Variationist Sociolinguistics E-Book

Sali A. Tagliamonte

• Herausgeber: John Wiley & Sons
• Kategorie: Geisteswissenschaft
• Serie: Language in Society
• Sprache: Englisch

Seitenzahl: 850

Veröffentlichungsjahr: 2011

Beschreibung

Variationist Sociolinguistics: Change, Observation, Interpretation presents a comprehensive, intermediate level examination of Language Variation and Change, the branch of sociolinguistics concerned with linguistic variation in spoken and written language.

• Represents the most up-to-date coverage of the history, developments, and methodologies of variationist sociolinguistics
• Addresses all aspects of linguistic variation, including areas not usually covered in introductory texts, e.g. the phonological, morpho-syntactic, discourse/pragmatic
• Outlines comparative sociolinguistic approach, data collection, methodological issues; and addresses state-of-the-art contemporary quantitative methods and statistical practice
• Features cutting-edge research at an appropriate level to facilitate student learning
• Engages students throughout with a variety of pedagogical features, including Mini Quizzes to test comprehension, extensive Exercises at the end of each chapter, the opportunity to do hands-on quantitative analysis of a never-before published data set, and Notes and Tips that offer insight into conducting sociolinguistic research. Extra materials and answers to the exercises are available at www.wiley.com/go/tagliamonte

Leseprobe

Contents

Acknowledgments

Foreword

Series Editor’s Preface

Preface

What this Book is About

Figures

Tables

1 Sociolinguistics as Language Variation and Change

Sociolinguistics

The Linguistic Variable

Linguistic Change

The Principle of Accountability

Circumscribing the Variable Context

Evolution of the Linguistic Variable

The Importance of Accountability

Language Variation and Change and Linguistic Theory

Exercises

2 Social Patterns

Social Class

Sex (or Gender)

Style and Register

Mobility in Space and Mobility in Class

Social Network, Communities of Practice

Ethnicity and Culture

The Mass Media

Age

Types of Change

Principles of Linguistic Change

Summary

Exercises

3 Linguistic Patterns

Sound Change

Morphological Change

Syntactic Change

Semantic Change

Grammaticalization

Lexical Effects

Exemplar Theory

Exercises

4 Data and Method

The Speech Community

Corpus Building

Creating Sociolinguistic Corpora

The Individual and the Group

Constructing an LVC Study

Research Ethics

The Gold – Your Data

The Real World

5 Quantitative Analysis

The Quantitative Paradigm

Distributional Analysis

Statistical Modeling

The Three Lines of Evidence

The Case Study – Variable (that)

Goldvarb Logistic Regression

Challenging the Variable Rule Program

Drawbacks to the Variable Rule Program

New Toolkits for Variationist Sociolinguistics

Summary

Exercises

6 Comparative Sociolinguistics

Comparison

The Comparative Method

Comparison in Origins

Comparison in Language Contact

Standards for Comparison

Variable (did)

Exercises

7 Phonological Variables

Variable (t,d)

Variable (ing)

Tips for Studying Phonological Variables

Exercises

8 Morpho-Syntactic Variables

Verbal (s)

Adverb (ly)

Modal (have to)

Studying Morpho-Syntactic Variables

Exercises

9 Discourse/Pragmatic Features

Quotative (be like)

General Extenders

Studying the Discourse/Pragmatic Variable

Exercises

10 Tense/Aspect Variables

Grammaticalization and Tense/Aspect Variables

Future (going to)

Perfect (have)

Studying Tense/Aspect Variables

Exercises

11 Other Variables

Variable (come)

Variable (Intensifiers)

Language and the Internet

Studying Unusual Variables

Exercises

12 Sociolinguistic Explanations

What Are the Constraints on Change?

How Does Language Change?

How Is a Change Embedded in Social and Linguistic Systems?

Evaluation of a Change

Statistical Modeling

Traditional Explanations

The Principle of Interaction

Appendix A Corpora Cited

Appendix B Time Periods in the History of English

References

Subject Index

Index of Linguistic Variables

Language in Society

GENERAL EDITOR

Peter Trudgill, Chair of English Linguistics, University of Fribourg

ADVISORY EDITORS

J. K. Chambers, Professor of Linguistics, University of Toronto

Ralph Fasold, Professor of Linguistics, Georgetown University

William Labov, Professor of Linguistics, University of Pennsylvania

Lesley Milroy, Professor of Linguistics, University of Michigan, Ann Arbor

Launched in 1980, Language in Society is now established as probably the premiere series in the broad field of sociolinguistics, dialectology, and variation studies. The series includes both textbooks and monographs by Ralph Fasold, Suzanne Romaine, Peter Trudgill, Lesley Milroy, Michael Stubbs, and other leading researchers.

1 Language and Social Psychology, edited by Howard Giles and Robert N. St Clair

2 Language and Social Networks (second edition), Lesley Milroy

3 The Ethnography of Communication (third edition), Muriel Saville-Troike

4 Discourse Analysis, Michael Stubbs

5 The Sociolinguistics of Society: Introduction to Sociolinguistics, Volume I, Ralph Fasold

6 The Sociolinguistics of Language: Introduction to Sociolinguistics, Volume II, Ralph Fasold

7 The Language of Children and Adolescents: The Acquisition of Communicative Competence, Suzanne Romaine

8 Language, the Sexes and Society, Philip M. Smith

9 The Language of Advertising, Torben Vestergaard and Kim Schrøder

10 Dialects in Contact, Peter Trudgill

11 Pidgin and Creole Linguistics, Peter Mühlhäusler

12 Observing and Analysing Natural Language: A Critical Account of Sociolinguistic Method, Lesley Milroy

13 Bilingualism (second edition), Suzanne Romaine

14 Sociolinguistics and Second Language Acquisition, Dennis R. Preston

15 Pronouns and People: The Linguistic Construction of Social and Personal Identity, Peter Mühlhäusler and Rom Harré

16 Politically Speaking, John Wilson

17 The Language of the News Media, Allan Bell

18 Language, Society and the Elderly: Discourse, Identity and Ageing, Nikolas Coupland, Justine Coupland, and Howard Giles

19 Linguistic Variation and Change, James Milroy

20 Principles of Linguistic Change, Volume I: Internal Factors, William Labov

21 Intercultural Communication: A Discourse Approach (second edition), Ron Scollon and Suzanne Wong Scollon

22 Sociolinguistic Theory: Language Variation and Its Social Significance (second edition), J. K. Chambers

23 Text and Corpus Analysis: Computer-assisted Studies of Language and Culture, Michael Stubbs

24 Anthropological Linguistics, William Foley

25 American English: Dialects and Variation (second edition), Walt Wolfram and Natalie Schilling-Estes

26 African American Vernacular English: Features, Evolution, Educational Implications, John R. Rickford

27 Linguistic Variation as Social Practice: The Linguistic Construction of Identity in Belten High, Penelope Eckert

28 The English History of African American English, edited by Shana Poplack

29 Principles of Linguistic Change, Volume II: Social Factors, William Labov

30 African American English in the Diaspora, Shana Poplack and Sali Tagliamonte

31 The Development of African American English, Walt Wolfram and Erik R. Thomas

32 Forensic Linguistics: An Introduction to Language in the Justice System, John Gibbons

33 An Introduction to Contact Linguistics, Donald Winford

34 Sociolinguistics: Method and Interpretation, Lesley Milroy and Matthew Gordon

35 Text, Context, Pretext: Critical Issues in Discourse Analysis, H. G. Widdowson

36 Clinical Sociolinguistics, Martin J. Ball

37 Conversation Analysis: An Introduction, Jack Sidnell

38 Talk in Action: Interactions, Identities, and Institutions, John Heritage and Steven Clayman

39 Principles of Linguistic Change, Volume III: Cognitive and Cultural Factors, William Labov

40 Variationist Sociolinguistics: Change, Observation, Interpretation, Sali A. Tagliamonte

This edition first published 2012© 2012 Sali A. Tagliamonte

Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell.

Registered OfficeJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Offices350 Main Street, Malden, MA 02148-5020, USA9600 Garsington Road, Oxford, OX4 2DQ, UKThe Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell.

The right of Sali A. Tagliamonte to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Library of Congress Cataloging-in-Publication Data

Tagliamonte, Sali.Variationist sociolinguistics : change, observation, interpretation / Sali A. Tagliamonte.p. cm.Includes bibliographical references and index.

ISBN 978-1-4051-3590-0 (alk. paper) – ISBN 978-1-4051-3591-7 (pbk. : alk. paper)1. Language and languages–Variation. 2. Sociolinguistics–Research. I. Title.P120.V37T348 2012306.44–dc22

2011010578

A catalogue record for this book is available from the British Library.

This book is published in the following electronic formats: ePDFs 9781444344448; Wiley Online Library 9781444344479; ePub 9781444344455; mobi: 9781444344462.

For

Anna Blanche Lawson

1930–2001

Love you forever Mum,

Sali

Acknowledgments

A person only ever stands somewhere along the ladder of life. I am indebted to many great minds and generous spirits who have helped me in my work. My students, my mentors, my colleagues, my friends, they are often the same people with no clear distinction among them. This is one of the truly gratifying aspects of doing sociolinguistics – you become part of a social network, a practice, a community.

My students have always been my best critics. Let them know that each one of them has helped immeasurably with this book. Derek Denis, Bridget Jankowski, Dylan Uscher, and Cathleen Waters: every question we considered over the past few years has made its way into these pages. My students in LIN1256, Advanced Language Variation and Change, January–April 2011 deserve special mention for their critical input to the prepublication version of the manuscript. Marisa Brook, Julian Brooke, Matthew Gardner, Heidi Haefale, Chris Harvey, Madeline Shellgren, and Jim Smith have shown me, yet again, how much teaching embeds learning.

My past has also woven its way through the chapters, as I have returned to my early research to integrate the present state of the field with its foundations. I am blessed by having been mentored by some of the greatest contributors to the field. Shana Poplack, David Sankoff, Jack Chambers, Peter Trudgill, and Jenny Cheshire: this book exists only because I have been able to stand on your shoulders. I am also lucky to have had a knowledgeable and attentive set of critics who scoured the draft manuscript and offered their insights, including four anonymous Wiley-Blackwell reviewers, a savvy team of Wiley editors, my new neighbour Victor Kuperman, my pal Paul Foulkes, and even the General Editor, Peter Trudgill, himself.

No field advances without change. Over the last ten years statistical methods have undergone a veritable renaissance. Chapter 5 evolved over several years of consultation on the state of the art in statistical methods in Variationist Sociolinguistics. I am thankful to Harald Baayen, Daniel Ezra Johnson, and John Paolillo for helping me in my ongoing efforts to model linguistic variation and change in ways that are not only insightful, but also statistically sound.

A sociolinguist is never alone in their research. I am lucky to have had a superb group of lab assistants and project coordinators. The latest, Michael Ritter, has the astonishing ability to manage, organize, interview, transcribe, extract, correct, code, copy, copy-edit, run Goldvarb, R, and Ant-Conc, and everything else I need doing.

I am immensely proud of my own academic progeny who have become my friends; Jen Smith and Alex D’Arcy, where would I be without your ongoing collaborations, savvy insights, and unabashed prodding? My wonderful colleagues Elizabeth Cowper, Elaine Gold, Alana Johns, Keren Rice, and Diane Massam: you have been exuberance, friendship, and community to me since I arrived at UofT in 2001. I have found in your model guidance and sanity. My confidant and best drinking buddy Anthony Warner listens and advises and tells me when I am being silly (this is a more important quality than you might think).

Since my last book, three of my children have become teenagers, the youngest one has started primary school, and I have gained a stepson in the early years of his professional life. This is a great learning ground for a sociolinguist. Dazzian, Freya, Shaman, Tara, and Adrian have taught me much of what I know about age grading, innovation, and incrementation. I am so very thankful to be part of the perpetual state of variation and change, love and commotion we live in. And to Duncan, who is the bedrock of my life, I am eternally grateful to have found in one man, husband, lover, gardener, and friend.

Finally, I would like to thank my mom. What I have been able to accomplish in my life was fostered in the love and support and many other intangible gifts she gave me.

Foreword

My grandparents lived in a small town in Southern Ontario. It was a farming hamlet in one of the oldest settled areas of Ontario, Canada, called Maple Station. They owned the general store, gas station, and post office. The store was always filled with locals. When I visited as a child, I would race to the store every time someone came in, trailing behind the adults to eavesdrop on the conversations. In the evenings, my great-aunts and -uncles would visit. Coming from farming stock, the families were huge. My grandfather had eight brothers and sisters and my grandmother had nine. There were people around all the time. They often talked long into the evening, playing Euchre or Crib. I can still hear the lilting cadence of those voices in my mind. This was a world of regularized past participles, double negation, all kinds of variation in vowels and diphthongs, and strange words and expressions. Little did I know of all that then! At the time, I only listened and marveled at how different they sounded.

My mother, who had grown up in that world, became a teacher, a specialist in early childhood education. Yet there were always aspects of her speech that were very different from the Canadian norms in my surroundings. When she talked to my grandparents or my aunts and uncles (her brother and sisters) on the phone, her voice would shift back toward the speech patterns I heard in Maple Station. Sometimes, when telling stories, I would even hear her use the occasional I says or He come. And when she quoted anyone in her family, her voice always changed.

While I sound just like any other Canadian, there are still parts of my speech that reflect my mother’s vernacular, words like “wee” for “small”, expressions like “it’s a good job” for “it’s a good thing”. Even today, when my children make fun of some of the words I use and my pronunciations (“tiger” [tægr], “Saturday” [sɛrde:], “southern” [sʌwðrn]) I blame my old-line Southern Ontario roots.

These are the realities of language variation and change. Our life histories are a study of continuities and changes, of ancestry and origins, of time and space, of uncommon similarities across time and remarkable differences across generation gaps. Our heritage follows us wherever we go and throughout our lives. For me, the world came alive when I discovered sociolinguistics because it made my experience make sense. The linguistic difference and variety around me had regularity and meaning, system and explanation. May this book make sociolinguistics – and the world of variation around you – more comprehensible to you.

Sali A. TagliamonteToronto, Ontario

Series Editor’s Preface

It is not often that one looks at a book and says “this is it.” That, however, is precisely what I found myself thinking when I first received the text of Sali Tagliamonte’s Sociolinguistics: Variation, Change and Interpretation. This really isit – this is the book that linguistic variation theory has been waiting for. It has not, however, been waiting too long. Now is exactly the right time for this book to appear; and, I like to think, the Blackwell’s Language in Society series is exactly the right place for it to appear. The study of “Language Variation and Change” (LVC) has been with us now, as Professor Tagliamonte says, for about 40 years. My own first encounter with the field, at that time still without a name, was at what I believe to be the first ever academic meeting devoted to the topic, the Colloquium on New Ways of Analysing Variation in English held at Georgetown University in the USA in October 1972. This turned out to be the first of a series of annual NWAV conferences which continue to be held to this day – though without the word “English” in the title now – and indeed at the time of this writing, the next meeting is going to be the fortieth. I don’t know what Sali was doing in October 1972, but she was certainly not nearly old enough to be at the meeting. In spite of her comparative youth, however, we are very grateful that it has fallen to her to produce in this book a distillation of all the advances that have been made and all the wisdom that has accrued in our now mature field over the last four decades. She is perhaps uniquely qualified – in terms of her erudition, her field-work experience, her analytical innovations, and the large amounts of data and the wide range of language varieties she has worked on – to write the first book which is truly an introduction to LVC, a summary of its main goals and achievements, and a springboard for future progress. She has done this, moreover, in a masterly fashion: not only will the reader of this volume learn how work in LVC is done, they will also learn why we do it, and what the benefits are. All languages are variable – variability is an essential component of human language. But it is only in the last 40 years that we have fully understood the degree to which this is so, have investigated the patterning in which variation is involved, and have developed the concepts and techniques for dealing with it – developments which Sali herself has played a very major role in advancing. As this book shows, any linguistic work which attempts to shed light on the nature of the human language faculty and on the nature of linguistic change, without taking account of language’s inherent variability, will inevitably fall short.

Peter Trudgill

Figures

1.1Rates of avoir and être usage with “tomber” per thousand lines of transcription2.1Idealized pattern for sharp stratification by social class2.2Idealized pattern for gradient stratification by social class2.3Idealized pattern for stratification by social class and style – indicator2.4Idealized pattern for stratification by social class and style – marker2.5Curvilinear pattern for social class when change originates from the middle class2.6Idealized pattern of stratification by sex and social class2.7Idealized pattern of female-led linguistic change2.8Frequency of phonological and grammatical variables2.9S-curve of linguistic change2.10An idealized pattern of linguistic change in progress (generational change)2.11Overall distribution of quotatives by age in Toronto English, c. 2002–20042.12An idealized pattern of age-graded change2.13An idealized pattern of the adolescent peak2.14Ch-lenition in Panama c. 1969 and 1982–842.15Pattern of a feature increasing in use over 60 years in real time for Jane Doe and apparent time for the speech community2.16An idealized pattern of a stable linguistic variable2.17An idealized pattern of linguistic change from across the branches of the family tree, i.e. from outside the community2.18Distribution of variable (h), York English, c. 19972.19Distribution of variable (that), Toronto, c. 2003–20042.20Distribution of [f], [t], and Ø variants as opposed to [θ]2.21Distribution of non-RP variants for three linguistic variables in Norwich English, c. 19723.1Idealized pattern for sound change via weakening3.2Idealized pattern for morphological change via analogical extension – leveling3.3Idealized change in progress that exhibits the Constant Rate Effect3.4Idealized functional effect3.5Idealized pattern of a grammaticalizing feature according to a relevant linguistic context3.6Three variants in apparent time in Texas, USA, c. 1980s3.7Frequency of have got for possession by nature of the complement in real time3.8Frequency of have to for deontic modality by verb type in apparent time in York English4.1Distribution of intensifiers by speaker sex, Toronto, c. 2003–20044.2Distribution of intensifiers by individual males in Toronto, c. 2003–20044.3Distribution of intensifiers by individual females in Toronto, c. 2003–20044.4Distribution of zero copula in real time by year and month – Shaman5.1R, random forest, linguistic, and social factors, all verbs5.2R, conditional inference tree, social factors7.1Constraint ranking of morphological categories on variable (t,d)7.2Overall distribution of simplified clusters of variable (t,d) in York, UK (c. 1997) and Toronto, Canada (c. 2003–2004)7.3Distribution of simplified clusters for variable (t,d) by gender and age in Toronto, Canada 1847.4Distribution of simplified clusters for variable (t,d) by gender and age in York, UK7.5Distribution of simplified clusters for variable (t,d) by education in Toronto, Canada7.6Distribution of simplified clusters for variable (t,d) by education in York, UK7.7Comparison of frequency of simplified clusters by grammatical category across communities7.8Overall distribution of alveolar variants of variable (ing) in York, UK (c. 1997) and Toronto, Canada (c. 2003–2004)7.9Distribution of alveolar variants for variable (ing) by gender and age in Toronto, Canada7.10Distribution of alveolar variants for variable (ing) by gender and age in York, UK7.11Distribution of alveolar variants for variable (t,d) by education in Toronto, Canada7.12Distribution of alveolar variants for variable (ing) by education in York, UK7.13Comparison of frequency of alveolar variants of variable (ing) by grammatical category across communities7.14Pattern of alveolar variants of variable (ing) among nouns compared with indefinite pronouns in York and Toronto7.15Distribution of simplified clusters for just and all other contexts by following phonological segment7.16Distribution of simplified clusters in apparent time in Toronto English8.1Constraint ranking for verbal -s in pronouns vs. NP contexts8.2Distribution of variable (s) by grammatical person, Devon and Samaná8.3Inter-variety comparison of the type of subject constraint8.4Inter-variety comparison of the subject type constraint, including northern Englishes8.5Overall distribution of -Ø adverbs in the history of English8.6Distribution of zero adverbs in York English by age8.7Distribution of zero adverbs in York English by age, sex, and education 2258.8Distribution of zero variants by adverb semantics and age in York English8.9Overall distribution of deontic modal forms across dialects8.10Distribution forms for deontic modality in contexts of subjective obligation by generation in York8.11Distribution of deontic have to in Toronto, Canada and York, UK in apparent time8.12Factor weights for the probability of deontic have to by type of modality across varieties, York, Buckie, Wheatley Hill, and Toronto9.1Distribution of quotatives in Canadian English, c. 1995 and c. 2002–20039.2Overall distribution of quotatives across the generations in Toronto English9.3Cross-study comparison of GE frequency per 10 000 words9.4Test of decategorization of and things like that9.5Distribution of GE types in apparent time9.6Comparison of token count and proportion count for like9.7Distribution of quotative like in real time (Clara) and in apparent time (all 9–19 year olds in Toronto)10.1Pathway for grammaticization of going to10.2Distribution of the major future variants in each of the communities10.3 Distribution of future variants in York, UK, c. 1997 by generation10.4Hierarchy of constraints for semantic function across varieties11.1Use of past reference come11.2Distribution of past reference come by age and sex11.3Distribution of major intensifiers in York in apparent time11.4Distribution of intensifier so by sex in Friends11.5Distribution of major intensifiers in Toronto in apparent time11.6Distribution of so and pretty by sex of the speaker in apparent time, Toronto11.7Delexicalization process11.8Distribution of so by adjective type across generations in Toronto11.9Distribution of so by emotional value of the adjective11.10Distribution of laughter variants among adolescents in IM, c. 2004–2006

Tables

1.1Count of all quotative types with be like as a quotative1.2Distribution of be like according to type of quotative, i.e. viewed as a proportion of the total of each type2.1Difficulty of acquisition of linguistic variables2.2The gender paradox3.1Leveled paradigm for past tense “to be”3.2Distribution of zero marking of tense on stative verbs across varieties3.3Frequency of unmarked verbs in Class I across varieties3.4Bybee’s model for types of change underlying lexical diffusion4.1York English Corpus (c. 1997)4.2Sampling strategy for Toronto neighborhoods4.3Toronto English Archive of Spoken Materials (c. 2003–2010)5.1Explanations for that/Ø variability5.2Logistic regression of the linguistic factors conditioning zero complementizers in York English5.3Logistic regression of the social factors conditioning zero in York English5.4Non-orthogonal factor groups. Worst case scenario5.5Non-orthogonal factor groups. Likely case scenario5.6Idealized logistic regression showing nonindependence of factor groups5.7Rbrul modeling menu5.8Rbrul, mixed effects model, individual random, age continuous5.9R, comparison of marginals for the zero complementizer5.10R, mixed effects model, individual random, age as continuous5.11Test of interaction between education and occupation5.12Comparison of calculations for matrix verb as a 4-way categorical factor group5.13Five logistic regression analyses of the contribution of factors selected as significant to the probability that strong verbs will surface as stems (all factor groups selected as significant)5.14Wald statistics for prepositional dative of the verb “give” – New Zealand English6.1Standards for comparison6.2Logistic regression of internal factors contributing to the probability of did in past habitual contexts in Somerset and Samaná6.3Comparison of similarities and differences in internal linguistic features in Samaná and Somerset6.4Logistic regression analyses of nonstandard marking on Ukrainian and English origin nouns in monolingual Ukrainian conversations6.5Comparison of constraints on absence of free N/pronoun objects in Tamambo and Bislama7.1Distribution of simplified clusters for just in comparison with grammatical categories in York, UK and Toronto, Canada7.2Distribution of simplified clusters in just and other preceding [s] words7.3Distribution of simplified clusters in just compared to all other contexts for following phonological context7.4Logistic regression analyses of the contribution of social factors selected as significant to the probability of [ʔ]8.1Variable -s in Tiverton, southwest England8.2Outline of the development of deontic modality variants8.3Logistic regression analysis for nonstandard “be” (is/was) in nonexistentials in Early New Zealand English9.1Predictions for increasing grammaticalization of be like9.2Overall distribution of quotative verbs, university students9.3Logistic regression of the contribution of internal and external factors to the probability of be like vs. all other quotative verbs. Factor weights not selected as significant in square brackets9.4Overall distribution of quotative verbs, Canadian youth, 9–19 years of age, c. 2002–20039.5Logistic regression of the contribution of internal and external factors to the probability of be like, Canadian English, Toronto, Canada, 2003–2004, 9–39 years of age9.6Contribution of external and internal factors on the use of be like in Toronto English, 17–39 years of age, 2003–2004 reordered9.7Timeline of earliest attestation of general extenders (OED)9.8Overall distribution of GEs and fixed expressions in Toronto, c. 2003–20049.9Summary of findings for distributional tests of grammaticalization9.10Four independent logistic regression analyses of the main GEs in Toronto English9.11Variable rule analysis of the contribution of speaker sex to the probability of different quotatives10.1Logistic regression analysis of going to in five North American varieties. Factor groups not significant in square brackets10.2Logistic regression analysis of going to in York English. Factor groups not significant in square brackets10.3Overall distribution of surface forms in past temporal reference in Samaná English10.4Overall distribution of surface forms in present perfect contexts in Samaná English10.5Overall distribution of surface forms found in present perfect contexts across corpora10.6Three logistic regression analyses showing the results for discourse position in habitual past contexts where all three forms are possible11.1Logistic regression analyses of the contribution of factors selected as significant to the probability of past reference come11.2Frequency of intensifiers in York, UK, c. 1997 (N ≥ 10)11.3Three logistic regression analyses of the contribution of factors to the probability of really in York English11.4Frequency of intensifiers in Friends, 20 year olds, USA, c. 1990s11.5Frequency of intensifiers in Toronto, c. 2003–2004 (N ≥ 10)11.6Rough overview of intensifiers in the history of English11.7Distribution of characteristic IM forms among Canadian teenagers, c. 2004–2006

1

Sociolinguistics as Language Variation and Change

Not all variability and heterogeneity in language structure involves change; but all change involves variability and heterogeneity. (Weinreich, Labov, and Herzog 1968: 188)

In this chapter I introduce fundamental concepts and key constructs of the study of Variationist Sociolinguistics that will be detailed in later chapters. Why approach the study of language from this perspective? What can be learned from this method that cannot be learned from other sociolinguistic methods? A major component of this approach to language is that it is linguistic, but also social and statistical. Why is a combined socioquantitative method useful and desirable?

Sociolinguistics

Sociolinguistics has its roots in dialectology, historical linguistics, and language contact with considerable influence from sociology and psychology (Koerner 1991: 65). This is why it has evolved into an exceptionally broad field. An all-encompassing definition would be that the domain of inquiry of sociolinguistics is the interaction between language, culture, and society. Depending on the focus, virtually any study of language implicates a social connection because without this human component language itself would not exist. However, the scope of sociolinguistics in this expansive interpretation is vast. Sociolinguistics has as many different facets as its roots. Some areas of the discipline put more emphasis on one area (culture); some disciplines put more emphasis on another (education). There is no one sociolinguistics other than the overarching unity of language in use. Depending on which aspect of language in use comes to the fore, sociolinguistics diverges into innumerable subdisciplines.

Every day we speak and write and use a complex, structured system to communicate but at the same time that system is evolving. The fundamental LVC (Language Variation and Change) question is, How does this happen? Weinreich, Labov, and Herzog (1968: 100–101) answered the question by saying, “the key to a rational conception of language change is the possibility of describing orderly differentiation.” This order, yet differentiation, as the normal state of affairs (Labov 1982: 17), the idea that variation is an inherent part of language (Labov 1969: 728), is the foundational maxim of the LVC approach. Differentiation, anomalies, and nonstandard features are easy to spot. In fact, just about everyone likes to talk about the wacky, weird, and/or reprehensible bits of language.

The normal condition of the speech community is a heterogeneous one … Moreover this heterogeneity is an integral part of the linguistic economy of the community, necessary to satisfy the linguistic demands of every-day life. (Weinreich, Labov, and Herzog 1968: 17)

Variability [is] not.. a nuisance but is a universal and functional design feature of language. (Foulkes, 2006)

Variation in language is most readily observed in the vernacular of everyday life. For example, a teenager says: “that were like sick”; an elderly man recounting a story to his granddaughter says: “you was always workin’ in them days.” Are these utterances mistakes? Are they slang? Are they instances of dialect? An LVC-oriented sociolinguist views such instances of language in use as an indication of the variable but rule-governed behavior typical of all natural speech varieties. The vernacular was first defined as “the style in which the minimum attention is given to the monitoring of speech” (Labov 1972c: 208). Later discussions affirmed that the ideal target of sociolinguistic investigation is “everyday speech” (Sankoff 1974, 1980b: 54), “real language in use” (Milroy 1992: 66). Variation in language can be observed just about everywhere from a conversation you overhear on the street to a story you read in the newspaper. Sociolinguists notice such variations too. In undertaking sustained analysis, what they discover is that people will use one form and then another for more or less the same meaning all the time the language varies. The harder part is to find the order, or the system, in the variation chaos. The way LVC undertakes this is by means of the “linguistic variable.” A linguistic variable is the alternation of forms, or “layering” of forms, in language. A basic definition is “two or more ways of saying the same thing.” A more nuanced, early, definition also mentions that linguistic variables should be structural and “integrated into a larger system of functioning units” (Labov 1972: 8).

Linguistic variables in a given speech community, whether morphosyntactic, phonological, lexical, or discursive, do not vary haphazardly, but systematically. Because it is systematic, this behavior can be quantitatively modeled (Labov 1963, 1969). Analyses of heterogeneous structures within the speech community rest on the assumption that whenever a choice exists among two (or more) alternatives in the course of linguistic performance, and where that choice may have been influenced by any number of factors, then it is appropriate to invoke statistical techniques (Sankoff 1988a: 2). The statistical tools used in the study of variation will be discussed in Chapter 5.

The combination of methods employed in Variationist Sociolinguistics forms part of the “descriptive-interpretative” strand of modern linguistic research (Sankoff 1988a: 142–143). Large-scale studies of variation in speech communities from New York to Norwich have produced extensive bodies of data. The descriptive component requires detailed, critical observation of variation and change. The patterns that have emerged from these undertakings have demonstrated that linguistic change is not only the result of universal principles but is also shaped by the social context in which it occurs (Labov 1963: 74). This is where the interpretive component of LVC has proven critical. Descriptions of variation can only be understood in context. While sociolinguistic principles prevail wherever you go, each situation provides a unique interpretation. In the case studies in Chapters 7–11 I will demonstrate how the study of different types of linguistic variables must take into account historical, contemporary, and social facts to explain language use.

The Linguistic Variable

LVC research begins with the observation that language is inherently variable. Speakers make choices when they speak and they alternate among these choices. Take, for example, the use of forms which strikes the ear as nonstandard, unusual, dialectal, or new, as in the examples in Example 1.1.1

Example 1.1

(a) And then next mornin’ [In] they were all brought back again. (YRK/002)

(b) Our car was like seven miles from where the entrance was. (TOR/021)

(c) There was a supply boat Ø came down to our cottage everyday. (TOR/036)

(d) He was like so funny and so nice. (TOR/054)

These features can only be fully understood if they are examined alongside the relatively unremarkable alternates with which they vary, as in Example 1.2.

Example 1.2

(a) And I started work on an evening [ŋ]. (YRK/012)

(b) We were oh probably about six miles from it. (TOR/054)

(c) The people that did it were brainwashed. (TOR/069)

(d) She’s really funny, and I think she’s really pretty too. (TOR/021)

Some variables may even have three or more alternates, as in Example 1.3.

Example 1.3

(a) I can’t remember what that building [in] is called. (TOR/008)

(b) I was on vacation for approximately six weeks. (TOR/038)

(c) I’m only exposed to the people who speak the same way that I do. (TOR/016)

(d) He’s very funny; he’s very generous. (TOR/023)

In other words, speakers may vary among various pronunciations of “ing” at the end of words. They may signal approximation with like or about or approximately. They may choose among relative pronouns that or who or leave it out entirely. They may select so or really or very to intensify an adjective. These choices are potential “linguistic variables.”

NOTE Linguistic variables are typically referred to by inserting the phoneme or morpheme or word that is variable inside parenthesis, i.e. variable (ing), (ly), (that), (so), etc. Phonetic realizations are represented inside square brackets, e.g. [n]. Phonemes are represented inside forward slashes, e.g. /n/.

A linguistic variable in its most basic definition is two or more ways of saying the same thing. An important question is, What does it mean to say two things mean the same thing? One time a student asked this question: what is the difference between a synonym and a linguistic variable? Let us explore this distinction. Synonyms are different lexemes with the same referential meaning as in Example 1.4:

Example 1.4

(a) car, automobile, vehicle, wheels

(b) girl, lass, chick, sheila, babe, doll, skirt

A more restrictive definition of synonymy would require that two synonyms are completely interchangeable in every possible context. In reality, most are not. For example, lass is primarily used in Scotland and northern England, chick is used in North America, sheila in Australia, whereas girl is not confined to a particular variety of English. For many practical purposes, such as with the production of dictionaries, it is customary to adopt a looser kind of definition for synonym. Near synonyms are lexemes that share an essential part of their sense, as in Example 1.5:

Example 1.5

(a) interesting, intriguing, fascinating, absorbing, spellbinding, engrossing

(b) striking, arresting, unusual, out of the ordinary, remarkable, salient

But this is not the whole story. Linguistic variables must also be alternatives (i.e. options) within the same grammatical system which have the same referential value (meaning) in running discourse (Sankoff 1988a: 142–143). Although some variants may differ subtly in meaning and distribution, if they are part of a linguistic variable they will be members of a structured set in the grammar. Moreover, the choice of one variant or the other must vary in a systematic way – this is what is meant by structured heterogeneity. There is difference, but there is structure to it. Different ways of saying more or less the same thing may occur at every level of grammar in a language, in every variety of a language, in every style, dialect, and register of a language, in every speaker, often even in the same discourse in the same sentence. In fact, variation is everywhere, all the time. This is why it is referred to as “inherent” variation (Labov 1969: 728). Now, consider a more in-depth definition of the linguistic variable:

two different ways of saying the same thing;an abstraction;made up of variants;comprising a linguistically defined set of some type: a phonemea lexical itema structural categorya natural class of unitsa syntactic relationshipthe permutation or placement of itemsalthough its delineation can be at any level of the grammar, the variants of the variable must have a structurally defined relationship in the grammar;they must also co-vary, correlating with patterns of social and/or linguistic phenomena.

Synonyms could be a linguistic variable. However a linguistic variable is more than simply a synonym. Deciding which forms co-vary meaningfully in language is actually a lot trickier than you would think.

Mini Quiz 1.1

Q1 How would a variationist sociolinguist explain the following example?

“There was two of us. Yeah, that’s right there were two of us.”

(a) Alternation in styles.

(b) Free variation.

(c) Linguistic variation.

(d) Random differentiation.

(e) Bad grammar.

Q2 Which of the following provides an example of two variants of a linguistic variable?

(a) And we said, “if you join the club, you must go to church.”

(b) He’d light a furnace for to wash the clothes.

(c) He was awful homesick, you know, my Uncle Jim.

(d) To prove I could do it, I had to prove that I could do it.

(e) There’s two girls on my street who have pink hats.

To this point this discussion has focused on the technical description of the linguistic variable. However, there is an entirely different side to linguistic variation that does not come from the mechanics of the linguistic system but involves issues of stigma and salience that come from the external evaluation of language by its users – us humans. There is no reason for a velar sound to be superior to an alveolar sound. There is no reason for a synthetic construction to be better than an analytic one. There is no inherently terrible thing about a double negative. However, there is an absolutely insidious view that certain ways of saying things are better than others. This comes down to the social interpretation of language use.

Most people are convinced that linguistic features are good or bad. For example, here is Sara Kempt, aged 49, in Toronto, Canada (c. 2003), in Example 1.6.

Example 1.6

… and I think the natural inclination of anybody is to get lazy and sloppy and not think. So I th– there’s more and more slang, and people dropping their Gs and things like that, just that … frankly grates on me. I hate it! Then again, I find myself doing it sometimes. (TOR/027)

Another fascinating thing about linguistic variables is that people are often completely unaware that they use them, particularly when certain of the variants are not part of the standard language. For example, this is Gabrielle Prusskin, aged 55, in Toronto, Canada (c. 2003). The interviewer has just asked her what she thinks about the word like, as in Example 1.7.

Example 1.7

It’s usually young females um when every other word is “like” and it drives me insane. I just like I hate it. (TOR/054)

TIP One way to find a linguistic variable is to look for the words that occur most frequently in data. Are there other ways of saying the same thing? If language is always in flux, then it is just a question of finding out what is on the move in a particular place and time.

Linguistic variables inevitably involve variants that have social meaning. These are typically called “sociolinguistic variables.” Sociolinguistic variables are those which can be correlated with “some nonlinguistic variable of the social context: of the speaker, the addressee, the audience, the setting, etc.” (Labov 1972c: 237). One variant might have overt social stigma, e.g. “I ain’t got it”, another might entail authority, e.g. “You must listen”, or prestige, e.g. “I shall tell you a story.” Yet another variant may be neutral, e.g. “I have it.” These social evaluations may differ markedly from one community to the next, from one country to the next, from one variety to the next, from one social situation to the next. It may even be the case that one person’s admired pronunciation will be another person’s loathed one. The patterns of a linguistic variable in the speech community tell the story of how the speech community evaluates the variants of the variable and in so doing this reveals how society is organized and structured. Which groups talk to each other? Which groups do not? How a linguistic feature is socially evaluated often has to do with its history as well. Which groups have been in the community a long time? Which groups are new? Language use is a reflection of the society in which it is embedded and the time period in which it occurs.

NOTE One time I went to a conference in the United States with my then current group of British graduate students. One of them had a strong accent from a variety of somewhat modest prestige in the United Kingdom. She was shocked to be told, repeatedly, how lovely her accent was. Similarly, I was chagrined to discover that my own middle-class Canadian accent – unremarkable in Canada – was heard as an entirely unbecoming American accent in the United Kingdom.

The primary empirical task of Variationist Sociolinguistics is to correlate linguistic variation as the dependent variable with independent variables (Chambers 2003: 17). The dependent variables are the features of the linguistic system that vary (e.g. the varying pronunciations of the same phoneme, the choice of relative pronoun, the selection of an intensifying adverb). Independent variables are the features associated with the variation. They can be external to the grammar, out in the world, relating to aspects of the social context, situation, community setting, or register. They can also be internal to the grammar, relating to the linguistic environment such as the grammatical category of the word, the type of subject in the clause, or its function.

Patterns in language are observed using a two-part undertaking: (1) find socially and linguistically significant factors that impact variation, and (2) correlate them with general social forces (Labov 1972c: 42). The patterns that arise are used by the analyst to interpret and explain the phenomenon under investigation. The fact that linguistic differentiation in communities has been consistent for different linguistic features and that these patterns repeat themselves across different situations in time and space have given rise to a series of “classic” sociolinguistic patterns from which Variationist Sociolinguistic inquiry has sprung. These patterns provided a baseline for all subsequent research and have informed several new generations of research-based study.

The study of sociolinguistics as LVC is unique in sociolinguistics in two ways: (1) its overriding goal and (2) its methodology. LVC research attempts to solve one of the great paradoxes of language in use – the fact that language is always changing.

The basic LVC procedure is the following:

Observation – hear and/or see variation in language use;Identification – select the linguistic variable for study;Reconnaissance – determine if the variation occurs and where;Systematic Exploratory Observation: What is the inventory of forms?What are the patterns?When does the variation occur and under what circumstances?Who uses the variation and how?Test hypotheses, claims, and observations;Interpret and explain the variable patterns, social and linguistic.

To discover the relevant factors (social and linguistic) which give rise to “speakers or writers’ sustained and repeated exercise of their linguistic facilities in producing large numbers of sentences” (Sankoff 1988a), the data are analyzed using statistical modeling. This method enables the analyst to ask and answer the following questions:

Which factors are statistically significant (i.e. not due to chance)?What is the relative contribution of the independent factors tested in the model, i.e. which factor group is most significant or least?What is the order (from more to less) of factors within the independent factors (predictors), the constraints or constraint hierarchy?Does this order reflect the direction predicted by one or the other of the hypotheses being tested?

Notice that Variationist Sociolinguistics has an essentially multiplex nature: on the one hand, empirical and data-based; on the other hand, scientific methods and statistical testing; but there is a third component. Linguistic patterns can only be understood through interpretation. Explanation in sociolinguistics can only happen when statistics are used in conjunction with a strong interpretive component, grounded in real-world language use.

Mini Quiz 1.2

Q1 The primary empirical task of Variationist Sociolinguistics is to:

(a) define linguistic variables

(b) relate linguistic variables with each other

(c) correlate independent variables with each other

(d) correlate linguistic variation as the dependent variable with independent variables

(e) correlate linguistic variation as the independent variable with dependent variables in society.

Linguistic Change

If I talk to, say, my grandfather, like I talk to one of my friends, he’d just be like, “what?” (TOR/023)

One of the driving forces of Variationist Sociolinguistics is the search for general principles that govern linguistic change. If one form appears to be replacing the other, either in time or along some economic, demographic, or geographic dimension (Sankoff and Thibault 1981: 213) then this may be an indication of language change in progress. Consider the way people talk about the weather. It is often the case that it is either cold or hot outside. When the temperature is extreme in one direction or another people will typically intensify their descriptions. For example, if it is particularly cold a person might say, “It’s very cold today!” But would a young person say it the same way? Probably not. A younger person (at least in Canada in the early twenty-first century) is more likely to say: “It’s so cold today!” In contrast a middle-aged person is more likely to say: “It’s really cold!” If these observations can be substantiated across a wide number of people (e.g. Tagliamonte 2008b) (see Chapter 9), this may be evidence for ongoing evolution of a subsystem of grammar – generational change. This is why linguistic data from different age groups in the same speech community, or different communities in the same country, or even communities in different countries in different circumstances, provide important evidence for understanding how language change may be happening.

All languages change through time. We do not really know why this is, but it is a characteristic of all human languages. They also change in different ways in different places. (Trudgill 2003: 7)

Linguistic change typically proceeds in “an ordered set of shifts in the frequency of application of the rule in each environment” (Labov 1982: 75). What this means is that the rate of use of a particular form, e.g. very, really or so, is not the most important observation. Instead, the contexts in which these forms occur – their patterns of use – is the key element in tapping into linguistic change. For example, very and really can occur as attributive or predicative intensifiers, as in Example 1.8a, but so can only occur as predicative, at least in current prescriptive accounts of English.

Example 1.8

(a) It was a really hot day [attributive] and like on the way there I started to feel really, really weak [predicative]. (TOR/011)

(b) I was so hungry [predicative]. (TOR/013)

These patterns of use are the fundamental units by which linguistic change occur (Labov 1982: 75) (see Chapter 3). Moreover, if the relationship of these environmental contexts can be captured this provides a critical measure for comparison – referred to as a “constraint” on variation. In this case the constraint is the difference between predicative vs. attributive position. Similarities and differences in the significance, strength, and ordering of constraints (the constraint ranking) offer a microscopic view of the grammar from which we can infer the structure (and possible interaction) of different grammars. The various statistical techniques in the LVC toolkit enable the analyst to assess and evaluate the competing influences and in so doing interpret the path of development of language through time and space and social structure (see Chapter 5).

TIP What is the difference between a “variable” and a “factor” or “factor group”? In sociolinguistics “variable” is reserved for the “linguistic variable”, the feature that varies and that is under investigation (i.e. the dependent variable). “Factors” are the aspects of the social or linguistic context that influence the variable phenomena (i.e. the independent variables). In statistics these are referred to as “predictors.” To avoid confusion I will continue to use the field’s current standard term “factor group” or “factor.”

LVC analysis asks the question, How can a variable linguistic phenomenon be explained? In this type of analysis the critical component is that the data come from the recurrent choices speakers make in the course of production. In this way, each choice is viewed not simply as an instance or token of use, but as a choice made within the context of the grammar from which it comes. When a large body of repeating tokens is part of the analysis, the choices can be assessed statistically so as to uncover the meaningful patterns of use (Cedergren and Sankoff 1974; Labov 1969; Poplack and Tagliamonte 2001: 89). The choices are taken to represent the (underlying) variable grammar of the speaker as well as the grammar of the speech community to which she belongs (Poplack and Tagliamonte 2001: 94). The goal to investigate language use in the context of language structure is what makes an LVC analysis “accountable.”

The Principle of Accountability

A foundational concept in the Variationist Sociolinguistic approach and one that sets it apart from other methods is the “principle of accountability” (Labov 1966: 49; 1969:737–738, fn. 20; 1972c: 72). This is where the analysis begins. Say the analyst is interested in the use of the relative pronoun who. The principle of accountability dictates that in addition to examining who