working heroes, with the characters of the Weaver in Lancashire and Yorkshire, the pitman and keelman in the northeast, embodying the symbolic virtues of the ‘gradely’ or the ‘canny lad’. (1991: 172)
Joyce’s “symbolic working heroes” here correspond to Agha’s “characterological figures” and the texts in which these iconic characters appear in association with northern dialect features provide evidence of enregisterment. Beal (2000, 2009) Cooper (2013) and Beal and Cooper (2015) provide evidence for the enregisterment of two northern dialects in the nineteenth century: “Geordie” from the North-east of England and more specifically Newcastle upon Tyne; and Yorkshire. These two areas were the most prolific with regard to the production of dialect texts in the nineteenth century. In the Newcastle area, these texts most
Northern English and enregisterment
33
typically took the form of songs and recitations performed in the music halls,10 whilst in Yorkshire the most popular format was the almanac. In both cases, the texts were intended for local consumption and celebrated local stereotypes. The very existence of popular names for citizens of these regions, subsequently used with reference to the corresponding dialect, is proof in itself of enregisterment. As can be seen from tables 2.1, 2.2 and 2.3 below, Beal and Cooper have also found the frequent and consistent use of sets of features enregistered as “Geordie” or “Yorkshire” respectively, which constitute Agha’s “linguistic repertoire differentiable within a language as a socially recognised register” indexing “speaker status linked to a specific scheme of cultural values” (2003: 231). Table 2.1: Repertoire of nineteenth-century “Geordie” features (adapted from Beal 2009) ‘Geordie’ spelling/ pronunciation
Standard English spelling/ RP pronunciation
Keyword (Wells 1982)
Examples
/uː/
, /au/
MOUTH
oot, broon, doon, mooth
/iː/
/ai/
PRICE
neet, (aal)reet
/ɔː/
, , , /әː/
NURSE
Borth, forst, surface, porsonal
/ɪә/
, /ei/
FACE
Fyes, Gyteshead Knaa, walk, aal(ways)
/aː/
, /ou/ ~ , /ɔː/
THOUGHT
, /dɪvɪnt/
, /dount/
NA
div/ divvint
, /ɔʀ/
, /ә/
lettER
Beggor, scarpor, remember, nivvor
Table 2.2: Repertoire of nineteenth-century Yorkshire features (after Cooper 2013: 266)11 Corpus proportion (DL:LD)
% texts with direct commentary
Examples
Feature
Definition
% texts with tokens
DAR
N/A
83
70:30
65
t’barber teld me
sen
self
72
60:40
40
az weel az mesen
nowt
nothing
70
55:45
70
thear nowt bur a kletch o hired jobbers
mun
must
64
40:60
60
we mun nah be off ta wark
owd/ oud
old
57
85:15
40
it wor owd Nick
gan/gang
go
43
60:40
50
I’se ganging theear myself
owt
anything
40
80:20
50
hestha browt owt to’t market?
10 See Hermeston (2011) for further discussion of the music-hall tradition on Tyneside. 11 Here DL refers to ‘dialect literature’, i.e. texts written mainly or wholly in dialect, whereas LD refers to ‘Literary Dialect’, texts written mainly in Standard English, but with extracts in dialect. The representation of Joseph’s speech in Emily Brontë’s Wuthering Heights would be an example of the latter.
34
Joan C. Beal
Table 2.3: Repertoire of modern Yorkshire dialect features (after Cooper 2013: 256)12
Feature
Definition
% texts with tokens
Corpus proportion (DL:LD)
% texts with direct commentary
DAR
N/A
90
85:15
40
Nowt
nothing
90
60:40
100
Examples Watch out for’t boggarts Nowt o’t’ sooart
Owt
anything
70
40:60
85
Doin’ owt this evening?
Summat
something
60
60:40
85
Tha nesh or summat?
h-dropping
N/A
70
80:20
60
Put t’ wood in t’oil (= hole)
reight
Really/right
50
70:30
70
Ah’m reight glad
sen
self
60
90:10
60
Missen, thissen
Tha/thee
You/your
90
65:35
70
Can’t tha fit us in somewheeare?
Beal (2000, 2009) and Cooper (2013) derive their repertoires from the analysis of corpora of dialect material. Cooper’s is the larger and more systematic study, based as it is on a specially constructed corpus made up of dialect literature (DL), literary dialect (LD) and metapragmatic discourse taken from sources such as introductions to dialect dictionaries, travel guides, etc. In the case of both Geordie and Yorkshire, there is some continuity in the repertoire of enregistered features from the nineteenth century to the present day. Beal (2000) notes that all the features in Table 2.1 except for the FACE vowel still occur in late twentieth-century representations of Geordie, such as the adult comic Viz13. This feature has been demonstrated to be recessive in present-day Tyneside English (Watt and Milroy 1999), and so could be in the process of being deregistered as an indexical feature of Tyneside English. The Yorkshire repertoire shows more change, as can be seen from a comparison of Table 2.2 with Table 2.3. A comparison of tables 2.2 and 2.3 reveals that DAR, nowt, owt and sen continue to be strongly enregistered as Yorkshire features. Gan, owd and mun seem to have dropped out of the repertoire, whereas summat, reight, tha/thee and h-dropping have been added. It is likely that gan and mun are now obsolescent in Yorkshire (gan continues to be used in Geordie dialect literature, reflecting the northward recession of this feature). Tha/thee and h-dropping 12 Features in Table 2. 3 are not listed in order of frequency alone. This is because Cooper’s original table (2013: 256) also ranks them according to the level assigned to them by respondents to a survey. 13 Viz is a humorous and scatological comic, including ‘characterological figures’ such as ‘Sid the Sexist’, a caricature of the unreconstructed northern male.
Northern English and enregisterment
35
were almost certainly present in nineteenth-century Yorkshire dialect speech, but not yet indexed as local. Along with the disappearance of for the FACE vowel in the Geordie repertoire, this suggests that enregistered repertoires can be subject to change. According to Labov (1972: 178–80) variants that have become stereotypes may eventually drop out of use altogether. Within the framework of indexicality and enregisterment, we might say that, at this point (or perhaps a little later), they will become deregistered. A good example of this process is the demise of v/w exchange as an enregistered Cockney feature. Walker (1791: xii) lists this as a “fault” of the Cockneys, and by the time Dickens used it in the speech of characters such as Sam Weller, it was already a stereotype, but it is completely absent from twentieth-century representations of Cockney. Although the pronunciation represented by for the FACE vowel has not yet completely died out in Tyneside English, it appears to have already been deregistered. What we see from the above discussion is that between the seventeenth and the twentieth centuries, the enregisterment of northern English evolves. At first, there is a general enregisterment of a northern repertoire associated with the long-standing stereotype of Northerners as, on the one hand rough and outlandish, and on the other hand pure and authentic. The Industrial Revolution creates new working-class communities in growing and emergent towns and cities in the North, and in some of these communities, such as Tyneside and Yorkshire, more specific northern dialects become enregistered and associated with local characters such as the Geordie miner/ keelman or the Yorkshire weaver. The enregisterment of these dialects is transmitted by texts which are both produced and consumed locally: the repertoires of music-hall entertainers and the outputs of local presses. On a national level, the enregisterment is facilitated by literary dialect and by metalinguistic and metapragmatic discourse in texts such as dialect dictionaries. These texts serve for us as witnesses to enregisterment, but there would also have been face-to-face communication involved in the process, given the increase in mobility and therefore dialect contact in the Late Modern period14. In the twenty-first century, the enregisterment of northern varieties of English continues to evolve as mobility increases even further and new forms of communication allow messages about linguistic repertoires and their associated characterological figures to “go viral”. We also see the commodification of northern dialects in the form of advertising campaigns, T-shirts and other goods 14 Indeed, as Johnstone et al (2006) point out, mobility and contact are pre-requisites for enregisterment: until we encounter a way of speaking that is different from our own, we have no way of knowing that our own variety has any significance
36
Joan C. Beal
trading on the cultural capital of northern varieties of English. In a post-industrial age, working-class heroes such as the miner and weaver of nineteenth-century literature and song have given way to more general associations of northern varieties with characteristics such as honesty, resilience and a sense of humour in the face of adversity. Examples of this include the advertising campaign by the Sheffield-based internet service provider PlusNet. In their television advertising, ironically-presented nostalgic stereotypes of northern life such as cobbled streets and ferrets are shown with a voice-over proclaiming in a recognisably northern accent that PlusNet offers “good, honest broadband from Yorkshire”. Their print campaign includes a banner placed on Sheffield buses advertising “a call centre down t’road”, using Definite Article Reduction, which is identified in tables 2.2 and 2.3 as the most commonly-occurring feature in representations of Yorkshire dialect from the nineteenth century onwards.15
7 Conclusion In this chapter, I have attempted to trace the historical enregisterment of northern English(es). One thing that emerges from this account is that as we get closer to the present day, evidence of enregisterment becomes more abundant. Of course, one reason for this is that printed and, more recently, electronic and audio-visual material of all kinds becomes more abundant, but the continuing if not increasing evidence for awareness of dialectal distinctiveness seems paradoxical given the prevalence of studies demonstrating the effects of dialect levelling in the late twentieth and early twenty-first centuries. However, as argued by Johnstone et al. (2006) and Beal (2010), enregisterment and levelling are two sides of the same coin, since both depend on mobility and dialect contact. The history of the English language runs alongside an external history in which social and geographical mobility increase and in which the means and media of communication proliferate as we move towards the present day. Since enregisterment depends on the transmission of messages in which linguistic features are associated together as recognisable repertoires associated with cultural personae, the availability of media which enable the rapid and expansive dissemination of these messages facilitates enregisterment. What, then, is the value of applying the framework of indexicality and enregisterment to the historical study of northern English? I would argue that
15 For further discussion of the enregisterment and commodification of Sheffield dialect, see Beal (2017).
Northern English and enregisterment
37
such an approach puts into context the current debates within sociolinguistics concerning dialect levelling. As I have pointed out elsewhere (Beal 2010), expressions of concern about the loss of dialectal distinctiveness and heightened awareness of this distinctiveness went hand in hand in the nineteenth century as they do today. The rural dialects whose demise was feared by nineteenthcentury commentators have been replaced by urban dialects, which may well be replaced in due course by varieties like Multicultural London English (Cheshire et al. 2011), but each new variety is in turn enregistered. Likewise, individual features may become deregistered, as happened with the representation of the FACE vowel in Tyneside English, but newly enregistered features take their place. The historical study of enregisterment can also act as a warning about the weight attached to the often fragmentary evidence for the origins of present-day variants. As I have argued (Beal 2007), the absence of commentary on a feature does not constitute evidence for its non-existence at the time. Such an absence merely indicates that it was not enregistered or indexed above the first, or nth order. Conversely, metalinguistic and metapragmatic comments on a feature may lag behind its first appearance and may outlive its vitality. At the higher orders of indexicality, corresponding to the Labovian stereotype, a feature may appear like a supernova, burning most brightly after its demise. What we need for a complete picture of the history and the future of northern Englishes is a combination of scientific studies (which will in themselves provide proof of nth order indexicality for future scholars) with a thorough understanding of the significance of the metalinguistic and metapragmatic commentary provided by literary texts and the media.
References Agha, Asif. 2003. The social life of cultural value. Language and Communication, 23, 231–273. Agha, Asif. 2005. Voice, footing, enregisterment. Journal of Linguistic Anthropology. 15: 1, 38–59. Barber, Charles, Beal, Joan C. and Shaw, Philip A. 2010. The English language: A historical introduction. 2nd edn. Cambridge: Cambridge University Press. Beal, Joan C. 2000. From Geordie Ridley to Viz: Popular literature in Tyneside English. Language and Literature 9. 4. 343–359. Beal, Joan C. 2009. Enregisterment, commodification and historical context: ‘Geordie’ versus ‘Sheffieldish’. American Speech 84 (2) 138–156. Beal, Joan C. 2007. To Explain the present: 18th and 19th-century antecedents of 21st-century levelling and diffusion. In Jorge L Bueno Alonso, Dolores González Álvarez, J. Pérez-Guerra and E. Rama Martínez (eds). ‘Of varying language and opposing creed’: New insights into Late Modern English, 25–46. Bern: Peter Lang (Linguistic Insights Series).
38
Joan C. Beal
Beal, Joan C. 2012a. À la mode de Paris: Linguistic patriotism and francophobia in 18th-century Britain. In Carol Percy and Mary C. Davidson (eds.) The languages of nation: attitudes and norms, 141–154. Cleveden: Multilingual Matters. Beal, Joan C. 2012b. “By those provincials mispronounced”: the STRUT vowel in eighteenthcentury pronouncing dictionaries. Language and History 55.1, 5–17. Beal, Joan C. 2016. Enregistering the North: the dialect of Mendicus in William Bullein’s dialogue against the fever pestilence. In Anita Auer, Victorina Gonzales-Diaz, Jane Hodson and Violeta Sotirova (eds.) Linguistics and literary history. In honour of Sylvia Adamson, 13–30.Amsterdam: John Benjamins. Beal, Joan C. 2017. “Strong and northern”: the enregisterment and commodification of Sheffieldish. In Anderwald, Lieselotte and Jarich Hoekstra (eds.) Enregisterment – zur sozialen Bedeutung sprachlicher Variation, 29–45. Frankfurt am Mein: Peter Lang. Beal, Joan C. and Paul Cooper. 2015. The enregisterment of northern English. In Raymond Hickey (ed.) Researching northern English, 27–50. Amsterdam: John Benjamins. Blank, Paula 1996. Broken English. Dialects and the politics of language in renaissance writings. London: Routledge. Cheshire, Jenny, Paul Kerswill, Sue Fox and Eivind Torgersen. 2011. Contact, the feature pool and the speech community: The emergence of Multicultural London English. Journal of Sociolinguistics 15:2, 151–196. Child, Francis J. 1882–1898. The English and Scottish popular ballads. New York: Houghton, Mifflin & Co. Cooper, Paul. 2013. Enregisterment in historical contexts: A framework. Sheffield: University of Sheffield PhD Thesis. Defoe, Daniel. 1748. A Tour thro’ the whole island of Great Britain. 4th edn. London: S. Birt, T. Osborne, J. Hodges, J. Osborne, A. Millar and J. Robinson. Gill, Alexander. 1619. Logonomia Anglica. London: Johannes Beale Gneuss, Helmut. 1972. The origin of standard Old English and Aethelwold’s school at Winchester. Anglo-Saxon England 1: 63–83. Gretsch, Mechthild. 2006. A key to Ælfric’s standard Old English. Leeds Studies in English 37: 161–77. Hermeston, Rod. 2011. “The Blaydon Races”: lads and lasses, song tradition and the evolution of an anthem. Language and Literature 20, 4: 269–282. Hogg, Richard. 2002. An introduction to Old English. Edinburgh: Edinburgh University Press. Holmberg, Björe. 1964. On the concept of standard English and the history of modern English pronunciation. Lund: Gleerup. Jewell, Helen. 1994. The North-South divide: the origins of northern consciousness in England. Manchester: Manchester University Press. Johnstone, Barbara (forthcoming). Enregisterment: linguistic form and meaning in time and space. To appear in Busse, Beatrix and Ingo H. Warnke (eds.) Sprache im urbanen Raum/ Language in urban space. Berlin: de Gruyter. Johnstone, Barbara, Jennifer Andrus and Andrew E. Danielson. 2006. Mobility, indexicality, and the enregisterment of “Pittsburghese”. Journal of English Linguistics 34: 2, 77–104. Jones, Charles. 2006. English pronunciation in the eighteenth and nineteenth centuries. Basingstoke: Palgrave Macmillan. Joyce, Patrick. 1991. Visions of the people: Industrial England and the question of class 1848– 1914. Cambridge: Cambridge University Press.
Northern English and enregisterment
39
Kenrick, William. 1773. A new dictionary of the English language. London: John and Francis Rivington, William Johnston et al. Kirkby, John. 1746. A new English grammar, or guide to the English tongue. London: R. Manby and H. S. Cox. Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. Lamont, Claire. 1997. Shakespeare’s Henry IV and the “old song of Percy and Douglas”. In John Batchelor, Tom Cain & Claire Lamont (eds.) Shakespearean continuities: essays in honour of E. A. J. Honigmann, 56–73. London: Macmillan. Lenker, Ursula. 2000. The monasteries of the Benedictine reform and the “Winchester School”: model cases of social networks in Anglo-Saxon England? European Journal of English Studies 4 (3): 225–38. Mair, G. H. (ed.) 1909. Wilson’s arte of rhetorique 1560. Oxford: Clarendon Press. Milroy, Lesley. 2000. Two nations divided by the same language (and different language ideologies). Journal of Linguistic Anthropology 9: 1, 1–34. Montgomery, Chris. 2007. Northern English dialects: A perceptual approach. Sheffield: University of Sheffield PhD thesis. Onions, Charles T. (ed.) 1959. Sweet’s Anglo-Saxon reader in prose and verse. 14th edn. Oxford: Clarendon Press. Påhlsson, Christer. 1972. The Northumbrian burr. Lund: Gleerup. Percy, Thomas. 1765. Reliques of ancient English poetry. London: J. Dodsley. Ruano-Garcia, Javier. 2010. Early Modern northern English lexis: A literary corpus-based study. Bern: Peter Lang. Scott, Walter. 1802. The minstrelsy of the Scottish border. Kelso: James Ballantyne. Sheridan, Thomas. 1761. A dissertation on the causes of the difficulties, which occur, in learning the English tongue. London: R. and J. Dodsley. Sheridan Thomas. 1762. A course of lectures on elocution. London: W. Strachan. Sheridan, Thomas. 1780. A general dictionary of the English language. London: J. Dodsley, C. Dilly and J. Wilkie. Silverstein, Michael. 1976. Shifters, linguistic categories and cultural description. In Basso, Keith H. and Henry A Selby (eds.). Meaning in anthropology, 11–55. Albuquerque: University of New Mexico Press. Silverstein, Michael. 2003. Indexical order and the dialectics of sociolinguistic life. Language and Communication 23: 193–229. Stewart, Lauren M. 2011. ‘The representation of northern English and Scots in seventeenth century drama’. Edinburgh: University of Edinburgh PhD thesis. Wales, Katie. 1999. North and south: a linguistic divide? Reporter 439. Wales, Katie. 2000. North and south: an English linguistic divide? English Today 61, 4–15. Wales, Katie. 2006. Northern English: a social and cultural history. Cambridge: Cambridge University Press. Walker, John. 1791. A critical pronouncing dictionary. London: G. G. J. and J. Robinson and T. Cadell. Watson, Godfrey. 1974. The border reivers. London: Robert Hale & Co. Watt, Dominic and Lesley Milroy. 1999. Patterns of variation and change in three Newcastle vowels: is this dialect levelling? In Foulkes, Paul and Gerard Docherty (eds.) Urban voices: accent studies in the British Isles, 25–46. London: Arnold. Wells, John. 1982. Accents of English. Cambridge: Cambridge University Press.
Katie Wales
3 Dickens and northern English: stereotyping and ‘authenticity’ reconsidered Abstract: It is now over forty years ago since Stanley Gerson produced the only detailed analysis of sound and spelling in the works of Charles Dickens as a whole. In this chapter I take a fresh look at Dickens’ handling of northern British dialect speech in particular, and including morphology, syntax and lexis as well as pronunciation, focussing on two of his major novels, Nicholas Nickleby (1838– 9) and Hard Times (1854). I take an approach to literary dialect representation and linguistic stereotyping that has so far been under-used in this context, namely social cognition theory, which concerns mental scripts and attitudes. Looking at Dickens’ northern characters we can see the process of literary formation in the mind of the novelist. Far from being either one-dimensional or hotchpotch of random features, they are constructed intertextually through schemas and social ideologies and also cultural practices, to a large extent different for each novel. Dickens is not so much ‘representing’ northern speech as creating it. This particular approach also raises an important issue to do with the notion of ‘authenticity’, a value judgment much contested traditionally in literary dialect representation. I shall argue instead for the significance for Dickens of ‘authentication’ or ‘authenticating’ devices and effects. I conclude the discussion of each novel by considering the implications of Dickens’ portrayal in the larger literary historical context of northern dialect (re)presentation.
1 Introduction: ‘authenticity’ and ‘accuracy’ Nicholas Nickleby (1838–9) and Hard Times (1854) are two well-known novels of Charles Dickens, but are surprisingly neglected from a linguistic and stylistic point of view generally, and from a dialectal perspective in particular. The exceptions might be considered to be Brook (1970) rather briefly; and particularly Gerson (1967) (on whom Brook heavily relies), in his ambitious scrutiny of the representation of vowels and consonants in all the non-standard spellings in Katie Wales, University of Nottingham DOI 10.1515/9783110450903-003
42
Katie Wales
all of Dickens’ novels. Dickens’ treatment of non-London dialects has tended to be rather negatively evaluated. Even I myself, as late as Hickey (2010), considered the dialect in Nicholas Nickleby somewhat of a ‘hotchpotch’ (p. 68). In each novel there is a prominent northern character with apparent ‘northern’ speech patterns: John Browdie, a corn-factor from North Yorkshire in Nicholas Nickleby; and Stephen Blackpool, a cotton-mill worker in Lancashire, in Hard Times. Using the framework which I expound below in section 2, I now think that it is possible to account for apparent inconsistencies in dialectal representation; and I also think that the framework is replicable as a model by which other representations of non-standard speech can be analysed, in other novels apart from Dickens’. This particular approach also raises an important issue to do with the notion of ‘authenticity’, a value-judgment much contested traditionally in the analysis of literary dialect representations. I shall argue instead for the significance for Dickens of ‘authentication’ or ‘authenticating devices’ and ‘effects’. Although much discussed by sociolinguists, this perspective has, again, been little considered in the study of literary dialect. (For a very brief discussion of the possibilities, however, see Hodson & Broadhead 2013: 316.)1 In a special debate in an issue of the Journal of Sociolinguistics (2003), some ten years ago now, Eckert (392–7) describes “authenticity” as the “elephant in the room” for sociolinguists, as they strive to examine the genuine and trustworthy, unadulterated by fieldworker intrusion, in order to produce reliable data. (We could note here a similar concern some fifty years previously with the Orton-led Survey of English Dialects.) There is another, related, elephant in the room for the historical study of dialect, and that is ‘accuracy’: i.e. whether or not we are confronted by a correct or truthful representation of ‘real’ discourse. Hickey and his fellow contributors to his 2010 volume wrestle repeatedly with this issue, and this is reflected in the book’s sub-title: “The written word as linguistic evidence”. How reliable are written materials, on which we are totally dependent for earlier periods of linguistic history other than the modern day, especially if an author is not a native of the region; and is clearly in any case not a professional linguist? Now Dickens himself did indeed visit North Yorkshire and the Greta Bridge area precisely at the time of writing Nicholas Nickleby in monthly instalments, to learn about the infamous boarding schools near Bowes and Barnard Castle, which help to inform the plot of the first part of the novel. He also went to Preston, often (if contentiously) claimed as the setting of Hard Times, to observe 1 On the contentious issue of ‘authenticity’ in the context of EFL teaching, see Pinner (2014: 22–7) for a useful summary.
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
43
the cotton-weavers’ strike about wages (January 1854). His visits in both cases were fleeting. However, I shall come back to this point later. His famous, albeit exhausting, reading tours throughout the British Isles did not begin until after Hard Times; his theatrical tours began after 1848, helped by the expansion of railways to the North. At the time of finishing the last instalment of Nicholas Nickleby he had been to Liverpool once and Manchester twice. So Dickens could be pigeon-holed as an ‘outsider’, his representation of northern English ‘extrinsic’ in Hickey’s terms (2010: 9) and hence prone to ‘unreliability’, if viewed from the perspective of linguistic ‘accuracy’. Not surprisingly, perhaps, Gerson (1967) struggles in his judgments. He finds many spellings consistent with what is known of northern English dialects in the early to mid-nineteenth century; but his net, as it were, is cast very wide, all over the North, not just North Yorkshire and Lancashire. He acknowledges also that Dickens sometimes seems to misunderstand the ‘values’ of his spellings for Browdie and Blackpool (p. xvii). I myself am going to suggest that the notions of ‘authenticity’ and ‘accuracy’ for Dickens are irrelevant; or rather, agree with Eckert that it is an “ideological construct” (p. 392). I am also agreeing with Bucholtz in the same debate (398– 416) that authenticity should not be a goal, but open to dissection. Her own preference is for authentication rather: identity seen as the (often creative) negotiation of social practices, not given. I would add here literary practices. What is also relevant for my argument is her suggestion of authenticating effects, achieved through authenticating practices. First, however, I should like to approach literary dialect representation from yet another different angle. The attraction of social cognition theory is that ‘accuracy’ is not the most important issue: rather, it is the kinds of knowledge that writers must draw upon, whether ‘outside’ or indeed ‘inside’ a region.
2 Social cognition theory I first encountered social cognition theory in a short section in Culpeper’s book (2001) on characterisation in drama, where he summarises some aspects of research in this field since the 1970s. He himself is interested in what an actor on stage playing a ‘shrew’, for example, meant to the Elizabethan reader or spectator, in terms of cultural references. From the perspective of how we know, analyse and remember the world around us, stereotypes are not simply a ‘loose collection of beliefs’ (2001: 75); they are highly organised social categories that have the properties of schemas (p. 78–9): hence their durability. Schemas themselves can be defined as abstract cultural knowledge structures, stored in memory
44
Katie Wales
as mental packages: also termed ‘scripts’ or ‘frames’. (See also Macrae, Stangor & Hewstone (eds.) 1996, Pennington 2000, Stangor (ed.) 2000.) Put more simply, stereotypes are the pictures we have in our minds of the people in social groups around us, pictures that are consensually shared. Although stereotypes are conventionally seen as stable and long-enduring, they can change, as categories are added to and taken on by new speakers and readers. Certainly work on evaluation of accents and corresponding impression traits (e.g. by Giles & Powesland 1975) can be seen to be related to this kind of social cognitive approach, since attitudes can be seen as mental constructs. Generally, however, sociolinguistics in the Labovian tradition tends not to comment on the relations between linguistic features, social characteristics and cognition. Work in perceptual dialectology, however, is clearly relevant here: analysing people’s ‘mental maps’ of communities or areas, features and boundaries. Indeed, I would go so far as to say that perceptual dialectologists are precisely not interested in the ‘authentic’ or the ‘accurate’: rather, the opposite. The mismatch between perception and ‘reality’ is often very obvious, as I myself discovered when asking a group of people at random where they thought Barnard Castle was. But where exactly do these beliefs or perceptions come from? (Do perceptual dialectologists find out?) My main interest here is with the authorial perspective: where do the writer’s beliefs about character or personality types come from? How are these linked to speech habits, especially the use of dialect? In social cognition terms, knowledge to be drawn upon for the creation of personae and dialogue comes not only from direct social contact in the real world, but, more significantly, intertextually from indirect sources: cultural ideologies and practices, including those drawn from other literary reading (genre schemata). It has to be noted here that there is another kind of social cognition theory which has to do with the mutual construction of knowledge in the discourses of social interaction. (See Condor & Antaki 1997: 320–47). My own interest is clearly in the mentalist social cognition strand; however, as I shall indicate in section 4 below, there may be some relevance in the other strand for the construction of Hard Times in relation to the work of Elizabeth Gaskell.
3 Nicholas Nickleby The speeches of John Browdie actually cover six chapters of the novel. Far from revealing a hotch-potch of northern and non-northern linguistic features I am arguing here that his Bakhtinian, polyphonic discourse reflects quite consistently Dickens’ mental image of Browdie, which is based on several schemas, them-
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
45
selves deriving from common cultural knowledge or consensually shared myths, and also from Dickens’ own literary reading. It is readily discernible that Browdie’s dialogue generally, like that of other Dickens’ characters, is vibrant and animated. As with other novelists, the ‘base’ schema, making up the foundational ‘layer’, as it were, is simply the mental script of natural conversation, of characters talking artlessly to others: hence the overwhelming frequency of colloquial, informal features. The representation of John Browdie’s speech is no exception. We can note the many elisions, as in: (1)
what are sometimes termed ‘allegro spellings’2. On this general base is superimposed for Browdie himself the schema of the partially educated working man, to which the generic non-standard features testify, such as: (2)
sumat [something]; nowt [nothing];
and in particular the many ‘eye-dialect’ spellings, such as: (3)
hart, reddy, steddy, etc.
For Browdie there also comes into play the more significant image of a rural farmer or rustic peasant, rooted in the provincial shires rather than an urban landscape, and popularly associated with the South in drama and fiction (‘Mummerset’). Browdie, the son of a corn-factor, is engaged to a miller’s daughter. This explains, I believe, in Browdie’s speeh alone the frequent and consistent occurrence of what would otherwise be very puzzling be-forms: (4) bean’t she? There be; if she bean’t; thee be’est; thee bean’t; I be asheamed, etc. One might have expected here instead forms, as in I is, so common in literary representations of northern, especially north-eastern, English, since Chaucer’s Reeve’s Tale (Wales 2006: 72–81). But if ‘Mummerset’ has traditionally had northern forms as well (see Edgar’s linguistic disguise in King Lear, for instance), there is no reason for the situation not to be reversed. In Browdie’s speech also occur for spellings, representing the [ai] diphthong, as in: (5)
loight an’ toight, loike, foind, sploiced, noice, etc.
2 All quotations and references to Dickens’ works are taken from the multi-volume edition published by Hazell, Watson, Viney Ltd, London [n.d.]
46
Katie Wales
A rural schema seems more plausible an explanation than the idea that Dickens may have picked up the forms from Lancashire dialect literature; he certainly does not re-use them later in Hard Times, where they might more plausibly have been expected3. Semiotically speaking, it is interesting in relation to this particular variant of ‘stage(d) rural dialect’ and ‘Hodge’ figure that there is a dramatization of Nicholas Nickleby by Edward Stirling in 1838 – before Dickens himself had finished writing the novel – which includes a stage direction describing Browdie’s dress (although now a miller’s son): “Green countryman’s coat, flowered cotton waistcoat, corduroy breeches, worsted stockings, shoes and buckles, red wig, round black hat. . . .” (See also note 4.) In Dickens’ own later The Uncommercial Traveller (1867/8) the narrator recalls how as a child he had first seen on stage at Chatham “the funny countryman . . . in a flowered waistcoat, . . . pull off his coat, saying ‘Dom thee, squire, coom on with thy fistes then!’” The swearing and the proneness to fighting can be noted here, as well as the northern coom vowel, attributes to which I return below. Gerson (1967: 367) has the interesting comment that Dickens’ acquaintance Charles Mathews “imitated Yorkshire countrymen, popular figures on the London stage”; but I have not been able to track down any of his own performances in that role. The final schematic ‘layer’ in Browdie’s representation is the mental image of the typical Northerner, indexed linguistically by conventional features associated with northern English, and recreated in literary representations through the ages. Some features are associated with the North generally: e.g. the very frequent high back FOOT vowel, as in [u:] for [au] as in: (7)
aboot, noo, hoo;
the raising of the diphthong in: (8) feace [face], seame, neame, etc; and the long /oː/ for the GOAT vowel in: (9) boans, soa, etc.
3 for spellings are not found, however, in either Elizabeth Gaskell’s Mary Barton or in Tim Bobbin, both discussed in section 4 below; but they do feature prominently in a version of the popular ballad The Oldham Weaver which Gaskell herself reproduces in chapter 4 of Mary Barton. There is also the form moind for ‘mind’ in the speech of Joey the waggoner, in Smollett’s Roderick Random, discussed below. It is also to be noted that Gerson (1967: 135) sees it as a ‘sign of general rustic speech’.
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
47
There are also the lexical items: (10) lass, weel [well], Browdie’s very first word, and used by him 12 times; also sike [such], daft, mind [as a final tag], mun [must], etc. Other features, however, appear possibly to represent ‘Border’ English: (11) awa’ [away], wa’ [way]; even Scots: e.g.: (12)
bluid [blood]
Others suggest north-eastern English: e.g.: (13) divn’t [don’t, imperative]. There are no ‘far northern’ lang, strang forms,however. We certainly find very prominently the Yorkshire (and also Lancashire) English pre-R dentalisation, as in: (14) butther, matther, betther, pasthry, etc.; and the distinctive Yorkshire concord of subject and auxiliary verb in forms like: (15)
she/he wur; it wa’ me; thee warn’t, etc;
also ‘linking /v/’ in: (16) tiv’ee; iniv’en. Whilst the Greta Bridge area in North Yorkshire is in reality close to the boundaries of the counties of Durham and Westmorland, and quite remote, it is doubtful that Dickens was aiming for accuracy; these are features most likely to have been absorbed from his literary reading, but shaped into a consistent speech pattern for Browdie himself. Even his name has been carefully selected to suggest general ‘northernness’: the Scottish Brodie with Yorkshire diphthongisation. To this generic representation of northern English can be added the behavioural script of typical northern speakers as a ‘race’: bluff, no-nonsense, forceful if also cheerful. Both Browdie and the guard who transports Nicholas to the North (chapter 6), and whom we meet first, are prone to swearing (euphemistic) oaths such as: (17)
Dang my boans; Dang it; Ding; Dash ma’ wig, etc.
(I return to the guard below.) Browdie also uses a great many thou forms, suggesting both familarity and the egalitarian treatment of others. Dickens also
48
Katie Wales
exploits the script of what he himself terms “a touch of Yorkshire caution” (ch. 13), Browdie’s ‘canniness’ with money. So much for the schematic elements: but Dickens’ cunning (in the sense of clever) artistry is revealed in the way that selected features repeated become part of Browdie’s idiolect, or rather idio-regiolect as I shall term it; functioning in the same way as his repeated catchphrases and expressions of emotion. Take what look like to be examples of Yorkshire definite article reduction (DAR), actually omission, as in: (18) I heard (0) door shut; outside (0) door. Colligating with the (fronted and pre-dentalised) schoolmeasther, phrases recur frequently, like a motif or refrain, such as: (19) efther (0)schoolmeasther was banged; at (0) schoolmeasther’s; in (0) schoolmeasther’s bed; (0)schoolmeasther might come, etc (13 examples in all). Of course the ‘schoolmaster’ concerned is the infamous Wackford Squeers, and Browdie’s falling out with him and revenge is part of the (comic) plot; but there is no doubt that the repetition is deliberate, confirming his distinctive speech pattern. Overall, then, he is a comic, exaggerated creation- but we laugh with him, as much as at him. His linguistic over-generalization is entirely in keeping with his huge appetite, size and strength. Such a distinctive idio-regiolect provides, I think, the most plausible explanation of an otherwise puzzling feature: the very frequent for spelling in the demonstrative and conjunction thot. (It is also found in glod [glad]; but not in words like lass or chap). Not actually attested in northern English, although there is sometimes rounding before a nasal as in Browdie’s hond, ony, it might be there by analogy with these same nasal context uses. It might, however, represent a perception of the short ‘a’ as being somewhat different in quality north of the Trent. It certainly appears in Roderick Random [1748] by Smollett, one of Dickens’ self-confessed favourite authors: the wagon driver Joey, clearly meant to be a Northerner, says coptain at least three times. I shall return to Smollett more generally below; but there is no doubt that Dickens artfully foregrounds the usage here by associating it with these very frequently occuring grammatical words, completely unignorable, to make it part of Browdie’s idioregiolect. The result overall is a carefully managed, reasonably consistent and to a large extent predictable model of speech which Dickens could easily and quickly draw upon for the demands of monthly serialization: a stereotype close to the
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
49
original and literal sense of a type-setter’s frame. Later he drew upon this model in performance for the needs of a listening audience on his public reading tours: the majority of his readings from Nicholas Nickleby comprising those very scenes involving Browdie, his fiancee Mathilda, and Squeers’ daughter Fanny, her close friend. (See further Dickens 1877, Collins (ed.) 1975: 249f ). It is interesting that it is the dialect scenes that Dickens used in this way. They are indeed comic, but the comedy is not really directed at Browdie’s speech style, at least not entirely. Certainly the idiosyncracies and repetitions (e.g. of thot and schoolmeasther), like catchphrases, would be enhanced in performance, and register rapidly and mnemonically4. Contemporary comments on the dialect in the written version of the novel I have been unable to trace; as also contemporary local reactions to the public readings; but Dickens must have expected John Browdie’s scenes to go down well. Collins (1975: 251) does report, however, that outside the North, in Cheltenham, the “broad hearty speech of the Yorkshire miller [sic] was admirably rendered” (Cheltenham Examiner 8th January 1862); and the Birmingham Gazette (3rd April 1869) declared that Browdie “captivated the audience most”. Collins’ own summary of reactions includes the (unattributed) comment that Browdie’s “blunt heartiness” was expressed “in a Yorkshire accent [sic] that was at least found convincing in Lancashire” (1975: 251). It is tempting to see these positive reactions arising out of the audience’s ignorance of what ‘authentic’ North Yorkshire speech would sound like; but if we view Dickens’ representational practice from the audience’s perspective within the framework of social cognition – Culpeper’s take on this, as noted above in section 2 – then there is no reason to think that listeners and readers could not themselves be applying their own frames of reference based on a similar set of schemas to Dickens’ own: colloquial speech habits; the degree of education surmised for Browdie; his rural occupation; a typical Northerner; literary or dramatic precedents. As Jane Hodson has neatly summarised my own argument (p.c.), “dialect representation is both multi-layered in its construction and also multivalent in its perception”: with a high degree of consensus, I would add. Dickens did not only use his Browdie model in his readings. Even before he went on the road, just nine years after Nicholas Nickleby appeared, he wrote a 4 It is noteworthy that stage dramatizations of the novel also highlighted many of the same features of Browdie’s idio-regiolect. Stirling’s version, referred to above, lifts whole sentences from Dickens’ text, and has absorbed the zero-article plus ‘schoolmaster’- motif. But he intensifies the ‘Mummerset’ rural script by the non-Dickensian typical South-western subject for object pronoun feature (gi’ I; kiss I, etc). He also appears to have been ignorant of the meaning of the word clem: ‘starve’, not ‘feed’ (‘I had to go and clem the horses, and milk ‘tould cow. . .’ Act 1, v.) Dickens himself clearly liked the play, as a letter to John Forster reveals around the 23rd November 1838 ( in House & Storey [eds] 1965: 459).
50
Katie Wales
special new Preface for the first cheap edition (1848), describing his meeting at Barnard Castle with his “informant” on the Yorkshire schools, a “ jovial, ruddy, broad-faced” Yorkshire attorney5. After a couple of hours’ talk, the attorney concludes: (20) ’Weel Misther, we’ve been varra pleasant toogather, and ar’ll spak’ my moind tivee. Dinnot let the weedur send her lattle boy to yan o’ our schoolmeasthers, while there’s a harse to hoold in a’ Lunnon, or a goother to lie asleep in. Ar wouldn’t mak ill words amang my neeburs, and ar speak tiv’ee quite loike. But I’m dom’d if ar can . . . not tellee, for weedur’s sak. (cited Jaques 1915: 296–7, my italics). Dickens comments that he sometimes “imagines” that he descries a “faint reflection” of him in John Browdie. The attorney may well have provided an influence (and enough critics in the early twentieth century certainly tried to identify the ‘real’ John Browdie); but I think Dickens is being cunning here: it is much more likely that Browdie is the influence for his representation of the attorney’s speech pattern. It is a clever ploy. I would further argue that this Preface, with the recreation of Dickens’ initial meeting at Barnard Castle with his ‘informant’ is an ‘authentificating effect’ as described in section 1 above. It arises, of course, out of a ‘real’ visit to the area; hence the actual setting of part of the novel at Greta Bridge: two further authenticating effects. And the real visit was itself occcasioned by true reports of awful conditions in the schools. There is a further authenticating effect, I think, which I am basing on Coupland’s argument in the same sociolinguistic debate (2003), and that is the very (internal) consistency and coherence, on the whole, of Browdie’s representation. Coupland (p. 417) argues for ‘consistency’ and ‘coherence’ as two of the qualities of ‘authenticity’: here then Dickens’ representation is validated. However, as a whole, it is the process of authentication and authenticating effects, that matter to Dickens as a creative practice; not ‘authenticity’ as it is commonly perceived. Even as late as 1857, three years after Hard Times was published, Dickens’ model of John Browdie came into use once again. Dickens and the novelist Wilkie Collins went on a tour of Cumbria, which was written up as a Christmas story, The Lazy Tour of Two idle Apprentices. Cumbria is, of course, in the Northwest, but the landlady at Allanby happened to have lived at Greta Bridge at the 5 A professional London solicitor had given Dickens letters of introduction in an assumed name. The set up was that a ‘friend’ had been left a widow, and wanted to place her boy at a North Yorkshire school.
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
51
time of his Nicholas Nickleby visit. She brings him a night-light, and in one of his letters home (1857) he records: (21)
‘It’s joost . . . new come oop. It’ll burn awt hoors a’end, and no goother [’gutter’, channel of wax], nor ony sike a thing. . .’ (in Storey & Tillotson [eds] 1995: 441–2.)
We can note the distinctive FOOT-vowel in oop, joost, goother; sike for ‘such’; rounding before a nasal (ony); and
to represent pre-R dentalisation in goother. It is arguable that his representation of Browdie-speech has not only become so ingrained, but also that this is how, by this time, he actually perceives the speech of the North-east (sic) of England north of the Trent.6 It is interesting that Dickens, again, did actually visit the area, so his representation is again ‘authenticated’. I conclude my analysis of Nicholas Nickleby’s northern English with a consideration of its significance in the history of literary dialect representation generally, for I feel that the novel has been insufficiently appreciated in this respect. First of all, let us return to the guard, a “stout old Yorkshireman” as he is called in Chapter 5, as Nicholas makes his journey from London to the North. Just past Newark there is an accident, and the guard is given his own voice (Chapter 6), using a similar if smaller range of features as Browdie is to use: elisions, for example (genelman, wi’); long /oː/ (boons, hoold) also brokken, wakken; toight; dean’t [don’t], etc. The contemporary reader used to the tradition before 1838 of dialect representation in novels would probably recognise a minor character of a lower social order; and, more specifically, the sort of character found in Smollett’s Roderick Random, the jolly waggoner Joey, referred to above, with a similar fleeting functional role7. What this same reader, however, would
6 Some of the voices he hears at Doncaster, on a visit to the races, are recorded in the same Christmas story: “We mun aa’ gang toogither”; “Bock your opinion loike a man. Coom! Dang it coom”; “Hond us a gloss of sal volatile in wather, or soom dommed thing o’ thot sart”. As well as the swearing, the rounding in bock, thot and gloss can be particularly noted. 7 Although he only appears in two chapters (11 and 12), it may well be the case that Smollett’s Joey is the first known example of a Northerner in a novel. Although it is not explicitly stated that he is a Northerner, there is the construction I’se (for ‘I am’, ‘I shall’), and forms like oop, coodgel and knaw. It is to be noted that Smollett’s only other Northern character, Timothy Crabshaw the ‘squire’ in The Life and Adventures of Sir Launcelot Greaves (1762), is a distinctly unsavoury character, with a hunchback, pot belly and bandy legs. Despite spellings as in moother, coople, the rural schema of ‘Mummerset’ predominates, with for as in vor and vorty (this feature absent in Browdie). It can be generally noted that the title Nichol-as Nickle-by has the same alliteration and sound-play as Peregrine Pickle, and the alliteration of Roderick Random, and the plot in the same picaresque tradition as Smollett’s novels.
52
Katie Wales
not be expecting is a character like John Browdie, with a major presence and force in the plot: it is he, after all, who helps in the escape of Smike, and Squeers’ downfall. In their discussion of their database of British fiction (1800– 36), Hodson & Broadhead (2013) give a table (p. 327) for the percentage of novels that feature at least one ‘Regional English’ – speaking character who speaks for more than ten lines. The percentage is never more than 10%; and in 1836, just before the publication of Nicholas Nickleby, only 9%: and that includes all regional varieties, not just Northern English. It will be 1842 before another Northerner is given a prominent role, in Emily Bronte’s Wuthering Heights. It is also the case that this same contemporary reader would also not be expecting the sheer creativity of Browdie’s idio-regiolect, as I have termed it. However, in terms of Dickens’ own development as a novelist, Browdie can recognisably now be placed in the gallery of idiosyncratic characters and their repetitive distinctive speech styles, begun with Sam Weller and Alfred Jingle in Pickwick Papers (1836–7), and no doubt still fresh in Dickens’ mind.
4 Hard Times Dickens’ ‘mental map’ and ‘dialect map’ of Coketown and its inhabitants in Hard Times, ostensibly placed across the Pennines in the North-west of the country, are quite different from that of Greta Bridge and Browdie, and deliberately so I believe, although the processes that make up the composition of Stephen Blackpool’s speech show some degree of overlap. A slightly different pattern of schemas is to be discerned, and hence a different texture and a different mood. There is, however, the same consistency and coherence to the representation. Stephen Blackpool is an (urban) cotton-mill worker, not a (rural) yeoman, so Dickens does not therefore invoke a rusticity schema for his character. As ‘Northerners’, Blackpool and Browdie do share common features of general ‘northern’ English, notably in phonology: so salient are forms like: Blackpool is more prone to L-vocalisation in words like: (23) towd, owd, gowd but similar is his short in: (24) mak, tak and also the lexicalised weel for ‘well’. There is a subtle distinction between them in definite article reduction (DAR): Blackpool favouring the reduced forms: (25) th’ bed, th’ kind, th’ pit, th’ road
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
53
rather than Browdie’s complete omission. They also share as Northerners the same script of bluntness and common decency. I return to this below. However, Dickens has tried to distinguish Blackpool from Browdie in several very distinctive ways, despite their common northernness. He appears to have made more of an effort at focussing on a specific dialect area, however broadly defined. There was certainly a strong literary tradition of Lancashire dialect representation. His friend Harrison Ainsworth had written The Lancashire Witches in 1848, itself drawing upon seventeenth-century attempts by playwrights such as Thomas Heywood and Richard Brome (1634) and Thomas Shadwell (1681) on the same subject (Wales 2006: 77–8). But Dickens, interestingly, does not try to present particular striking Lancashire pronunciations represented in spellings such as keaw; sooa; loook (also acknowledged by Gerson 1967: 339); nor the marked hoo for the pronoun ‘she’. One specific literary source can certainly be identified for Blackpool’s distinctive vocabulary, namely Tim Bobbin, a series of dialogues and a glossary published some one hundred years before Hard Times (in 1746), by John Collier of Urmston on the Lancashire-Cheshire border- and so, significantly, probably out of date, reflecting archaic lexis, by the midnineteenth century. (Certainly spinning and weaving had progressed from family to factory.) According to Stonehouse (1935) Dickens had his own copy of this celebrated work, the 1818 edition, and he could certainly have picked out lexical items from the Glossary. We can note in Stephen’s speech words like: (26) dree [dreary], fewtrils [little things], hetter [violent], hottering [raving], hummorbee [bumble bee], moydert [muddled], hey-go-mad [enthusiastic], fratch [quarrel]. Whilst it was over thirty years ago that Easson (1976) identified Tim Bobbin as the probable source of Blackpool’s vocabulary, it has not been noted elsewhere to my knowledge that the very first page of the ‘Observations’ to that same work are likeliest to be the inspiration for a distinctive repeated feature of Blackpool’s morphology – and certainly not a feature of Browdie’s – namely -en participle forms: and often regardless of historical ‘accuracy’. In Tim Bobbin’s own words, these are the “Saxon Termination”: hence (26) ha’ growen; ha’ getn; has been broughten. Note the remarkable over-generalised or hypercorrect string in: (27) I that ha’ ett’n and droonken . . . an’ seet’n . . . and toil’n and lov’n . . . Remarkable indeed is the visual homography with the negative as in: (28) would’n ha’ took’n; shouldn’ ha’ had’n
54
Katie Wales
and also the visual homography with the reduced form of the present participle in forms like: (29) ett’n, setten, fitten, looken, want’n, likens, etc. The homography is unignorable and surely deliberate: and I think is meant to suggest symbolically an idio-regiolect, and moreover one which is out-of-date, like Tim Bobbin itself. This in itself symbolizes Stephen’s old-fashioned ways and homely values, which make him out of step, out of tune, with the fasterpaced times. Awl(s) a muddle is his frequent refrain: everything a muddle, moydert8. But in terms of my larger argument, what Tim Bobbin also does is ‘authenticate’ Dickens’ representation; as indeed, again, does his own visit to the North-west and to Preston, described in his serial Household Words in February 1857. Noteworthy too and again unignorable on the printed page is the greater frequency than in Browdie’s speeches of eye-dialect forms: (30) widder, hansom, forrard and especially of colloquial elisions or allegro forms, graphemically marked by frequent apostrophes: (31) ha’ [have], an’, o’ [of], wi’. with an apostrophe is very frequent also: representing two distinct forms: the pronoun it before an auxiliary verb as in: (32) ‘tis, ‘twould, ‘twas, ‘twill, ‘tan’t; and the preposition to as in: (33) t’ sew, t’ turn, t’night. Again, Collier’s ‘Observations’ note how a “great many letters, and even words” are replaced by apostrophe and run together. As with Browdie, however, the 8 This ‘catchphrase’ certainly caught the contemporary imagination. As soon as the novel had appeared, Frederick Fox Cooper produced his play version in three acts, where the phrase appears several times, if in Standard English (the rest of the dialogue being prone to rustic stage-dialect). A reviewer Richard Simpson in October 1854 summed up the whole novel as “aw a muddle! Fro’ first to last, a muddle!” (cited in Collins [ed] 1971: 303–4). In a collection of parodies and lampoons (1857) edited by a good friend of Dickens, E.H. Yates, and also a contributor to Household Words, appeared a parody of Hard Times by Robert Brough. As Stephen lies in the pit he tells Rachel that his head is “awlus a muddle”; and tells his friends who come to rescue him “wi’ regard to law – it’s awlus a muddle”. ( At least in this version, as in Cooper’s, Stephen does not die, but slips away to marry Rachel.)
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
55
subliminal meaning is surely that such characters are or-ate rather than liter-ate, ill at ease in formal rather than informal situations, however articulate, indeed eloquent, in their own domains. Blackpool’s lack of standard concord signifies, like Browdie’s his lack of education befitting his social status; but in this much more serious novel crucially his powerlessness: (34) people’s leaders is bad; they taks; the mills is awlus a-goin’9. The prosody of the non-standard and of the colloquial as a schematic ‘base’ can be seen, however, to symbolise the generic worker, regardless of region: what Collins (1979 ) nicely terms “an industrial Everyman” (xv)10. We can note in this context Dickens’ letter to Peter Cunningham printed in the Illustrated London News (11th March (1854), denying that the workers’ strike in Preston lay behind the title of the novel: “It localizes . . . a story which has a direct purpose in reference to the working people all over England” (cited in Storey et al (eds) 1993: 291). And clearly the setting ‘Coketown’ is symbolic, non-referential, not a ‘real’ place in the way that Greta Bridge is. Blackpool’s own lack of social standing, however, is of direct relevance to the plot, sowing the seeds of his own tragedy. Blackpool is even more important a character in the novel than Browdie in his. He appears in eight chapters of a much shorter novel than Nicholas Nickleby; and his dilemma of a marriage he would like to be free from mirrors the main plot involving Gradgrind’s daughter Louisa. I would also here like to argue that the sheer frequency of the colloquial spellings, which distinguish Blackpool’s idio-regiolect from Browdie’s, also owes a great deal to a frame of literary reference, which comes from Dickens’ readings of his contemporary and fellow contributor to his own journal Household Words, Elizabeth Gaskell. Specifically significant is her novel Mary Barton (1848), set in Manchester amongst the mill weavers and spinners, which she had sent Dickens to read and which he was impressed with (see Schlicke [ed] 2000: 25). The dialogues of the workers and their families are not noticeably dialectal or idiosyncratic like Stephen Blackpool’s, but they do, especially Mary Barton herself, rely heavily on the similar representation of mostly colloquial features (e.g. th’, a’, wi’, o’,) with a similar low-key tone11. I would argue here that Dickens recognised, from his reading of Gaskell, the need for linguistic restraint, rather 9 As one French reviewer puts it more grandly (Jacques Carre 1973), cited in Manning (1984: 77): “the dialect signifies the ontological impossibility of the proletariat’s ever becoming master of its own destiny”. 10 Elsewhere Collins (1980: 662) is not so complimentary about Stephen, speaking of his “lingo”. 11 There are some dialectal words, and these are glossed, especially in the early chapters, but Melchers (1978: 117) thinks that the omission of consonants is “the novel’s most striking feature”.
56
Katie Wales
than his habitual exuberance, in his portrayal of Blackpool, to reflect the seriousness of his theme rather than any comedy. Morever, precisely because Gaskell had spent seventeen years living in Manchester, and so was close to being an ‘insider’ rather than an ‘outsider’, any use of Mary Barton could be seen to be an ‘authenticating’ device. By a remarkable coincidence Gaskell’s novel North and South was due to be serialized in Household Words after Hard Times (the latter, April to August 1854), which similarly concerned discontents amongst the (Manchester) mill-workers. Dickens had certainly discussed it (January 1854), even possibly suggesting the title; and he had possibly also read some chapters by the beginning of July, although this is difficult to be sure of. As letters reveal (cited in Schlicke [ed] 2000: 254) Gaskell herself was certainly worried about the possibility of overlapping themes between them; and she had, in fact, started her novel before Dickens began his. Her complete manuscript did not, arrive, however, until the summer was advanced, appearing from the beginning of September onwards. Dickens’ scholars on the whole rather play down the possible influence of this novel on Dickens; but knowing of its existence may well at least have prompted Dickens to take more pains with his speech representation, and to go back to Mary Barton. In turn, it is also possible that Gaskell learned something from Hard Times (hinted at by Carnall 1964: 44, n. 28). Here then could be seen the co-construction of models of characterization: social cognition in its other strand, referred to in the Introduction, namely of a discursive community of practice in the widest sense12. There was certainly by the 1850s a concentration of political, ideological and also literary focus on what is sometimes termed ‘the condition of England’, as result of a rapid industrialization north of Watford. Dickens and Gaskell shared this focus; but unlike say Disraeli in Sybil (1845) or Fanny Trollope in The Life and Adventures of Michael Armstrong (1840), they stand out for their attempts to render northern working-class speech. Morever, this shared interest in the representation of vernacular speech patterns means that Hard Times must be seen in the context of literary tradition, to be a significant novel; and notions of ‘insider’ versus ‘outsider’ irrelevant. Also, in their own literary ways, the northern characters of Gaskell’s novels and Hard Times invoke a kind of Wordsworthian ‘authenticity’, anticipating the demotic characterization of the later novels of Thomas Hardy. A year after Gaskell’s death Dickens wrote to a theatrical producer friend Charles Fechter (September 4th 2866), about a play by Boucicault based on Mary Barton, called 12 Both William and Elizabeth Gaskell were also strongly interested in the Anglo-Saxon roots of the Lancashire dialect.
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
57
The Long Strike, and similarly set in Manchester. Dickens makes a rare critical comment: If the characters did not speak in a terse and homely way, their idea and language would be inconsistent with their dress and station, and they would lose, as characters, before the audience. . . Throughout there is an honest, straight-to-the purpose ruggedness in it, like the real life and the real people (cited in Storey [ed] 1999: 240–1; my italics)
This Wordsworthian principle of decorum is, I think, at the heart of the representational practice of Hard Times for Blackpool, and what Dickens could have learned from Gaskell herself. It is interesting in this respect that a contemporary reviewer of Hard Times in the Westminster Review vol. 6. for 1854, albeit a London publication, admired the portrayal of Stephen Blackpool for his “rugged steadfastness, sturdy truth, upright bearing and fine Northern English dialect, smacking strongly of the old Saxon. . .” (cited in Collins [ed] 1971: 306–7): ‘Northern English’ (sic) clearly associated in this critic’s mental schema with virtue, not mockery, an association no doubt aided by Dickens’ own portrayal. It is also further of interest that this same reviewer linked Blackpool with Browdie, continuing with: “[he] is a noble addition to the gallery which already contained the bluff John Browdie”. For these two characters the popular nineteenth-century northern epithet, common in poems and ballads, gradely comes to mind. (See further Wales 2006: 133.) Meaning ‘decent, respectable, worthy’, etc it is first recorded in Tim Bobbin, and if most frequently found in the North-west, it is also recorded in Yorkshire. One irony, however, is that the kind of ‘Northern English’ the London critic admires in Hard Times is heavily based on a text a century old by then. But the archaic flavour which characterizes Stephen’s idio-regiolect may well have played a part in this critic’s (and others’) perception of Northern English anyway, as his praise of the hint of “old Saxon” (echoing Tim Bobbin) seems to confirm.[13]
5 Conclusion(s) In terms of my basic framework of social cognition, characters like John Browdie and Stephen Blackpool are not monological, single-voiced: their voices, with their values are constructed sympathetically and inter-textually through social ideologies, beliefs and cultural practices, including other literary texts. Dickens invites several frames of reference, based on: common social images of the under-classes; cultural knowledge and beliefs about Northerners shared by his readers and audiences; also his reading of earlier literary dialects and dialect
58
Katie Wales
materials, as well as the work of his contemporaries; and general knowledge of dramatic and novelistic conventions. His dialect-speaking characters as a result are indeed ‘stereotypes’, but not in any simple one-dimensional way: they have been built up incrementally. As a result, they are not predictable, like ‘stock’ characters in drama. What I have particularly come to appreciate by ‘unpicking’ Dickens’ method in these terms, by examining the processes of literary formation, is the way he succeeds in keeping the voices and also the whole ‘textures’ of the two novels distinct, by a careful, reasonably consistent selection of both generic and particular features, however oddly placed regionally speaking. There is no single stereotypical Northerner to serve both novels. Further, his dialect speech for his alleged ‘North Yorkshire’ and ‘Lancashire’ characters is a good deal more complex than might at first appear. This is no random collection of features. In an important sense Dickens is not ‘representing’ northern speech but creating it. Given the general critical praise for Dickens’ stylistic creativity, this should not be too much of a surprise. His idio-regiolects as I call them are based on their own selected schemas, they have their own ‘rules’, their own ‘gaps’, their own overgeneralizations or ‘motifs’. Mental modelling meets artistic modelling. I have also come to appreciate more and more Dickens’ achievement as an artist on the one hand; and as a ‘regional’ novelist on the other: together he is truly significant in the English novel tradition. Furthermore I have tried to make a case for the irrelevance of notions of ‘authenticity’ and ‘accuracy’ in the traditional senses; or rather, their transcendence in considering the work of a writer like Dickens. For the literary critic Trilling “the concept of authenticity can deny art itself, yet at the same time it figures as the dark source of art” (1972: 11). For Dickens, read ‘authentication’. I have been struck by the different ways in which he has succeeded in processes of ‘authentication’ or ‘authenticating effects’, which have also blurred the traditional distinctions made in criticism between ‘insider’ and ‘outsider’ status, and ‘intrinsic’ and ‘extrinsic’ presentation. In the future I should like to take both the social cognition framework and the idea of ‘authentication’ and look at other novelists’ dialect representations from 1700 to 1839; as well as Dickens’ own better known representations of London English or ‘Cockney’ as a posited ‘insider’ (London being his main residence). Novels, like plays, perpetuate images in readers’ minds: Blackpool and Browdie probably did contribute to the general literary schema of ‘northernness’; but it would be easier to find evidence of this kind of process for Dickens’ Cockney representations, because of the likelihood of more evidence of critical reception.
Dickens and northern English: stereotyping and ‘authenticity’ reconsidered
59
References Brook, G. L. 1970. The language of Dickens. London: Deutsch. Bucholtz, Mary. 2003. Sociolinguistic nostalgia and the authentication of identity. Journal of Sociolinguistics 7 (3), 398–416. Carnall, Geoffrey. 1964. Dickens, Mrs Gaskell, and the Preston strike. Victorian Studies 8: 31–48. Collier, John. [1746] 1798. The Miscellaneous works of Tim Bobbin, Esq. Containing his view of the Lancashire dialect. London: for A. Milar et al of York. Collins, Philip. (ed.) 1971. Charles Dickens: The critical heritage. London: Routledge. Collins, Philip. (ed.) 1975. Charles Dickens: The public readings. Oxford: Clarendon Press. Collins, Philip. (ed.) 1979. Hard times. London: Dent. Collins, Philip. 1980. Dickens and industrialism, Studies in English literature 20, 651–73. Condor, Susan & Charles Antaki. 1997. Social cognition and discourse. In Teun. A. van Dijk (ed.) Discourse as structure and process, 320–347. London: Sage. Cooper, Frederick F. [1854] 1886. Hard times: a domestic drama in three acts. (London: Dick’s Standard Plays vol.31 no 785) Coupland, Nikolas. 2003. Sociolinguistic authenticities, Journal of Sociolinguistics 7(3), 417–31. Culpeper, Jonathan. 2001. Language and characterisation in plays. Harlow: Longman. Dickens, Charles. [1867] 1877. The readings of Mr Charles Dickens, as condensed by himself. Boston: Lee & Shepard. Easson, Angus. 1976. Dialect in Dickens’ Hard Times, Notes & Queries n.s. 23 (9), 412–3. Eckert, Penelope. 2003. Sociolinguistics and authenticity: an elephant in the room, Journal of Sociolinguistics 7 (3), 392–397. Gerson, Stanley. 1967. Sound and symbol in the dialogue of the works of Charles Dickens. Stockholm: Almqvist & Wiksell. Giles, Howard & P. Powesland. 1975. Speech style and social evaluation. London: Academic Press. Hickey, Raymond. (ed.) 2010. Varieties of English in writing. Amsterdam: John Benjamins. Hodson, Jane & Alex Broadhead. 2013. Developments in literary dialect representation in British fiction 1800–36, Language and Literature 22 (4), 315–32. House, Madeline & Graham Storey (eds.) 1965. The Pilgrim edition of the letters of Charles Dickens, vol. 1. (1820–39). Oxford: Clarendon Press. Jaques, E. T. 1915. The original of John Browdie, The Dickensian, 11: 296–9. Macrae, C. Neil, Charles Stangor & Miles Hewstone (eds). 1996. Stereotypes and stereotyping. London: The Guildford Press. Manning, Sylvia B. 1984. Hard Times: An annotated bibliography. New York: Garland. Melchers, Gunnel. 1978. Mrs Gaskell and dialect, Studies in English Philology, Linguistics and Literature, 112–124. University of Stockholm; Stockholm Studies in English 46. Pennington, Donald C. 2000. Social cognition. London: Routledge. Pinner, Richard. 2014. The authenticity continuum: Towards a definition incorporating international voices, English Today 30 (4), 22–7. Schlicke, Paul. (ed.) 2000. The Oxford reader’s companion to Dickens. Oxford: Oxford University Press. Stangor, Charles. (ed). 2000. Stereotypes and prejudice: essential readings. Philadelphia: Taylor & Francis. Stirling, Edward. 1838. Nicholas Nickleby: A farce. London: Chapman Hall.
60
Katie Wales
Stonehouse, John H. 1935. A catalogue of the library of Charles Dickens from Gadshill. London: Fountain Press. Storey, Graham, Kathleen Tillotson & Angus Easson (eds.) 1993. The Pilgrim edition of the letters of Charles Dickens, vol.7. (1853–5). Oxford: Clarendon. Storey, Graham & Kathleen Tillotson (eds.) 1995. The Pilgrim edition of the letters of Charles Dickens, vol. 8. (1856–8). Oxford: Clarendon Press. Storey, Graham. (ed.) 1999. The Pilgrim edition of the letters of Charles Dickens, vol. 11. (1865–7). Oxford: The Clarendon Press. Trilling, Lionel. 1972. Sincerity and authenticity Oxford: Oxford University Press. Wales, Katie. 2006. Northern English: A social and cultural history. Cambridge: Cambridge University Press. Wales, Katie. 2010. Northern English in writing. In Raymond Hickey (ed.), Varieties of English in writing, 61–80. Amsterdam: John Benjamins. Yates, Edmund H. & Robert B. Brough (eds.) 1857. Our miscellany. London: no publisher given.
Michael Pearce
4 The linguistic landscape of north-east England Abstract: Research on language attitudes and perceptual dialectology has shown that north-east English is one of the most widely recognized and positively evaluated varieties in Britain. There is also a rich tradition of dialect writing associated with the region, and a long history of both ‘folk’ and scholarly attention to local forms of language – for example, it is the only part of England to have a major corpus devoted to it – the Diachronic Electronic Corpus of Tyneside English (University of Newcastle). However, one element is missing from the otherwise well-charted dialectological terrain: an account of localized forms in the ‘linguistic landscape’ (that is, the public display and representation of written language on road signs, advertisements, house names, vehicles, and so on). This chapter is a first foray into an under-researched aspect of the region’s local linguistic ecology. It describes and contextualizes a corpus of signs compiled in 2014–15, showing how they draw extensively on a set of features which previous studies have revealed to be enregistered as part of north-east dialect. The implications of the findings are discussed and the – perhaps surprisingly – infrequent appearance of such forms in the linguistic landscape is addressed.
1 Introduction The origins of north-east England as a coherent region can be traced back to the establishment of the kingdom of Northumbria by Anglian settlers in the seventh century. At its greatest extent it stretched from the Humber in the South to the Firth of Forth in the North, bounded to the West by the Pennines and the East by the sea. Today’s north-east England remains a “distinct and well defined geographical, administrative, historical, economic and cultural region” (Duke et al. 2006: 5) lying mostly within the territory of the historic counties of Durham and Northumberland, and dominated by the urban areas at the mouths of its three main rivers: the Tyne, the Wear and the Tees. One of the region’s most wellknown contributions to English culture is linguistic. With the exception, perhaps, of Merseyside, the English spoken in the North-east is the most widely recognized
Michael Pearce, University of Sunderland DOI 10.1515/9783110450903-004
62
Michael Pearce
by British people (Montgomery 2012: 173). Research on language attitudes and perceptual dialectology has also revealed that a stable set of stereotypes has developed concerning north-east speakers, reinforcing broad cultural generalizations. For example, Coupland and Bishop report that in a study of informants’ attitudinal responses to a set of 34 “relatively unambiguous regional or national accent labels” the “Newcastle” accent is ranked higher for social attractiveness than for prestige (2007: 77–80), reflecting widely held and generally positive beliefs about the culture of working-class sociability and solidarity associated with the region (see Tomaney 2010: 87–88). The cultural importance of northeast English is also reflected in its position in academic dialectology. Scholarly attention to local language has a long history in the region, with the earliest lexicographic work purporting to cover northern English – Brockett’s North Country Words – having a strong “Northumbrian” focus (1825, a second enlarged edition appeared in 1829). Towards the end of the nineteenth-century, a number of works on the English of the region were published by the English Dialect Society, including Heslop’s Northumbrian Words (1892–3) and Palgrave’s Words and Phrases in Every-day Use by the Natives of Hetton-Le-Hole in the County of Durham (1894–5). These were not simply glossaries; they also contained linguistic descriptions of the local folk-speech. Scholarly interest was maintained into the twentieth century with a wealth of academic publications appearing from the 1970s onwards (see Beal et al. 2012: 24–26 for an overview). It remains the only English region which has a major corpus devoted to it: the Diachronic Electronic Corpus of Tyneside English (DECTE). However, there is a missing element in the well-charted linguistic terroir of the region: an account of the use of localized forms in the “linguistic landscape”. Shohamy and Ben Rafael (2015) define the linguistic landscape (LL) as the display and representation of languages in public spaces – mainly, but not exclusively in the form of writing on road and shop signs, advertisements, house names, graffiti, and so on. Some of these signs are more or less permanently fixed in place; others are temporary. And some might be categorized as “noise” (Blommaert 2013: 53), passing through the landscape on cars, lorries, and boats, or on discarded packaging, on placards in political demonstrations, or printed on items of clothing and other forms of merchandise such as greetings cards.1 Traditionally, linguistic landscape studies has focused on the interface between the LL and multilingualism, where the visibility of a language and the role given
1 Online linguistic landscapes have also begun to attract the attention of researchers, although this domain lies outside the scope of the chapter.
The linguistic landscape of north-east England
63
to it is regarded as indicative of the status of that language and its ethnolinguistic vitality (Johnstone 2011: 215–16). Often, LL research has a quantitative element, where the proportion of signage devoted to a particular language in a given territory is calculated. In theory, such an approach can also be adopted in relation to the visibility of varieties within a language. However, in practice, for reasons outlined in the next section, this is not always an effective method. In this chapter I report on a study of local forms of English in the linguistic landscape carried out in 2014–15. I describe my corpus and its contexts, showing how the signs draw extensively on a set of features which previous studies have revealed to be enregistered as part of north-east dialect. I consider the implications of these findings, as well as addressing the – perhaps surprisingly – infrequent appearance of these forms.
2 Local features in the linguistic landscape The scarcity of north-east English in the linguistic landscape was made clear in a survey I undertook in 2014 of “The Nook” in South Shields, Tyne and Wear.2 Anyone on a situationist-inspired dérive past the Nook’s 85 retail outlets will be struck by the contrast between the linguistic soundscape, which is overwhelmingly local and vernacular, and the linguistic landscape, which is not. Only one business used a north-east form in its signage: Computa-z.3 How do we account for the rarity of written local dialect on public display? One answer lies with the “standard language ideology” (Milroy 2007): a set of value-judgments whereby characteristics such as correctness, authority, prestige and legitimacy are associated with one particular sort of English (Standard English), and not with others (Garrett 2010: 34). In the UK almost all edited and printed texts designed for a broad circulation are in Standard English. Nonstandard writing tends to be limited to “dialect literature” (literary work written in a non-standard variety, largely for the local consumption of people familiar with the dialect) and “literary dialect” (where representations of dialect speech 2 “The Nook” is a purpose built shopping precinct, serving the people on the southern outskirts of the town since the 1930s. It consists of a mixture of independent and chain shops, and is situated in a largely working-class area. Interestingly, according to the English Dialect Dictionary (EDD), nook is a dialect term to describe a remote, out of the way place or an extremity of anything (Wright 1905: 294). 3 This spelling reflects the quality of the vowel in the final syllable of words such as letter and comma (see 3.2).
64
Michael Pearce
occur within a Standard English matrix) (Shorrocks 1996: 386). An additional context is informal personal writing; these days often to be seen on internet messageboards and social network sites (see Pearce 2011a and 2015). In England, the most public domain of all – that part of the Earth’s surface which lies within an observer’s field of vision – remains dominated by Standard English. This domination is maintained despite well-founded claims that the hegemony of the standard is unravelling somewhat in the conditions of late modernity. Coupland has labelled this process vernacularization: “a realignment of ideological values and social norms” which shifted the balance between the standard and “urban British vernacular”. Vernacularization involves both “a weakening or restriction of standard language ideology” and “a more positive valorization of vernacularity” (Coupland 2014: 85–86). These tendencies have been in tension ever since the process of standardization began: “valorization of vernacularity” is not a uniquely contemporary phenomenon. However, late modernity has exacerbated this process by accelerating mobility and bringing speakers of different languages and varieties into contact, resulting in phenomena such as levelling, supra-localization and dialect erosion (Heller, 2005; Britain, 2010; Johnstone, 2010). As Johnstone points out (2010: 391), the conditions which reduce linguistic difference are also those which, somewhat paradoxically, “foster dialect and language awareness”: when speakers of different varieties come into contact, the result is “increased popular attention to variation”. In north-east England, this is reflected in some domains but not others. For example, as Beal (2009) has shown, there is considerable commodification of features of north-east English on a range of consumer items such as clothing, kitchenware and greetings cards. But these features are not often seen in the wider linguistic landscape. Perhaps this is due to the relatively stable position and status of northeast dialect. In contexts where local vernaculars are not so secure there seems to be more use of local features in the LL. Evidence for this lies in work done on “Low German” in East Frisia in the North-west of Germany. Reershemius (2009 and 2011) claims that after a period of relatively stable diglossia in which Low German was the spoken variety and High German the standard variety until the middle of the twentieth century, “parents stopped speaking Low German with their children” because of fears of educational disadvantage and being left behind in a rapidly modernizing society, threatening the variety with extinction (2009: 132). Research published in 1984 indicated that only a third of the region’s population could be considered as “competent speakers” of Low German. Reershemius concludes her study of the linguistic landscape of East Frisia by pointing out that while Low German is quite visible in the LL, this does not mean that it reflects high levels of ethnolinguistic vitality. On the contrary, she
The linguistic landscape of north-east England
65
concludes that Low German is presented in a way that appeals to addressees who value “past orientation and the perceived quietness of a more traditional and less complex pre-modern life” (2011: 44). This extends to the use of Fraktur in Low German signs, a typeface popular in Germany between the 1920s and 1940s designed to evoke tradition. Reershemius also shows how many of the businesses using Low German are guest-houses, tea rooms and coffee shops: concerns with a largely tourist clientele. Here the LL has been commodified for external consumption. This is not the case in the North-east of England, where, as we shall see, few of the signs directly evoke the region’s past and most are designed to appeal to local people. The rarity of north-east elements possibly signals that speakers of northeast English do not feel that their varieties are under any particular “threat”. It also means that traditional LL methods such as the survey of a prescribed geographical area will not offer up much in the way of relevant data. For this study, a different method was employed. Initially, I started my own personal collection; I then enlisted the help of students and colleagues.4 By the end of the sampling period (March 2014 to November 2015) I had 57 images. The criteria for inclusion was that the picture should contain at least one “north-east element”. The inverted commas here indicate the problematic nature of identifying items associated with the English of a particular region, due to the well-known fact that there are no discrete boundaries between areas where different dialects can be heard. Such fuzziness is evident in some of the pictures which were submitted. For example, one showed a sign attached to a bus shelter with the imprecation Please keep our bus shelter clean and fresh – ta! When I asked my informant why he had sent me the image, he pointed to ta. What status does ta have here? Is it “north eastern”? Is it “northern”? Is it simply colloquial? Another photograph shows a catering van in a lay-by advertising CHARLTONS’ FAMOUS FREE-RANGE SAUSAGE BUTTY.5 The relevant element here is butty (“sandwich”), but like ta it is found widely across the UK, perhaps (originally) with a northern flavour – the EDD gives Yorkshire, Lancashire, Cheshire and Staffordshire (Wright 1898: 468). Nevertheless, almost all of the submitted items did indeed contain examples of north-east English; that is, features which have been identified by various scholars as only occurring or occurring with relatively higher frequency, in the region. It seems that the majority of contributors had quite a firm sense of what
4 The corpus can be found at . I would like to thank all those who contributed photographs, particularly Angela Smith, Lesley Savage, Clare Fairclough and Damien Hall. 5 Text from the data is presented in italics. The original’s use of lower- or upper-case is preserved, as is spelling and punctuation.
66
Michael Pearce
a north-east element might be. This is because dialects are ideological constructions which have arisen through the process of “enregisterment”, whereby a set of linguistic features is “represented collectively in the public imagination as a stable variety” (Beal and Cooper 2015: 28). This collective representation can be found anywhere there is metapragmatic activity or “talk about talk” (Johnstone 2009: 160). Sometimes this ‘talk’ takes place in the context of online discussions which often display nuanced awareness of the social salience of particular linguistic forms (see Pearce 2015). In addition, the region’s speech is the topic of numerous publications in the field of what Honeybone and Watson (2013) and Honeybone et al. (this volume) call Contemporary Humorous Localised Dialect Literature (CHLDL): brief, locally produced books celebrating linguistic and cultural distinctiveness and diversity (examples include Dobson 1986, Waddell 2008, Black 2008, Candlish 2006). The process of enregisterment in the region has a long history (see Beal 2009; Beal and Cooper 2015) and many of the features discussed in the next section have considerable historical pedigree. Yet, as I shall show, there is very little of the archaic about them: such items reflect a living and dynamic vernacular linguistic culture.
3 Corpus description and analysis 3.1 The corpus Table 4.1 contains all items archived up to November 2015. LL studies often proceed by categorizing material according to the particular domains with which signs are associated. One distinction that is made is between “top-down” and “bottom- up” (Backhaus 2007: 27) which contrasts official signs produced by local and national government authorities with all others. No signs in the corpus are top-down, with the marginal exception of those found on public transport. This reflects the potency of the standard language ideology in England where Standard English is the de facto dialect of officialdom. Signs with north-east elements are bottom-up, and the majority of these are associated with commercial enterprises. For example, nearly half are the names of shops and services (e.g. KLIPAZ; CANNY CARPET CLEAN) or slogans related to them (e.g. Div’nt Dunch us or a’ll gan to PREMIER). Of the 18 commercial premises with northeast elements in their name, 11 are food and catering outlets (e.g. MACKEM PIZZA; TOON Sarnie) five offer manual trades, and two are second-hand shops. And just as these businesses cluster in terms of the kinds of goods and services they provide, they are geographically concentrated in areas of comparatively high social deprivation. Nearly 40 percent are located in the 10 percent most deprived
67
The linguistic landscape of north-east England
Table 4.1: The LL corpus Index of social deprivation
Text
Sign type, context and placement
(1) #4GEE. IT’S CANNY FAST PET. SUPERFAST MOBILE #4GEE HAS LANDED.
Advertisement for mobile phone service in shop window
1
(2) A CANNY LOYALTY CARD
Business loyalty card distributed to customers of food/drink outlet
1
(3) Any Wedge or Stottie Sandwich
Advertisement on board outside food/drink outlet
1
(4) Av-A-Pizza
Business name on sign fixed to front of food/drink outlet
3
(5) AYE PHONES
Business name on sign fixed to front of second-hand shop
2
(6) AYE REET
Graffito on wall of underpass
4
(7) Bairns plodging
Text accompanying drawing in museum
1
(8) BAIT BOX
Business name on sign fixed to mobile catering van
1
(9) BONNIE PIT LAD
Business name on sign fixed to front of public house
2
(10) BONNIE TILER
Business name on sign fixed to van
1
(11) Calm Doon BONNY LAD
Text on greetings card on sale in shop
–
(12) Calm Doon BONNY LASS
Text on greetings card on sale in shop
–
(13) Calm Doon PET
Text on greetings card on sale in shop
–
(14) Canny Beds
Business name on sign fixed to van
–
(15) Canny Cakes
Business name on sign fixed to front of cake shop
1
(16) CANNY CARPET CLEAN
Business name on sign fixed to van
3
(17) Dinnit break me TEMPER
Text accompanying drawing in museum
1
(18) Div’nt Dunch us or a’ll gan to PREMIER
Business slogan for insurance firm on car sticker
–
(19) Eee hiya, where have you been hiding?
Advertisement for visitor attraction in train carriage
–
(20) eeeeh HINNY
Text on greetings card on sale in shop
–
(21) Gadgy Ganny
Text accompanying drawing in museum
1
(22) GAN CANNY WOR GRADUATES
Text on bag given away to Newcastle University graduates
5
68
Michael Pearce
(23) gannin’ on tha HOY
Text on greetings card on sale in shop
–
(24) GEORDIES ARE BLACK AND WHITE. BOOT OUT THE EDL AND NAZIS
Political sticker on bench in street
3
(25) HA’WAY THE LADS
Spelled out with white seats against background of red seats in football stadium
4
(26) HA’WAY THE LADS
Banner attached to scaffolding outside house
2
(27) Haddaway an’ SHITE
Text on greetings card on sale in shop
–
(28) Haway, MARRA!
Text on greetings card on sale in shop
–
(29) HOWAY, man, WOMAN
Text on greetings card on sale in shop
–
(30) HUGE RANGE OF MAM & NANA CARDS
Advertisement on board outside stationery shop
1
(31) I Love carpets me!
Business slogan fixed to front of carpet shop
1
(32) IT’S MINT. SOUTH TYNESIDE COLLEGE. FRESH THINKING.
Advertisement for FE College on bus shelter
1
(33) KLIPAZ
Business name on sign fixed to front of hairdressers
2
(34) Koffie-Te-Gaan
Business name on sign fixed to front of food/drink outlet
2
(35) mackem
Text on mug on sale in shop
–
(36) mackem mover
Business name on sign fixed to van
–
(37) MACKEM PIZZA
Business name on sign fixed to front of food/drink outlet
1
(38) NAY WET CLOTHING & PLEASE WIPE YA FEET
Notice fixed to entrance of visitor attraction
7
(39) OWT TA SELL GIZ A BELL!
Business slogan fixed to front of second-hand shop
2
(40) REET BOBBY DAZZLA
Text on greetings card on sale in shop
–
(41) Reet Petite
Business name on sign fixed to front of food/drink outlet
1
(42) Sand Dancer
Business name on sign fixed to front of public house
4
(43) Scran2Gan
Business name on sign fixed to front of food/drink outlet
2
(44) TAPHOOOUSE QUEEN
Name plate fixed to fishing boat
–
69
The linguistic landscape of north-east England
(45) THE TOON TAKEAWAY
Business name on sign fixed to front of food/drink outlet
4
(46) Toon
Slogan on side of bus
–
(47) TOON Sarnie
Business name on sign fixed to front of food/drink outlet
5
(48) WASS HAIR
Business name on sign fixed to front of hairdressers
4
(49) WATCH the WEANS!
Notice fixed to ice-cream van
–
(50) What’s gannin on?
Sign on notice-board in university bookshop
5
(51) What’s gannin’ on in the Toon?
Slogan on advertisement in university bookshop window
5
(52) Whey Aye Five O
Slogan on side of bus
–
(53) whey AYE PET
Text on greetings card on sale in shop
–
(54) WHY AYE, PET
Text on greetings card on sale in shop
–
(55) Wi-i-Fi it’s FREE man!
Banner advertising wi-fi in shopping centre
3
(56) WOR HOOSE
Charity name on sign fixed to house
1
(57) YOU’RE MINT
Text on greetings card on sale in shop
–
neighbourhoods in the country, with almost all in the 50 percent most deprived neighbourhoods.6 Only five of the signs occur in what might be regarded as “elite” contexts, where an audience of non-locals might be envisaged. Three are associated with Newcastle University.7 Two are located in a museum. When businesses use them in their signs it is presumably because features of north-east English are perceived as having some value in their section of the marketplace. Certainly, other LL studies have shown this imperative in action (see, for example, Papen (2012) on the symbolic use of English in Berlin shop signs). A similar motivation is presumably behind the use of north-east features
6 The Office for National Statistics ranks the 32,844 neighbourhoods in England according to an index of multiple deprivation. 1 = a neighbourhood in the 10 percent most deprived neighbourhoods; 10 = a neighbourhood amongst the 10 percent least deprived (2015 Indices of Deprivation Mapper: ). Table 1 only shows indices for permanently fixed signs. I include a selection of greetings cards since they were a popular category of submission, although strictly speaking because they are usually found indoors they are not part of the LL. 7 According to the Higher Education Statistics Agency, in 2014-15 only nine English universities had a lower proportion of state-educated students than Newcastle .
70
Michael Pearce
in the advertising material in the corpus (that is, promotional text not physically attached to a business premises or vehicle). For example, the broadband company (IT’S CANNY FAST PET), wi-fi provider (Wi-i-Fi it’s FREE man!), nature reserve (Eee hiya, where have you been hiding?) and FE college (IT’S MINT) must have wanted to evoke aspects of the local because they envisaged commercial returns from it. The incorporation of speech-like, vernacular styles in promotional writing is an element in the broader strategy of conversationalization, in which public discourse employs some of the features traditionally associated with informal speech (Fairclough 1996). This strategy is particularly stark in relation to merchandise featuring what we might call commodified dialect display, as seen in the greetings cards and tote bag in the LL corpus. Here, the language itself is the commodity. Table 4.2 lists all north-east elements found in the corpus showing how many times they occur. In what follows I explore the data, focusing on some of the most common elements and discussing them in the context of the dialectological literature. I begin with phonetics and phonology, before considering lexis, discourse features, and morphosyntax.
Table 4.2: North-east features represented in the LL corpus
feature pronunciations (respellings) [uː] in MOUTH lexical set
example
TAPHOOOSE QUEEN (‘taphouse’)
number of items in which feature occurs
9/57
[iː] in some words in PRICE lexical set ending in orthographic
Reet Petite (‘right’)
3
[ɑ] and [ɐ] occur in final syllable of words such as letter and comma
KLIPAZ (‘clippers’)
2
[h]-dropping
Av-A-Pizza (‘have’)
1
I pronounced [ɑ]
Div’nt Dunch us or a’ll gan to PREMIER
1
give us → [gɪz]
OWT TA SELL GIZ A BELL!
1
Lexis, discourse features, and morphosyntax GAN What’s gannin on? canny CANNY CARPET CLEAN aye AYE PHONES bonny/-ie BONNIE TILER ha’way/haway/howay HA’WAY THE LADS LAD Calm Doon BONNY LAD pet Calm Doon PET
7 6 5 4 4 4 4
71
The linguistic landscape of north-east England
Mackem whey aye/why aye eeh man mint wor us bait dinnit div dunch gadgy ganny Geordies haddaway hinny hoy lass me mam marra nay owt scran stottie wass weans
MACKEM PIZZA Whey Aye Five O Eee hiya, where have you been hiding? HOWAY, man, WOMAN YOU’RE MINT WOR HOOSE Div’nt Dunch us or a’ll gan to PREMIER BAIT BOX Dinnit break me TEMPER Div’nt Dunch us or a’ll gan to PREMIER Div’nt Dunch us or a’ll gan to PREMIER Gadgy Ganny Gadgy Ganny GEORDIES ARE BLACK AND WHITE Haddaway an’ SHITE eeeeh HINNY on tha HOY Calm Doon BONNY LASS I love carpets me! HUGE RANGE OF MAM & NANA CARDS Haway, MARRA! NAY WET CLOTHING OWT TA SELL GIZ A BELL! Scran2Gan Any Wedge or Stottie Sandwich WASS HAIR WATCH THE WEANS
3.2 Phonetics and phonology Because of the standardized nature of the orthographic system, written English usually gives no clues about accent variation. However, writers can use semiphonetic respellings to suggest particular pronunciations, a strategy with a long history in both dialect literature and literary dialect, which has been revitalized recently in the context of writing on social media platforms. The most common respelling in the corpus is of the vowel in words in the MOUTH lexical set. Down, house and town are spelled , , to represent the traditional pre-GVS pronunciation [uː] in these words (compare Old English dūn, hūs and tūn). While contemporary production evidence would suggest that the majority of north-east speakers use the diphthongs [əʊ], [aʊ] or the more localized [ɛʊ] for most words in this set (Pearce 2009: 179–184; Beal 2004: 124), [uː] can also be heard in the region, particularly amongst male working-class speakers on Tyneside. Indeed, there is a particular perceptual
72
Michael Pearce
linking of [uː] with Newcastle (Pearce 2012: 7), which is probably related to its prominence in a small set of words associated with Tyneside identity, where the pronunciation “has been lexicalised and reflected in the spelling” (Beal 2004: 134). The most well-known of these is Toon which is used not to refer to towns in general, but usually occurs in the phrase “the Toon” as an alternative label for Newcastle and as a nickname for Newcastle United Football Club. Newcastle Brown Ale is also widely referred to as Broon (Beal et al. 2012: 35). However, in contemporary speech this vowel can still be heard in a larger set of words (e.g. out, pound, round), and wider use is also evident in informal, personal writing online (see Pearce 2012: 10). Though in decline, it is by no means entirely “historical”, and it is well-known enough to have acquired the status of an enregistered feature. Like and , the second most commonly occurring respelling – exemplified in for right – represents a survival from an earlier period. The modern indicates that in words such as right, night and light there was once an /h/ phoneme, realised as [ç]. During the Middle Ages, the /h/ was lost, and the preceding vowel was lengthened in compensation. This means that spellings such as and reflect a pronunciation – [iː] – that was once general in English but which is now only preserved in certain dialects of northern England and in Scotland. Beal et al. (2012: 35) suggest that words such as night and right take [iː] “in more conservative varieties”, and this is reflected in DECTE, where it is used by older male speakers. However, it does occur more widely across the age-range in fixed phrases such as “the neet” (“tonight”) and “all reet” (“all right” as a greeting). The only other accent feature represented more than once in the corpus is the quality of the vowel in the final syllable of words such as letter and comma. This is present in the LL in (clippers) and (dazzler). Such spellings are perhaps attempts to suggest an open vowel (frequently [ɐ]) rather than the closer vowel which is found in most non-rhotic varieties of English (RP has [ə], for example). Wells describes this vowel as [ɑ~ɛ] (1982: 376) and labels it “Geordie”; and indeed perceptual studies have shown that it is often identified as a marker of Newcastle speech (Pearce 2012: 10). Like [uː] in MOUTH it might therefore be considered an enregistered characteristic of the accent. Only one consonantal feature is represented in the corpus. The first element in Av-A-Pizza (have) reflects [h] deletion in stressed syllables. There is geographical variation within the region in [h] retention in words other than unstressed function words, and the fact that this business does lie in a part of the region where [h] deletion has been documented (South Hetton – a village south of Sunderland) potentially makes this sign a symbol of intra-regional variation which has been documented as salient for some participants in perceptual studies (see Beal et al. 2012: 41–42, Pearce 2009: 184 and 2012: 11).
The linguistic landscape of north-east England
73
The spelling of the first person singular pronoun in a’ll reflects the northeast pronunciation [ɑ], which preserves an earlier stage in the history of the pronoun, after it lost the palatal/velar element but before it was diphthongized in some varieties of English. Other respellings attempt to show the unstressed quality of certain vowels in an imagined spoken performance of the expressions in which they occur (e.g. and for to and for you), presumably to give a lively, spontaneous feel to the message and so add to its conversationalized flavour. There is nothing especially “local” about the spellings arising from this strategy; but in combination with the other features they contribute to vernacular authenticity. Only four segmental features are represented through respelling in Table 4. 2. Why – when the most recent comprehensive description of north-east English covers 14 segmental features (Beal et al. 2012) – does such a limited range appear in the corpus? It cannot be because of a lack of awareness. In previous research in which respondents were asked to describe accent differences between speakers from different parts of the region (Pearce 2009 and 2012), respellings were used to show vowels in several lexical sets which do not appear in the LL: FOOT (e.g. ); GOOSE (e.g. ); FLEECE (e.g. , ); GOAT (e.g. , ); NURSE (e.g. , ), as well as the vowel in make and take (e.g. , ). Contemporary dialect writing and informal online writing also display a wide range of respellings (see, for example the “Cheryl Kerl” parody Twitter account and “The Bacons” comic strip in Viz).8 The main reason for the relatively infrequent representation of phonetic and phonological features of north-east English in the LL is that within the standard language ideology there are powerful constraints against “mis-spelling”, especially in public and professional contexts, and perhaps dialect respellings might sometimes be perceived as merely “sloppy” or “poor” English. Respelling can also be difficult for readers to process, which is inimical to the function of signs in the public domain. The incorporation of local lexical features, which I turn to in the next section, is somewhat less problematic from this perspective, because they do not run the risk of being dismissed as errors.
3.3 Lexis, discourse features, and morphosyntax As Beal et al. point out, “the traditional regional vocabulary has been, and continues to be, a very distinctive feature of dialects in the North-East” (2012: 70), and it features heavily in the corpus. Table 4.3 shows whether or not an item appears in any of three well-known works of lexicography charting the region’s 8 Cheryl Kerl https://twitter.com/CherylKerl (Cheryl Cole is a Newcastle-born singer and celebrity); Viz is an adult comic magazine, founded in Newcastle in 1979 .
74
Michael Pearce
Table 4.3: Features of the LL corpus attested in dictionaries and corpora9
aye bairn bait bonnie/bonny canny div eeh dunch gadgy gan ganny Geordie haddaway ha’way/haway/howay hinny hoy lad lass mam man Mackem marra mint nana nay owt pet plodge Sand Dancer scran stottie Toon wass weans whey aye/why aye wor
Brockett (1829)
Heslop (1892–3)
Griffiths (2005)
TLS (1960s–1970s)
PVC (1990s)
NECTE2 (2000s)
Z Z ? Z Z Z Y Z Y Z Y Z Z Z Z Z Z * Y * Y Z Y Y ? Z Z Z Y ? ? * Y Z ? Z
* Z Z Z Z Z Z Z Y Z Y Z Z Z Z Z Z Z Y Z Y Z Y Y ? Z * Z Y Z ? Z Y Y ? Z
Z Z Z Z Z Z Y Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Y ? Z Z Z Z Z Z Z Y Z Z Z
Z Z * Z Z Z Z * Y Z Y Z Z Z Z Z Z Z Z Z Y Y Y Y Z Z Z Y Y Y ? * Y Y Z Z
Z Z Z Z Z Z Z Y Z Z Y Z Y Z Z Z Z Z Z Z Z Y Z Z Y Z Y Z Y Y Z Y Y Y Z Z
Z Z Y Z Z Z Z Y Z Z Y Z Y Z Z Z Z Z Z Z Z * Z Z Y Z * Y Z Y Y Z Y Y Z Z
9 TLS = Tyneside Linguistic Survey (late 1960s, early 1970s); PVC = Phonological Variation and Change in Contemporary Spoken English (1991–1994); NECTE2 = Newcastle Electronic Corpus of Tyneside English 2 (2007-2010). indicates that the item is either a headword in a dictionary or is used un-selfconsciously in DECTE. means that the item appears in the dictionary, but not as a headword. It also indicates the word’s presence in DECTE, but only as the topic of metapragmatic discussion. indicates that the lexeme appears, but that the meaning is different from the one in the LL corpus (for example Brockett and Heslop do not have stottie as bread product, but they do have the verb stot – “bounce”). In some cases indicates, as in the case of nay, the presence of a closely related form with a different orthography.
The linguistic landscape of north-east England
75
vocabulary over a period of nearly two centuries (Brockett, Heslop and Griffiths). Its occurrence in the three components of DECTE is also indicated. These measures give a sense of an item’s longevity and saliency. Aye, bairn, canny, div, gan, Geordie, ha’way/haway/howay, hinny, hoy, lad, lass, man, owt and wor appear in all six sources, while aye, bairn, bonny, canny, div, gan, Geordie, haway/howay, hoy, lad, lass, mam, man, owt, why aye, and wor are used in semi-natural discourse in all three eras of DECTE, suggesting they have remained a part of north-east speech, despite their long history. Indeed, like some of the pronunciations discussed in the previous section, many of these words have been enregistered as a feature of the dialect for a considerable time. Ruano-García et al. (2015) in a study of the lexis of “northern texts” from the Salamanca Corpus list gang/gan, bonny, and lass amongst their “top ten northern words” in both the Early (1500–1700) and Late Modern English (1800–1900) sub-corpora, while lad, bairn and owt occur in the top ten Late Modern English list. RuanoGarcía et al. describe these items as “lexical pan-northernisms” and stress that because the corpus texts are “representative of literary dialects” they “were not necessarily written for a regional audience”, concluding that “writers balanced the number of dialectalisms used so as not to interfere with the readers’ understanding of their literary message, and very likely selected terms that readers might have been familiar with” (2015: 145). It would seem then, that as early as the sixteenth century some of the items present in the contemporary LL were undergoing enregisterment and have therefore long held special status in the region. There is certainly some nineteenth-century evidence to show that individual lexical items had acquired this status. For example, some of Heslop’s definitions of LL items include accounts of their cultural significance in the region. This excursus is in the definition of bairn (1892: 30): The power of a homely word is in no case more exemplified than in the use of the word bairn. It is full of affectionate tenderness, and whether used in old ballad or in the folkspeech of the present day it equally breathes a spirit of yearning love for the little folk. A bit bairn or a bairnie is a little child. The pronunciation is sometimes lengthened, and a mother is heard to call “Gan up to the barin!” or “Mind the baiorin!”
Canny, bonny and hinny receive similar treatment. Evidence for the special significance of some LL items is also revealed in recent academic studies. For example, I have explored the history and current use of canny in north-east England (Pearce 2013), showing its centrality as a “cultural keyword” and its recent acquisition of new functions unique to the region. Snell (2017) focuses on howay, paying close attention to how it is deployed strategically in the discourse of children on Teesside to construct a particular kind of working-class identity.
76
Michael Pearce
Durkin exemplifies the challenges faced by historical dictionary makers in documenting regional variation in lexis with reference to pet and the ethnonyms Mackem and Geordie. He describes pet as frequently functioning “as a marker of north-eastern identity” even though it is unlikely to have originated in the region (2015: 315), while the ethnonyms are described as “key terms in defining different types of north-eastern identity” (319). My own studies of the region’s ethnonymic field include Geordie and Mackem, and cover an additional label also found in the LL: Sand Dancer (Pearce 2014). Such terms reflect a desire to maintain distinctions of local identity and affiliation in the face of external homogenizing forces and attitudes, and their appearance in the LL is a public assertion of this desire. Two articles examine the cultural significance and sociolinguistics of lad and lass. They are shown to have a complex and shifting set of referents, and also to play an important role in expressions of local identity, in the context of working-class sociability in general, but in relation to football in particular – as in the exhortation Howay/Haway the Lads, where “the lads” are the favoured team (Hermeston 2011; Beal and Burbano-Elizondo 2012). Finally, my study of attitudes towards the choice between mam and mum in north-east England as displayed on an online messageboard (Pearce 2015) offers a detailed account of “folk” discourse about a significant north-east variable, illustrating one of the mechanisms – “talk about talk” – through which enregisterment is sustained. Respellings and lexis are perhaps the most obviously localized features in the LL, but morphosyntactic features also appear in the corpus. For example, there are a number of morphological variants present, including the first-person possessive determiner wor, and us as a singular first person object pronoun. Wor occurs twice in the corpus in two contrasting contexts. GAN CANNY WOR GRADUATES was printed on a bag given away as a branded piece of merchandise during Newcastle University’s graduation ceremonies in 2015. As we have seen, the university is one of the few “elite” contexts where LL items have been recorded. In contrast, WOR HOOSE appears on a sign on a house in one of the most deprived parts of Newcastle. The building hosts a local community project offering such facilities as a “hardship service” and “washing machine service”.10 The use of wor by the university is unusual from a socio-pragmatic perspective. In DECTE, wor displays a semantic preference for animate entities (especially close relatives and friends) or things/concepts close to the everyday concerns of, or belonging to the speaker, such as eye, breakfast, coffee, dinner, sandwiches, supper, sweets, room, house, holidays. Wor graduates, therefore, is a peculiar collocation, since the “speaker” here is a public organization rather than an 10 https://localgiving.org/charity/worhoose/.
The linguistic landscape of north-east England
77
individual, and graduates is certainly not drawn from the array of nouns typically possessed by wor. In contrast, house is a much more “natural” semantic preference for wor, especially in the context of an organization run by and for the use of members of a working-class community. Nevertheless, the incongruity of WOR GRADUATES is a calculated part of the message’s rhetorical effect, as is its combination with “Gan canny”, an idiomatic expression with nineteenth-century antecedents meaning something like “take it easy” (Heslop 1892: 131). Also worth considering in relation to morphosyntax are local forms of negated do (Beal et al. 2012: 63–65). These are represented in the corpus by the signs Div’nt Dunch us or a’ll gan to PREMIER and Dinnit break me TEMPER. Both negated forms originate in a northern form of DO with a fronted and unrounded vowel – usually represented orthographically as or and pronounced [di], [dɨ], or [dɪ]. Dinnit combines [di] with a reduced form of not and is normally pronounced [dinət] or [diːnt]. Div seems to occur in Tyneside English only when it precedes a vowel-initial syllable and only very rarely with positive polarity. Indeed, there are only two instances of div without the negation clitic in DECTE. Rowe suggests that this usage is very rare, and “seems to occur either among very conservative (particularly older) speakers or among youths” (2007: 361). Div’nt/divn’t/divvent etc. [dɪvənt] is more common, although those who use it also tend to be male and/or working class (Beal et al. 2012: 64). This association was recognized by Brockett in the early nineteenth century, whose terse dictionary entry reads: “DIV, for do. Very common among the vulgar” (1829: 97). Heslop (1892: 236) is much less censorious: DINNA, DINNET, DIVENT, do not. All these words are used with the same meaning, but euphony suggests their selection. This is an example of the richness of the dialect which may well be noted.
These forms can be regarded as evidence of intra-regional variation; Beal et al. suggest that “divvent is indexed as ‘Geordie’ and dinnet as ‘Mackem’” which means that it “could be a feature that, perceptually at least, differentiates the three major urban varieties in the North-East from each other” (2012: 64–65; see also Pearce 2012: 15–16). In the case of the two examples in the LL, div’nt is associated with a business in North Tyneside, while dinnit occurs in a display in a museum in Sunderland. The final feature I will consider lies on the border of syntax and discourse: right dislocation. There is one instance of this in the LL corpus in the slogan I Love carpets me! Here, a noun phrase tag (me) is co-referent with the subject of the preceding clause (Biber et al. 1999: 957). This feature is certainly not unique to the region, though it is probably used more in the north of England than the
78
Michael Pearce
south (Moore and Snell 2011). Nevertheless, there is some evidence to suggest that it has a particular regional significance. For example, Heslop includes right dislocation – with and without the operator – in his account of the dialect in Northumberland Words (1892: xxi): Another tendency is that of placing the subject of a sentence at the end of a phrase. “He’d getten a sair tumm’le, Jack had.” “They’ve come oot o’ skyul, the bairns hez.” “Th’or myestly a’ that colour, wor coos.”
Moore and Snell (2011) provide more evidence of the feature’s importance in north-east English by showing how it is used not only to mark emphasis or provide clarification but also to perform a wide range of interpersonal functions concerned mainly with positioning speakers within the social group and forging and maintaining social bonds. It is this aspect of its use which is being exploited in the slogan of the carpet shop.
4 Conclusion The analysis has shown the extent to which enregistered lexical, morphosyntactic and orthographic items with a long history of cultural salience capable of carrying out complex social work dominate the vernacular elements in the north-east linguistic landscape. Most of them have a considerable pedigree, yet remain a part of the everyday speech of people across the region. They do not – unlike some of the Low German items in Reershemius’s study – appear in quotation marks to set them apart from the rest of the (standard) co-text, implying “as they say” or “as they said” in an idealised past (2011: 40). They are there on their own terms – deployed rhetorically, mainly in the signage of local businesses (often cheap catering outlets, second-hand shops, and small firms offering manual trades) to index working-class values of sociability, humour, solidarity and tradition, and so form a bond between the texts’ producers and receivers. I have established that in the north-east such elements are rare, probably as a consequence of the strong standard language ideology that prevails in the UK, and possibly – and perhaps paradoxically – because of the ethnolinguistic vitality of north-east English. As a variety it is not widely perceived to be under threat, so its use in the visual public domain is not felt to be a particular priority. Nevertheless, these rare items, which have – so far – been somewhat underresearched, are a fascinating and revealing component of the linguistic ecology of the region.
The linguistic landscape of north-east England
79
References Backhaus, Peter. 2007. Linguistic landscapes: A comparative study of urban multilingualism in Tokyo. Clevedon: Multilingual Matters. Beal, Joan C. 2004. English dialects in the North of England: phonology. In Bernd Kortmann & Clive Upton (eds.), Varieties of English 1: The British Isles, 122–144. Berlin & New York: Mouton de Gruyter. Beal, Joan C. 2009. Enregisterment, commodification and historical context: “Geordie” versus “Sheffieldish”. American Speech 84(2): 138–156. Beal, Joan C, Lourdes Burbano-Elizondo & Carmen Llamas. 2012. Urban North-Eastern English: Tyneside to Teesside. Edinburgh: Edinburgh University Press. Beal, Joan C. & Lourdes Burbano-Elizondo. 2012. “All the lads and lasses”: Lexical variation in Tyne and Wear. English Today 28(4): 10–22. Beal, Joan C. & Paul Cooper. 2015. The enregisterment of Northern English. In Raymond Hickey (ed.), Researching Northern English, 27–50. Amsterdam & Philadelphia: John Benjamins. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad & Edward Finegan 1999. The Longman grammar of spoken and written English. Harlow: Longman. Black, Ian. 2008. Geordies Vs Mackems and Mackems Vs Geordies: Why Tyneside is better than Wearside and why Wearside is better than Tyneside. Edinburgh: Black and White Publishing. Blommaert, Jan. 2013. Ethnography, superdiversity and linguistic landscapes: Chronicles of Complexity. Bristol, Buffalo & Toronto: Multilingual Matters. Britain, David. 2010. Contact and dialectology. In R. Hickey (ed.), The handbook of language contact, 208–229. Oxford: Blackwell. Brockett, John Trotter. 1829. A glossary of north country words in use. 2nd edn. Newcastle upon Tyne: Emerson Charnley and Baldwin and Cradock. Candlish, Alan. 2006. Ha’way/howay the lads: A history of the rivalry between Newcastle United and Sunderland. Cheltenham: Sportsbooks. Corrigan, Karen, Isabelle Buchstaller, Adam Mearns & Hermann Moisl. 2012. The Diachronic Electronic Corpus of Tyneside English. Newcastle University. Coupland, Nikolas. 2014. Sociolinguistic change, vernacularization and broadcast British media. In Jannis Androutsopoulos (ed.), Mediatization and sociolinguistic change, 67–96. Berlin & Boston: Walter de Gruyter. Coupland, Nikolas & Hywel Bishop. 2007. Ideologised values for British accents. Journal of Sociolinguistics 11(1): 74–93. Dobson, Scott. 1986. Larn yersel’ Geordie. Morpeth: Butler Publishing. Duke, Chris, Robert Hassink, James Powell & Jaana Puuka. 2006. Supporting the contribution of higher education institutions to regional development. Peer review report: North East of England. Durkin, Philip. 2015. Mackems, Geordies and ram-raiders: documenting regional variation in historical dictionaries. English Language and Linguistics. 19(2): 313–326. Fairclough, Norman. 1996. Border crossings: Discourse and social change in contemporary societies. In H. Coleman & L. Cameron (eds.), Change and language, 3–17. Clevedon: Multilingual Matters. Garrett, Peter. 2010. Attitudes to language. Cambridge: Cambridge University Press.
80
Michael Pearce
Griffiths, Bill. 2005. A dictionary of North East dialect (2nd edition). Newcastle: Northumbria University Press. Hermeston, Rodney. 2011. “The Blaydon Races”: lads and lasses, song tradition, and the evolution of an anthem. Language and Literature 20(4): 269–282. Heller, Monica. 2005. Language and identity. In U. Ammon, N. Dittmar, K. J. Mattheier & P. Trudgill (eds.), Sociolinguistics/Soziolinguistik: An international handbook of the science of language and society/Ein internationals Handbuch zur Wissenschaft von Sprache und Gesellschaft Vol. 2, 1582–1586. (2nd edition) Berlin: Mouton de Gruyter. Heslop, Richard. 1892. Northumberland words: A glossary of words used in the county of Northumberland and on the Tyneside, Vol. 1. Oxford: Oxford University Press. Heslop, Richard. 1893. Northumberland words: A glossary of words used in the county of Northumberland and on the Tyneside, Vol. 2. Oxford: Oxford University Press. Honeybone, Patrick & Kevin Watson. 2013. Salience and the sociolinguistics of Scouse spelling: Exploring the phonology of the Contemporary Humorous Localised Dialect Literature of Liverpool. English World-Wide 34(3): 305–340. Johnstone, Barbara. 2009. Pittsburghese shirts: Commodification and the enregisterment of an urban dialect. American Speech 84(2): 157–175. Johnstone, Barbara. 2010. Indexing the local. In Nikolas Coupland (ed.), The handbook of language and globalization, 386–405. Oxford: Blackwell. Johnstone, Barbara. 2011. Language and place. In Rajend Mesthrie (ed.), The Cambridge handbook of sociolinguistics, 203–217. Cambridge: Cambridge University Press. Milroy, James. 2007. The ideology of standard language. In Carmen Llamas, Louise Mullany & Peter Stockwell (eds.), The Routledge companion to sociolinguistics, 133–139. London: Routledge. Montgomery, Chris. 2012. Mapping the perceptions of non-linguists in Northern England. In Sandra Hansen, Christian Schwarz, Philipp Stoeckle & Tobias Streck (eds.), Dialectological and folk dialectological concepts of space, 164–178. Berlin & Boston: Walter de Gruyter. Moore, Emma and Julia Snell. 2011. “Oh, they’re top, them”: right dislocated tags and interactional stance. In Frans Gregersen, Jeffrey Parrott & Pia Quist (eds.), Language variation: European perspectives III, 97–110. Amsterdam: John Benjamins. Palgrave, F.M.T. 1894–5. Words and phrases in every-day use by the natives of Hetton-Le-Hole in the county of Durham. London: Henry Frowde. Papen, Uta. 2012. Commercial discourses, gentrification and citizens’ protest: The linguistic landscape of Prenzlauer Berg, Berlin. Journal of Sociolinguistics. 16(1): 56–80. Pearce, Michael. 2009. A perceptual dialect map of North East England. Journal of English Linguistics 37(2): 162–192. Pearce, Michael. 2011a. Exploring a perceptual dialect boundary in North East England. Dialectologica et Geolinguistica 19: 3–22. Pearce, Michael. 2011a. “It isn’t geet good, like, but it’s canny”: A new(ish) dialect feature in North East England. English Today 27(3): 1–7. Pearce, Michael. 2012. Folk accounts of dialect differences in Tyne and Wear. Dialectologica et Geolinguistica 20: 5–25. Pearce, Michael. 2013. “That word so fraught with meaning”: The history, cultural significance and current use of canny in North East England. English Studies 94(5): 562–581. Pearce, Michael. 2014. “Not quite a Geordie”: the folk-ethnonyms of north-east England. Nomina 37: 1–34.
The linguistic landscape of north-east England
81
Pearce, Michael. 2015. Mam or mum? Sociolinguistic awareness and language-ideological debates online. Sociolinguistic Studies 9(1): 115–135. Reershemius, Gertrud. 2009. Post-vernacular language use in a Low German linguistic community. Journal of Germanic Linguistics 21(2): 131–147. Reershemius, Gertrud. 2011. Reconstructing the past? Low German and the creating of regional identity in public language display. Journal of Multilingual and Multicultural Development 32(1): 33–54. Rowe, Charley. 2007. He divn’t gan tiv a college ti di that, man! A study of do (and to) in Tyneside English. Language Sciences 29: 360–371. Ruano-García, Javier, Pilar Sánchez-García & Maria F. García-Bermejo Giner. 2015. Northern English: Historical lexis and spelling. In Raymond Hickey (ed.), Researching Northern English, 131–157. Amsterdam & Philadelphia: John Benjamins. Shohamy, Elana and Eliezer Ben Rafael. 2015. Linguistic landscape. Linguistic Landscape: An International Journal. 1 (1/2): 1–5. Shorrocks, Graham. 1996. Non-standard dialect literature and popular culture. In Juhani Klemola, Merja Kytö & Matti Rissanen (eds.), Speech past and present: Studies in English dialectology in memory of Ossi Ihalainenm 385–411. Frankfurt am Main: Peter Lang. Snell, Julia. (2017). Enregisterment, indexicality and the social meaning of ‘howay’: dialect and identity in north-east England. In Emma Moore and Chris Montgomery (eds.), Language and a sense of place 301–324. Cambridge: Cambridge University Press. Tomaney, John. 2010. Contemporary Britain and its regions. In Michael Higgins, Clarissa Smith & John Storey (eds.), The Cambridge companion to modern British culture, 79–95. Cambridge: Cambridge University Press. Waddell, Sid. 2008. Taak of the Toon: How to speak Geordie. Glasgow: Collins. Wells, John. 1982. Accents of English 2: The British Isles. Cambridge: Cambridge University Press. Wright, Joseph. 1898. The English dialect dictionary. Vol. 1. Oxford: Henry Frowde. Wright, Joseph. 1905. The English dialect dictionary. Vol. 4. Oxford: Henry Frowde.
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
5 Lenition and T-to-R are differently salient: the representation of competing realisations of /t/ in Liverpool English dialect literature Abstract: This article investigates variation in the way that phonological dialect features are represented orthographically in a corpus of dialect literature from Liverpool English. The texts considered are examples of Contemporary Humorous Localised Dialect Literature (CHLDL). We compare what is found in these texts with the extent to which phonological dialect features are represented in corpora of spoken Liverpool English, and we show that dialect literature can subtly represent the different degrees of salience that dialect features have. We focus on two phonological features which are well established in spoken corpora: ‘Liverpool Lenition’ (in which stops, including /t/ are affricated and spirantised) and T-to-R. We show that, although both are very common in spoken Liverpool English, and both could in principle be represented orthographically, only T-to-R is robustly represented in our corpus of dialect literature. We go on to show that this makes sense: phonological theory predicts that processes with certain types of characteristics should be salient and others should not, and we show that T-to-R has the characteristics that fit with being salient, while T-lenition does not.
1 Introduction The orthographic variation found in contemporary dialect literature is sometimes dismissed as ad hoc and unenlightening. Grant (2007: 157), for example, sees little value in “‘comical’ pronunciations or eye-dialect renditions of pronunciations of otherwise perfectly ordinary Standard English words, such as dem for ‘them’”. In this chapter we develop the argument (first set out in Honeybone & Watson 2013) that precisely the opposite can be true: in the corpus of dialect literature that we consider (which represents the same variety that Grant discusses), there
Patrick Honeybone, University of Edinburgh Kevin Watson and Sarah van Eyndhoven, University of Canterbury DOI 10.1515/9783110450903-005
84
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
are subtle patterns in the ways that dialect features are represented, which accurately reflect the different degrees of salience that we would expect of them. The notion of ‘salience’ will thus be important in what follows. Salience has long been discussed in linguistics (see Kerswill & Williams 2002 and Racz 2013 for overviews); at heart, it simply means how ‘noticeable’ a particular linguistic feature is to speakers. In sociolinguistics, salience is often operationalised in terms of Labov’s (1972) continuum of indicators (least salient) > markers > stereotypes (most salient), with the steps on the cline typically differentiated by speakers’ social evaluation of and commentary on particular linguistic features, and this ties in with a strand of argumentation in phonological theory which connects different degrees of salience with different derivational levels (as in, for example, Kiparsky 1982, 2015). Many characteristics have been said to contribute to the salience of a particular feature. Trudgill (1986: 11) for example, argues that a feature is salient if it is overtly stigmatised, in line with Labov’s (1972) definition of features which act as sociolinguistic ‘stereotypes’, and that salience can be affected by the contrastive status of the feature in question, in line with a phonological approach. In Honeybone & Watson (2013), we expand on this by arguing that salience is not only conditioned by lexical contrast, but that other phonological properties can also contribute to the salience of a linguistic feature. We elaborate on this idea throughout this piece, referring to it as the phonological salience of a linguistic feature. On the empirical level, we focus here on the ways in which two phonological features of Liverpool English are represented orthographically in contemporary dialect literature: the lenition of /t/ and what Wells (1982) calls ‘T-to-R’. These seem structurally similar at first glance but, as we will see, are spelled with very different frequencies, indicating that they have very different degrees of salience. This, we will argue, is predictable once we understand their phonological status. Section 2 of this piece describes the genre of dialect literature that we consider; section 3 describes the variety of English that provides our data and the corpus of dialect literature that we investigate; section 4 sets out our general framework for investigating dialect literature; section 5 describes the two phonological features that we focus on; section 6 sets out our new results and our explanation for them; and section 7 concludes.
2 Contemporary Humorous Localised Dialect Literature Although it can sometimes seem hidden from view, there is a substantial amount of published writing involving non-standard forms of English. There are a number of different genres of such material, and some have long traditions. For
Lenition and T-to-R are differently salient
85
example, ‘traditional’ dialect literature, often poetry, in which authors have sought to represent a non-standard dialect against a context of codified Standard English, stretches back to the 18th century. Such material has been assessed in a number of places for its potential as linguistic evidence for the dialects that it represents and their history (see, for example, Shorrocks 1996, Wales 2006, and the chapters gathered in Taavitsainen, Melchers & Pahta 1999 and in Hickey 2010). In this chapter, we consider a genre that we identified in Honeybone & Watson (2013) as ‘Contemporary Humorous Localised Dialect Literature’ (CHLDL). CHLDL is contemporary in that it is current, being published continuously since the 1960s; it is humorous in that it is intended to amuse; it is localised in that it is published by regional publishers and is often only available in the area where the dialect is spoken; and it is true ‘dialect literature’ (rather than ‘literary dialect’, cf. Shorrocks 1996) in that it is written by, and is meant in large part for, an audience who speaks the dialect that it represents. CHLDL texts are normally well-received, and are kept in print. They typically have the form of small or thin booklets, and often masquerade as ‘phrase books’ or ‘dictionaries’, which means that they involve a direct comparison of (extreme and constructed) dialect sentences with absurdly formal Standard English ‘translations’. Some of this can be seen in the title of the series of texts that we consider: Lern Yerself Scouse (which could be translated as ‘Teach Yourself Liverpool English’ – the full details of our corpus are given in section 3). CHLDL exists for many varieties of English, and texts of this type have been considered elsewhere in linguistic work (e.g. Schneider 1986, Beal 2000, 2009, Johnstone 2009, Bennett 2012, Jensen 2013), but were never investigated in a fully quantitative way before Honeybone & Watson (2013). While CHLDL can also give evidence about non-standard lexis and morphosyntax, our interest is in the extent to which CHLDL authors use respellings of General English words to represent the phonology of a non-standard variety and the ways in which it differs from the phonology of a standard/reference variety. We call the relevant standard/reference variety ‘General British’ (GB), following Cruttenden (2014). Such respellings rely on the fact that readers know the Standard English spellings of words, so that any divergence from this will be recognised and will mark out a form as ‘dialect’. It need not matter precisely how a word is respelled1 – if a feature has a non-standard spelling it will draw 1 The fact that respellings do not need to be phonologically transparent in order to mark out a particular phonological feature is not unusual in spelling – as Lass (2015: 107) points out: “[p]retty much any kind of ‘defective’ spelling will do, as long as the reader can be assumed to have some idea in advance of what a word is likely to be”.
86
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
attention to the fact that the form is pronounced differently in the dialect concerned. A reader who speaks the dialect represented will know this already, of course, and is just hoping to have the CHLDL text point this out. Nonetheless, such respellings are not random. Rather, they rely on speakers’ knowledge of the phon-to-graph2 correspondences that exist for GB, as these can be reemployed to represent the phonological features of a dialect which differ from those of GB (this thus also relies on readers having a fair knowledge of the phonology of GB). For example: typically spells [uː] in GB, as in tune, rude, fluke, so this grapheme can be used in dialect literature to represent [uː] (or the equivalent/corresponding vowel in the relevant dialect) in words in which it does not occur in GB, as in the retention of the long vowel in many northern English dialects in words like cuke ‘cook’ typically indicates that something is absent in English orthography (as in contractions like don’t, I’m, she’d), so this grapheme can be used to represent the absence of /h/, when a non-standard dialect is compared to GB, as in ’at ‘hat’ and typically spell an alveolar fortis stop in GB (as in [tʰ] or [t] in top, bat, matter), so these graphemes can be used to represent [t ]̪ in TH-stopping, as in nuttin’ ‘nothing’ – this works even though the TH-stopped form is dental (not alveolar, as in top, etc) because graphemes can be used to represent more than one sound – as, for example, in the GB cases of
representing both /θ/ and /ð/, representing both the tense/long/free vowel (henceforth simply ‘long’) in cart and the lax/short/checked vowel (henceforth ͡ simply ‘short’) in cat, and representing both /ɡ/ and /dʒ/
In sections 4 and 6, below, we investigate such respellings on the basis of a corpus of CHLDL for one dialect of English. The next section explains what that corpus and dialect are. As we will see, the above holds fundamentally true, but needs to be modified in the light of the differential phonological salience of particular dialect features.
3 Liverpool English and Liverpool CHLDL The variety of English that is represented in the texts that we consider is Liverpool English (LE). The dialect is known popularly as ‘Scouse’, and its sound system has been the subject of a number of studies, starting with Knowles (1973), and including Newbrook (1986), Watson (2007a), Clark & Watson (2016) and Watson & Clark (to forthcoming). Most of the features that distinguish the variety are phonological, and many of these have their roots in a process of new-dialect 2 “phon” here is intended to be ambiguous in terms of the phonetic/phonological distinction.
Lenition and T-to-R are differently salient
87
formation that occurred in Liverpool in the 19th century, due to vast in-migration which was driven by Liverpool’s status during that time as the site of Britain’s most important docks. The migrants came from nearby areas of northern England, further afield in England, Scotland and Wales, and in large numbers from Ireland, which is just across the Irish Sea from Liverpool and was then tightly connected to the city by sea links (see Honeybone 2007 and Cardoso 2015 for details of the formation of LE, and also the slightly different take in Crowley 2012). LE is tightly connected to the city of Liverpool and the nearby area, and, while it has many features which contextualise it as a ‘northern English’, it is quite distinct from neighbouring varieties (e.g. Lancashire English and Cheshire English). We discuss some of the phonological features that set it apart from all or most other varieties of English in the following sections. LE is a well-recognised dialect in Britain, always featuring near the top of lists of ‘accents in Britain’, but it is typically rated low in subjective ‘aesthetic’ rankings of British varieties, no doubt due to its urban status and stereotypes connected with the city of Liverpool (see such work as Montgomery 2007, Coupland & Bishop 2007 and Leach, Watson & Gnevsheva 2016). The flipside of this is that speakers closely identify with the dialect, and see it as a central part of Liverpool identity (see, for example, Liverpool Echo 2008). LE was in at the start of the contemporary wave of CHLDL texts for British dialects, which began in the 1960s, and there is a set of CHLDL volumes which are well-known in the city: the Lern Yerself Scouse books. This is a series of books which together form our corpus. They follow the common CHLDL ploy of pretending to be a phrase book for humorous purposes – for example, the phrase I’m werkin fer de Queen is ‘translated’ as ‘I am drawing unemployment benefit’ (it could be glossed as ‘I’m working for the Queen’). We do not consider the lexis, humour or attitudes portrayed in the volumes. Our focus is on the phonological knowledge about distinctly LE-related forms that is shown through the respellings used in the texts. For example, in the sentence just cited, the words werkin and de accurately represent aspects of LE vocalic and consonantal phonology respectively (as we discuss below); at the same time, the elision of in werkin and the spelling of for as represent common English forms which are perfectly accurate but which are not distinctively tied to Liverpool English (in British varieties of English it is unexceptional for unstressed -ing to be realised as [-ɪn] and for for to be reduced to [fə], and given that often spells a schwa in non-rhotic varieties, as in matter, winner, ladder, it is sensible to use that sequence to spell the reduced form of for). The full details of the volumes that form our corpus are as follows: – Shaw, Frank, Stan Kelly & Fritz Spiegl (1965/2000) Lern Yerself Scouse: How to talk proper in Liverpool. (Vol 1). Liverpool: Scouse Press. – Lane, Linacre (1966/2000) The ABZ of Scouse. (Lern Yerself Scouse: How to talk proper in Liverpool. Vol 2). Liverpool: Scouse Press. (ed. Spiegl, Fritz)
88 – – –
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
Minard, Brian (1972/2000) Wersia Sensa Yuma? (Lern Yerself Scouse: How to talk proper in Liverpool. Vol 3). Liverpool: Scouse Press. Spiegl, Fritz (ed) (1989) The Language of Laura Norder. (Lern Yerself Scouse: How to talk proper in Liverpool. Vol 4). Liverpool: Scouse Press. Spiegl, Fritz (ed) (2000) Scouse International: The Liverpool Dialect in Five Languages. Liverpool: Scouse Press.
In sections 4 and 6, we present results of a quantitative analysis of the dialectal respellings of a number of LE dialect phonological features that are found in our CHLDL corpus. We compare this with descriptions of the same features on the basis of corpora of LE speech, making reference to two sets of data: the material elicited from 16 adolescent speakers via elicitation tasks for Watson (2007b) and the material collected for OLIVE – the Origins of Liverpool English corpus3 (see Watson & Clark to forthcoming) – which holds spontaneous conversation and reading data representing over 100 years in apparent time. OLIVE consists of subcorpora with three age cohorts: younger speakers, born between 1992 and 1994, older speakers, born between 1918 and 1942, and archive speakers, born between 1890 and 1943; we report below on some data from the archive speakers. In order to produce the numbers for the quantitative analysis of Liverpool English CHLDL, the books were digitised, the Standard English ‘translations’ removed, and both standard and non-standard spellings in the LE text were manually annotated with two sets of tags. The first set provided an identifying label for words which are relevant to the particular dialect features in question and the second categorised the word as being spelled standardly or non-standardly – every potential occurrence of a non-standard orthographic form was thus counted (in compliance with Labov’s 1972 ‘principle of accountability’), to give a percentage figure for how frequently a particular dialect feature is respelled in the CHLDL corpus.
4 Representing dialect features in CHLDL and their phonological salience In Honeybone & Watson (2013), we report on the extent to which eleven LE phonological dialect features are represented in the CHLDL corpus mentioned 3 OLIVE was created thanks to the financial support of the Economic and Social Research Council, as part of a project entitled ‘Phonological levelling, diffusion and divergence in Liverpool and its hinterland’ (RES-061-25-0458). We also gratefully acknowledge support from the University of Canterbury’s summer scholarship scheme. Thanks are also due to the North West Sound Archive for donating the recordings for OLIVE’s Archive subcorpus.
Lenition and T-to-R are differently salient
89
above, showing that these features are accurately represented in this DL corpus in fundamental ways. One central finding is that groups of the eleven features are represented to different extents, and that this differential representation makes linguistic sense. In this section, we summarise some of the findings of Honeybone & Watson (2013), to provide the context for our introduction of new results and analysis in section 6. We consider here only three of the features discussed in Honeybone & Watson (2013), in order to keep our discussion focused. We refer to them as (d), (th), and NURSE /SQUARE , adopting the round bracket convention which is often used to represent sociolinguistic variables and the key words (in small capitals) proposed in Wells (1982) to describe English vowel distributions. The feature (d) refers to the fact that underlying stops in LE are subject to what is often called ‘Liverpool Lenition’. That is, there is a synchronic (variable) process through which the stops can be realised as affricates or fricatives in certain phonological environments – this involves the stops heading down a lenition trajectory (of the type discussed in many places including Lass 1984 – see Honeybone 2008 for a summary) for either one or two steps. Liverpool Lenition has been investigated in quite some work (such as Honeybone 2001, Sangster 2001, Watson 2007b). The full environmental patterning is complex, but some broad generalisations are clear: fricatives are common in final and (foot-/word-) medial positions (and affricates are also possible in these positions), while affricates but not fricatives are common in initial positions. Lenition is possible in stops at all places of articulation, but Watson (2007b) shows that it is most common for /t/, /d/ and /k/ (such that light can be realised as [laɪθ], lad as [lað], and lock as [lɒx]). The fricative results of the lenition of /t/ and /d/ involve a wide range of realisational possibilities (Watson 2007a, 2007b) – the phenomenon is phonetically gradient – what we transcribe here are common realisations: alveolar fricatives with a flat cross-sectional tongue shape (Pandeli et al. 1997). Figure 5.1 shows that this lenition, while variable, is common. It shows the realisations of utterance-final /d/ in the speech of 16 adolescents from Vauxhall (a working-class area of Liverpool), who were recorded for Watson (2007b). Only 14% of underlying stops in final position are realised as stops in the female group – the others are all lenited in some way, with over 50% realised as fricatives; only 29% of such stops are realised as stops in the male group, with 45% realised as fricatives.
90
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
Figure 5.1: Realisation of utterance-final /d/ (in environment V_##) in 16 adolescent speakers, adapted from Watson (2007b). N values are token counts.
Is this variability in (d) – that is, the lenition of /d/ – represented in the spelling of LE words in CHLDL? It can be. Examples include: Folly me leedzer (= ‘follow my leader’) where the medial /d/ in leader is written as an affricate, and laz ‘lad’, where the final /d/ is written as a fricative. The letter used in these spellings is , which could imply that the lenition involves neutralisation with /z/. This is not the case, however, as the product of the lenition of /d/ is not normally a canonical grooved fricative (Sangster 2001 is explicit about the lack of neutralisation). Nonetheless, is available in English spelling to represent a lenis alveolar fricative, and it is not unusual or problematic to use one grapheme to represent more than one phonological segment, as discussed above (in any case, is also unambiguously available to represent the outcome of lenition). We consider the extent to which the lenition of /d/ is represented in CHLDL once we have introduced the other two LE features that we compare here in this regard. The feature (th) refers to the fact that LE traditionally has ‘TH-stopping’. That is, words which have /θ/ in most varieties of English can be realised with [t ]̪ in LE (the same is true of the lenis congener, but we focus on the fortis segment here). This is not a phonological process as it can affect all forms which have [θ] in other varieties – in all phonological environments – it is a contextfree difference in terms of the realisation of a segment between LE and the reference variety. TH-stopping does not involve the neutralisation of a contrast which is available in other varieties: the stop is dental, so thin [t ̪ɪn] still contrasts with tin [tɪn], for example. TH-stopping is relatively robust in LE, but stop realisations are minority forms in all quantitative reports. Figure 5.2 shows the number of occurrences of stops in fortis ‘TH-words’ in the speech of eight speakers from OLIVE’s Archive subcorpus. Stops occur around 30% of the time.
Lenition and T-to-R are differently salient
91
Figure 5.2: Realisation of (th) in four male and four female speakers from OLIVE’s Archive subcorpus. N values are token counts.
The occurrence of stops for (th) is straightforwardly respellable, through the use of the Standard English conventions for the representation of /t/ (despite the fact that this collapses a contrast, as discussed above, because one grapheme can readily represent more than one phoneme). One example of the spelling of TH-stopping from our corpus is: T’ingy ‘Anything or anyone whose name escapes the speaker’ (= ‘Thingy’), which in fact combines and , the latter implying that the of Standard English spelling is omitted but also creates a grapheme which manages to preserve the contrast in spelling. The use of de in the above-mentioned example I’m werkin fer de Queen (and indeed the use of dem for ‘them’ mentioned at the very start of this piece) are also spellings intended to indicate dental fricative stopping, but of the lenis equivalent (that is, of (dh), not the (th) which is in focus here – (dh) and (th) pattern in the same way). The NURSE /SQUARE feature refers to the fact that LE lacks a vowel contrast that virtually all other English varieties of English have: LE speakers typically use the same vowel in both the NURSE and SQUARE lexical sets, so that, for example, fur and fair are homophones. Because LE, like virtually all English varieties of English, is non-rhotic, (which means that there are no rhotics in rhymes at the surface), it is only the vowels which are relevant to discussion here (we set aside analyses which assume an underlying coda /r/ in non-rhotic dialects in order to keep our discussion focused). In GB and related varieties, the NURSE vowel is central [ɜː], and the SQUARE vowel is fundamentally front, either as a long monophthong [ɛː] or a front-starting centring diphthong [ɛə]. In LE, the same phonetic range (from front to central vowels) can be encountered, but
92
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
the distribution of these vowel types is not determined lexically, so there is no contrast between them. Some speakers tend to use front vowels in the relevant words and some speakers tend to use central vowels (it is also clear that some use both, and also likely that there has been change over recent time in terms of the commonness of central vs front vowels). Figure 5.3 shows vowel plots of all monophthongs for eight speakers from OLIVE’s Archive subcorpus. In this group, the four females have front vowels in NURSE and SQUARE and the four males have central vowels, but both groups show an absence of contrast, with the two vowels clustering together.
Figure 5.3: Vowel plots from 4 female speakers (F, left pane) and 4 male speakers (M, right pane) from OLIVE’s Archive subcorpus
The absence of contrast in the NURSE and SQUARE vowels can be straightforwardly represented in DL due to the fact that a number of graphemes are used to spell them in GB: for example, in person, where, in turn, in bird, in hair, in care. If any of these graphemes are used to spell a word which uses a different grapheme in its Standard English form, attention is drawn to the vowel, which allows recognition of the fact that the absence of contrast between NURSE and SQUARE is a characteristic of LE which distinguishes it from GB (and almost all other dialects). This is shown in the use of werk in the above-mentioned example I’m werkin fer de Queen, and also in such spellings as gerl ‘girl’, furs ‘fares’, tirn ‘turn’ and shairt ‘shirt’, which are all found in our corpus.
Lenition and T-to-R are differently salient
93
4.1 To what extent are different features represented in CHLDL? It is crucial for our argument here that the three LE dialect features just discussed are respelled to different degrees in CHLDL. Honeybone & Watson (2013 – henceforth H&W) present the results summarised in Figure 5.4. This shows that (d) is respelled very infrequently (only 6% of the time), whereas (th) is respelled considerably more (just over 30% of the time), and NURSE /SQUARE is respelled quite commonly (60% of the time). The figures for (d) combine the occurrence of /d/ in all environments (initial, medial and final), and while lenition is least common word-initially, it is still possible there, and even if we only consider word-final /d/, the figures are essentially the same: 7% of words with word-final occurrences of /d/ are respelled.
Figure 5.4: CHLDL respellings: (d), (th) and NURSE /SQUARE . Adapted from Honeybone & Watson (2013). N values are token counts.
Why are the three features represented to such different extents? It cannot be because authors are not able to represent them orthographically – our discussion above shows that all three are quite straightforwardly representable. It also does not seem to be related to the frequency of particular variants in speech – this may seem to be the case for (th), where approximately 30% of tokens of ‘TH-words’ in speech have stops (see Figure 5.2) and approximately 30% of tokens of ‘TH-words’ in CHLDL are respelled (Figure 5.4), but this is presumably a coincidence, because there does not seem to be a correlation for NURSE /SQUARE , as there is no evidence that the contrast is variably present, but relevant words are respelled only 60% of the time (Figure 5.4), and, furthermore,
94
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
lenition in final position in (d) is very common, at up to 50% fricatives and over 80% for any kind of lenition in utterance-final position (Figure 5.1), but words with final /d/ are very rarely respelled (Figure 5.4 and associated numbers). H&W present two main explanations for the differing degrees to which dialect features are represented in CHLDL. One is that the localisedness of a feature can influence the extent to which it is represented in DL. For example, the absence of contrast in LE in NURSE /SQUARE when compared to GB is phonologically the same as the absence of contrast in FOOT /STRUT when compared to GB; however, FOOT /STRUT is much less commonly represented in our corpus of CHLDL (H&W show that FOOT /STRUT words are respelled just 11% of the time, compared to the 60% representation rate for NURSE /SQUARE ). We interpret this difference as due to the fact that the NURSE /SQUARE feature is much more localised to Liverpool than is FOOT /STRUT (so it is more important to represent it in these Liverpool-focused texts). The FOOT /STRUT difference between LE and GB is one which is shared with all northern dialects – it is not a characteristic feature of LE, so is not that salient; on the other hand, the NURSE /SQUARE difference between LE and GB is shared with only one other variety that LE speakers are likely to be aware of 4 – Lancashire English – all other relevant dialects have a contrast, like GB; this means that the NURSE /SQUARE feature is more localised to Liverpool, and is more salient. This explanation (which accounts for other things in H&W, too) does not help with the three features discussed above as they all have approximately similar localisedness: lenition of /d/ is not really found in any other dialect, TH-stopping is only found in one other locally-relevant variety (Irish English), and, as noted, NURSE /SQUARE is also only found in one other locally-relevant variety (Lancashire English). This does not predict the features’ different degrees of representation in CHLDL – on this basis, (d) should be most commonly respelled. The other explanatory factor that we propose in H&W, as introduced above and further discussed below, is that different degrees of phonological salience between dialect features can account for differences in the extent to which these features are represented in DL. This builds on ideas in work such as Trudgill (1986), McMahon (1991, 1994), Kerswill & Williams (2002), and much else, that speakers are aware of phonological phenomena to different degrees. The fundamental difference that we hang this on is the distinction between ‘early phonology’ and ‘late phonology’, which builds on the feed-forward metaphor in derivational phonology. This model assumes that underlying representations are the earliest 4 It may also be found in Middlesbrough English, where NURSE can have [ɛː] (Beal, BurbanoElizondo & Llamas 2012) and in Hull English, but these varieties are not well known outside of the North-east of England and it is likely that LE speakers are not aware of this.
Lenition and T-to-R are differently salient
95
stage in a derivation, and that underlying representations in part encode contrasts, which are phonological differences that are easily noticeable; and that phonological processes can intervene in a derivation in some kind of order, concluding in ‘late’ processes, which derive things that are typically not noticed by speakers. This makes sense of the different degrees to which the three LE features are represented in our corpus of CHLDL: – NURSE /SQUARE involves a systemic difference of contrast between dialects, at the earliest stage of phonology, and so is salient to speakers and should be a good candidate for representation in DL – (d) involves differences in the realisation of a segment in certain specific, context-dependent phonological environments – a low-level, exceptionless phonological process – and so is not salient to speakers and hence is not a good candidate for representation in DL – (th) involves a context-free difference of realisation between dialects – this does not involve a difference of contrast, but will always involve a difference between varieties, so may be expected to be placed between systemic differences and process-related differences in terms of its salience It is clear that the two criteria (localisedness and phonological salience) could in principle interact, but it is also clear that phonological salience trumps localisedness – only if a phonological dialect feature is phonologically salient can it be a candidate for orthographic representation, and only then can localisedness play a role in determining the extent to which features are represented. In the rest of this chapter, we consider a dialect feature which was not fully considered in H&W, and compare it with a further case that we did consider in H&W. The two features are ‘T-to-R’ and the lenition of /t/ (which is closely related to the just-discussed lenition of /d/).
5 Doing different things to a /t/ in LE – T-to-R and T-lenition Like other dialects of English from the north of England, LE features what Wells (1982) calls ‘T-to-R’. This is a process in which underlying /t/, in certain specific environments, is realised in the same way as a dialect’s underlying rhotic (which we transcribe as /r/). In many varieties this is [ɹ], and such approximant realisations do occur in LE, but LE also has [ɾ] (a tap) as a common realisation of /r/, so T-to-R can involve /t/ being realised as [ɹ] or [ɾ] in LE. Wells (1982) describes T-to-R as typically applying cross-lexically, with a word-final /t/ in an intervocalic
96
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
environment (with a vowel-initial word providing the following vowel), in contexts like get off and shut up. Subsequent investigations have confirmed this: Clark & Watson (2011) find only 1.5% of occurrences of T-to-R in a corpus of LE were in word-medial context, and Buchstaller et al. (2013) find wordmedial T-to-R massively dispreferred in comparison to word-final application in an investigation into the intuitions of speakers (this considered T-to-R in a different dialect, but all T-to-R dialects seem to pattern alike in this regard). In what follows we therefore consider T-to-R only with reference to the word-final environment, where it is robust. In our present context an obvious question arises: to what extent, if at all, is T-to-R represented in the respellings found in CHLDL? We investigate this in section 6, after a more detailed consideration of the nature of T-to-R, and of a phenomenon which it competes with in LE in terms of the realisation of /t/. LE also features another process which competes with T-to-R as a way of realising word-final /t/. This is related to the phenomenon discussed as (d), above – as mentioned there, Liverpool Lenition also affects /t/, meaning that if T-to-R does not apply to derive a rhotic, /t/ can be realised as an affricate or fricative. As with /d/, lenition of /t/ (henceforth T-lenition) is very common in speech. Watson (2007b) finds that less than 10% of occurrences of utterancefinal /t/ are realised as a stop, as shown in Figure 5.5. Affricates and fricatives are both as common in the female group, and fricatives are preferred in the male group (T-to-R cannot occur in utterance-final position, of course, so Figure 5.5 shows how common T-lenition can be expected to be in environments where T-to-R cannot occur).
Figure 5.5: Realisation of utterance-final /t/ (in environment V_##) in 16 adolescent speakers. Adapted from Watson (2007b). N values are token counts.
Lenition and T-to-R are differently salient
97
In the environment in which it can occur, T-to-R is also common. In order to see this we need to recognise a fundamental characteristic of T-to-R: it is lexicallyspecific. This means that there are only certain words in which it can occur – other words which have fundamentally the same phonological shape as those which allow T-to-R do not permit it. Caffrey (2011) employed the methodology developed for Buchstaller et al. (2013) to consider some aspects of T-to-R in LE. This methodology probes the intuitions of speakers using a questionnaire, asking speakers indirectly whether it is possible to pronounce specific T-final words with a rhotic. All 24 informants agreed that it would be normal for them to pronounce the /t/ in the word not as a rhotic if it occurs in the phrase ‘oh no – not again’, and all informants also agreed that they would never pronounce the /t/ in the word knot as a rhotic if it occurs in the phrase ‘oh no – he’s tied it in a knot again’. This shows that it cannot be any aspect of the phonology of the word which determines whether it can undergo T-to-R, as not and knot are homophones – it must be marked in some way in the lexical entry of the word. The full details of the methodology employed to get such results are given in Buchstaller et al. (2013), and more details of the results of Caffrey (2011), along with a phonological analysis, are given in Honeybone (forthcoming), but the basic point is clear: we can distinguish between ‘T-to-R words’ and ‘non-T-to-Rwords’. Clark & Watson (2011) considered a considerable spoken corpus of LE and found rhotics in only twelve T-final words: let, bit, not, put, at, that, what, lot, get, got, it, but. These words crop up again and again in descriptions of T-to-R (such as Broadbent 2008 and Asprey 2008), so it is clear that all dialects with T-to-R share a core set of T-to-R words, although different dialects may have slightly different inventories of words which undergo the process: Buchstaller et al. (2013) find that fit is also a T-to-R word in Newcastle upon Tyne, for example. All of these words comply with Wells’ (1982) other basic claim about the process: that the vowel before the /t/ needs to be short. The results of a search of the OLIVE Archive sub-corpus for the purposes of this piece are given in Figure 5.6, which shows that T-to-R occurs (in a V__#V environment) to different extents in different words (nearly 70% of the time in got, for example, and nearly 20% of the time in it), but except for a few of these words (discussed below), it is robust (even though it only occurs 18% of the time in it, for example, that still involves 18 occurrences).
98
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
Figure 5.6: Frequency of occurrence of T-to-R (in environment: V__#V) in four male and four female speakers from OLIVE’s Archive subcorpus. N values are token counts.
This shows the same twelve words just mentioned, but also a few occurrences in four other words: out, about, sort, start. These latter words may require further consideration as they all have long diphthongs or monophthongs lexically, which is surprising given most previous descriptions of T-to-R. The words out and about both occur with a rhotic four times in the corpus, so it is unlikely that this is a misinterpretation of the data; about is also reported to be a T-to-R word in Black Country English in Asprey (2008), and was tested by Caffrey (2011) for LE, who found that all informants agree that T-to-R is possible in the word. In start, a rhotic-like realisation is only attested once, in an utterance which is not completely clear, so it may be that a rhotic was not intended and thus, in fact, start is not a T-to-R word; sort is analysed as having a rhotic three times in the OLIVE materials, so it cannot be dismissed so easily. A full consideration of this is beyond the scope of this chapter, however. As it is at least possible that all these words have T-to-R in the OLIVE archive subcorpus, which contains recordings from speakers who were adults around the time of the first publication of the DL considered here, we take these 16 words as the set of ‘T-to-R’ words for LE. Although it is lexically-specific, T-to-R is a productive process. Buchstaller et al. (2013) show that speakers find it just as acceptable in an infrequent collocation (likely never encountered before), such as ‘get Ethel’ as they do in a
Lenition and T-to-R are differently salient
99
normal collocation such as ‘get about’, indicating that it is productive with those words that allow it. Both T-to-R and T-lenition are thus phonological processes found in LE, and both count as dialect features, as neither occur in GB. If we focus on the word-final environment in which both can occur (in order to compare them), then T-to-R can be understood as a generalisation of the sort given in (1), which assumes that the derived rhotic is realised in the same way as the underlying rhotic (due to a separate generalisation), and note that the rule only applies to certain words (see Honeybone, forthcoming, for a nonarbitrary way of restricting the process to specific words); and the relevant part of T-lenition (its full environment is broader) can be understood as a generalisation of the sort given in (2), which uses [θ] as a cover-symbol for all lenited realisations. Their applicability is summarised in (3). (1)
t → r / V__#V
(2)
t → θ / __#
(3)
(a)
if a T-to-R word is followed by a vowel (__#V), T-to-R can occur
(b)
if a T-to-R word is followed by a consonant (__#C) or is in utterancefinal position (__##), T-lenition can occur
(c)
if a non-T-to-R word is followed by a vowel (__#V), or a consonant (__#C) or is utterance-final (__##), T-lenition can occur
Are these two LE dialect features represented in CHLDL? Both are representable in spelling, and respellings are found in the CHLDL corpus for both. The rhotic product of T-to-R can be straightforwardly spelled using the letter (or , using a doubled consonant to indicate shortness in the preceding vowel), given that it is definitional for the phenomenon that it neutralises /t/ with /r/. One example found in our corpus is: Ee azzin gorra potter piss in ‘That gentleman’s economic status leaves a lot to be desired’ (= ‘He hasn’t got a pot to piss in’), where got – a canonical T-to-R word – is spelled in a way that clearly indicates T-to-R. The product of T-lenition can be spelled in an analogous way to that used for the lenited realisations of (d), discussed above – using the graphemes that are employed in Standard English spelling to represent the fortis alveolar fricative /s/, because the product of T-lenition is also a fortis alveolar fricative. This is indeed attested, in such cases as oh rice ‘very well then’ (= ‘oh right’), where is used. Are T-to-R and T-lenition represented to the same extent? T-to-R is not strongly localised as it is found throughout the English North, and down into the Midlands (see Buchstaller et al. 2013), while T-lenition is quite highly localised,
100
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
found only in Liverpool and a few other dialects (Irish English, Middlesbrough English – see Jones & Llamas 2008), so this would predict that T-lenition should be represented in orthographic forms more commonly than T-to-R. We suggested above, though, that phonolog]ical salience can trump localisedness. These two phenomena are both phonological processes, however, so might be expected to score equally in this regard. We show in the next section that there are good reasons why this is not the case.
6 Representations of word-final /t/ in CHLDL We can consider the spelling of all occurrences of all words with final /t/ that occur in the corpus of CHLDL described above. The results are very clear. They are presented in Figure 5.7.
Figure 5.7: Spelling of /t/ in the LE corpus of CHLDL in three phonological environments: utterance final (__##), word-final preconsonantal (__#C) and word-final prevocalic (__#V). The spelling combines and . N values are token counts.
The immediate impression is that the results are not the same in all the situations set out for T-to-R and T-lenition in (3a), (3b) and (3c). In the environments shared by (3b) and (3c), where T-lenition is possible (and indeed common in speech, as shown in Figure 5.5), the overwhelming spelling is with , as in Standard English. In (__##), 93.9% of the spellings are with , and in (__#C),
Lenition and T-to-R are differently salient
101
96.1% of the spellings are with . T-lenition is not written in CHLDL, except for in 1.5% of the cases in (__##), where is used, as in the example above. In the environment which is relevant to (3a), spellings with are common (this includes spellings with ). In the overall (__#V) environment, 55.6% of words with final /t/ are written as a rhotic. This compares to 43.7% of the words which are written with either or . There are also tiny numbers of cases where other graphemes are used (0.5% overall have and 1.3% overall have no final consonant transcribed), which we do not consider further here. Figure 5.7 includes all /t/-final words in the three environments, but (3a) refers to T-to-R words – as we have shown, not all words allow T-to-R. If we split /t/-final words into T-to-R words (as defined in section 5) and non-T-to-R words, the results are clearer still. Figure 5.8 shows the results for the three environments split between T-to-R words and non-T-to-R words.
Figure 5.8: Spelling of /t/ in Liverpool CHLDL split by words which exhibit T-to-R in OLIVE’s Archive subcorpus and words which do not, in three phonological environments: utterance final (__##), word-final preconsonantal (__#C) and word-final prevocalic (__#V)
Two things are clear from Figure 5.8: T-to-R is frequently spelled in CHLDL where it is possible in speech – 73.3% of T-to-R words in (__#V) are spelled as a rhotic; and the CHLDL authors are highly accurate in using to spell word-final /t/ only in words which allow it phonologically – the only exception is one occurrence with might. It may be that might is a T-to-R word for at least some LE speakers, or it may be a mistake on the part of the author – if so, this pales in comparison to the accurate spellings.
102
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
To summarise: T-lenition is represented vanishingly rarely in CHLDL, even though it is very common in speech and is in principle respellable: in just 1.5% of the cases in (__##) and never in (__#C) or (__#V); whereas T-to-R is represented very commonly where it would be expected phonologically: in 73.3% of cases of a T-to-R word occurring in (__#V).
6.1 Why is there a difference? Lenition and T-to-R are differently salient It may seem surprising that T-lenition and T-to-R should be represented orthographically in CHLDL to such different extents. It was argued in section 4 that different types of phonological phenomena should be expected to show different degrees of salience (that is, noticeability by speakers), and hence CHLDL authors would be differently able to indicate them in spelling. In decreasing degrees of salience, we differentiated between: – systemically-relevant phenomena which involve underlying contrasts – context-free differences in the realisation of segments – context-determined phenomena which involve phonological processes Both T-to-R and T-lenition fit into the last of these categories – they are formalisable as phonological rules (or could be expressed through ranked constraints or in any other phonological formalism) which affect the same underlying segment and derive different segments in specific (overlapping) phonological environments. Why should there then be the vast difference in representability between T-lenition and T-to-R? We hint at a rationale in H&W and explore it in detail here. The explanation comes from the fact that T-to-R and T-lenition (the latter as part of Liverpool Lenition more generally) have different characteristics as phonological processes – they respectively display traits of ‘early’ phonological processes and ‘late’ phonological processes, to use Coetzee & Pater’s (2011) terminology. All processes come after the earliest aspect of phonology (the underlying forms themselves) and the idea that there are different types of phonological processes is not new – it is visible in the structuralist idea that there is a distinct morphophonology and then a phonology proper, and it is fundamental in the model of Lexical Phonology (also called Stratal Phonology – see Kiparsky 1982, 2015; Bermúdez-Otero, forthcoming) where it is instantiated as a difference between lexical and postlexical processes. Coetzee & Pater (2011) set out five “typical characteristics assigned to each” of the two types of processes, reproduced here as (4i-v), and McMahon (1991, 1994), summarising work such as Kiparsky (1982,
Lenition and T-to-R are differently salient
103
1988), lists all these – apart from (v) – and also adds two more, reproduced here (one in the wording from 1991 and one in the wording from 1994) as (4vi–vii). (4)
Early Phonology
Late Phonology
(i)
Sensitive to morphology
Insensitive to morphology
(ii)
May have exceptions
Exceptionless
(iii)
Makes only categorical changes
Can introduce non-categorical changes
(iv)
Word bounded
Sensitive to cross-word contexts
(v)
Insensitive to factors like speech rate
Sensitive to factors like speech rate
(vi)
Observable/categorisable
Speakers unaware
(vii)
Operate on and introduce only contrastive units
May introduce novel segments and features
T-to-R and T-lenition differ on many of these criteria, with T-to-R showing hallmarks of early phonology and Liverpool Lenition hallmarks of late phonology. Thus, to consider the relevant criteria: (ii) T-to-R has a large number of exceptions given that it is word-specific, whereas T-lenition is exceptionless; (iii) T-to-R makes a categorical change involving either a [t] (if it does not apply) or an [r] (if it does), whereas T-lenition can introduce a gradient range of fricatives, meaning that it is non-categorical; (v) links to the products of T-lenition being somewhat sensitive to speech rate as they vary in line with many such factors, whereas T-to-R does not show clear evidence of being sensitive to speech rate; (vii) T-to-R is neutralising, and in some sense ‘structure preserving’ (Kiparsky 1982) in that it only involves units which are contrastive and which exist at the underlying level (/t/ and /r/), whereas T-lenition creates [θ] and other novel fricatives which do not exist at the underlying level. Criteria (i) and (iv) do not distinguish the two processes, and criterion (vi) is, in fact, what our results show to be the case: speakers are aware of T-to-R and are able to spell it in CHLDL, whereas speakers do not seem to be aware of T-lenition and therefore do not try to spell it in CHLDL5 In general, there is a good correlation between T-to-R and the properties expected of an early phonological process and Liverpool Lenition and the properties expected of a late phonological process; it may also be relevant that
5 This is in part no surprise – it is clear on other grounds that T-to-R is above the level of consciousness: our own intuitions tell us so, and the success of the questionnaire method of investigation into T-to-R in Buchstaller et al. (2013) and Caffrey (2011) also show it to be the case.
104
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
the two processes would in any case be ordered this way through an ‘elsewhere relationship’ – this would see T-to-R, with a more specific environment, ordered before T-lenition, with a more general environment, due to the Elsewhere Condition. At the start of this section, we set out the three degrees of phonological salience that we had previously recognised – it is now clear that we need to refine this: although both T-to-R and T-lenition are context-determined phonological processes, we have shown that we need to differentiate between early and late phonological processes in terms of their degree of salience. This has been tied to the basic characteristics recognised for the two types of phonological process, and it correlates with our findings. If early phonological processes are more ‘noticeable’ than later processes, and if differential phonological salience determines the extent to which a dialect feature can be spelled in dialect literature, all falls into place: we should expect only early phonological processes to lead to frequent respellings in dialect literature, and this is what we find in our corpus of CHLDL.6 This retains the general idea that ‘early phonology’ is more salient than ‘late phonology’ (with underlying contrasts being the ‘earliest’ aspect of phonology overall) but also extends our understanding by differentiating between different levels of phonological salience for different types/levels/ strata of phonological processes. While localisedness can play a role, as shown in H&W, that role seems to be secondary, such that it may make a difference when two phonological phenomena tie in terms of phonological salience, but it can also be substantially overridden by different degrees of phonological salience.
7 Conclusions This piece has reconsidered a number of issues that we discussed in Honeybone & Watson (2013) and has added to the results and rationale considered there. We have presented new detail in the representation of T-to-R in written LE, showing it to be well represented in orthographic forms in our corpus of CHLDL. We considered the relationship between T-to-R and another process which can also affect /t/, but which is more localised: Liverpool Lenition. We have shown T-to-R to be an early phonological process, and as such to be somewhat phonologically salient, like differences which involve underlying contrasts. This explains why it 6 As Auer, Barden & Grosskopf (1998: 184) put it: “postlexical processes should be less salient than lexical regularities” (we can to a fair extent map ‘postlexical’ to ‘late’, and ‘lexical’ to ‘early’). They go on to say that this is “a prediction which remains to be tested” – our results provide some evidence that the prediction is met.
Lenition and T-to-R are differently salient
105
is so much better represented in CHLDL than T-lenition: T-lenition is not salient to speakers of LE because it is a low-level, late phonological process. As in H&W, CHLDL has been shown to represent dialect features accurately, and, the notion of phonological salience has provided a way to understand the difference in the extent to which native speakers notice dialect features, and hence can represent them in dialect literature.
References Asprey, Esther. 2008. The sociolinguistic stratification of a connected speech process – the case of the T to R rule in the Black Country. Leeds Working Papers in Linguistics and Phonetics 13, 109–140. Beal, Joan C. 2000. From Geordie Ridley to Viz: popular literature in Tyneside English. Language and Literature 9: 343–359. Beal, Joan C. 2009. Enregisterment, commodification and historical context: “Geordie” versus “Sheffieldish”. American Speech 84: 138–156. Beal, Joan C., Lourdes Burbano-Elizondo and Carmen Llamas. 2012. Urban North-Eastern English: Tyneside to Teesside. Edinburgh: Edinburgh University Press. Bennett, Joe. 2012. “And what comes out may be a kind of screeching”: the stylisation of chavspeak in contemporary Britain. Journal of Sociolinguistics 16, 5–27. Bermúdez-Otero, Ricardo. (forthcoming). Stratal optimality theory. Oxford: Oxford University Press. Broadbent, Judith M. 2008. t-to-r in West Yorkshire English. English Language and Linguistics 12, 141–168. Buchstaller, Isabelle, Karen Corrigan, Anders Holmberg, Patrick Honeybone and Warren Maguire. 2013. T-to-R and the Northern Subject Rule: questionnaire-based spatial, social and structural linguistics’ English Language and Linguistics 17, 85–128. Caffrey, Catherine. 2011. T-to-R in Liverpool English. Edinburgh: University of Edinburgh MA (Hons) dissertation. Cardoso, Amanda. 2015. Dialectology, phonology, diachrony: Liverpool English realisations of PRICE and MOUTH . Edinburgh: University of Edinburgh PhD thesis. Clark, Lynn and Kevin Watson. 2011. Testing claims of a usage-based phonology with Liverpool English t-to-r. English Language and Linguistics 15, 523–547. Clark, Lynn & Kevin Watson. 2016. Phonological leveling, diffusion, and divergence: lenition in Liverpool and its hinterland. Language Variation and Change 28: 31–62. Coetzee, Andries W. and Joe Pater. 2011. The place of variation in phonological theory. In John Goldsmith, Jason Riggle & Alan Yu (eds.), The handbook of phonological theory, 401–434. Oxford: Blackwell. Coupland, Nikolas and Hywel Bishop. 2007. Ideologised values for British accents. Journal of Sociolinguistics 11, 74–103. Crowley, Tony. 2012. Scouse: a social and cultural history. Liverpool: Liverpool University Press. Cruttenden, Alan. 2014. Gimson’s pronunciation of English. Eighth edition. Abingdon: Routledge. Grant, Anthony. 2007. Looking (literally) at Liverpool English: thoughts on the popular (and less popular) documentation of Scouse lexicon. In Grant, Anthony and Clive Grey (eds.), The Mersey sound: Liverpool’s language, people and places, 141–163. Liverpool: Open House Press.
106
Patrick Honeybone, Kevin Watson and Sarah van Eyndhoven
Hickey, Raymond (ed.). 2010. Varieties of English in writing: the written word as linguistic evidence. Amsterdam: Benjamins. Honeybone, Patrick. 2001. Lenition inhibition in Liverpool English. English Language and Linguistics 5, 213–249. Honeybone, Patrick. 2007. New dialect formation in nineteenth century Liverpool: a brief history of Scouse. In Grant, Anthony and Clive Grey (eds.), The Mersey sound: Liverpool’s language, people and places, 106–140. Liverpool: Open House Press. Honeybone, Patrick. 2008. Lenition, weakening and consonantal strength: tracing concepts through the history of phonology. In Brandão de Carvalho, Joachim, Tobias Scheer and Philippe Ségéral (eds.), Lenition and fortition, 9–93. Berlin: Mouton de Gruyter. Honeybone, Patrick. (forthcoming). Categorical frequency effects favour formal phonology: T-to-R in northern British English. Honeybone, Patrick and Kevin Watson. 2013. Salience and the sociolinguistics of Scouse spelling: Exploring the phonology of the Contemporary Humorous Localised Dialect Literature of Liverpool. English World-Wide 34, 305–340. Jensen, Marie. 2013. Salience in language change: a socio-cognitive study of Tyneside English. Newcastle: University of Northumbria at Newcastle PhD thesis. Johnstone, Barbara. 2009. Pittsburghese shirts: commodification and the enregisterment of an urban dialect. American Speech 84, 157–175. Jones, Mark J. and Carmen Llamas. 2008. Fricative realisations of /t/ in Dublin and Middlesbrough English: an acoustic study of plosive frication rates and surface fricative contrasts. English Language and Linguistics 12, 419–443. Kerswill, Paul and Ann Williams. 2002. ‘Salience’ as an explanatory factor in language change: evidence from dialect levelling in urban England. In Mari C. Jones and Edith Esch. (eds.), Language change. The interplay of internal, external and extra-linguistic factors, 81–110. Berlin, New York: Mouton de Gruyter. Kiparsky, Paul. 1982. Lexical morphology and phonology. In In-Seok Yang, (ed.), Linguistics in the morning calm, 3–91. Seoul: Hanshin. Kiparsky, Paul. 1988. Phonological change. In Frederick Newmeyer (ed.), The Cambridge survey of linguistics. Vol. 1, 363–415. Cambridge: Cambridge University Press. Kiparsky, Paul. 2015. Phonologization. In Patrick Honeybone and Joseph Salmons (eds.), The Oxford handbook of historical phonology 563–579. Oxford: Oxford University Press. Knowles, Gerald. 1973. Scouse: the urban dialect of Liverpool. Leeds: University of Leeds PhD thesis. Labov, William. 1972. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press. Lass, Roger. 1984. Phonology: an introduction to basic concepts. Cambridge: Cambridge University Press. Lass, Roger. 2015. Interpreting alphabetic orthographies: Early Middle English spelling. In Patrick Honeybone and Joseph Salmons (eds.), The Oxford handbook of historical phonology, 100–120. Oxford: Oxford University Press. Leach, Hannah, Kevin Watson and Ksenia Gnevsheva. 2016. Perceptual dialectology in northern England: Accent recognition, geographical proximity and cultural prominence. Journal of Sociolinguistics, 20: 192–211. Liverpool Echo. 2008. Why the Liverpool accent is as sound as a pound. Liverpool Echo http:// www.liverpoolecho.co.uk/news/liverpool-news/liverpool-accent-sound-pound-3478590. McMahon, April. 1991. Lexical phonology and sound change: the case of the Scottish vowel length rule. Journal of Linguistics 27, 29–53.
Lenition and T-to-R are differently salient
107
McMahon, April. 1994. Understanding language change. Cambridge: Cambridge University Press. Montgomery, Chris. 2007. Northern English dialects: a perceptual approach. Sheffield: University of Sheffield PhD thesis. Newbrook, Mark. 1986. Sociolinguistic reflexes of dialect interference in West Wirral. Frankfurt am Main: Peter Lang. Pandeli, Helen, Joseph F. Eska, Martin J. Ball and Joan Rahilly. 1997. Problems of phonetic transcription: the case of the Hiberno-English slit-t. Journal of the International Phonetic Association 27, 65–75. Racz, Peter. 2013. Salience in sociolinguistics: A quantitative approach. Berlin, New York: Mouton de Gruyter. Sangster, Catherine M. 2001. Lenition of alveolar stops in Liverpool English. Journal of Sociolinguistics 5, 401–12. Schneider, Edgar W. 1986. “How to speak Southern”? An American English dialect stereotyped. Amerikastudien / American Studies 31, 425–39. Shorrocks, Graham. 1996. Non-standard dialect literature and popular culture’ In Juhani Klemola, Merja Kytö and Matti Rissanen (eds.), Speech past and present: studies in English dialectology in memory of Ossi Ihalainen, 385–411. Frankfurt: Peter Lang. Taavitsainen, Irma, Gunnel Melchers and Päivi Pahta (eds.). 1999. Writing in non-standard English. Amsterdam, Philadelphia: Benjamins. Trudgill, Peter. 1986. Dialects in contact. Oxford: Blackwell. Wales, Katie. 2006. Northern English: a social and cultural history. Cambridge: Cambridge University Press. Watson, Kevin. 2007a. Liverpool English. Journal of the International Phonetic Association 37, 351–60. Watson, Kevin. 2007b. The phonetics and phonology of plosive lenition in Liverpool English. Ormskirk: Edge Hill Collee and Lancaster University PhD thesis. Watson, Kevin and Lynn Clark. (forthcoming). The origins of Liverpool English. In Hickey, Raymond (ed.), Listening to the past: audio records of accents of English. Cambridge: Cambridge University Press. Wells, J.C. 1982. Accents of English. Cambridge: Cambridge University Press.
II Phonology
Sandra Jansen
6 External and internal factors in a levelling process: Prevocalic (r) in Carlisle English Abstract: This chapter addresses the language-internal and external factors in the levelling of (r) in Carlisle English, an urban community in the far northwest of England. Based on a quantitative analysis of oral history recordings and sociolinguistic interviews, which span 30 years in real time and 100 years in apparent time, this chapter concentrates on the distribution of this variant in different prevocalic environments. In particular, I explore internal and external constraints in the loss of taps, a formerly superregional feature which is now quite restricted geographically in the north of England. I argue that the different environments of prevocalic (r) need to be investigated separately because different mechanisms are at work in each environment and that languageinternal factors play an important part in this change. I also argue that the loss of PreR-dentalisation is linked to the loss of taps.
1 Introduction Extensive research on (r) exists in the fields of phonetics, phonology and sociolinguistics, mainly because of the astonishing variation in this feature across and within languages which van Hout and van de Velde (2001: 1) try to capture: “The /r/ appears to cover an impressive range of sounds between trills, taps, fricatives and vocalic realisations, ranging in place of articulation from labial to uvular”.1 In addition to the place of articulation, the phonotactic position of (r) is highly significant. The variant can be separated into two broad categories: prevocalic and postvocalic.
Acknowledgement: I would like to thank Sylvie Hancil for inviting me to the Northern British Conference and the two anonymous reviewers for their helpful comments. Of course, all errors remain my own. 1 One of the reviewers pointed out that (r) can also be realised in pharyngeal and glottal places of articulation, e.g., [h] as pronunciation of (r) in Brazilian Portuguese. Sandra Jansen, University of Paderborn DOI 10.1515/9783110450903-006
112
Sandra Jansen
In particular, variation of and loss in postvocalic (r) in England – in the pockets of rhoticity that still exist in Lancashire and the south-west – have been discussed in recent years. Piercy (2007) investigates postvocalic (r) in four communities in Dorset and reports that rhoticity is in decline. She suggests that this decline runs along an urban hierarchical trajectory from the most urban to the most rural community. She also proposes that the fragmentation of very rural communities has led young speakers to abandon rhoticity faster than in other communities. In another paper, Piercy (2012) focuses in particular on linguistic constraints in her study of the loss of rhoticity and identifies the preceding vowel environment, the word context, stress, and lexical frequency as significant factors in this change, while social factors such as speaker sex and age are found not to be significant. Moreover, she compares her results with those of Nagy and Irwin (2010), who investigate constraints on r-fullness, i.e., the increase of rhoticity in Boston and neighbouring communities in New Hampshire. Piercy finds remarkable commonalities in the linguistic constraints lexical frequency, stress, preceding vowel, and word context. She comes to the conclusion that “this suggests that linguistic factors, which may apply across all varieties of English could have universal effects in the use of /r/” (Piercy 2012: 85). Piercy’s findings show that internal constraints must play an essential role in this language change. Similar to Piercy’s study, Barras (forthcoming) compares the loss of rhoticity in Lancashire with the loss of this feature in rural Oxfordshire. In the first decade of the 21st century, attention was predominantly paid to variation in prevocalic (r) when it was studied mainly in terms of the diffusion of the labiodental form [ʋ] across England (cf. Foulkes & Docherty 2000; Llamas 2001; Kerswill 2003; Marsden 2006), and even though taps are mentioned in some studies, the focus tends to be on [ʋ] (Kerswill 2003: 231). Similar to other consonantal changes that are observed around the country (cf. Britain 2005), the change toward [ʋ] led by working class (WC) male speakers is an example of covert prestige (Foulkes & Docherty 2001: 39). In Carlisle, the use of [ʋ] is still restricted (cf. Jansen 2012) and hence, the findings are only tentative but it seems that a similar pattern to other northern locations (cf. Foulkes & Docherty 2000, 2001; Llamas 2001) is emerging in this city. The results for [ʋ] are briefly reviewed in the discussion section. The change from taps to postalveolar approximant [ɹ] in prevocalic position, on the other hand, has been neglected in the discussion of levelling, i.e., “the reduction or attrition of marked variants” (Trudgill 1986, 98; emphasis in original). In general, it seems that while in Scottish English taps are commented on (cf. Lawson, Scobbie & Stuart-Smith 2011; Stuart-Smith 2007, 2008, Jauriberry, Sock & Hamm 2015), the use of taps is quite understudied in English English.
External and internal factors in a levelling process
113
Only infrequently is the use of taps mentioned. At the beginning of the 1980s, Wells (1982: 368) claimed that taps are still fairly common after voiceless interdental fricatives [θ] and labials [f] and in intervocalic position in the North, and indeed, Shorrocks (1998: 390–392) found taps in intervocalic position in Bolton. Similarly, Coupland (1980: 5) commented that “the tapped variant is most common intervocalically” in Cardiff. Watt, Llamas and Johnson (2014) discussed the distribution of (r) in pre- and postvocalic positions in four communities along the Scottish-English border, among others, Carlisle. They show that taps are produced in only low numbers in Carlisle speech and come to the conclusion that “overall, [. . .], the alveolar approximant [. . .] is clearly the default form of overt (r) among speakers from this region of Britain” (Watt, Llamas & Johnson 2014: 95). Even though the literature on the use of taps in English varieties is sparse, the combination of voiceless interdental fricatives and taps is frequently commented on, and Maguire (2012: 373), who draws on SED data, claims that an R-Realisation Effect (RRE) might exist in the north of England, i.e., taps and trills are preferred over postalveolar approximants after [θ] and intervocalically. The studies which explore the use of [ʋ] in varieties of English English are part of the large body of studies of geolinguistic processes in England. The analysis of these changes often revolves around internal and external motivations that trigger changes. Hickey (2014: 388) defines internally-motivated change as “any change which can be traced to structural considerations in a language and which is independent of sociolinguistic factors”, while externally-motivated change is “any variation and change in a language which can be connected with the community or society using that language” (2014: 389). However, he rejects the idea of a simple binary division. It is obvious that the simple labels ‘internally-motivated’ and ‘externally-motivated’ language change do not do justice to the complex and intricate relationship between how speakers act linguistically in their community and the postulated abstract level of structure which is taken to provide the basis for their behaviour. It is clear, then, that linguistic reality is too complex to be fully captured by a simple binary division of change types into ‘internal’ and ‘external’ (Hickey 2014: 388).
This sociolinguistic complexity is discussed by Torgersen and Kerswill (2004: 26) who investigate a levelling process, the short front vowel shift, in several places in the south of England. They examine the tension between internal and external factors in language change and come to the conclusion that both internal and external factors can trigger change. However, they emphasise that “contact and extra-lingusitic factors ultimately have the capacity to override natural motivations” (Torgersen & Kerswill 2004: 47). Tagliamonte (2012a) highlights that social
114
Sandra Jansen
and linguistic factors are of importance. “One of the prevailing conclusions of research in [Language Variation and Change] from the 1990s and 2000s has been the affirmation that variables are sensitive to both linguistic and social factors” (Tagliamonte 2012a: 73). Hickey (2014) takes an extreme view in this discussion by ruling out the possibility that language-internal factors can trigger a change but that language external factors always play a role in the initiation of a change. The loss of taps in prevocalic (r) position in Carlisle English is an example of the complex nature of a change and the interplay of external and internal motivations. Jansen (2015a) suggests that an externally-motivated change triggered the (internally-motivated) loss of taps in onset clusters. The decrease of PreRdentalisation, i.e., the historically attested dental pronunciation of the alveolar consonant [t ]̪ before /r/ (cf. Maguire 2012), seems to have had a knock-on effect on the loss of taps in CrV position. We do not know the exact reasons for the loss of PreR-dentalisation but Maguire (2012) shows that the loss of this feature was diffusing northwards, i.e., this must have been a contact-induced change. Jansen (2015a) then suggests that the loss of PreR-dentalisation then triggered the language-internal change from taps to alveolar approximants in Carlisle English. This chapter aims to provide more insight into the loss of [ɾ] for (r) in Carlisle English. The trajectory of change differs in the different phonotactic positions of prevocalic (r) and will be discussed more closely. My points of departure in this analysis are: (1) to investigate the distribution of (r) in prevocalic position in Carlisle English; (2) to consider the developmental trajectory of (r) in apparent time and real time, and the underlying mechanisms that may be guiding the pathway to the loss of taps; and (3) to discuss the impact of internal and external motivations on the change. The chapter is structured as follows. In section 2, I describe the sociolinguistic background of Carlisle, before turning to the methodology used in this study in section 3. In section 4, I present the data analysis by discussing the overall results before examining the different phonotactic environments, which is followed by the discussion in section 5.
2 Sociolinguistic background information on Carlisle Carlisle is a city in the far north-west of England. Figure 6.1 shows the geographical position of the city. It has a population of 101,000, which makes it the largest conurbation in Cumbria. Carlisle is also known as the Border City. The title reflects its proximity to the Scottish border, only 16 km away.
External and internal factors in a levelling process
115
The city’s surrounding area is dominated by agriculture and the closest urban area is Tyneside (Newcastle), some 90 km away. To the south-west of the city, about 50 km from Carlisle, lies the Lake District. To the east, the Pennines, which extend down the country, are less steep than elsewhere, making the north-east relatively accessible. The Borderlands, as well as the area west of Carlisle which stretches out to the Irish Sea (West Cumbria), are sparsely populated regions. No other urban area can be found in close proximity to the city. Hence, Carlisle is a regional centre where people commute from the surrounding areas (Coombs 1995).
Figure 6.1: Location of Carlisle
Two phases of incoming waves of people need to be distinguished. In the second half of the 18th century, when the situation around the Scottish-English border stabilized, the number of people living in Carlisle grew from about 2,000 in 1759 to 33,000 in 1831. From the end of the 18th century, migrants from Scotland and Ireland arrived in Carlisle (MacRaild 1998: 30) and from the beginning of the 19th century, up until around 1960, Carlisle saw a steady increase in population. It is no surprise that the Carlisle dialect has undergone levelling over the late 19th and 20th centuries ( Jansen 2012). Similarly, features that are spreading across the country, for example, T-glottalling, TH -fronting and
116
Sandra Jansen
R-labiodentalisation (e.g., Foulkes & Docherty 1999; Britain 2009), are now also attested in Carlisle English (Jansen 2012).
3 Methodology The data stem from two sources: from oral history recordings and from recordings of sociolinguistic interviews with people from Carlisle. My fieldwork took place in two major stages in February/March 2007 and February/March 2008, with the addition of three individual recordings in September 2009 and two further recordings in September 2010. In the sociolinguistic interviews, three styles were recorded: interview, text and sentence list. In September 2011, I also digitised sixteen oral history recordings of interviews which were conducted in Carlisle in the early 1980s.2 All of the oral history recording speakers were born and bred in Carlisle and had worked in one of the local factories or in one case, as a nanny. Eight of these recordings are used for real-time change analysis here. For obvious reasons, for participants in the oral history recordings only interview data are available. The participants in the oral history recordings were born between 1890 and 1918, while the speakers in the sociolinguistic interviews were born between 1936 and 1990, which makes the sample span one century. The data are stratified by age group, speaker sex, and social class.3 Table 6.1 provides an overview of the sample. All prevocalic (r) tokens for the text and sentence list were analysed. The interviews were transcribed and an impressionistic-auditory analysis of the first 40 instances of prevocalic /r/ produced after minute ten was conducted. Table 6.1: Participant sample in the study
Young 15–29 Middle-aged 30–64 Retired 65+ Oral History Recordings Total
Male
Female
Total
4 5 3 4 16
4 5 5 4 18
8 10 8 8 34
2 My sincere thanks go to Edwin Rutherford, Keeper of Human History at Tullie House Museum in Carlisle. 3 I need to point out that social class is included but needs to be handled with caution as the speakers from the oral history recordings all belong to the working class.
External and internal factors in a levelling process
117
The statistical model used here for the overall variation in prevocalic (r) includes the social factors of age group, speaker sex, and social class, and the languageinternal factor environment. The environment around (r) is so diverse that I coded for a range of independent variables for each environment. First, the overall distribution will be analysed; then each environment will be discussed separately. A mixed-effect multiple logistic regression using Rbrul (Johnson 2014) was conducted, in which the tap variant was treated as the application value. The advantage of this statistical tool is to include random factors such as individual and lexical item as a continuous variable.
4 Data analysis Five main variants were identified in the sample: – a voiced postalveolar approximant [ɹ]. This is the unmarked variant which is now found in all varieties of English in England. – an alveolar tap [ɾ] or trill [r]: the latter is used only on a handful of occasions but tap and trill share articulatory characteristics and therefore, both sounds are counted towards the tap variant here. – a zero (non-/r/) realisation: in Vr#V position, a hiatus between the two vowels is observed. Foulkes (1997: 78) mentions that in some cases, glottal stops are inserted instead of linking /r/ in Newcastle English, which is also the case in Carlisle English. This possible realisation is also classified as the zero form. This zero variant is also found in low numbers in CrV position. – a fricative that occurs after /t, d/4. The IPA symbol [ʑ] is used as an umbrella term for the different degrees of frication in CrV position. – labiodental [ʋ]: this variant is a fairly recent innovation in Carlisle English (cf. Jansen 2012). For – – – – –
prevocalic position, five environments were identified: intervocalic position VrV (as in very); word-initial consonant clusters: CrV (as in crowd); word-initial position I: C#rV (as in quite regular); word-initial position II: V#rV (as in to run); phrase-initial (r) (#rV as in Read!).5
4 This fricative is not discussed any further as the quality varies quite dramatically. Further acoustic investigation is needed to understand the gradual differences. 5 The dataset contained only 24 tokens of (r) in this environment (20 tokens are realized as approximants, two as the tap and two tokens as [ʋ]) but due to the low numbers, this environment will not be discussed in detail.
118
Sandra Jansen
In addition, tokens were collected from intervocalic position across wordboundaries (as in far away). The data extraction provided a sample of 2,438 tokens. Table 6.2 provides an overview of the distribution of variants of preceding (r) in the sample according to age group and sex. Across apparent time, a decrease in taps appears and the data suggest that they are replaced by the unmarked form. The two variants [ʑ] and [Ø] are phonotactically restricted, i.e., neither variant is found in VrV position and the former variant is not found in V#rV position. Table 6.2: Variants of prevocalic (r) in % in the present study ɾ
ɹ
ʑ
Ø
ʋ
other
Total (N)
Oral History Recordings
M F
56.3 66.3
37.5 28.8
1.9 3.1
4.4 1.9
0 0
0 0
160 160
Retired
M F
49.6 38.2
40.7 52.0
7.7 8.6
1.6 1.2
0.4 0
0 0
248 408
Middle-aged
M F
25.0 20.7
58.0 68.7
11.2 8.6
5.3 1.7
0 0
0.5 0.2
412 406
Young
M F
10.0 7.5
61.1 71.7
9.7 10.9
6.0 5.9
13.2 4.0
0 0
319 320
29.5
56.0
8.6
3.5
2.3
Total
0.01
2438
Table 6.3 summarises the distribution of taps in the dataset according to age group, sex and style. As expected, we observe a steady decline in taps across the age groups. Even though some kind of style shifting occurs in the different age groups, a clear stratification of the data according to style is not observable. Table 6.3: Distribution of taps in % according to age group, sex and style Style
Interview
Text
Sentence List
Total N
Oral History Recordings
M F
100.0 100.0
– –
– –
90 106
Retired
M F
53.7 56.4
19.5 18.6
26.8 25.0
123 156
Middle-aged
M F
53.4 45.2
16.5 22.6
30.1 32.1
103 84
Young
M F
50.0 58.3
28.1 16.7
21.9 25.0
32 24
65.9
14.2
19.9
718
Total
External and internal factors in a levelling process
119
The results from the oral history recordings are included in the descriptive statistical analysis but are excluded from the statistical models presented here. This is due to the fact that only interview data exist for this age group and that mixing real time and apparent time data in a statistical model needs to be approached with caution. Nevertheless, the data show a continuing decrease in taps across the four age groups and hence, across both real and apparent time. The statistical results for taps in prevocalic position for the retired, middleaged and young speakers are presented in Table 6.4. Social class, age group and environment are all significant factors. Sex and style are not significant predictors in this change and will therefore not be discussed any further. Table 6.4: Rbrul results for (r) in prevocalic position. Application value: tap (oral history data are not included)
Social Class p = 0.00931 Working Class Middle Class Age Group p < 0.001 Retired Middle-aged Young Environment p < 0.001 VrV Vr#V V#rV CrV C#rV #rV
Weight
N
Proportion of [ɾ]
.64 .36
1141 972
.316 .167
.80 .52 .19
656 818 639
.425 .229 .088
.90 .90 .54 .35 .12 .12
384 251 237 1021 199 21
.536 .534 .211 .121 .035 .048
deviance: 1537.15 df: 10 speaker ID random SD: 0.987
Age The results from the oral history recordings, in addition to the results from the data collected in 2007–2010, provide a real time view of the decrease in taps in the community, which confirms Watt, Llamas and Johnson’s (2014) findings. The results show that in the oral history recordings, taps were only used about half of the time, i.e., variation between taps and other forms was already quite common when the speakers from the oral history recordings were young at the end of the 19th/beginning of the 20th century.
120
Sandra Jansen
The three age groups included in the statistical model show significant differences in the distribution of (r) variants (p < 0.001). We see a continuous decrease across the three age groups and the change away from taps is almost complete for young speakers. Moreover, the use of taps is now restricted to certain phonotactic environments, which will be discussed later. First, we turn to the external factors sex, social class and style.
Social Class Social Class is a significant factor in this model (p = 0.00931). WC speakers use taps more often than MC speakers. However, no interaction between social class and sex and between social class and age group is found. The results reflect the common finding that WC speakers retain a greater number of traditional features longer than MC speakers.
Environments The different environments are statistically significant (p < 0.001) with VrV, Vr#V and V#rV favouring taps, while CrV, C#rV and #rV disfavour this variant. Phonotactic environments have been shown to have significant influence in this change previously (cf. Jansen 2015a, with a slightly different sample) and while the constraints for CrV position have been discussed before, factors in other environments have not been looked at in detail. In the following, I provide an overview of the distribution of taps in each environment and discuss the statistical results in the different environments in more detail. The order of the environments presented here follows the constraint order as provided in Table 6.4. Depending on the environment, I coded for a number of language internal factors such as preceding or following vowel height or stress, which were in most cases not statistically significant and are therefore not discussed.
VrV Though they are decreasing, the use of taps is retained longer in this environment than in others. In Figure 6.2, the distribution of (r) is presented and Table 6.5 provides an overview of the statistical results for VrV with taps as application values. This effect is mentioned in other varieties as well; for example, in Bolton, taps are only found in this environment and after dental fricatives (Shorrocks 1998: 390–392). Maguire (2012: 373) claims that there is an R-Realisation Effect in these two environments. In his study on style shifting in the workplace in Cardiff, Coupland (1980) reports that the use of taps in VrV is much higher than in word-initial position. A similar distribution is found in Carlisle. Maguire
External and internal factors in a levelling process
121
(2012) comments on the “deliberate transcription policies” in the Survey of English Dialects (SED) and he suggests that taps would probably have been found in other environments besides CrV. Table 6.5: Rbrul results for (r) in VrV position. Application value: tap (Oral History data are not included)
Social Class p = 0.00407 Working Class Middle Class Age Group p < 0.001 Retired Middle-aged Young Primary Stress p < 0.001 No Yes
Weight
N
Proportion of [ɾ]
.76 .24
203 178
.66 .40
.88 .59 .09
117 158 106
.78 .55 .26
.86 .14
330 51
.60 .14
deviance: 327.124 df: 7 speaker ID random sd: 1.665 Not significant factors: sex, style, height of following vowel, height of preceding vowel, frontback preceding vowel, front-back following vowel Random intercepts: speaker, word
Predictors of change in the statistical model are age (p < 0.001), the absence of primary stress (p < 0.001), and social class (p = 0.004). Again, sex and style are not predictors of change and other than stress, internal factors do not have significant effects in this model. This lack of further internal factors is somewhat surprising as stronger phonological conditioning was expected in this environment, given that Maguire (2012) suggests an R-Realisation Effect.
V#rV The decrease in taps in V#rV position is quite dramatic (see Figure 6.4). However, none of sex, social class and style as external factors are significant influences in this change. The only significant language-internal factor is the height of the preceding vowel. This is somewhat surprising, as the tongue needs to cover quite a distance to the alveolar ridge where the tap is produced. High vowels, on the other hand, disfavour taps. Overall, only three variants occur in this environment: taps, postalveolar approximants and [ʋ].
122
Sandra Jansen
Figure 6.2: Distribution of (r) in VrV position
Figure 6.3: Distribution of (r) in Vr#V position
External and internal factors in a levelling process
Figure 6.4: Distribution of (r) in V#rV position
Figure 6.5: Distribution of (r) in CrV position
123
124
Sandra Jansen
Figure 6.6: Distribution of (r) in C#rV position
Here the position of (r) might be the important factor, as it occurs in the highly salient word-initial position (similar to C#rV). But there are indications that the salience of this position only became important well into the overall change of prevocalic (r). The change away from taps in word-initial position following a consonant (C#rV) is much more advanced than in V#rV position, which leads to the conclusion that the change either started earlier or that the change in C#rV position happened more rapidly.] Table 6.6: Rbrul results for (r) in V#rV position. Application value: tap (Oral History data are not included) Weight Age Group p < 0.01 Retired Middle-aged Young Preceding Vowel Height p = 0.0379 Low Mid High
N
Proportion of [ɾ]
.84 .47 .18
79 88 70
.44 .14 .04
.73 .46 .31
36 156 45
.39 .20 .11
deviance: 185.41 df: 7 speaker ID random sd: .211 Not significant factors: front-back of the preceding vowel, following vowel height, style, frontback of the following vowel, sex, number of morphemes, social class
External and internal factors in a levelling process
125
In both the word-initial positions C#rV and V#rV, the innovative variant [ʋ] is increasing rapidly (see Figure 6.2 and Figure 6.6). In these environments, over 20% of this variant are used by the group of young speakers, while the frequency of this variant is still very limited in the other prevocalic (r) positions.
CrV The data show that in CrV position, [t] is rarely followed by taps, while taps are preceded – to varying degrees – by other consonantal sounds. At least three sounds are phonologically conditioned in this environment. The lack of taps in this environment after [t] can be observed6, but also [θ] which (still) highly favours taps. As mentioned before, Maguire (2012) assumes that an RRealisation Effect is in place with this preceding consonant. I argue here that the use of taps after [θ] is also phonologically conditioned and the R-Realisation Effect is found in this environment in Carlisle English. However, external forces seem to be at work as well which lead to the decrease in taps. Hence, taps are retained longer in this environment but are also decreasing across apparent time. Table 6.7: Rbrul results for (r) in CrV position. Application value: tap (Oral History data are not included) Weight Preceding Consonant p < 0.01 [θ] [ɡ] [b] [k] [f] [v] [p] [t] Age Group p < 0.01 Retired Middle-aged Young Social Class p = 0.0179 Working Class Middle Class
N
Proportion of [ɾ]
.92 .78 .75 .62 .62 .21 .16 .06
33 71 166 51 168 31 157 248
.61 .34 .20 .26 .14 .13 .05 .02
.82 .51 .17
282 359 284
.29 .12 .03
.66 .34
503 422
.20 .07
Deviance: 493.433 Df: 13 Speaker ID random sd: .969 Not significant factor: style, stress, sex 6 Because there were no instances of taps after [d], this variant was not included in the statistical model.
126
Sandra Jansen
A third phonological conditioning is the frication of (r) after /t, d/ (see Figure 6.5). The data show that there is a tendency to produce fricatives instead of postalveolar approximants. This tendency is found after /t, d/. The use of fricatives in CrV position increases across apparent time and the fricative quality differs in that the friction seems to increase across apparent time. This is clearly an area for further study. Another variant that is used infrequently but also mentioned by Watt, Llamas and Johnson (2014) is the zero form in this environment. This variant also sporadically appears in VrV position.
C#rV In C#rV position, taps are now obsolete in Carlisle English (Figure 6.6). The use of taps decreases across real and apparent time. However, the variation becomes more complex for the young speakers. In the three older age groups, postalveolar approximants are replacing taps. In the group of young speakers, [ʋ] is increasing at the expense of taps and postalveolar approximants. The statistical model does not provide valuable insights as only the randomized factors lexical item and speaker are significant. This statistical result is probably due to the low number of taps across the age groups.
5 Discussion The data analysis has shown that the use of taps has decreased and has been replaced by postalveolar approximants more and more in Carlisle English. Even though very little research has been conducted which focuses, in particular, on the diffusion of [ɹ] (see Glauser 2000; Watt, Llamas & Johnson 2014, for the situation on the Scottish-English border) and at the same time the erosion of taps in the north of England, the sparse comments on the use of taps in this area lead to the assumption that this is indeed a change that has progressed northwards. The data for Carlisle show that it was more or less the last chance to study taps in this variety. Hence, geolinguistic processes seem to have been in place and the question arises of whether this northward movement of increase of postalveolar approximants and with it the loss of taps is only due to languageexternal factors or whether the picture is more complex for this change. One finding from the data analysis which is quite striking is that taps are not represented after /t, d/ in the data. Hickey (2004: 38) points out that a preceding alveolar stop is less likely to be followed by a tap due to articulatory complexity. He mentions that dental stops [t ̪] instead of alveolar stops are used in front of taps in Northern Irish English. The articulatory complexity, which
External and internal factors in a levelling process
127
Hickey describes, might indeed be the reason why Carlisle English speakers – who do not use [t ]̪ – hardly use taps in this environment. However, dental stops are not restricted to (Northern) Irish English. They seem to have had a wider currency in England in the 19th century and earlier. Maguire (2012) argues that PreR-dentalisation is indeed a traditional feature of English dialects and that it does not originate from dialect contact with (Northern) Irish English. Ellis (1889) described the feature for most of the north of England. By the time of the SED in the middle of the 20th century, the variant had declined further and was only attested in Cumberland, Westmorland, Lancashire, Durham, northern Yorkshire, and the Isle of Man (Maguire 2012: 368). Ellis (1889: 543) comments on the use of dental sounds for Edenside, the area which includes Carlisle: “After t, d the r is invariably dental”, and the SED data for Cumberland indicate that in the majority of cases, [t ̪, d̪] occurred in combination with taps. Wright (1977: 14) also mentions that dental sounds are produced before taps in Carlisle: “Phonetically speaking we have [. . .] some dental t’s so that e.g. ‘tractor’ sounds like ‘thractor’”. All of this points to a former use of [t ̪, d̪] in Carlisle English, but even though the sources do not provide much information about the quantitative distribution of [t ̪, d̪] at the time, the choice of words in both sources, i.e., “invariably” and “some”, could hint at some kind of decrease in this feature over the decades between these two mentions. Assuming that PreR-dentalisation existed in Carlisle English in the not too distant past, the present status of (r) after /t, d/ might provide us with information about change processes in this variety. While we can assume that taps were used parallel to PreR-dentalisation (cf. Maguire 2012), taps are hardly found after /t, d/ which replaced the dental sounds. Hence, the data might indicate that the shift from a dental to an alveolar sound in PreR-position had a knock-on effect on the use of [ɾ] and thus the variation in (r). This also means that the decrease in taps could – at least partially – be accounted for by articulatory complexity in CrV position which arose due to a separate sound change in progress, i.e., phonological conditioning led to a comparatively fast focussing of (r) after /t, d/, the process by which the systemic interspeaker variability is reduced. Britain (1997) shows that the Fenland raising, the phonological conditioning of the PRICE vowel, is a stable variation in the different communities in the Fens, whereas a change that is not phonologically conditioned takes longer in order to focus. Jansen (2015a) similarly suggests that taps after /t, d/ rapidly change when the place of articulation moves from dental to alveolar, which leads to a quick focussing of (r) in this specific environment. We might want to ask what the sociolinguistic situation is like in Carlisle. Jansen (2015b) mentions several features that are still described for older speakers in the 1970s but which are no longer part of the Carlisle English inventory in the
128
Sandra Jansen
first decade of the 21st century. Carlisle is a place where (at least) three groups of people with significantly different accents come into contact due to Carlisle’s role as the biggest city in a radius of 100 km: i.e., (West) Cumbrians, often with an accent that still contains many traditional Cumbrian dialect features, people from the western part of Northumberland, and people from the south-west of Scotland. All of these groups have quite distinctive accents. As a local centre, Carlisle attracts many people who work in the city; people come for shopping and Carlisle has for a long time been – and still is – an important junction between north-south and east-west. Carlisle was also a military base for a long time and even though this group is not represented in the city anymore, in the first half of the 20th century the presence of soldiers from around the country added to the complexity of the dialect contact situations. These contact situations are not as easily quantifiable as, for example, group migrations such as in Corby (Dyer 2010), where the number of people of Scottish origin who settled in this community is recorded. However, we can assume that the contact situations described for Carlisle must have had a long-standing effect on the dialect. Indeed, strong levelling has taken place in this community (cf. Jansen 2012) which might have resulted in the linguistic insecurity Montgomery (2006) describes for Carlisle speakers. If levelling is so widespread in this community, the question arises of why [ɾ] has not eroded further in Carlisle. The tap was a feature that was and is shared between the different dialects mentioned above, as it used to be a supraregional feature in the far north of England and it is still a quite frequent feature in Scottish English (cf. Stuart-Smith 2008; Watt, Llamas & Johnson 2014), i.e., the majority of people who came into contact in Carlisle shared this feature and therefore, accommodation toward a less marked feature was not necessary. Hence, dialect contact between speakers of these dialects would probably not have led to the decrease in taps. However, the decrease in this variant in Carlisle English does not imply that variation is decreasing in this variety – on the contrary. Røyneland (2009) explains that decreasing contrast does not mean immediate loss of more local forms, but that local and supraregional forms can co-occur for some time. While we see a decrease in [ɾ] to varying degrees according to environment, in particular in the second half of the 20th century, a more complex picture of variant distribution emerges in the 1990s when the frequency of [ʋ] increases. The number of speakers in the sample who use this variant is still very small with only three of the younger speakers. People in Carlisle seem to pick up the variant later than in Newcastle (cf. Foulkes & Docherty 2001) which suggests that the feature is diffusing, possibly from the north east. At the same time, a similar social
External and internal factors in a levelling process
129
spread as in other varieties is observable with WC speakers favouring this feature (cf. Foulkes & Docherty 2001). On a supraregional scale, the loss of taps corresponds to what Kerswill (2003) sees as a countrywide levelling process. Taps need to be seen as a supralocal feature which is more and more restricted geographically due to levelling. Even though taps used to be a northern norm, speakers seem to abandon this feature now. Foulkes and Docherty (1999: 13–14) discuss the issue of disloyalty: Watt, by way of illustration, summarises the motivation behind ongoing changes in the Newcastle vowel system in terms of younger speakers aiming to ‘dispel the “cloth cap and clogs” image’, and to ‘sound like northerners, but modern northerners’. Speakers can achieve these aims by avoiding variants which they perceive to be particularly indicative of their local roots, while at the same time adopting some features which are perceived to be non-local. It seems to be important, too, that the incoming features do not signal any other particularly well-defined variety, because of the potential signaling of disloyalty to local norms.
The tap, even though geographically widespread in the north of England, belongs to the group of features which is being abandoned in Carlisle English. A possibility would be that speakers in Carlisle either avoid this feature because they want to sound ‘modern’ or they do not associate taps exclusively with northernness, since the feature is also found in Scotland (cf. Stuart-Smith 2008; Watt, Llamas & Johnson 2014). In the case of Carlisle, both factors might play a role. On the other hand, language external factors only seem to play a minor role in this change. While age and social class are significant influencing factors in the change, sex and style do not reach significance in the overall distribution and in the different environments. Working Class speakers retain the more traditional feature longer, but the lack of (statistical) sex differences is slightly surprising. Women are said to use more standard forms in the process of change (cf. Labov 2001) and this is true for this data set to some extent; for example, women use fewer taps in the interviews, but this stratification breaks down in the results for different styles. This lack of clear distinction between styles might have to do with the interview protocol. The first task for the speakers was to read a text passage, which was then followed by an interview. However, since the protocol was followed throughout the fieldwork, it could not explain the varying tendencies of style shifting in each group. This overall lack of strong sex and style distinctions might also have to do with the language-internal trigger of this change. Internal factors seem to play a more important role than external factors. In the overall distribution, the different phonotactic environments are highly significant. Going beyond this level and analysing the use of (r) within the different
130
Sandra Jansen
environments provides us with even more information about the change. As discussed elsewhere (cf. Jansen 2015a), in CrV position in particular, internal constraints are at work. However, this internal change was probably triggered by another external change, the loss of PreR-dentalisation, which had a knockon effect on the loss of taps in CrV position. Articulatory complexity was eased by replacing taps with postalveolar approximants. The SED provides us with the information that there were hardly any in between forms for /t, d/ in CrV so that this phonologically conditioned change seems to have happened fairly quickly. The trajectory of change for (r) in CrV with PreR-position is a good example for the link between externally- and internally-motivated change mentioned by Hickey (2014). He proposes that external motivations always trigger a change; otherwise, the question of why a change occurs at a certain point in time could not be answered. The interplay of language-internal and external motivations is a very complex process but it seems to be the tendency that external factors indeed trigger a change which then leads to related, language-internal changes, e.g., in chain shifts. Looking at Torgersen and Kerswill’s (2004) claim that external factors can overrule language-internal factors in levelling processes, there are two examples in the data which confirm this claim. As discussed beforehand, in VrV position and [θ] in CrV position, internal motivations seem to work against the decline in taps. Maguire (2012: 373) speaks of an R-Realisation Effect which spans across varieties and can be found in Carlisle English as well. However, the use of taps is merely sustained for longer in these environments, as the data analysis has shown. These environments seem to be resisting the change longer than other environments but the articulatory ease of tap production in those two positions does not withstand the pressure to abandon the variant caused by external forces. Hence, the operating external pressures override internal motivations.
6 Conclusion The present chapter discussed factors affecting the levelling of (r) in Carlisle in the 20th century, a process in which the postalveolar approximant replaces the tap as the ‘new’ supraregional variant. In other words, the unmarked variant is spreading at the cost of a more conservative, now regionally restricted variant. It is a complex process that is to some extent related to the decline in [t ]̪ in this variety. The chapter has shown that internal and external factors are crucial for the change from tap to postalveolar approximant, and in the different environments, different forces are at work. Yet external motivation overrides internal motivation, in particular seen in VrV position.
External and internal factors in a levelling process
131
It is not really satisfying that the most crucial point in the argument, i.e., an external factor triggering the internally-motivated change, needs to fall back on indirect evidence. A possible solution would be to investigate more isolated varieties in the north of England, such as Holy Island (cf. Maguire 2014) or Maryport (cf. Jansen, forthcoming). More peripheral areas are said to retain local features longer (cf. Tagliamonte 2012b; Britain 2014). If we can still find PreRdentalisation in these communities, we could improve our understanding of the loss of taps. In addition, the fricative realisation of (r) after alveolar stops needs to be investigated in more detail. Variation seems to exist and an acoustic analysis of this environment will shed more light on this phenomenon.
References Barras, William. 2011. Sociophonology of rhoticity and r-sandhi in East Lancashire English. Edinburgh: University of Edinburgh PhD thesis. Barras, William. Forthcoming. Residual rhoticity and emergent r-sandhi in the north-west and south-west of England: Different approaches to hiatus-resolution? In Natalie Braber & Sandra Jansen (eds.), Sociolinguistics in England. Basingstoke: Palgrave Macmillan. Britain, David. 1997. Dialect contact, focusing and phonological rule complexity: the koineisation of Fenland English. University of Pennsylvania Working Papers in Linguistics 4(1). 141–170. Britain, David. 2005. The dying dialects of England? In Antonio Bertacca (ed.), Historical linguistic studies of spoken English, 35–46. Pisa: Edizioni Plus. Britain, David. 2009. One foot in the grave?: Dialect death, dialect contact and dialect birth in England. International Journal of the Sociology of Language 196/197. 121–155. Britain, David. 2014. Linguistic diffusion and the social heterogeneity of space and mobility. Plenary talk at the International Society for the Linguistics of English (ISLE), University of Zürich, 24–27 August. Coombs, Mike. 1995. The impact of international boundaries on labour market area Definitions. Area 27(1). 46–52. Coupland, Nikolas. 1980. Style shifting in a Cardiff work setting. Language in Society 9(1). 1–12. Dyer, Judy. 2010. Migration, national identity and reallocation of Forms. In Carmen Llamas & Dominic Watt (eds.), Language and identities, 205–216. Edinburgh: Edinburgh University Press. Ellis, Alexander J. 1889. On Early English pronunciation. Available online: http://www.archive. org/details/onearlyenglishpr00elliuoft. Foulkes, Paul. 1997. English [r]-sandhi – a sociolinguistic perspective. Histoire, Epistémologie, Language 19. 73–96. Foulkes, Paul & Gerard J. Docherty (eds.). 1999. Urban voices: Accent studies in the British Isles. London: Edward Arnold. Foulkes, Paul & Gerard J. Docherty. 1999. Urban voices: overview. In Paul Foulkes & Gerard J. Docherty (eds.), Urban voices: Accent studies in the British Isles, 1–24. London: Edward Arnold.
132
Sandra Jansen
Foulkes, Paul & Gerard J. Docherty. 2000. Another chapter in the story of /r/: ‘Labiodental’ variants in British English. Journal of Sociolinguistics 4. 30–59. Foulkes, Paul & Gerard J. Docherty. 2001. Variation and change in British English /r/. In Hans Van de Velde & Roeland van Hout (eds.), ‘r-atics: Sociolinguistics, phonetic and phonological Characteristics of /r/, 27–43. Special Issue of Études et Travaux. Brussels: ILVP. Giegerich, Heinz J. 1999. Lexical strata in English: morphological causes, phonological effects. Cambridge: Cambridge University Press. Hickey, Raymond. 2004. A sound atlas of Irish English. Berlin: Mouton de Gruyter. Hickey, Raymond. 2014. Internally- and externally-motivated language change. In Juan M. Hernández-Campoy & J. Camilo Conde-Silvestre (eds.), The handbook of historical sociolinguistics, 387–407. Malden/Oxford: Wiley Blackwell. Jansen, Sandra. 2012. Variation and change in the Cumbrian city dialect of Carlisle. Essen: University of Duisburg-Essen PhD thesis. Jansen, Sandra. 2015a. A century of change in (r) in Carlisle English. Internal constraints in a levelling process. In Eivind Torgersen, Stian Hårstad, Brit Mæhlum & Unn Røyneland (eds.), Language variation – European perspectives V. Selected papers from the 7th International Conference on Language Variation in Europe (ICLaVE 7), Trondheim, June 2013. Amsterdam/New York: John Benjamins, 129-144. Jansen, Sandra. 2015b. Cumbria and Carlisle. In Raymond Hickey (ed.), Researching northern English 205–225. Amsterdam: John Benjamins. Jansen, Sandra. Forthcoming. A crossdialect view on levelling: (r) in Carlisle and Maryport English. Jauriberry, Thomas, Rudolph Sock & Albert Hamm. 2015. Phonetic variation in Standard Scottish English: Rhotics in Dundee. Proceedings of the International congress of phonetic sciences. Johnson, Daniel E. 2014. Rbrul. Available online: http://www.danielezrajohnson.com/rbrul.html (accessed 5 January 2015). Kerswill, Paul. 2003. Dialect levelling and geographical diffusion in British English. In David Britain & Jenny Cheshire (eds.), Social dialectology: In honour of Peter Trudgill, 223–243. Amsterdam: John Benjamins. Labov, William. 1966/2006. The social stratification of English in New York City. Cambridge: Cambridge University Press. Labov, William. 2001. Principles of linguistic change. Volume II: Social factors. Oxford: Blackwell. Lawson, Eleanor, James M. Scobbie & Jane Stuart-Smith. 2011. The social stratification of tongue shape for postvocalic /r/ in Scottish English. Journal of Sociolinguistics 15. 256–268. Llamas, Carmen. 2001. The sociolinguistic profiling of (r) in Middlesbrough English. In Hans Van de Velde & Roeland van Hout (eds.), ‘r-atics: Sociolinguistics, phonetic and phonological characteristics of /r/, 123–140. Special Issue of Études et Travaux. Brussels: ILVP. MacRaild, Donald. 1998. Culture, conflict and migration: The Irish in Victorian Cumbria. Liverpool: Liverpool University Press. Maguire, Warren. 2008. What is a merger, and can it be reversed? The origin, status and reversal of the ‘NURSE – NORTH merger’ in Tyneside English. Newcastle-upon-Tyne: University of Newcastle PhD thesis. Maguire, Warren. 2012. Pre-R dentalisation in northern England. English Language and Linguistics 16(3). 361–384. Maguire, Warren. 2014. Variation and change in a traditional northern English rural dialect. Plenary talk at the 6th Northern Englishes Workshop (NEW6), University of Lancaster, 16– 17 April.
External and internal factors in a levelling process
133
Marsden, Sharon. 2006. A sociophonetic study of labiodental /r/ in Leeds. Leeds Working Papers in Linguistics and Phonetics 11. 153–172. Montgomery, Chris. 2006. Northern English dialects: a perceptual approach. Sheffield: University of Sheffield PhD thesis. Nagy, Naomi & Patricia Irwin. 2010. Boston (r): Neighbor(r)s nea(r) and fa(r). Language Variation and Change 22. 241–278. Piercy, Caroline. 2007. A quantitative analysis of rhoticity in Dorset: evidence from four locations of an urban to rural hierarchy of change. CamLing 2007. 199–206. Piercy, Caroline. 2012. A transatlantic cross-dialectal comparison of non-prevocalic /r/. Selected Papers from NWAV 40 18(2). 77–86. Røyneland, Unn. 2009. Dialects in Norway: catching up with the rest of Europe? International Journal of the Sociology of Language 196–197. 7–30. Sankoff, Gillian & Hélène Blondeau. 2007. Language change across the life span. /r/ in Montreal French. Language 83(3). 560–588. Shorrocks, Graham. 1998. A Grammar of the dialect of the Bolton area, Part 1: Introduction. Phonology. Frankfurt am Main: Peter Lang. Stuart-Smith, Jane. 2007. A sociophonetic investigation of postvocalic /r/ in Glaswegian adolescents. In Jürgen Trouvain & William J. Barry (eds.), Proceedings of the 16th International Congress of Phonetic Sciences, 1449–1452. Saarbrücken: Universität des Saarlandes. Stuart-Smith, Jane. 2008. Scottish English: Phonology. In Bernd Kortmann & Clive Upton (eds.), Varieties of English. The British Isles, 48–70. Berlin: Mouton de Gruyter. Tagliamonte, Sali. 2012a. Variationist sociolinguistics: Change, observation, interpretation. Malden/Oxford: Wiley-Blackwell. Tagliamonte, Sali. 2012b. The roots of English. Cambridge: Cambridge University Press. Torgersen, Eivind & Paul Kerswill. 2004. Internal and external motivation in phonetic change: Dialect levelling outcomes for an English vowel shift. Journal of Sociolinguistics 8(1). 23–53. Trudgill, Peter. 1986. Dialects in contact. Oxford: Blackwell. Van de Velde, Hans & Roland van Hout (eds.) 2001. r-atics: Sociolinguistic, phonetic and phonological characteristics of /r/. Special issue of Études et Travaux. Van Hout, Roland & Hans van de Velde. 2001. Patterns of /r/ variation. In Hans Van de Velde & Roeland van Hout (eds.) ‘r-atics: Sociolinguistics, phonetic and phonological characteristics of /r/, 1–9. Special Issue of Études et Travaux. Brussels: ILVP. Watt, Dominic, Carmen Llamas & Daniel E. Johnson. 2014. Sociolinguistic variation on the Scottish-English border. In Robert Lawson (ed.), Sociolinguistics in Scotland, 79–102. Basingstoke: Palgrave. Wells, John C. 1982. Accents of English. Cambridge: Cambridge University Press. Wiese, Richard. 2001. The unity and variation of (German) /r/. In Hans Van de Velde & Roeland van Hout (eds.), ‘r-atics: Sociolinguistics, phonetic and phonological characteristics of /r/ 11–26. Special Issue of Études et Travaux. Brussels: ILVP. Wright, Peter. 1977 (1978 edn.). How they talk in Carlisle. The Journal of the Lakeland Dialect Society. 7–15.
Marten Juskan
7 Scouse NURSE and northern happY : vowel change in Liverpool English Abstract: This chapter investigates change in the NURSE and happY vowels in Liverpool English (Scouse) across 3 generations of speakers and discusses if and how the results might be connected to questions of salience, local identity, and Liverpool’s changing fortunes in the 20th and the 21st century. Based on a sample of 20 sociolinguistic interviews, this study finds that younger speakers use more local variants of the NURSE-SQUARE merger, a highly salient variable (Honeybone and Watson 2013, Watson and Clark 2013), than their parents’ or grandparents’ generation. Realisations of happY, on the other hand, become laxer, which is a change away from the (tense) traditional local norm, and towards the majority of the other varieties spoken in northern England (Trudgill 1999). These changes in production are linked up with qualitative data from the interviews, which indicate that younger Liverpudlians not only readily express pride in their city and its accent, but that they also feel a strong connection to the north of England more generally. Phonetic change in the two vowels under scrutiny is interpreted as being governed by a combination of salience and questions of identity: younger speakers use Scouse variants of the socially salient NURSE vowel to express their ‘primary’ identity as Liverpudlians, and laxer realisations of less-salient happY to also associate themselves with other towns and cities in the North – a strategy which allows them to simultaneously express both their local, and their regional identity linguistically.
1 Introduction Scouse1, the variety of English spoken in and around the city of Liverpool, is not only one of the most widely known accents in the UK (Trudgill 1999), but also one of the most heavily stigmatised ones (Montgomery 2007). To a degree at least, this stigmatisation can surely be linked to the economic and social changes Note: This chapter is a revised excerpt taken from my PhD dissertation. The full analysis can be found in Juskan (2016). 1 presumably from lobscouse, a sailor’s dish Marten Juskan, University of Freiburg DOI 10.1515/9783110450903-007
136
Marten Juskan
that the city went through in the second half of the last century. After a brief post-war revival and world-wide fame and popularity thanks to the Beatles and other bands of the Merseybeat era, Liverpool went into economic and physical decline. In the 1970s and 80s, the city (and its accent) had become associated primarily with mass unemployment, poverty, and crime (Belchem 2006a). In the last two decades negative external perceptions seem to have persisted despite the fact that Liverpool has experienced a considerable degree of urban regeneration. Against this backdrop, one could expect Liverpool English to be a prime candidate for dialect levelling, a process which has been amply documented in the UK and by which regional accents lose distinctive features and become more similar to the standard or surrounding non-standard varieties (Kerswill 2003). However, this does not seem to be the case. Watson (2007) has investigated the production of a number of Liverpool features and found that, in most cases, young speakers either retain traditional variants or even use them more frequently than older Liverpudlians. This led him to conclude that, far from levelling out, Scouse is actually “getting Scouser” (Watson 2007: 237). The ‘why’ is, however, still open to debate. As potential explanations for his results Watson mentions covert prestige, salience, and the fact that all his young speakers had a working class background. He also stresses that further research is needed to answer this question. Juskan (2015) has already found that not all local features of Liverpool English are necessarily used more frequently by younger speakers. The apparent time study discussed in this chapter provides further data taken from a socially more balanced sample and concerning two additional variables not investigated by Watson (2007): happY-tensing and the NURSE - SQUARE merger. The former pertains to the final vowel in words like city, or baby. With respect to RP, Harrington (2006: 441) writes that in the 1950s, the vowel used in this position was “phonetically closer to [ɪ] in KIT than to [i:] in FLEECE ”, i.e. happy was pronounced [hæpɪ], not [hæpi]. In the late 20th century, however, happY has undergone tensing in RP. The phonetic realisation is now [i] for most speakers, and dictionaries generally use /i/ to represent this vowel. HappYtensing has now spread to most parts of England, with the exception of “[t]he Central North, Central Lancashire, Northwest Midlands and Central Midlands areas”, where the older pronunciation [ɪ] is still retained. Only the port cities Liverpool, Hull and Newcastle are notable exceptions to this (cf. Trudgill 1999: 62). Beal (2000: 494) argues that happY-tensing is “neither so recent, nor so southern” as implied by the traditional account and provides evidence that happY-tensing was present in Tyneside English as early as 1775. However, the history of the feature is only of secondary interest for this chapter and there is no argument about the present-day situation: Liverpool, or rather the whole of
Scouse NURSE and northern happY: vowel change in Liverpool English
137
Merseyside (and parts of Cheshire) is “an ‘ee’-pronouncing island surrounded by a sea of accents which do not (yet) have this feature” (Trudgill 1999: 72). Notwithstanding its usefulness as a feature that distinguishes Liverpool English from surrounding non-standard accents, happY-tensing seems to have low salience and is not the subject of comments about Scouse in Liverpool – possibly because it does not diverge from the modern standard. Salience of the NURSE - SQUARE merger, on the other hand, is rather high. For many speakers in the city words such as fair and fur are (near-)homophones (cf. Trudgill 1999: 72), with [ɛ:] usually being the target of the merger (cf. Honeybone 2007: 127). While not unique to Liverpool, this merger is rare enough to be generally perceived as one of the most characteristic (or even defining) and most salient features of Scouse (cf. Trudgill 1999: 73). The social salience of this variable has been corroborated by research on humorous dialect literature (Honeybone and Watson 2013) and perception tests with listeners from inside and outside of Liverpool (Watson and Clark 2013; Juskan 2016). Since SQUARE is usually the (more) stable element of the merger, analysis will focus primarily on NURSE . In the remainder of this chapter I will investigate production of these two variables in three generations of speakers and I will try to give some tentative explanations of the changes that seem to be happening.
2 Method 2.1 Interview structure Production data were obtained through one-on-one interviews conducted by the author. They consisted of a free speech section where subjects were asked a number of questions about the area of the city they grew up in, changes in the city, football and other sports, Liverpool’s image in the UK and the rivalry with Manchester. Furthermore, subjects were questioned about their understanding of a number of identity labels (Scouser, Liverpudlian, Northerner. . .). They also indicated which of these labels they would use with respect to themselves. Towards the end of the interview, participants read out a text and a word list. Next, subjects were asked to read out the reading passage a second time using their strongest Scouse accent. In the present study, these different registers will only be referred to towards the end of section 3.2. Finally, subjects were asked a number of questions concerning perceived change in Scouse, typical features, and attitudes towards the variety (these meta-linguistic comments are briefly summarised in section 3.3). Interviews lasted between 50 and 60 minutes and
138
Marten Juskan
were conducted in September/October 2012 and April/May 2013. All interviews were recorded using a Roland Edirol R-09HR MP3/Wave recorder.
2.2 Participants A total of 20 subjects were interviewed and provided almost 19 hours of recorded material. All participants were born and/or had grown up in the Liverpool Urban Area since age 12 or younger. Both men and women were interviewed and a rough socio-economic distinction into working class or middle class was made based on people’s occupation and level of education. All participants were White British and monolingual speakers of English. The age range was 19–84, with people being classified as belonging to one of three age groups (19–29, 30–55, and 56–85) to mirror social, economic, and cultural change in Liverpool. With the boundaries set as they are the formative years (roughly up to and including the 20s) of most of the participants in the respective group fall together with one of the three phases of the city’s development in the latter half of the 20th century: 50s and 60s (Merseybeat era) for the oldest, 70s and 80s (economic depression) for the middle-aged, and 90s and 2000s (regeneration) for the youngest speakers. Subjects came from a number of different districts of Liverpool, but all age groups are more or less evenly spread across the city. Table 7.1 shows how participants are distributed across the categories outlined above. Table 7.1: number of speakers by age, gender, and social class 19–29
WC MC
30–55
56–85
F
M
F
M
F
M
2 2
2 2
2 2
2 2
1 1
1 1
2.3 Measurement and statistics All interviews were transcribed orthographically in Praat (Boersma and Weenink 2015), followed by manual marking of relevant vowel segments. The first two formants were then measured automatically at the midpoints of vowels using a Praat script written by Mietta Lennes2 and modified by the author.
2 http://www.helsinki.fi/~lennes/praat-scripts/, last accessed 29/01/2013
Scouse NURSE and northern happY: vowel change in Liverpool English
139
In addition to the three test variables happY, NURSE , and SQUARE (of which all instances were included), between 10 and 25 tokens each of FLEECE and TRAP were also measured as input for normalisation. These were taken from the reading passage and word list sections of the interviews since these contexts were considered most likely to produce the most ‘extreme’ realisations (in terms of the periphery of speakers’ vowel spaces). FLEECE measurements were additionally included in the calculation of Pillai scores for happY. Pillai scores (cf. Hall-Lew 2010) are a measure of overlap of distributions, with values ranging from 0 (complete overlap) to 1 (perfectly distinct). The Watt and Fabricius (2002) algorithm was used to normalise formant measurements. All models and figures reported are based on these normalised F1 and F2 values (coded as “F1W” and “F2W” respectively). All statistical tests were performed using the R software (R Core Team 2015). Mixed linear effects models (including carrier word as a random intercept) were computed with the help of ‘lmerTest’ (Kuznetsova, Bruun Brockhoff, and Haubo Bojesen Christensen 2015), an R package which builds on ‘lme4’ (Bates et al. 2015), but adds p-values calculated on the basis of F statistics, with degrees of freedom derived from Satterthwaite’s approximation. Sum coding, instead of R’s default treatment coding, was used for all regressions. As a replacement for the goodness-of-fit measures known from linear regression models this chapter reports the R2 of a linear model that regresses the observed values on the fitted ones from the linear mixed effects model (cf. r-sig-mixed-models mailing list 2015).
3 Results In total, 7950 vowels were measured. 4565 of these are observations of happY, 1770 concern NURSE , and 882 SQUARE (the remainder are instances of TRAP and FLEECE ). The two dependent variables F1 and F2 are analysed side by side, since it is quite possible that change manifests itself with respect to one dimension but not the other (cf. Harrington 2006). I will also briefly look at the two dimensions in conjunction, largely on the basis of F1-F2 vowel plots which are familiar from many works on phonetics and phonology. All figures (except the F1-F2 plots just mentioned) are designed in such a way that higher values on the y-axis represent ‘more local’ variants, i.e. realisations that diverge from those of the neighbouring varieties (and from RP in the case of NURSE ). When present, p-values in the boxplots (rounded to three decimal places) correspond to t-tests comparing the oldest to the middle-aged,
140
Marten Juskan
and the middle-aged to the youngest group. The value that is horizontally centred is always taken from the t-test comparing means in the oldest and the youngest group. The notches in the boxplots correspond to confidence intervals of the medians (thick horizontal bars), arithmetic means are marked by black dots. All plots were generated with ‘ggplot2’ (Wickham 2009).
3.1 happY The minimal adequate models for F1 and F2 of happY are reprinted as Table 7.2. Table 7.2: happY : mixed linear effects regression models (extracts) F1