UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A socio-dialectology survey of the English spoken in Ottawa : a study of sociological and stylistic variation.. Woods, Howard B. Mar 18, 1979

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-UBC_1979_A1 W67.pdf [ 21.28MB ]
Metadata
JSON: 831-1.0100260.json
JSON-LD: 831-1.0100260-ld.json
RDF/XML (Pretty): 831-1.0100260-rdf.xml
RDF/JSON: 831-1.0100260-rdf.json
Turtle: 831-1.0100260-turtle.txt
N-Triples: 831-1.0100260-rdf-ntriples.txt
Original Record: 831-1.0100260-source.json
Full Text
831-1.0100260-fulltext.txt
Citation
831-1.0100260.ris

Full Text

A SOCIO-DIALECTOLOGY SURVEY OF THE ENGLISH SPOKEN IN OTTAWA: A STUDY OF SOCIOLOGICAL AND STYLISTIC VARIATION IN CANADIAN ENGLISH by HOWARD B. WOODS B.Sc., M.A. A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY . - IN-THB FACULTY OF GRADUATE STUDIES (Department"^ Linguistics) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October, 1979 Q Howard B. Woods,, 1979 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of t^J* The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 DE-6 BP 75-51 1 E ii A Socio-dialectology Survey of the English Spoken in Ottawa: A Study of Sociological and Stylistic Variation in Canadian English. ABSTRACT This study is a response to the long-standing need -within the field of Applied Linguistics for a "better understanding of General Canadian English and for a quantitative documentation of its usage. This dissertation presents an analysis of a sample of this national language as spoken "by persons whose mother tongue is English and who were born and raised in the city of Ottawa. The analysis demonstrates that the•informants vary their speech patterns according to the linguistic tasks which they are asked to perform. E-urther, the analysis demonstrates that variations in usage are correlated to age, sex, ariW'-socio-economic status. ' An hour-long interview was conducted with one hundred informants by means of a questionnaire which was designed to elicit phonological, morphological, syntactic, semantic, and lexical responses while the informants were performing a.number.of language related tasks. Unlike several urban socio-dialectology surveys conducted recently, this survey analyses the speech characteristics of a broad range of the socio-economic structure from skid row through the middle classes to the lower upper class. It is thereby the first such survey dealing with Worth American English which includes the mainstream of society as well as the minority ethnic groups. Prior to presenting the extensive data and evaluative comments on the co-variation of stylistic and sociological variation of Canadian iii English usage, a number of background subjects are dealt with. First, a review and evaluation of previous dialectology studies is presented. Second, the background of Ottawa's settlement patterns and present situation is.discussed. Third, the development of Canadian English is traced through time so that this national.dialect can be placed in relation to other dialects of the.English language family. This is followed by a detailed account of those characteristics which distin guish Canadian English from Northern American, its closest relative. The fourth subject dealt with is the methodology of the study itself. ' Chapter 5 presents an analysis of the co-variation of 27 phonological segments and sociological and stylistic parameters. Our data forcefully prove our hypothesis.of phonological and stylistic co-variation, with 20 of the 27 items demonstrating variation directly related to the degree of formality of the task performed by the informant. Similarly, our data prove our hypothesis of phonological and socio-economic co variation. In this case, 10 of the 27 items demonstrate ordered stratification. The upper classes use a much broader range of styles than do the lower classes. Our analysis also reveals important findings of sex and age differentiation in addition to a number of important phonological dis coveries . Chapter 6 analyses the co-variation of grammatical, pronunciation, and vocabulary forms and sociological factors. The data demonstrate that I3 of 15 grammatical items, 25 of hQ pronunciation items, and 2 of 8 vocabulary items have clear and ordered socio-economic stratification. Prestige and stigmatized forms were designated according to usage patterns. These data also demonstrate sex and age differentiation; our female iv informants maintained the strongest class hierarchy, had the lowest frequency of stigmatized forms, and the highest frequency of prestige forms. Both Chapters 5 and 6 have a summary section which contains the important findings.of their respective analyses. The Conclusion contains a.summary of the significant findings and suggestions for future studies. Appendices.A to G contain the question naire, the complete computer printout of the data, and tabular presen tations of paralinguistic and attitudinal observations. v: TABLE OF CONTENTS page ABSTRACT 11 TABLE OP CONTENTS. •••• DEDICATION ^ PREFACE... viii CHAPTER 1: INTRODUCTION 1 1. Purpose2. The Value of Sociolinguistic Urban Surveys 5 3. Previous Studies. .' 10 Footnotes 15 CHAPTER 2: OTTAWA 21. Ottawa, Its Settlement and Setting 20 2. Ottawa Valley Urban Centres 24 Footnotes 26 -CHAPTER 3: CANADIAN ENGLISH 7 1. Canadian English in Relation to Other Dialects 22. Linguistic Features 33 Footnotes 57 CHAPTER 4: METHODOLOGY 9 1. The Sociological Parameters 52. The Contextual Styles... 66 3. The Sample 71 4. The Questionnaire 83 5. Interviews 4 6. Transcription and Analysis 85 7. Limitations 6 Footnotes 90 vi i Page CHAPTER 5: THE CO-VARIATION OF THE PHONOLOGICAL VARIABLES WITH SOCIOLOGICAL AND STYLISTIC PARAMETERS 93 1. Measurement of Co-variation 92. Evaluation and Summary 178 3. Somers' D Statistical Analysis 194 Footnotes 205 CHAPTER 6: THE CO-VARIATION OF GRAMMATICAL, PRONUNCIATION, AND VOCABULARY VARIABLES WITH SOCIOLOGICAL PARAMETERS 212 1. Grammar and Syntax 213 2. Pronunciation 226 3. Vocabulary 262 4. Evaluation and Summary 270 5. Somers' D Analysis 286. Comparison with SCE 7 Footnotes 289 CHAPTER 7: CONCLUSION 292 1. Methodology2. The Co-variation of Phonological Variables with Sociological and Stylistic Parameters 294 3. The Co-variation of Grammatical, Pronunciation, and Vocabulary Variables with Sociological Parameters 301 Footnotes .' 308 BIBLIOGRAPHY 310 APPENDIX A: OTTAWA ENGLISH QUESTIONNAIRE 316 APPENDIX B: ITEM BY ITEM PRINT OUT 345 APPENDIX C: DISTINCTIVE ITEMS OF OTTAWA VALLEY URBAN CENTRES 407 APPENDIX D: CANADIAN INDEX 412 APPENDIX E: LANGUAGE ATTITUDES 4i5 APPENDIX F: PARA-LINGUISTIC'.PHENOMENA 421 APPENDIX G: MAPS 424 vii DEDICATION This dissertation is dedicated to Drs. Robert J. Gregg and Ruth E. McConnell both of whom have devoted their careers to the study of Canadian English. viii PREFACE This study is a response to the long-standing need within the field of applied linguistics for a better understanding of General Canadian English and for a quantitative documentation of its usage. This disser tation will present an analysis of a sample of this national language as spoken by persons whose mother tongue is English and who were born and raised within the urban boundaries of the city of Ottawa. The analysis will demonstrate that the informants vary their speech patterns according to the tasks which they are asked to perform. Further, the analysis will demonstrate that variations in usage are correlated to age, sex, socio-economic status, ethnic background, rural background and sometimes to the number of generations the informants family has been in Canada. An hour-long interview was conducted with one hundred informants by means of a questionnaire which was designed to elicit phonological, mor phological, syntactic, semantic, and lexical responses while the informants were performing a number of language related tasks. There was, of course, no way of eliminating completely the influences of the interview situation,which generally causes speech to be more formal than the relaxed, unguarded style of casual conversation. However, a number of techniques were employed in order to enable the informant to relax and thereby speak more freely. ix Unlike several urban socio-dialectology surveys conducted previously, this survey will analyse the speech characteristics of a broad range of the socio-economic structure, including the lower, working, lower middle, middle middle, upper middle, and lower upper, classes. It is therefore the first such survey dealing with North American English which analyses the speech characteristics of the mainstream of society as well as those of the-minority ethnic groups. Chapter 5 will present an analysis of the co-variation of 27 phono logical segments and sociological and stylistic parameters. Our data will forcefully prove the\hypothesis of phonological and stylistic co-variation, with 20 of the 27 items demonstrating variation directly related to the degree of formality of the task performed by the infor mant. Similarly, the data will prove the hypothesis of phonological and socio-economic co-variation. In this case, 10 of the 27 items will demonstrate ordered stratification. 'The data also reveal that, the upper classes use a much broader range of styles than do the lower classes. Our analysis will also reveal interesting findings of sex and age dif ferentiation in addition to a number of important phonological dis coveries. Chapter 6 will analyse the co-variation of grammatical, pronuncia tion, and vocabulary forms and sociological factors. The data will demonstrate that 12 of 15 grammatical items, 25 of 48 pronunciation items, and two of 8 vocabulary items have clear and ordered socio economic stratification. We will.designate particular forms to be prestige and stigmatized forms according to observed usage patterns. Our data will also demonstrate interesting sex and age differentiation; for example,our female informants maintained the strongest class X hierarchy, had the lowest frequency of stigmatized forms, and the highest frequency of prestige forms. Chapters 5 and 6 both have an evaluation and summary section which contains the important findings of their respective analyses, a com parison of data and conclusions from previous studies and a Somers' D statistical description of our results. Prior to presenting the extensive data and evaluative comments on the co-variation of stylistic and sociological variation of Canadian English usage, which is to be found in Chapters 5 and 6, a number of background subjects will be dealt with. First, a review and evaluation of the previous dialectology studies from which the present study is adapted will be presented. Second, the background of Ottawa's settle ment patterns and present situation will be discussed. Third, the development of Canadian English will be traced through time so that this national dialect can be placed in relation to other dialects of the English language family. This historical sketch will be followed by a detailed account of those characteristics which distinguish Canadian English from Northern American, its closest relative. Linguistic comparisons in this study will be made most often with reference to Northern American and General American rather than with Standard Southern British. The fourth subject to be dealt with will be the methodology of the study itself. The conclusion will contain a summary 7of the important findings, a comparison of data from previous studies, and suggestions for future studies. The Appendices A to F will contain the questionnaire and the complete computer printout of the data, linguistic data from Ottawa TP? Valley urban centres,and tabular and graphic presentations of para-linguistic and attitudinal observations. 1 CHAPTER 1 INTRODUCTION 1. Purpose The purpose of this dissertation is to investigate the nature and extent of the co-variation of certain linguistic features with sociological and stylistic parameters in the speech of the people of Ottawa, Ontario. This study, which makes use of special techniques to elicit stylistic, sociological, and regional dialectological variants of our linguistic variables, is the first large-scale effort to describe the speech characteristics of an urban area of English Canada."'" It is also the first work in North American urban dialectology which focusses on the mainstream, of the population, including a large range of social classes, many ethnic backgrounds, and all sections of the city and its residential areas. In recent times because of the great social need, linguists in the United States have focussed their urban studies on minority groups within the inner-city cores of the largest American 2 metropolises. Until fifteen years ago, dialectologists were mostly concerned with regional dialectology. One of their major interests was to determine the geographic boundaries of the dialects of their respective languages. This interest caused most of their work to take place in rural settings, which in turn led the dialectologists to search for old relic forms, before those forms disappeared forever. The many projects contributing 2 to the Linguistic Atlas of the United States and Canada are excellent 3 examples of this type of dialectology. Given their aims it is under standable that little was done to describe or analyse the speech characteristics of the people who lived in the urban centres, and even less was done to determine the correlation of variation of linguistic material with sociological or stylistic parameters. In 1966, the situation was altered considerably by W. Labov's sociological and stylistic study of the English spoken in the Lower East Side of Manhattan Island in New York City, entitled The Social Stratification of English 4 in New York City. As a result of this breakthrough, dialectology now has two major areas of investigation, sociological urban dialectology and traditional regional dialectology. This present study finds its theoretical base and organizing principles in sociological urban dialec tology specifically from the Labov and Trudgill studies,"' while at the same time using a large number of linguistic items which have been shown to be characteristically Canadian as opposed to Northern American in previous regional studies. Most of the purely rural Canadianisms were excluded from this study. In order to accomplish the investigation, we constructed a questionnaire of 752 items characteristic of Canadian English which we suspected would show stylistic and sociological variation. This questionnaire was then administered orally to one hundred native-born anglophone Ottawans. Care was taken to ensure that all major socio logical sub-divisions were represented in numbers large enough that meaningful comparisons among these groups could be made. The sociological groups were based on: age, sex, socio-economic class, ethnic background, urban/rural background, and the number of generations the informant's 3 family had been in Canada. It is a major hypothesis of this dissertation that certain items in the linguistic material will show co-variation with these sociological parameters in a systematic and non-random manner. In addition to the co-variation of the linguistic phenomena and social differentiation, a second sociological dimension, style or register as it is also called by many linguists, was investigated. Stylistic variation generally occurs as a result of changes in the social context, e.g. the formality of the occasion, the roles of the individuals, or the relative age and sex of those communicating. How ever, in the interview situation, the social context is relatively constant. The problem, then, was to construct interview situations which would elicit as full a range of stylistic variation as possible. Two previous studies already mentioned have developed and tested language related tasks for interview situations which elicit a large range of styles. These styles have been shown to approximate the range 6 of styles found in most social contexts. Ranging from the most formal to the most casual, the tasks which the informant was required to do in this survey, were: 1) reading of minimal pairs,. 2) reading of word lists,, 3) identifying objects by pictures, 4) reading a story, and 5) speaking freely about Ottawa, Canadian English, his/her closest encounter with death, and creating a narrative from a sequence of pictures. The second major hypothesis of this dissertation is that certain items in the linguistic material will demonstrate co-variation with the language related tasks in a systematic manner. The linguistic items systematically incorporated in the question naire and investigated in this study are divided into two major categories. The first group is made up of phonological segments, 4 either separately or in groups from which thousands of words can be generated. Liaison patterns also form part of this group. The pro nunciation of the twenty-seven items in this group will be analysed thoroughly with relation to the stylistic tasks and sociological para meters referred to above. The second major category of linguistic variables investigated in this study is comprised of seventy-one grammatical, word pronunciation, and vocabulary items. Much of the data of this portion of our study can be compared directly with the results of The Survey of Canadian  English by M.H. Scargill and H.J. Warkentyne.^ Although our study is restricted to mainly one city and does not have the scope of all ten provinces as the SCE does, we are able to analyse our data with refer ence to more interesting sociological parameters, including especially socio-economic classes, ethnic background, rural/urban background, and the number of generations an informant's family has been in Canada. The fact that Canadian English is a dialect derived from and much influenced by American and British dialects is a major theme of reference throughout this study. Indeed, most of the variables in this study are limited to either British or American usage choices. In addition to a limited number of indigenous usage items which distinguish Canadian English from other dialects of English, Canadian English is most easily characterized by its unique blend of American and British usage. There fore, the co-variation of sociological and stylistic parameters with linguistic phenomena were discussed frequently with reference to usage patterns from these two dialects with which Canadian English is in contact. In order to deal with and understand this 'dialects in contact' situation, we first present the history of Canadian English; second, we 5 present the linguistic features which characterize Canadian English; and finally, we determine the frequency break-down of the variants of the variables. In addition to the several hundred linguistic items elicited and analysed in this study, a number of paralinguistic phenomena were investigated; these included: utterances of agreement, pulmonic ingressives, reading time, hesitation phenomena, stuttering, swearing, and language attitudes. These data are to be found in Appendix C. 2. The Value of Sociolinguistic Urban Surveys A New Body of Data One of the most obvious results of this type of survey is the accumulation of a new body of linguistic data. These data can provide a base upon which linguistic theory can be developed and verified. Because of a lack of such data the generative-transformationalists of the 1960's, for example, frequently experienced considerable difficulty g while trying to decide what was acceptable English usage. The Value to Dialectology In a circular manner, each new survey in the field of dialectology can help dialectologists re-evaluate their methods and techniques for surveying language usage. Sociolinguistic urban surveys are a relatively new activity and the stylistic and sociological parameters used in correlation to the linguistic phenomena need to be tested and refined. Similarly, the eliciting and sampling techniques need re-evaluation and development. This study is innovative in relation 6 to recent urban studies in its use of pictures for eliciting isolated 9 words or phrases, in the use of sequential pictures for eliciting casual speech,"^ in the use of a casual reading passage for eliciting liaison features and fairly relaxed speech, and in setting out to observe a number of para-linguistic phenomena. This study will also prove to have broken new ground in its setting out to elicit linguistic material from informants in all levels of society, including the upper middle and lower upper classes. W. Labov's study appears to be an assessment of a truncated portion of American society located on the Lower East Side of Manhattan Island whose income is not differentiated after $4,500.00 per annum,''""'" and P. Trudgill's index of income for his informants in Norwich, England stops differentiating upwardly at approximately 12 $5,000.00 annually. Our informants had annual incomes ranging from $0.00 to over $35,000.00 and it would appear both socio-economically and linguistically that the informants of these two other studies would be situated in our middle-middle class downward to our lower or working classes. The Value to English Education On the practical side, a study such as this which provides a large data base of current usage is of great value to those in the language education profession and more specifically those involved in the teaching of English. For how is one to know what to teach and what problems to teach to if one does not know the present status and standard of English usage in Canada and especially in Canadian cities. Ever since 1876 when Georg:; Wenker started his survey of the Rhine dialects of Germany, linguists the world over have been conducting 7 surveys and building up data on their languages. Their main focus, however, has been to survey the rural areas so as to understand the linguistic substratum from which most urban speech was derived and to record the most archaic types of rural speech before that speech disappeared. Their work frequently has not been immediately relevant to the school systems in the cities. Only recently have linguists turned to urban surveys in any large or systematic way. The value and justification for urban surveys in Canada seem immediately apparent. 13 Canada is now 76 percent urbanized and only 24 percent rural. The proportion of people living in urbanfareas has shifted drastically from the time when dialect surveys started, and, of course, usage surveys should take place where the people live. Not only are urban surveys important because of the number of people living in cities, but they are important because the cities are cultural and commercial centres from which language styles and standards emanate. People from outside the urban areas generally adjust their language habits to conform more closely with urban patterns in order to obtain higher education, gain employment, or raise their social status. The sociological aspect of such urban studies is important because of the fact that only a certain segment of the urban society sets the linguistic standard. This linguistic standard is not legislated or even consciously thought out; it is mainly the composite of the language habits of the upper-upper, lower upper, and upper middle classes. This segment of the population, which comprises about 10 percent of the nation, exerts a disproportionate influence on the legal, governmental, educational, media, and commercial Institutions of Canada, and only a study of this segment of society and its linguistic interaction with the 8 other classes will give us an understanding of the characteristic features of Canadian English. The Value of Data Most educationalists find it desirable to teach what they would call the standard language. Even those most fearful of the negative effects of such a policy advocate the teaching of the standard language as a second dialect to non-standard speakers late in their educational development. Unfortunately, failure to acquire a standard dialect tends to hinder an individual's opportunities for development and advancement. If the standard dialect is to be taught, then it must be described. There are two major ways of going astray when describing standard language usage. One way is to describe the manner in which most people speak, thereby ignoring the impact of the upper classes on language standards. The other is to prescribe usage according to rules of logic, classical Latin grammar, and past British usage, thereby denying the arbitrary nature of language and the fact that language is continuously changing. Sociolihguistic studies such as this are designed to avoid the pitfalls of these two extremes and to provide a data base from which an adequate description of the language can be made. The Value to E.S.L. Such sociolinguistic surveys are also of great benefit to those involved in teaching English as a second language (E.S.L.). In E.S.L., the teacher is concerned with teaching usage registers for the purposes of casual communication with colleagues and friends as well as teaching 9 the formal registers for job interviews, oral tests, speeches, etc. The availability of a large data base of stylistic and sociological varia tion will be a great aid to a teacher's intuition. Such studies should be made available as reference material for teacher training programmes and teacher resource libraries. Curriculum developers and textbook writers must ensure"that their examples, exercises, dialogues, activities, etc., reflect appropriate usage. And appropriate usage should be related •to'-.'.the situations in which the students are likely to be placed. Language data which show usage patterns according to sex, age, social class, stylistic level, etc., are precisely what is needed. Moreover, if the textbooks are designed to train students for life in Canada, then the data base should be from Canadian English usage. At the present time, many American texts are used in the E.S.L. classrooms in Canada frequently giving models of English inappropriate for the Canadian context. (For examples of the difference between Canadian and American usage see Section 2 of Chapter 3 entitled 'Linguistic Features'). Equally unfortunate is the fact that many Canadian E.S.L. textbooks are based on research on American English. This study and others.llke it help alter this situ ation by providing a large data base of Canadian usage. The Value to Other Fields Knowledge of appropriate Canadian usage is also a concern of those involved in radio and television broadcasting. Announcers, producers, writers, and their support staff are often made painfully aware of any usage which their listeners find unusual or incorrect. The CBC's 14 monthly publication entitled You Don't Say and its CBC News Style Book help its personnel to avoid offending the taxpayers' ears. Other stations often request advice from university departments of English and Linguistics when they feel uncertain about a particular usage. Further, members of other professions such as playwrights, novelists, sit-com writers, actors, people in advertising, and customs officers are all known to have made observations and inquiries about Canadian speech characteristics. Surveys such as this could be of value to these professions as well. Sociolinguistic studies can also be of value to sociology and sociologists, as linguistic behavior is one of the prime indicators of social class, age, and sex groupings. Finally it is hoped that individuals from the United States, Britain, and elsewhere who have moved to Canada and become teachers of English, E.S.L., modern languages and linguistics will want to use the data of such surveys to better understand Canadian English and the people who speak and write it. 3. Previous Studies Regional Studies . The earliest studies of North American English dialectology have been concerned with rural dialectology. As stated previously, most of these projects were associated with or integral parts of the research 16 for The Linguistic Atlas of the United States and Canada. The aim of these studies legitimately has been concerned with: 1) determining the geographic areas and boundaries of the major regional dialects, 2) finding the linguistic items which are most characteristic of the dialects, and 3) recording old dialect forms before they have disappeared. The works of Kurath, Marckwardt, Allen, andC. Reed are perhaps most relevant to this study, as they define the American dialects with which Canadian English will be compared. Urban Studies In the past decade a number of urban studies have been conducted ' in the United States These studies were carried out with the primary goal of describing the dialects of minority groups found in inner-city areas of large American.cities. The motivation for many of these studies was to gain recognition for these dialects as languages of instruction for those students who spoke them. Labov's work, The Social 18 Stratification of English in New York City, is the first study which presents the correlation between linguistic variation on the one hand and stylistic and sociological variation on the other. His work is a major reference for all sociolinguistic studies which have followed. A few years later in Britain, the excellent survey in sociological urban dialectology The Social Differentiation of English in Norwich by P. 19 Trudgill was published presenting the speech characteristics of the people of that city. Although the speech characteristics and social class structures of both Manhattan and Norwich differ'greatly from those Ottawa, these two studies have provided a model in theory and planning from which, we constructed our study. Other important contributions in the area of sociological urban dialectology came from Field Techniques 20 in an Urban Language Study by Shuy, Wolfram, and Riley and from "Sample Survey Methods and Computer Assisted Analysis in the Study of 21 Grammatical Variation" by Sankoff and Sankoff. Canadian English Usage Many works provided needed data on Canadian English usage; those which dealt with speech differences along the Canadian-American border 22 were Allen, 1959; Avis 1954, 1955, and 1956; and Reed, 1957 and 1961. Those which dealt with particular sounds in Canadian speech were: 23 Chambers, 1975; Gregg, 1973a, 1975; Joos, 1942. There are several local/regional surveys extant including Avis, 1975; Chambers, 1974; Gregg, 1973b; Hamilton, 1958; Kinloch, 1972; Poison, 1969; Wanamaker, 24 1974; and Wilson, 1975. General background and reference books are: Avis, 1967, 1973; Chambers, 1975; Lovell, 1955; McConnell, 1979; McDavid, 1967, 1971; Scargill, 1957, 1977; and von Baeyer, 1976, 1977.25 Finally, there is The Survey of Canadian English sponsored by the Canadian Council of Teachers of English from 1970 to 1974. This survey fostered several publications including Kinloch, 1971, 1972; Rodman, 1974; Scargill, 1974 (the major work); Scargill and Warkentyne, 1972 26 (the report); and Warkentyne, 1971. A few of the above works merit special mention. Walter Avis' works in 1954, 1955, 1956 dealing with speech differences along the Ontario-27 U.S. border inspired personal interest in the topic and provided a number of items for this survey. Martin Joos' article "A Phonological Dilemma in Canadian English" is important because it is the first to single out the special characteristics of the Canadian diphthongs. M.H. Scargill's Modern Canadian English Usage, referred to as the SCE, is perhaps the most useful resource^material with precise statistics on Canadian English usage for our study. The SCE is a report on the responses to over 30,000 questionnaires sent out to grade 9 students and their parents in all ten provinces of Canada. The questionnaire 13 was made up- ;of questions dealing with morphology and syntax, pronun ciation, spelling, and vocabulary. The many tables give us a quick view of the differences in usage (or perhaps the differences in attitudes towards usage) by age group, sex, and province. These tables also sub stantiate quite conclusively that females conform more readily to what is "correct", and that a great deal of language levelling is taking place among young students. Certain disadvantages were inherent in the SCE study, however. They were: (1) the study could report only what informants claimed they said, and it assumed that the informants could make the necessary distinctions, particularly in phonology. Scargill states on page 70, "With a mailed questionnaire, it is difficult to interpret the answers given about the pronunciation of calm, caught, father, bother." One could also feel some doubt about Questions 27 and 110, i.e. vase and guarantee; (2) informants gave one answer only and no attempt was made to take stylistic variation into account. In response to questions like Scargill's Question 49 butter (which reads; "Does the -tt- in butter sound like the -dd- in shudder?"), we will test the stylistic variation of several words which contain a medial t. Some of our informants actually ranged from 100 percent medial [t] in their formal register to 100 percent medial [d] in their casual speech. It can be safely assumed that most informants claimed usage consistent with their more formal registers of speech; and (3) the survey avoided categorizing people according to social or educational groups. Instead, the survey grouped the informants by province. That decision may have been politically motivated, consciously or subconsciously, in order to avoid the taboo of social class. This study will attempt to show that for Ontario and west, stylistic and socio-economic factors are far greater determining factors in usage variation .than are geographic factors. These limita tions notwithstanding, the SCE is the major work on Canadian English usage, and the one which provided the impetus and desire to begin this present study. R.J. Gregg's latest work, the as yet unpublished monograph on 28 Canadian English for The Commonwealth Series on English gives an excellent summary of the development of Canadian English and describes the major systematic elements of Canadian English with comparison made mainly to British English. This manuscript was unavailable during most of this present study. However the knowledge involved had been con veyed by Gregg over several years of seminars on Canadian English. Of most immediate benefit to this survey was the Urban Dialect Survey of the City of Vancouver: Pilot Project. This project headed by Professor Gregg and assisted by Margaret Murdoch, Gaelan de Wolf, and Erika Ludt is a survey of sociological and stylistic variation of 29 Vancouver. During the summer of 1977, I was fortunate enough to have the opportunity to work on this pilot project. Most of the design of the Ottawa questionnaire is a development from that pilot project, including the reading" passage by Murdoch. In return the newly begun Vancouver Project is now modelling a number of concepts after the Ottawa Survey. 15 Chapter I: Footnotes ''"David and Gillian Sankoff conducted a survey on Canadian French in Montreal. For a report on their methodology see D. and G. Sankoff, "Sample Survey Methods and Computer Assisted Analysis in the Study of Grammatical Variation," Canadian Language in their Social Context, ed. Regna Darnell, (Edmonton: Edmonton Linguistic Research, 1973), pp.12-23. 2 The leading resource materials available on this topic are: William Labov, The Social Stratification of English in New York City, (Washington, D.C: Center for Applied Linguistics, 1966), pp.1-655; William Labov et al., A Study of the Non-Standard English of Negro and  Puerto Rican Speakers in New York City, (Philadelphia: U.S. Regional Survey, 1968), vol.1:1-375, vol.2:1-357; Roger W. Shuy, Walter A. Wolfram, and William K. Riley, Linguistic Correlates of Social  Stratification in Detroit Speech, Final Report, Cooperative Research Project 6-1347, (Washington: U.S. Office of Education, 1967); Roger W. Shuy, Walter A. Wolfram, and William K. Riley, Field Techniques in  an Urban Language Study, (Washington, D.C: Center for Applied Linguistics, 1968), pp.1-128; Walter Wolfram, Sociolinguistic Aspects  of Assimilation; Puerto Rican English in New York City, (Arlington, Va.: Center for Applied Linguistics, 1974), pp.1-241; Walter A. Wolfram and Ralph W. Fasold, The Study of Social Dialects in American English, (Englewood Cliffs, N.J.: Prentice Hall, 1974), pp.1-239. 3 There are scores of branching projects within this project; for an anthology of these works see: Readings in American Dialectology, eds. Harold B. Allen and Gary N. Underwood, (New York: Appleton, Century, Croft, 1971), pp.1-283 (hereafter cited as RAD) and A Various  Language, eds. Juanita V. Williamson and Virginia M. Burke, (New York: Holt, Rinehart and Winston, 1977), pp.1-706. 4 W. Labov, The Social Stratification of English in New York City, (Washington, D.C: Center for Applied Linguistics, 1966), pp.1-655. ^Peter Trudgill, The Social Differentiation of English in Norwich, (Cambridge: Cambridge University Press, 1974), pp.1-211. 6Labov, 1966, pp.90-135 and Trudgill, 1974, pp.45-54. ^M.H. Scargill and H.J. Warkentyne, "The Survey of Canadian English: A Report," English Quarterly, Volume 5, No.3, (Autumn 1972), pp.47-104. Reprinted separately with permission from the Canadian Council of Teachers of English and revised and published as a book: M.H. Scargill, Modern Canadian English Usage: Linguistic Change and 16 Reconstruction, (Toronto: McClelland and Stewart in cooperation with the Canadian Council of Teachers of English, 1974), pp.1-143. Transformationalists tried to come to terms with what was and was not grammatical. Following is a typical footnote: "11 The same seems to hold true for English, also, at least for some dialects. Observe the following sentences: (i) A girl that John knew deceived a boy that he disliked, (ii) A girl that he knew deceived a boy that John disliked, (iii) It is the boy that John disliked that the girl that he ^knew deceived. (iv)?It is the boy that he disliked that the girl that John knew deceived... There seems to be a wide fluctuation in native speakers' intuition on the grammaticality or ungrammaticality of these and related sentences." The above was taken from Susumu Kuno "The Position of Locatives in Existential Sentences," Linguistic Inquiry, Vol.11, No.3 (Summer 1971), pp.343. W. Labov commented on this difficulty that transformationalists had: "When challenges to data arise on the floor of a linguistic meeting, the author usually defends himself by stating that there are many 'dialects' and that the systematic argument he was presenting held good for his own 'dialect'. This is an odd use of the term, and it raises the question as to what the object of linguistic description can or should be." Quote taken from "The Study of Language in its Social Context," Studium  Generale, Vol.23 (1970), p.37. 9 We agree with the evaluation of the use of pictures which A.M. Kinloch put forth in his article, "The Use of Pictures in Elicitation," American Speech, Vol.46, (1971), pp.38-46. ^We made use of the sequence of pictures in P.R. Hawkins; Social  Class, the Nominal Group and Verbal Strategies, (London: Routledge and Kegan Paul, 1977), pp.56-57. 13"Labov, 1977, p.215. 12Trudgill, 1974, pp.60-61. 13 "Population: Geographic Distribution Urban and Rural Distribution," 1976 Census of Canada, (Statistics Canada, Ottawa, 1976), p.7-1. ^"You Don't Say," unpublished monthly bulletins compiled and dis seminated by the Office of Broadcast Language, eds. Lamont Tilden and George Rich, (CBC, Toronto, 1975-78), volume 1, issues 1-20; volume 2, issues 1-20. 17 15CEC News Style Book, (CBC, Toronto, 1971), pp.1-83. 16 See footnote 3 and the following works: Hans Kurath, Handbook of  the Linguistic Geography of New England, 2nd ed., (New York: AMS Press, 1973), pp.1-527; Albert H. Marckwardt, "Principal and Subsidiary Dialect Areas in the North-Central States," in RAD, pp.74-82; Harold B. Allen, The Linguistic Atlas of the Upper Midwest, (Minneapolis: University of Minnesota Press, 1973-6), vol.1:1-425, vol.2:1-92, vol.3:1-362; Harold B. Allen, "The Primary Dialect Areas of the Upper Midwest," in RAD, pp.83-93; Harold B.Allen, "The Minor Dialect Areas of the Upper Midwest," in RAD, pp.94-104; David W. Reed, "Eastern Dialect Words in California," in RAD, pp.105-114; Carroll E. Reed, "The Pronunciation of English in the Pacific Northwest," in RAD, pp.115-121. "^See footnote 2. 18 Labov, 1966, op.cit. 19Trudgill, 1974, op.cit. 20 (Washington, D.C: Center for Applied Linguistics, 1968) ,pp. 1-128. 21 See footnote 1. 22 H.B. Allen, "Canadian-American Speech Differences along the Middle Border," JCLA, vol.5, no.1, (1959), pp.17-24; W.S. Avis, "Speech differ ences along the Ontario-United States border, 1 vocabulary," JCLA, vol.1, no.l, (1954), pp.13-17; W.S. Avis, "Speech differences along the Ontario-United States border, II grammar and syntax," JCLA, vol.1, no.l, (1955), pp.14-19; W.S. Avis, "Speech differences along the Ontario-United States border, III pronunciation," JCLA, vol.2, no.2, (1956), pp.41-59; and CF. Reed, "Word Geography in the Pacific Northwest," Orbis, vol.6, (1957), pp.86-93; CF. Reed, "The pronunciation of English in the Pacific North west," Language, vol.37, (1961), pp.559-564. 23 J.K. Chambers, "Canadian Raising," in Canadian English: Origins  and Structures, ed. J.K. Chambers (Toronto: Methuen, 1975), pp.83-100; R.J. Gregg, "Neutralisation and Fusion of Vocalic Phonemes in Canadian English as Spoken in the Vancouver Area," JCLA, vol.3, (1957), pp.78-83; R.J. Gregg, "The Diphthong ai and ai in Scotland, Scotch-Irish and Canadian English," CJL, vol.18, no.2, (1973), pp.136-145; R.J. Gregg, "The Phonology of Canadian English as Spoken in the Area of Vancouver, British Columbia," Canadian English, ed. J.K. Chambers, (Toronto: Methuen, 1975), pp.133-144; -M. Joos, "A Phonological Dilemma in Canadian English," Language, vol.18, (1942), pp.141-144; D.E. Hamilton, "Standard Canadian English: Pronunciation," Proceedings of the Ninth  International Congress of Linguists, ed. H.C Lunt, (The Hague: Mouton, 1964), pp.456-459. 18 W.S. Avis, "The Phonemic Segments of an Edmonton Idiolect," Canadian English: Origins and Structures, ed. J.K. Chambers, (Toronto: Methuen, 1975), pp.118-128; J.K. Chambers, "The Ottawa Valley 'twang'," Canadian English: Origins and Structures, ed. J.K. Chambers, (Toronto: Methuen, 1975), pp.55-59; R.J. Gregg, "The Linguistic Survey of British Columbia: the Kootenay Region," Canadian Languages in their Social  Context, ed. Regna Darnell, (Edmonton: Edmonton Linguistic Research, 1973) , pp.105-116; D.E. Hamilton, "The English Spoken in Montreal: A Pilot Study," (unpublished M.A. thesis, Universite de Montreal, 1958), pp.1-74; A.M. Kinloch, "The Survey of Canadian English: Possible Evidence for Pronunciation," EQ, Vol.4, No.4, (1971), pp.59-66; A.M. Kinloch, "The Survey of Canadian English: a first look at New Brunswick results," EQ, Vol.4, No.4, (1972-3), pp.41-54; J. Poison, "A Linguistic Question naire for British Columbia," (unpublished M.A. thesis, University of British Columbia, 1969), pp.1-122; M.G. Wanamaker, "Survey of Canadian English—focus on Manitoba," Classmate (Official publication of Manitoba Association of Teachers of English), Vol.4, No.2 (1974), pp.37-43; H.R. Wilson, "Lunenburg Dutch: Fact and Folklore," Canadian English: Origins  and Structures, ed. J.K. Chambers (Toronto: Methuen, 1975), pp.40-44. 25 W.S. Avis et al., Dictionary of Canadianisms on Historical  Principles, (Toronto: Gage, 1967), pp. 1-926; W.S. Avis, "The English language in Canada," Current Trends in Linguistics, ed. T.A. Sebeok, Vol.10, No.l, (1973), pp.40-74; J.K. Chambers, ed., Canadian English:  Origins and Structures, (Toronto: Methuen, 1975), pp.1-144 (hereafter cited as Canadian English); C.J. Lovell, "Lexicographical Challenges in Canadian English," JCLA, Vol.1, No.l, (1955), pp.2-5; R.E. McConnell, Our Own Voice: Canadian English and how it is Studied, (Toronto: Gage Educational Publishing, 1979), pp.1-275, H1-H48; R.I. McDavid, "Linguistic Geography in Canada: an introduction," JCLA, Vol.1, No.l, (1954), pp.3-8; R.I. McDavid, "Canadian English," AS, Vol.46, (1971), pp.287-289; M.H. Scargill, "The Sources of Canadian English," JEGP, Vol.56, (1957), pp.610-614, reprinted in Canadian English, pp.12-15; M.H. Scargill, A Short History of Canadian English, (Victoria: Sono Nis, 1977), pp.7-63; Cornelius von Baeyer, Talking about Canadian  English, (Ottawa: Supply and Services Canada, 1976), pp.1-74; Cornelius von Baeyer, The Ancestry of Canadian English, (Ottawa: Supply and Services Canada, 1977), pp.1-62. M.A. Kinloch, "The Survey of Canadian English: Possible Evidence for Pronunciation," English Quarterly,' Vol.4, No.4 (Winter, 1971) , pp. 59-65; M.A. Kinloch, "The Survey of Canadian English: A First Look at New Brunswick Results," English Quarterly, Vol.5, No.4, (1972-1973), pp.41-51; L. Rodman, "Characteristics of B.C. English," EQ_, Vol.7, No.4, (1974-5), pp.49-82; M.H. Scargill, Modern Canadian English Usage:  Linguistic Change and Reconstruction, (Toronto: McClelland and Stewart, 1974) , pp.7-143; M.H. Scargill and H. Warkentyne, "The Survey of Canadian English: A Report," EQ_, Vol.5, No.3, (1972), pp.47-104; H.J. Warkentyne, "Contemporary Canadian English," American Speech, Vol.46, (1971), pp. 193, 199; 19 I grew up on that border in Port Huron, Michigan opposite Sarnia, Ontario, son of a Canadian National Railway family, with my father and grandfather having worked many years on both sides of the border. 28 R.J. Gregg, "Canadian English," Varieties.,of English: Common wealth English Series, ed. Y. Matsumura, (Japan, forthcoming), MS. pp.3-12. 29 To date two papers have been read publicly from the pilot survey; they are: R.J. Gregg, "Urban Dialectology: a pilot survey of.the English spoken in the city of Vancouver, B.C.," read at the Learned Societies Conference, Fredericton, 1977, pp.1-9; M. Murdoch, "Reading Passages and Informal Speech," read at the Third International Conference of Dialectology Methods, London, Ontario, .1978, pp.1-7. Abbreviations for names of journals: AS American Speech EQ English Quarterly CJL Canadian Journal of Linguistics (1961- ) JCLA Journal of the Canadian Linguistics Association (1954-60). CHAPTER 2 OTTAWA 1. Ottawa, its Settlement and Setting This work in Canadian English dialectology and sociolinguistics takes the form of a study of the English spoken in the city of Ottawa, Ontario, the political capital of Canada and the nation's fourth largest city. The population of the metropolitan area is growing rapidly and is presently at 693,288 inhabitants, with 302,341 actually within the boundaries of the city proper."'" In addition, the village of Rockcliffe Park, population 2,017, and the city of Vanier, population 2 18,237, are surrounded by the city of Ottawa. Ottawa is situated on the Ontario-Quebec border on the south bank of the Ottawa River where the Rideau River joins it, some 100 miles upstream from the St. Lawrence River, 114 miles west of Montreal, and 220 miles northeast of Toronto. Ottawa, Hull, and adjacent suburbs on both sides of the Ottawa River form the metropolitan area of Ottawa-Hull and a somewhat larger area forms the vague entity called the National Capital Region. Bytown, as Ottawa was first called, was a construction site for the building of the Rideau Canal and the centre for lumber trade. Queen Victoria selected it to be the capital of Upper and Lower Canada in 1857, and its capital status has been its main characteristic ever since. This factor has important linguistic consequences. The population of Ottawa has become in many ways a representative cross-section of the country as a whole. Firstly, since World War II thousands of individ uals with specialized education and skills have been hired annually from across Canada to take on responsibilities within the Federal Government, and naturally the political representatives and their staffs come from all regions of Canada. This movement to Ottawa has brought about a linguistic mixing and levelling process which tends to decrease the importance of the local speech characteristics. It is not only the number of these newcomers but also the fact that these people hold important positions which has linguistic relevance. Secondly, and possibly by chance, the Francophone population of Ottawa is very close 3 to the national average of 27 percent. Ottawa is also typical of many Canadian cities in that it has acted as a goal for in-migration and commuting from the surrounding rural area. This rural area is, however, not linguistically typical of General Canadian and needs to be inves tigated in order to understand the local sub-stratum of urban Ottawa. Ottawa in the Ottawa Valley No summary of the settlement and setting of Ottawa can be made in any linguistic work without mentioning the surrounding Ottawa Valley. The Ottawa Valley here is defined as the agricultural region of the Ottawa River watershed in Quebec as well as Ontario. This region, settled mainly by the Irish and the Scots and later by Quebecers and some Poles and Germans, forms one of the most distinctive rural dialect areas in Canada. (See Map 1.) Some relic linguistic influences of the 'Ottawa Valley Twang' can be found in the city of Ottawa today. This study, however, will comment on only those items which were recorded during our urban survey. For an in-depth study of the Ottawa Valley dialect one should consult the work of Enoch Padolski and Ian Pringle now in progress."' The city of Ottawa serves as a cultural, trade, and employment centre for much of the Ottawa valley. Socio-economic Considerations in Ottawa Within the city of Ottawa the socio-economic differentiation is not as clearly a west-to-east gradation of rich to poor as in a city such as Vancouver. Fine houses have been built along the canal, rivers, parks, and parkways, but only two blocks removed from these houses one often finds quite humble dwellings. This pattern tends to make the official census tracts quite heterogeneous and therefore very misleading. Personal income and salaries are the highest per capita in the nation. Furthermore, Ottawa, among Canadian cities, has by far the highest percentage of university graduates. There may also be a cer tain inflation of title and position which may influence any socio economic comparisons with other Canadian cities. Communications and Culture Ottawa has air, rail, bus, and expressway links to Montreal and Toronto and, mainly through these cities, to the rest of the world. Ottawa has two English evening newspapers and one French. The Globe and  Mail from Toronto serves as the morning paper. Ottawa's English tele vision stations are the CBC, CTV, Global, and Ontario Educational, and via cablevision the American networks NBC and CBS are relayed from Rochester, New York and the Public Broadcasting System (Educational) from Watertown, New York. Ottawa's cultural opportunities both in the field of enjoyment and 23 Table 1. 'Income, Education, and Migration Statistics for Canada's' largest Cities. Census Average University In-migrants Immigrant Metropolitan Family Graduation from the rest Population Area Income $ 1971 % of Canada Total % Calgary 10,943 7.3 78,410 20.5 Chicoutimi- 9,162 3.9 10,535 1.4 Jonquiere Edmonton 10,660 6.4 80,450 18.3 Halifax 10,176 6.8 32,605 7.2 Hamilton 10,757 4.3 45,755 26.7 Kitchener 10,661 4.6 32,890 21.8 London 10,763 5.6 42.565 20.0 Montreal 10,292 5.5 160,390 14.8 Ottawa-Hull 12,010 9.5 85,560 12.5 Quebec 10,159 5.7 52,150 2.2 Regina 9,637 5.7 25.465 13.1 St. Catharines- 9,997 3.4 23,245 22.9 Niagara St. John's 8,488 3.6 14.435 3.0 Saint John 8,821 3.5 9,850 4.9 Saskatoon 9,479 7.0 27,240 13.9 Sudbury 11,739 3.6 22,825 12.4 Thunder Bay 10,165 3.4 10.620 21.1 Toronto 11,841 6.3 185,530 34.0 Vancouver 10,664 5.7 131.555 26.5 Victoria 9,921 5.4 35,650 24.7 Windsor 11,281 3.9 18,600 21.5 Winnipeg 9,989 5.7 58,590 19.9 Metropolitan 10,788 5.8 20.8 Canada Nonmetropolitan 8,062 2.7 8.5 Canada Canada 9,600 4.4 15.3 From Canadian Urban Trends: Metropolitan Perspective, Vol.2, ed. D. Michael Ray, (Toronto: Copp Clark Publishing, 1977), Tables 4.1, 4.3, 1.6 and 5.1; pp.7, 40, 43, and 71. employment are much greater than are found in other Canadian cities of comparable size. Quite naturally, because Ottawa is the capital of the confederation, it is the home of such institutions as the National Art Gallery, The National Museum of Man, The National Library and Public Archives, The National Arts Centre and the National Museum of Science and Technology. Ottawa also contains Carleton University, The University of Ottawa, St. Paul's University, and Algonquin College. 2. Ottawa Valley Urban Centres Eleven informants from urban centres in the Ottawa Valley were interviewed in order that we could compare the speech characteristics of native Ottawans with the speech characteristics of people from nearby towns, namely Renfrew and Smiths Falls. We were also interested in ascertaining to what extent the Ottawa Valley 'twang' could still be found among townspeople in that area. J. K. Chambers in his article "The Ottawa Valley 'twang'" states: A visitor in Carp or Arnprior or Killaloe is likely to spend a lot of time there nowadays before he comes upon a native who speaks much differently than he himself does. The folklore is even much less credible, of course, as it applies to the urbanized centres of the Ottawa Valley like Ottawa, Hull, Renfrew and Pembroke. Residents of southern Ontario who move to those centres, as hundred do annually, may very well live the rest of their days in them without contacting a speaker of the twang. It was not always thus. The folklore has its basis in fact, and even a generation ago might have been termed generally true. Now the Ottawa Valley dialect survives only in what ever isolated rural communities remain. Along the main routes, the dominant dialect has come to be the dialect of heartland Canada.? We hypothesize that people from these towns are speakers of General Canadian English with perhaps a few vocabulary items remaining from the 'twang'. We will find some of these items of 'Valley Talk' extant in the towns. We will determine to what extent these items are in the speech of native-Ottawans. Renfrew g Renfrew, a town of 8,530 inhabitants, is an agricultural centre which also has some secondary industry in electronics and machine parts. It is situated on the CN, the CPR, and the Trans-Canada Highway, sixty miles west of Ottawa. Smiths Falls Smiths Falls, situated fifty miles southwest of Ottawa on the Rideau Canal, highways 15 and 29, and CN and CPR lines, is a commercial and service centre for the surrounding agricultural lands and the Rideau Lakes. It also has some secondary industry. Unfortunately, a number of secondary industries have failed or moved recently. The 9 population is 9,149. 26 Chapter 2: Footnotes ^"Ottawa, Statistics Canada, 1976 Census of Canada: Population:  Preliminary Counts, (1976), pp.50, 52, 54. 2 Ottawa, Statistics Canada, Census Tracts, Ottawa-Hull, (1971), pp.1-10. 3Ibid, pp.1-10. S.K. Chambers, "The Ottawa Valley 'twang'," Canadian English:  Origins and Structures, ed. J.K. Chambers (Toronto: Methuen, 1975), pp.55-59. The above article offers a short summary of the dialect situation in the Valley. ^E. Padolsky and I. Pringle, "Reflexes of M.E. Vowels before /r/ in Ottawa Valley Dialects of Hiberno-English Types," unpublished article read at the Learned Societies Convention at Fredericton, N.B., (1977), pp.1-13. Both researchers are professors at Carleton University, Ottawa. ^Canadian Urban Trends: Metropolitan Perspective, Vol.2, ed. D.M. Ray, (Toronto: Copp Clark Publishing, 1977), pp.39-42. (See our Table 1 compiled from the above source.) ^Chambers, op.cit., p.55. Population: Preliminary Counts: 1976 Census of Canada, (Ottawa: Statistics Canada, 1976), p.54. Ibid., p.50. 27 CHAPTER 3 CANADIAN ENGLISH 1. Canadian English in Relation to Other Dialects In order to understand the variety of English with which we are dealing, it is necessary that the dialect, Canadian English, be placed within the larger framework of the English language family. There are two main branches of the English language, the British branch and the North American branch. The British Branch The British branch includes the English dialects of the British Isles, Australia, New Zealand, Southern Africa, and a number of smaller colonies and former colonies. Approximately 91 million people are native speakers of this branch of the language.''" Some elements common to most of the dialects within this branch and distinguishable from North American English are: 1) the pronunciation of [a] in words such as dance, can't, half, and grass where the vowel does not immediately precede a stop, 2) the pronunciation of ft] in medial position in words like city or dirty, 3) the 'r less' pronunciation, i.e. the deletion of /r/ in word and morpheme final position and pre-consonantal position, e.g. car and party, 4) the syntactic combination of modal verb followed by the auxiliary verb do_ in short form sentences such as We would do. and 5) the use of got as the past participle of get. Most people throughout the world still choose Standard Southern British, SSB, the 28 prestigious dialect within this branch, as their model when learning English as a Second Language. The North American Branch The North American Branch is much larger; approximately 211 million 2 people speak it natively. Although conservative in some respects, this branch has developed gradually away from its source, which was,of course,. the English spoken in Britain in the seventeenth and eighteenth centuries. The American colonies were the first area to be settled, and at that time, 1620-1775, trans-oceanic communications and overland transportation 3 were primitive. It was quite natural that the two branches grew apart. At that time, innovations were made to the language on the new continent mainly in the area of vocabulary because of the presence of previously unknown flora and fauna and as a result of the contact with the Indians. In addition, the inevitable internal change of language usage over time was happening in the mother country as well as in the colonies. London and the surrounding areas of Southern England, for example, were in the process of losing final and pre-consonantal /r/. With the onset of the Revolutionary War, 1776, and independence in 1781, it was only natural that within the United States linguistic diversity would develop even more rapidly and that any difference in usage would be sought out, magnified, and institutionalized. The Settlement of Canada It was in 1783 and 1784, the year immediately after the Revolutionary War, that Canada first became settled in any significant 4 number by English speaking people. These settlers, wishing to remain 29 under the British Crown and the rule of law and order, were called United Empire Loyalists; some moved from New England and further south onto the coastline of Nova Scotia, a region later to be called the provinces of New Brunswick, Nova Scotia, and Prince Edward Island. Others settled in Quebec City and Montreal. From Vermont, up-state New York, western Pennsylvania, and elsewhere, United Empire Loyalists of generally humbler status and a more frontier background moved over land to Sorel, Quebec,where they were given food and shelter in a refugee camp for two years while the initial land survey was being conducted. These Loyalists eventually settled along the northern bank of the St. Lawrence River and the northern shores of Lakes Ontario and Erie in what is now Ontario."' (See Map 3.) There were, perhaps, only 10,000 United Empire Loyalists all told who settled Ontario.^ Many Loyalists who first had settled in Nova Scotia later moved to Upper Canada upon hearing reports of a more favourable climate and better soil conditions. A somewhat larger number of settlers moved from the northern states to Upper Canada a decade or two later (hence their name 'Late Loyalists'), taking up land concessions in a program sponsored by the Crown. These two groups together, the Loyalists and the late Loyalists, formed in Upper Canada a population of about 100,000 by 1812.7 Linguis tically, this was the most important settlement in British North America, because it was Ontario which was destined to become the rich and densely populated heartland of Canada. Could the sparse settlement of 100,000 form the basis of a national language by each summer assimilating ship loads of scores of thousands of British immigrants? During the first half of the nineteenth century, from 1812 to 1850, 800,000 immigrants from Britain, mainly Scots and g Irish, settled in the same areas of Upper Canada. It remains a con troversy today whether the American immigrants or the British immigrants 9 laid the foundation for Canadian English. What seems important in summing up the development of Canadian English in those early years from 1782 to 1850 before Canada existed is: 1) that both American and British immigration and settlement contributed jointly to the develop ment of Canadian English, 2) that the British and American dialects must have been less divergent than they are today,3) that each major dialect, i.e. American and British, must have been spoken in Canada by hundreds of thousands of speakersand 4) that Canadian English was developing independently of the two by picking and choosing what it preferred and by innovating on its own when it did not like the choice or when there was no question of choice. After 1850, we see continued immigration from Britain and the United States and a growing prosperity in southern Ontario and Montreal. As western Canada is being settled, we see three major English speaking groups settling the prairies and influencing the language; these are the,Americans and the British again but also for the first time the Canadians, mostly Ontarians. The Canadian style of English prevails, and this pattern continues as Canada expands westward and northward. Southern Ontario continues to grow and dominates the rest of Canada industrially, commercially, and politically, and as a consequence of dominating in these three fields, it sets the linguistic standard. Present Situation If one looks at the language situation across Canada today, one will see that the Maritime provinces still retain a style of speech substantially different fromsthat of<the rest of Canada, a result of the influx of settlers from coastal New England since the 1760's and from continuous cultural and economic ties with New England ever since. Newfoundland too has a unique style of English based mainly on the speech characteristics of settlers from Ireland and from the southwest of England. The speech of individuals who live in the outports is the 12 most different from General Canadian that one will encounter in Canada. The speech heard in the cities, however, resembles more and more the speech heard across Canada. In fact, recent interviews of Newfoundland high school students sounded almost indistinguishable from their counterparts in suburban Toronto. This same language levelling among young people is also very noticeable not only in the Maritime provinces but also in rural dialect pockets across Canada. English in rural regions of Quebec is almost non-existent except for the Eastern Townships and Pontiac County. The Eastern Townships were settled by Loyalists and Late Loyalists from western New England and up-state New York. French is now replacing English in this area. Pontiac County is part of the Ottawa Valley and has speech charac teristics which conform to that region. Anglophone Montrealers sound 13 very much like speakers from the rest of Canada. As stated previously, overland settlers from western New England, up-state New York, New Jersey and Pennsylvania moved into Ontario. Of those items which are of American derivation in Canadian English a predominant number are from the Northern American dialect - the prestige dialect of the United States and the major contributor to General American. The geographic region of the Northern American dialect is 32 also the rich industrial heartland of the United States. The Northern American dialect area is important to Ontario English not only because it was the source of immigration from the United States, but also because it is the area of the United States with which Ontario has maintained the closest ties. All this is not to say that Midland America did not play a meaningful role, for Midland speakers also immigrated to southern Ontario. This explains why we have Midland forms in Canadian speech, e.g. blinds, dew worm, and coal oil. These forms were most easily accepted when British immigrant usage coincided, as in the case of blinds. The prairies were settled by farmers moving west from Ontario, the American Mid-West (mainly Northern and some Midland speakers again), the Great Plains states, and from Britain. There also were a large number of Ukrainians, French Canadians, Germans and Scandinavians. Educated Englishmen often obtained, as they do today, positions of authority and prestige as administrators, clergy, officers, educators, civil servants, etc. This may be one of the most significant differ ences between the linguistic situation of Canada and the United States today, namely that British English enjoys a position of prestige and respect throughout Canada. This advantage is afforded to one generation only as the immigrants' children inevitably speak Canadian. The coastal region and the Okanagan region of British Columbia had a much larger number and proportion of settlers from Britain than did the Prairies, but even here, the Canadians from the Maritimes, Ontario, and the Prairies moved in and later assimilated them. Much of the interior of British Columbia was settled by miners from Utah and Idaho whose ancestors came from Midland areas. These miners.were later assimilated, too. 33 Canadian English thus is very uniform from Ontario to British Columbia and northward, and is becoming increasingly so in Atlantic Canada. The influences, Northern American, British, and Midland American were continuous and sometimes reconverging for over 200 years of development. As a result, a uniform Canadian dialect covers a larger land mass than any other one dialect in the world. Canadian English today is most definitely a North American dialect of English, be it from the Loyalists or from 200 years of constant contact across some 4,000 miles of international border. To illustrate how similar Canadian English and Northern American are, we can report that it takes most Britons a few years in Canada to distinguish the difference, and that some Canadian university students have attended lectures for one year without knowing whether their professor was American or not. These instances notwithstanding, Canadian English is a distinctive dialect within the North American branch of the English language family, and it enjoys the sociopolitical status of a national language, something which Northern, Midland, Southern or Western American English can not. Having examined the development and present state of Canadian English, let us now investigate those linguistic characteristics which distinguish it from educated Northern American, the dialect most closely 14 related to Canadian English. 2. Linguistic Features As a result of 1) listening to Canadian and Northern American English, 2) noticing how foreigners categorize Canadian English and 3) analysing our survey results which indicate that Canadians and Northern Americans notice little difference between each other, it is clear that Canadian English and Northern American (also General American) are very similar. Acknowledging this similarity and using it as a reference point, the next section will outline the disting uishing characteristics of Canadian English. We will now take the major linguistic categories of language and describe the differences which exist. Those items which are unique to Canadian speech or which may be suspected of being in a state of change will then be the basis for the stylistic and sociological survey which is to follow. Phonology It is at the level of phonology that we find the greatest systematic difference between Canadian English and Northern American. Any difference here in the respective systems could mean a difference in pronunciation of hundreds of words. We will now look at the seg mental units of these two dialects and compare them. The Consonants The consonantal systems of Canadian English and Northern American are phonemically identical. In fact all other major English dialects seem to have a consonantal system which is tabulated below. 35 Labial Dental Alveolar Palatal Velar Glottal Stops P b t d k g-Affricates tj d^ Nasals m n Fricatives f V e 5 s z J' 3 h Lateral 1 Frictionless Continuant r Glides w j It is the allophonic realization of these phonemes and their combina tions which differentiate the dialects. 1. /tj, dj , nj/ Perhaps the most noticeable speech characteristic associated with the consonants is the pronunciation by many Canadians of the yod, /j/, when followed by an /u/ and preceded by a It/, /d/, or /n/. Canadians realize this pronunciation frequently when the word containing the sound is pronounced in isolation or when the word is in a stressed position within a sentence. Americans do not have the pronunciation with yod as a goal. Word Canadian Northern American 1. tube [tjub] [tub] 2. Tuesday [tjuzdei] [tuzdei] 3. tune [tjun] [tun] 36 Word Canadian Northern American 4. student [stjudant] [student] 5. stupid [stjupad] [stuped] 1. dew [dju] [du] 2. dual [djual] [dual] 3. due [dju] [du-:]. 4. duke [djuk] [duk] 5. dune [djun] [dun] 6. dupe [djup] [dup] 7. duplex [djupleks] [dupleks] 1. new [nju] [nu] 2. nude [njud] [nud] 3. avenue [svanju] [asvanu] 2. /hw/ Similarly, many Canadians pronounce a voiceless version of /w/, i.e., /hw/ or /M/, in words like whether, where, what, which, why, whine, white, etc., when these words are stressed. Americans typically do not hold this as a goal. These two differences between Canadian and American speech within the area of consonantal usage will be investigated thoroughly in the stylistic variation portion of the survey. The Vowels 1. The Vocalic System The vocalic systems of Canadian and Northern American are somewhat different. The Canadian system is illustrated below. Illustration 3.1 beat The Canadian Vocalic System As one can see, there are 10 stressed vowel phonemes plus the schwa which is an allophonic variant of any vowel sound in an unstressed position. The Northern American vocalic system differs in one item only: where Canadian English has the back, open, rounded phoneme /n/, Northern American has the additional unrounded /a/. 38 cot 1 caught a [ kat ] • • 1 t> [ kDt ]_ Northern American thus has a contrast in sound and meaning between cot and caught; and between caller [knler] and collar [kdler]. Most Canadians actually use both these sounds but in free variation, 1. e. without making any distinction in meaning. They seem to make their choice lexically or at random. We will analyse the nature of this free variation in the study. 2. Canadian Diphthongs There are three diphthongs common to the vocalic systems of Canadian English, Northern American and all other major English dialects; these are: 1. /ai/ as in buy, 2. /ao/ as in bough, and 3: /oi/ as in boy. 39 Illustration 3.2 English Diphthongs The allophonic distribution of the second of these three diphthongs is what differentiates Canadian English most markedly from North American, and it is responsible for the shibboleth around and about the house. In the environment where the 'ou' diphthong is followed by a voiceless con sonant it is pronounced [Au ] or sometimes [eu], e.g. south [SAUG] and out [Aut]. This diphthong with a higher stressed element may also occur, but certainly less frequently, in environments where a voiceless con sonant is the underlying form, but where the surface form is realized^as voiced, e.g. as in shouted where the underlying form is [J/uted] but where the surface form is [jAudad], applying the medial /t/ voicing rule. Here we have an interesting case of rule ordering, i.e., does one select the diphthong with the higher stressed element first, then voice the /t/, 40 or does one first voice the /t/ and then choose the. diphthong with the lower stressed element unraised? ..Some manage to say [Jauded]. Illustration 3.3 Canadian Diphthongs with High Stressed Elements The other diphthong which has characteristic Canadian allophonic distribution is the /a 1/ diphthong. Parallel to the [AU] distribution, the [si] realization occurs when followed by a voiceless consonant, e.g. like [laik], night [nsit], or when the underlying form of the consonant is voiceless, e.g., writer is frequently pronounced [raider]. Canadians and Americans seem unaware of this difference in their speech charac teristics but subconsciously may include it in their evaluation that Canadian speech is more "clipped" and "crisp". In the study we will investigate these diphthongs thoroughly in many environments and styles. 3. Phonemic Reduction before /r/ Canadian as well as Northern American speakers have reduced the number of stressed vowel phonemes occurring before /r/ from the ten or eleven possible down to five: 1. pier [pir] 3. purr [per] 4. poor [por]^ 2. pear [per] 5. par [par] Some Canadians, however, pronounce an over-rounded [o] before /r/ in words like porridge, Dorothy, orange, and sorry. Also, many Canadians tend to retain a difference between: Mary ] \ [meYi ] merryJ marry) [mari ] Most Northern Americans have merged all three words to [meVi]. 4. Liaison Northern Americans frequently do not apply the second half of a prescriptive rule which states that one should say a. [a], the [6e], _to [te], etc., before words beginning with consonant sounds and an [an], the [3i]» _t° [tu]5 etc. before words beginning with vowel sounds. We will investigate these liaison features in Ottawa speech. Pronunciation of Words In addition to the phonological systems which have been described and contrasted above, there is the simpler factor of the choice of phoneme to be used in an individual word or set of words. The choice of phoneme used in a word gives us some of the most easily recognizable differences between Canadian and American English. In ;the list that follows, the pronunciation cited is not necessarily the most frequent variant, but it does represent a stylistic variant which most strikingly 16 distinguishes a Canadian from an American and vice versa. There are a few affixes with which we will deal first. Affix 1. -ile a. f. g-h. Examples) agile fertile futile hostile missile mobile proj ectile virile Canadian [ail] Northern American [ai] Comment The differentiation is high, i.e. most Canadians say [at I ] and most Americans say [al], cf. CEU, pp.80, 81, and Will. 17 2. anti-multi-semi-3. -ing [ant i ] [mAlti] Is£m'\ ] [IQ] [in] I eent a L ] [ mA* I ta i ] [semai] [iQ] [an] The differentiation is high, cf. CEU, pp.60, 61. Some differentiation, You Don't'Say, Vol. one, Issue 5, dis cusses this usage in Canada. 43 Word absurd again Canadian [abzard] Northern American Comment asphalt [ffisfnlt] [agei n] [absard] [fflJfDlt] [agin ] Algonquin [aelgnnkin] [ael gAnkw i n ] apricots [eiprakbts] [ffiprekats] Some di-fferentation, cf. DCE and Will. Some differentiation, cf. DCE and Will. Americans rarely say [agei n]; see CEU, p.72 for Canadian percentages. See DCE and Will. See CEU, p.54, and Gregg, 1973, pp. 109-113. balcony been blouse Datsun decal eh either [ br> I kan i ] [ baa I kan i ] [bin] [bIaoz] [ditsan] [d£kal ] [ei] [ai5ar] [be Ikan i ] [bin] [bIaos ] caramel [ keerame I ] [ karma I ] [ d'cttsan ] [dake* I ] [ d f ke I ] [hA] [f5ar] See DCE and Will. Americans rarely say [bi n]. See DCE for Canadian usage. Americans never say [blaoz]. See DCE for preferred usage. High differentiation, see CEU, pp.67, 68. High differentiation. High differentiation. High differentiation in tag sounds. See CEU, pp.78, 79 for Canadian per centages . eleven [ al e'van ] [ i I e'van ] Some differentiation. garage [gard^] [gar&d^] Some differentiation 44 Word Canadian Northern American Comment H 1. hoof [huf ] [hof ] Canadians rarely say [hof]. See CEU, p.75, for Canadian percentages of roof. Iroquois [ irakwd] [ Crakwo i] High differentiation, cf. DCE and Will. K, khaki [ k&rki] [ kaeki ] Americans never say [kdrki]. See DCE I and Will. leisure 2. lever [Ii?er] [ I fv.sr] 4. lilac [lailak] [ If^r] [ I ever] 3. lieutenant [ I a 11 ask ] See CEU, p.74, and Gregg, 1973:, pp.111, 112, for Canadian percentages. See CEU and Gregg, 1973', pp.112, 113 for Canadian per centages . Americans never say [Ieftenant]. See CEU p.7.3 for Canadian percentage. High differentiation, cf. DCE and Will. M 1. neither [na tSar ] [n'Sar] Usage percentages seem to approximate those for either, except in the phrase me neither. 0 45 Word Canadian Northern American p 1. produce n. [proudus] [prddus] 2. progress [prougres] [prdgres] 3. process [prouses] [prases] Q 1. Renault [renou] [ renalt] 2. roof [ruf] [ rof ] 3. root [rut] [rot] route [rut] [raot] Sl. schedule [J^d^ul] [skerd^ul] 2. senile [s^nail] [s'nail] 3. shone [J^n] [Jon] T 1. tomato [temseto] [tameido] Comment Some differentiation, see CEU, p.86, 87, for Canadian per centages . High differentiation. High differentiation, see CEU, p.75, for Canadian percentages. High differentiation, see Gregg, 1973, pp. 108-112 for Kootenay percentage. See CEU, pp.87, 88 and Gregg, 1973, pp. 108-113 for Canadian percentages. Americans never say [Jed^ul ]. See CEU pp.55,56, and Gregg, 1973, pp.109-113 for Canadian percentages. Some differentiation, cf. DCE and Will. Americans never say [Jnn ]; Canadians almost always do. See Gregg, 1973, pp. 108,113, for Canadian percentages. Americans never say [t ameeto ]; Canadians sometimes do. See CEU, pp.65, 66, and Gregg, 1973, pp.108, 113 for Canadian percentages. 46 Word Canadian Northern American Comment U w vase 1. were 2. weren't [ vnz] or [vaz ] or [ vei z ] [ wer] [wernt] [ vei s ] [wer ] [warnt ] Some differentiation, see CEU, pp.58, 59, and Gregg, 1973, pp. 109-113 for Canadian percentages. Some differentiation. Some differentiation. X Y Jl. Z 2. zebra [zed] [zebra ] [zi] [ zfbra ] High differentiation, see CEU, pp.59, 60, for Canadian per centages . Americans never say [zdbra]; Canadians sometimes do. Gregg, 197.3, 113, for B.C. centages. See pp.108-per-3. zero [ ziYou ] [zi:rou ] Some differentiation. Grammar: Morphology and Syntax The grammatical structure is the most conservative element of any language. The phonology and lexicon of various dialects of any language vary widely one from another, but it is at the levels of morphology and syntax that dialects are most similar. The items which differentiate are: 47 Canadian Northern American Comment Do you not. . Are you not. Don't you.. Aren't you. Tuesday next Tuesday week Air Canada Health and Welfare Canada Labour Canada Statistics Canada Sport Canada Transport Canada Lotto Canada etc. eh (as a substitute for question tags, etc.) next Tuesday a week from this coming Tuesday Canadians may use Don't you and Aren't you but Americans rarely use the Do you not and Are you not forms. Tuesday next is normal British usage. Tuesday next is normal British usage. This type of word order, started in governmentese, is patterned after French and is one of the results of the bilingualism policy and language contact. Americans use eh when teasingly reprimanding a friend and other wise only extremely rarely. Canadians use huh only extremely rarely. Canadians may prefer Have you got... more frequently than do Americans. See Avis, 1954, pp.13-17. In addition, there may be a few Standard Southern British forms and usages which are used more frequently by Canadians than by Americans. huh Have you got.. Do you have. SSB USA 1. got gotten (as in the past particple of get) 48 SSB USA 2. Don't let's~rr Let's not... 3. proved proven (as in the past participle of prove) Lexicon It is at the level of lexicon that we find the greatest number of items which differentiate Canadian English from Northern American. Moreover, it is at this level more than any other that one can see that Canadian English has developed independently. Many Canadian words and phrases relate to uniquely Canadian experiences and therefore will be found nowhere else but Canada. The Lexicographical Centre for Canadian English located at the University of Victoria under the direction of Professor Scargill has produced two books which deal specifically with 18 this subject. These books are A Dictionary of Canadianisms edited by 19 W. Avis and A Short History of Canadian English by M.H. Scargill. Below, we list some common Canadian items which are part of the vocabulary of everyday life and which differentiate Canadians from Northern Americans. The claim is not that all Canadians use a particular term 100 percent of the time, but rather that that usage coupled with other suohousages will distinguish a Canadian from a Northern American. The items are presented according to semantic area. 49 Canadian Grade 1, 2, etc. someone in grade 3 Grade 13 secondary school senior secondary junior secondary elementary school matriculated Separate Schools brush (blackboard) supply teacher elastic (band) zed grades invigilate (tests) invigilator residence calendar (bulletin) EDUCATION Northern American first grade, etc. third grader junior high school and high school high school junior high school grade school graduated Parochial or Catholic Schools eraser substitute teacher rubber band zee marks monitor dorm(itory) catalog Comment 100% differentiation High differentiation Grade 13 does not exist in the United States and some provinces. Americans never say collegiate institute. Some Canadians say (junior) high school Some Canadians say (senior) high school Some Canadians say (junior) high school Some differentiation Matriculated is not heard in the U.S. Parochial Schools or Catholic Schools is not heard in Canada. Brush is not heard in the U.S. High differentiation High differentiation Americans never say zed. High differentiation High differentiation High differentiation High differentiation collegiate institute high school 50 Canadian 19. university Canadian 1. bag (grocery) blinds Northern American college (in a phrase like 'gone off to _. ') HOUSEHOLD Northern American sack shades 3. braces 4. budgie 5. chesterfield 6. cutlery flatware 7. hydro bill hydro pole etc. 8. porridge 9. serviette 10. tap(s) suspenders parakeet davenport silverware electricity bill electricity or telephone pole oatmeal napkin faucet Comment High differentiation, Canadians usage is university in such phrases. Comment High differentiation, see Avis, 1954, p.13. Some differentiation, see Gregg, 1974, pp. 108-114 for Canadian and American per centages . Some differentiation. Some differentiation High differentiation, see CEU, pp.106, 107, and Gregg, 1974, pp. 108-112. Low differentiation Americans do not have this usage. High differentiation Low differentiation, see CEU, pp.116, 117. Faucet is gaining in usage in Canada for one fixture which combines hot and cold water. Tap is used in both countries for the outside fixture. See CEU, pp.107-108, and Gregg, 1974, pp. 108-115 for Canadian and American percent ages . 51 Canadian 11. veranda 12. wallet Northern American porch billfold Comment Some differentiation. Few Canadians say billfold. 4. 5. 6. 7. Canadian elected by acclamation alderman concession concession road concession line county town reeve (of a municipality) fire hall postal code LOCAL POLITICS, PUBLIC AFFAIRS Northern American Comment postie elected without opposition councilman land grant county seat mayor fire station zip code mailman social insurance number (SIN number) social security number This usage is not found in the States. See DCE and The Dictionary of  Canadianisms. Councilman is not used in Canada. Councillor is used in P.E.I. Concession is rarely used in the States. High differentiation. Reeve is rarely used in the States. Some differentiation. Some differentiation, postal code is pat terned after French and is a result of the bilingualism policy and language contact. Zip code is heard very frequently in Canada. Some differentiation. Postie is not heard in the States. Mailman is most common in Canada. High differentiation. 52 10. 11. 12. Canadian garbage returning officer (an official in charge of an election in a constituency. scrutineer (one who examines votes during an election) Northern American garbage (food products) junk (non-food products) trash (non-food products) Comment High differentiation. Returning officer is not used in the States. Scrutineer is not used in the States, 13. hustings (a platform from which a candidate makes speeches) Hustings is ^seldom used in the States. 14. riding (a political divis ion represented by an M.P., M.L.A., M.P.P., etc.) 15. backbencher (an ordinary member of Parliament or of a legislative assembly) congressional district, state district Riding is not used in the States. Backbencher is not used in the States. 16. rate payer (one who pays municipal taxes) Rate payer is not used in the States. 17. enumerator (one who, prior to an election, registers eligible voters) 18. white paper green paper (government reports) Enumerator is not used in the States. This usage is seldom found in the States. 53 Canadian 19. constable 20. Baby Bonus ($20.00/month/child) Northern American policeman Comment Constable is rarely used in the northern states. This usage is not found in the States. Canadian Anglophone Francophone dew worm reserve (for Indians) the ex(hibition) CNE, CCE, PNE, etc. the old country (Europe or most often Britain is how being extended to Asia) the line (meaning the U.S.-Canada border) the 49th parallel (meaning the U.S. Canada border even in N.B., Que., Ont., Man. and B.C.) down East (meaning the Maritimes) MISCELLANEOUS Northern American Comment An English speaking Anglophone is unknown person in the States. A French speaking Unknown in the States, person angle worm night crawler reservation state fair High differentiation. High differentiation. High differentiation. High differentiation. High differentiation. High differentiation. High differentiation. 54 Canadian back East (anywhere in Canada East of the Lakehead) the States beer parlor (a room in a hotel where beer is sold) mickey (a small whisky bottle shaped to fit in the hip pocket) rye red ribbon (first place) Northern-American America the U.S.(A) Stateside beer garden flask whiskey blue ribbon Comment High differentiation High differentiation High differentiation, see DCE. Mickey is not used in the States. This meaning of rye is uniquely Canadian. A red ribbon is the award for second place in the States. Girl Guides biscuit Girl Scouts crackers cookies High differentiation Some differentiation, there is some over lapping of meaning of these three words. chocolate bar cache ( a place for storing supplies; to store away) candy bar stampede Landed Immigrant (one admitted to Canada as a settler and potential citizen) rodeo (Legal) Alien Some differentiation, candy bar is seldom heard in Canada. This meaning of cache is uniquely Canadian. This meaning of stampede is uniquely Canadian. Landed Immigrant is less intimidating and uniquely Canadian. 55 Canadian 22. New Canadian (one born in another country living permanently in Canada) 23. ethnic groups Canadians etc. (euphemism for foreign, i.e. not of the two founding peoples) 24. pink slip (proof of third party liability insurance) 25. running shoes 26. August Civic Holiday 27. Boxing Day 28. Remembrance Day 29. ramp (expressway) Northern American Comment Naturalized Citizen New Canadian is obviously unique. This usage of ethnic is unique. tennis shoes sneakers first Monday in Augus t December 26 Armistice Day "Veterans Day expressway exit and entrance This usage is uniquely Canadian. Some differentiation. This holiday does not exist in the United States. High differentiation. Websters III omits Canada from list of countries which celebrate this holiday. High differentiation. This usage is probably to accommodate French rampe. In addition to the above listed words and terms which help to dis tinguish a Canadian from an American, Canadian English contains many other usages, mainly of British origin but also French Canadian, which are little known or foreign to American English; some fairly common items included in the group are: balaclava, bands (of Indians), bank holiday, Bay Street, bilingualism, biscuit, bloody, bloke, buckshee, 56 bugger, chap, chips, Donnybrook, in future, Grits, holidays (vacation), the liquor store, muck about, queue, The Reserve (National Guard), a round about, second (verb), shadow cabinet, superannuation, Tory, tuque, 20 twit, the Van Doos (vingt-deux), waffle (verb), and write an exam. A large number of the 752 variables investigated in this study were chosen from among the phonological, pronunciation, grammatical, and lexical items presented in this chapter, thus giving the study a definite focus on Canadian English as it is differentiated from Northern American. 57 Chapter 3: Footnotes ^"The United Nations Statistical Yearbook, (New York: United Nations, 1977), pp.67-73. 2 Ibid., pp.67-73. 3 The coastal areas of North America did accept some linguistic innovation from the mother country, e.g. the 'r lessness' of coastal cities from Boston to Savannah. 4 Prior to the American Revolutionary War, there were a few thousand troops garrisoned in Halifax, Annapolis Royal, Quebec City, Montreal, etc. and there were some 7,000 New Englanders, 2,000 British, 3,000 Germans, and 8,000 Acadians living in Nova Scotia. N. MacDonald, Canada,  1763-1841 Immigration arid Settlement, (Toronto: Longmans, Green and Co., 1939), pp.41-73. ^The Loyalists drew lots, by status and rank, for their wilderness land, and they suffered greatly until they re-established themselves. N. Mika, H. Mika, United Empire Loyalists: Pioneers of Upper Canada, (Belleville: Mika Publishing, 1976), p.154. 6J.B. Brebner, "The Arrival of the Loyalists," The United Empire  Loyalists: Men and Myths, ed. L.F.S. Upton, (Toronto: Copp Clark, 1967)p.92. ^Helen I. Cowan, British Emigration to British North America, (Toronto: University of Toronto Press, 1961), p.12. Helen I. Cowan, British Immigration Before Confederation, (Ottawa: Canadian Historical Association, 1968), p.16. 9 The following sources give varying opinions about this question: Morton Bloomfield, "Canadian English: its Relation to Eighteenth Century American Speech," Journal of English and Germanic Philology, (JEGP), Vol. 47, (1948), pp.58-63; M.H. Scargill, "Sources of Canadian English," JEGP, Vol.56, (1957), pp.610-614; Walter S. Avis, "The English Language in Canada," Current Trends in Linguistics, Vol.10, part 1, ed. T. Sebeok, (1970), pp.40-68; M.H. Scargill, A Short History of Canadian English, (Victoria: Nis Sono Press, 1977), Chapts.l, 7, 8, 9; Cornelius von Baeyer, The Ancestry of Canadian English, (Ottawa: Supply and Services Canada, 1977), pp.1-7; R.J. Gregg, "Canadian English" Varieties of English:  Commonwealth English Series, ed. Y. Matsumura (Japan, forthcoming), MS. pp. 3-12. 58 "^For an historical perspective of this divergence and a detailed description of one phonological item, see Gregg, op.cit., pp.3-13. "'""'"This situation of two concurrent dialects has continued to the present day. One result of this fact is that most Canadians are bi-dialectal (British/North American), especially in listening and reading comprehension,to a much higher degree than their American counterparts. Another result may be that in the formal end of many Canadians' stylistic range, the British usage may be the goal. 12 General Canadian refers to a standard dialect which is spoken in most parts of Canada from the Ottawa River to the Pacific; it is roughly the dialect of broadcasters on the national networks and of the university educated. Increasingly, it is the majority dialect of all Canadian cities. 13 D.E. Hamilton, "The English Spoken in Montreal: A Pilot Study," unpublished masters dissertation, University of Montreal, Montreal. 14 For a detailed description of.how Canadian English and Standard Southern British differ, see Gregg, (forthcoming), op.cit., pp.13-40. "''"'Some Canadians and Americans pronounce words like poor, [per] and Moore, [nor]. Canadians seem to do this much more frequently than do Americans; see the Dictionary of Canadian English, eds. W.S. Avis, R.J. Gregg, et al., (Toronto: Gage, 1975). "^Such Canadian forms will be referred to as Canadianisms through out this study. ^CEU is the abbreviation for M.H. Scargill's Modern Canadian  English Usage, (Toronto: McClelland and Stewart, 1974). All data in that book are identical to data in M.H. Scargill and H.J. Warkentyne's "The Survey of Canadian English," op.cit., however the comments, con- • elusion, and page numbers are different. The abbreviation Will refers to Websters Third Internation Dictionary. DCE stands for the Dictionary of Canadian English, op.cit. 18(Toronto: Gage, 1967), pp.1-927. 19 (Victoria, B.C.: Sono Nis Press, 1977), pp.9-60. Most words and phrases cited in both these books, though of course of historical interest, are generally little known to most Canadians. 20 The British items in this list are a reflection of the fact that Canadians are somewhat bi-dialectal understanding both North American and British English. 59 CHAPTER 4 METHODOLOGY 1. The Sociological Parameters When measuring usage of a language, it becomes increasingly apparent that one is actually measuring variation, and that variation can be measured with reference to only a few parameters, namely temporal, geographical, sociological, and stylistic. Because our survey has been conducted within the city limits of Ottawa and since our survey is by design synchronic, we will not be dealing directly with the first two parameters."'' The stylistic parameters are presented in Section 2, the section immediately following this. We shall, therefore, discuss the sociological parameters which will be correlated with the linguistic phenomena. The questionnaire contains more questions pertaining to the informant's background than we will probably want to use; however, if an unusual usage occurs, we may desire to trace it back to its probable origin, for example, the Ottawa Valley, the ethnic background, etc. The sociological parameters which we will use systematically are: sex, age, and social class. Sex and age are straightforward factual infor mation. Age The informants did not hesitate to give their year of birth, and the information given appeared to be true. We suspect that the speech characteristics of Ottawa changed markedly during and after the Second World War, when life styles were changing rapidly everywhere in the industrialized world. For Ottawa, this was the first time that thousands of people from across Canada were hired to work for the Federal Government. Ottawa would no longer be such a closed society as it had been previously. This was also the time when married women were first allowed to work in the Civil Service, and the time when Ottawans switched from having their large dinners at home at noon to having lunches in their offices. We hypothesize that individuals raised in Ottawa before this period of change will have certain usage patterns significantly different from those used by people raised in Ottawa after the change. The cut off year of birth will probably be around 1937. We will check this and other age group factors. 2 Social Class Social class is a much more complicated matter, for at least the three following reasons. First, although segments of society can be grouped fairly easily through averaging a very large number of people, it does not seem fully adequate to simply take the average of an 3 individual's sociological indicators. Consider the difficulties of categorizing someone who has "dropped out" and is living off the land, or a skilled labourer who is earning more money than a university music professor, or someone who has recently married into or made a great deal of money. Secondly, the concept of class is an area of taboo for many middle class and upper middle class persons. Thirdly, language usage is in itself an important feature of class differentiation; thus for this study, in order not to become circular, linguistic features must be kept out of the social index. The Socio-economic Index The goal in creating a socio-economic class index is to have a tool by which to measure objectively the sociological situation of each informant. Each informant can then be given a score, and groups can be formed from these scores. The linguistic variation can then be plotted with reference to these social class groupings. The type of index developed for this study was a multi-item index modelled after Labov's and Trudgill's indices but modified greatly for 4 the Canadian and Ottawa context. The index was specifically extended to include the upper middle class and the lower upper class, thus affording us with the first non-truncated socio-economic study of urban dialectology. The upper upper class was inaccessible. The index, as shown below, has seven social indicators which can be employed separ ately or conjointly for correlations with linguistic variation. Each of the se-ve n indicators has a scale assigned to it with points ranging in each case from 1 to 6 so that each informant could be assigned total points ranging from 7 to 42.^ A discussion of each of the indicators appears below the chart. Occupation The informants were given scores ranging from 1 to 6 according to the occupation in which they were presently involved. The scores and relative positions of occupations are based on Barber, 1957, pp.102-104^ and'an intimate knowledge of the job-ranking hierarchy of the Federal Public Service. Retired persons are rated according to their occupation before retirement. Housewives, househusbands and widows are rated as to Table 4.1 SOCIO-ECONOMIC CLASS INDEX Points Occupation Father's Occupation Income Education Spouse's Education House Value Location •% 1 seasonal manual seasonal manual 0-9,999 UIC-Welfare grade 7 grade 7 <40,000 Rented lower town 10 2 manual blue collar semi-skilled manual blue collar semi-skilled 10,000-14,999 grade 11 grade 11 40,000-49,999 subsidized lower town East End 20 3 skilled worker foreman blue collar skilled worker foreman blue collar 15,000-17,999 grade 12 or 13 grade 12 commercial 50,000-59,999 bungalow areas 20 4 white collar clerk,- teacher semi-profes sional, sales man skilled worker foreman blue collar 18,000-23,999 high school complete & some Univ.-Coll.training high school complete 60,000-79,999 $60,000.00 house value areas 40 5 careerman in profession, branch and unit managers white collar clerk, teacher semi-profes sional, sales man 24,000 34,99.9 B.A,, B,Sc., some grad. studies university degree 80,000-119,999 Glebe, Alta Vista, West End 8 6 high level manager, employer, professional careerman in profession, branch and unit managers >35,000 graduate pro fessional degree,private schools university degree plus >120,000 fireplaces, library,den, central air conditioning, landscaping Rockcliffe along canal Island Park 2 0-10.5 points - lower 24.5-31.5 points - middle middle 10.5-17.5 points - working 31.5-38.0 points - upper middle 17.5-24.5 points - middle lower >38 points - lower upper ON t-o 63 their spouse's occupation. Students were given an undiscriminating four points. Father's Occupation The point system for this indicator is similar to that of Occupation, but some provision was made for upward mobility in these recent decades of affluence and increased opportunity. Therefore Father's Occupation offers higher points for a lower ranking occupation in three boxes. A comparison of the first two indicators of the index may show a case of upward mobility. Instances of upward mobility are of great interest to linguistic studies such as this one because a large number of linguistic changes tend to co-occur with the sociological changes.^ Income This indicator is based on the salary, dividends, etc. of the major bread winner of the household, not on the combined income of husband and wife, for two lower middle class workers with.good salaries do not assume the manners, power, and prestige of a lower upper class family with only one person working externally. The income scale is based on the salary scales for the Ottawa Fire Fighters, Ontario University Employees, and the Federal Public Service, and on U.I.C., welfare, and old-age pensions, all of which are public information. Housewives, househusbands, and children still living at home were given points commensurate with the income of the family's major income earner. Education The score for this indicator is probably the easiest to assign as 64 one either did or did not complete a grade or degree. Unlike Occupation and Income the points for education are not transferable to dependents. Two additional points are given for private school attendance and one additional point for further studies after once having been in the work force for some years. Spouse's Education People tend to marry members of the opposite sex of the same social class or of immediately adjacent social classes. Perhaps one of the best modes of evaluating who a person thinks he is, is to look at the person whom he or she marries. There is a strong relationship between the amount and quality of schooling and the preparation for leadership roles in society. House Value One can acquire a fairly adequate knowledge of house and property value by studying the real estate pages of the local papers. The Census Tract data help in this respect too. It does not seem to matter if a residence is rented or not, as the value differentiates sufficiently. House Location Where one resides within a city indicates a great deal about one's social status in life. In every city, there are certain areas where the poor tend to live and others where the rich tend to live. The well-to-do may prefer certain locations according to interests such as recrea tional, e.g., proximity to parks, golf course, marinas, etc.; cultural, e.g. proximity to the National Arts Centre; historical, e.g. the prestige of heritage houses.; occupational, e.g. the proximity to a university, the External Affairs Building, etc. Identical houses in different parts of town can sell for vastly different prices. This difference in price is a reflection of society's desire for prestige and association. In the past certain races and ethnic groups were excluded from certain areas of greater Ottawa. Although discrimination is no longer legally possible, some of the pattern may still exist. This concludes the exploration of the indicators for our Socio economic Class Index. Other sociological factors sometimes used in this survey are: Ethnic Background Certain usages in Canada are known to have their origins in British, American, Scottish, Irish, French, German, and Scandinavian g Dialects. When items are suspected of coming from other languages we will analyze the variation patterns with reference to the ethnic background of the informants. The ethnic background of the informant was asked directly in question number 18. Rural/Urban Background Although all informants interviewed in Ottawa were born and raised in Ottawa, and therefore are to be considered urbanites,. some had rural backgrounds. An informant was categorized to have a rural background if the informant's mother, father or spouse had lived outside a city, or if the informant had lived in the country for a short period of time. All informants from Renfrew and Smith's Falls were categorized to have rural backgrounds. Ottawa Valley This category is identical to Rural Background, as all our infor mants with rural backgrounds happened to have Ottawa Valley Backgrounds. New Canadian/Several Generation Canadian Informants who were born in another country or whose mother or father was . born in another country were categorized as New Canadians; all others were classified Several Generation Canadians sometimes shortened to Old Canadians. 2. The Contextual Styles The co-variation of linguistic material with sociological parameters can be seen along two main dimensions. The dimension already dealt with in this chapter is that of social differentiation, e.g. age, sex, ethnic background, rural or urban background, education, and social class. The second dimension, the one dealt with here, is stylistic variation. Stylistic variation is the response, conscious or subconscious, on the part of the individual speaker to the.social situation in which the speech act occurs. Factors which may influence an individual's style could be: 1) the formality of the occasion, 2) the role of the individual on that occasion, 3) the relative age and sex, and 4) the social ranking of those present. The survey was designed to elicit from each informant as full a range of stylistic variation as possible by means of one structured questionnaire. Of course, one interviewer with a tape recorder is not able to create the full gamut of human experience, but by means of giving the informants specific language related tasks, one is able to observe a wide range of stylistic formality and informality. We will first describe careful speech and casual speech after which we will label and define those tasks which were incorporated into the question naire in order to elicit the five styles used throughout the analysis. Careful Speech Careful speech is likely to be the style present when individuals do not know one another or when their roles separate them socially one from another, e.g. in response to formal school testing, job interview situations, linguistic questionnaires, addresses to large groups, or first encounters. During the survey interview, formal speech would .. most likely, be .elicited at the beginning when family background is being asked about, in the grammar section, and in those sections which require that individual words be spoken in isolation, i.e. the minimal pairs, word list, and pictures sections. As a result of the fact that it took several minutes to persuade many Ottawans to submit to the interview, some informants were often very relaxed and chatty by the time we started the interview at Section One. This led to an uneven comparison of performance in that section, and therefore this section was excluded from the analysis. Casual Speech Casual, unguarded speech is most likely to occur in association with friends and family at informal occasions, e.g., at home, at work during breaks, and in restaurants and bars. The situation of a stranger entering someone's house, placing a tape-recorder in front of that person, and asking more than one hundred 68 questions might lead one to expect that only careful, guarded speech and not casual speech would be forthcoming. Fortunately that is not the case. There are many techniques in interviewing which help elicit casual speech. These include asking the informant to: 1) recite lists known from childhood, 2) count by fives or tens, 3) tell a funny story, 4) talk about his city, 5) recount a dangerous situation, 6) create a story from a sequence of pictures, 7) read an informal story, 8) digress from any topic. Furthermore, family members and friends were encouraged to remain in the room while the interview was taking place. The presence of these people tended to relax the informant,enabling him to speak more casually. 9 The Five Styles Minimal Pairs This task elicits the most careful speech by presenting the infor mant with a list of pairs of words which are pronounced alike or almost alike. The similarity of the words entices the informant to make phonemic distinctions which are not normally part of his idiolect or dialect. What the informant pronounces here is therefore very important because the task elicits the underlying form, the form which he thinks is most 'correct'. This may be the closest look we ever get at a person's competence, for when a person is talking normally it is most often only his performance level that we can observe. Word List For this task, the informant is asked to read a list of 120 words. As is the case with the other tasks, most of these words contain at least one of the 27 linguistic items which we are investigating system atically. This task, however, has the disadvantage of often eliciting a reading pronunciation for a number of words. This task will most likely elicit the second most careful style of speech. Pictures This task requires that the informant identify pictures of objects presented to him.*''"'" This task reduces the chances of reading pronun ciations. Because the pictures distract the informant from the linguistic nature of the interview, in this task and the Word List task we will find a large number of Canadian lexical markers. It is hypo thesized that this task will elicit the third most careful style of speech. Reading Here we have the first task which requires the informant to speak connected discourse, albeit read, for normal lengths of time using natural catenation or liaison patterns. In a reading passage there are so many linguistic items to control that one's habitual speech charac teristics will prevail most of the time. The. reading passage was intentionally put in an informal setting involving young people and a mother in conversation at home. Although Labov and Trudgill include their reading styles, Style C and Reading Style respectively, as part 12 of formal speech, we will demonstrate that our reading passage elicits a casual style, arid that for some items, the reading style may be more 13 casual than the so-called casual style. We will investigate this aspect in depth. It seems that many readers take on a role when 7G reading which may be more casual than their own casual speech. Free Speech We incorporated a few questions in order to elicit free speech; these were: 1) have you ever been in serious danger or what was your closest encounter with death, 2) has something happened to you that was strange or funny, 3) could you tell me about a recent trip, and 4) what do you think of living in Ottawa? In most cases, the topic and follow up questions dis tracted the informant from the linguistic interview situation. Further we Incorporated a new sub-task into the questionnaire in order to elicit more controllable casual speech. From Hawkins (1977), 14 p.56, we took a series of four pictures and asked the informant to make up a story following the sequence of pictures. In two instances, the informant was asked to say exactly what the person in the picture sequence would say. This picture sequence sub-task afforded us the chance to direct all informants to a common experience and more often than not, to the same vocabulary items. There are a number of occasions outside the formal interview proper when casual speech is apt to occur. These occasions often offer us the opportunity to record speech in its most unguarded, natural style. Such occasions presented themselves before and after the interview and during interruptions in the Interview. If the interview took place in the informant's home, as they most often did, frequent interruptions would occur, for example the telephone would ring, a baby would need changing, the children and/or spouse would have a comment to make or bring coffee, a neighbour would drop in, etc. In this survey, care was taken to record these happenings of 71 real life. Another occasion which offers casual speech we will label digression. This occurs when some informants make a weak link between a question and their favorite topic then continue for minutes; this, too, was encouraged and recorded. 3. The Sample In taking the sample for the Ottawa survey, the following principles were adhered to-: 1) all segments of the English speaking society should be represented; 2) a broad geographic distribution of the census tracts should be represented; 3) there should be external motivation for the person contacted to agree to be interviewed; 4) every member of the English speaking population should have a random, near equal, and non zero chance of representing his segment of society; consequently 5) the method of selection should in no way be related to the primary variables of the survey; and 6) no chain or network of friends, family, or colleagues should be allowed to form a group of informants. Sampling Procedure We decided that the sampling universe was to consist of all individuals age 16 or over who were born and raised in Ottawa, life long residents of Ottawa, and native speakers of English. For practical purposes for this survey a native Ottawan was defined as a person who was born in Ottawa or anyone who had moved to Ottawa before starting kindergarten, i.e. age five. A life-long resident was defined as anyone who has lived all his life in Ottawa, or one who had left the city only in order to attend university or serve during the wars. In addition, a person who had left the city for less than two years and had returned long ago was considered a life-long resident. No statistics are available on born and raised anglophone Ottawans; however, it is common knowledge that they do not represent a cross section of the Ottawan population. Figure 4.9 in Canadian Urban Trends, volume 2, entitled "Income disparition"^ clearly shows that the French and south European sections of Ottawa have the lowest incomes. Non-native Ottawans who migrated to Ottawa from other parts of Canada seemed to be the majority of the anglophone population. Sixty-three percent of those contacted who did not meet our criteria were in this category. The Census Tract Bulletin: Ottawa-Hull shows 70,975 in-migrants to 16 Ottawa proper in the five year period before the 1971 census. This number, which is an indicator of a mobile population, is larger than the in-migrant number for cities of comparable size. For a comparison of Canadian cities, see our Table 1, p.23. The process of selecting an address of a potential informant was based on the sociological make-up of the census tracts. From the Ottawa-Hull Census Tract Bulletin"*"^ we chose census tracts according to the following criteria: 1. Highest employment income (male). 2. Highest employment income (female). 3. Highest income (single). 4. Highest house value. 5. Highest cash rent. 6. Mid level employment income (male). 7. Mid level employment income (female). 8 Mid level income (single). 9. Mid level house value. 73 10. Mid level cash rent. •11. Lowest employment income (male). 12. Lowest employment income (female). 13. Lowest income (single). 14. Lowest house value. 15. Lowest cash rent. We canvassed the city according to these census tracts with a view to finding informants of various sociological backgrounds who would fill our predetermined sociological cells. See our Tables 4.3.1. to "4.3.13..for a presentation of the social characteristics of the sample. After designating the appropriate census tract, we then determined the geographical midpoint of that census tract, went to the north side of the street of that location, knocked at the door, and explained the survey. If the person met our criteria, we asked for an appointment, and in some cases we interviewed there and then. If the person did not meet our criteria or refused (there were nine refusals) we would con tinue trying successively higher street numbers staying on the north side of the street until we found a person who met our requirements and who was willing to be interviewed. After completing an interview with one informant in a particular mid level census tract, we would move to another mid level census tract, where we would repeat the search and interview procedures. In dealing with the non-middle class census tracts of Ottawa, one is faced with certain sociological limitations. First, there are only 62 census tracts in Ottawa, and several of these are predominantly francophone districts. Secondly, a few tracts are almost exclusively governmental, commercial or industrial zones. Thirdly, there are only a few areas in Ottawa where the lower-upper class live. These areas are Rockcliffe Park and small sections of Alta Vista, Island Park, and the Glebe. Similarly, there are only a few areas where lower class and working class anglophones live. These districts are parts of Centre Town, LeBreton Flats, and eastern Overbrook. Thus the census tracts of these non-middle classes had to be surveyed more intensely than those tracts which contained the middle classes. We set a limit of no more than 6 interviews for these census tracts. Filling the age and sex quotas was left to the interviewer, who would try to make up for the unavailability of working age males. For example, the interviewer would ask to interview the father if all members of the family qualified as native-born anglophone.Ottawans. See Appendix B number 527 for an analysis of the distribution of the informants by census tract. Map 2 on page 426 shows the geographical location of the census tracts. A tally sheet account of people con tacted but who did not meet our requirements is as follows: 1. Non-Ottawa born Canadian anglophones - 63%. 2. Francophones - 11%. 3. Foreign born anglophones - 6%. 4. Foreign born foreign mother tongue - 19%. (Italian, Portuguese, Chinese, German, Lebanese, Greek, unknown) In the middle class census tracts, the percentage of people unreached, e.g. not home, indisposed, or did not answer, was about 20% even though the interviewers returned a few times. In the non-middle class census tracts, the unreached percentage was less than 10% as the interviewer returned to the same street and address on several occasions. 75 Nearly 1,000 addresses were approached in order to obtain the 100 interviews. Informant Motivation One may wonder what would motivate an individual to submit to an interview such as ours. The following motives were stated: 1. Monetary—each person who met our requirements was offered three dollars to take part in the interview. This offer motivated some of the poorer people and some of the younger people: it had the added effect that it convinced many that this was not a magazine sales promotion. 2. Canadian Unity—during the explanation of the survey, we mentioned that we would be comparing data of speech characteristics of different cities across Canada. This motivated many individuals to feel that they should do the interview for the sake of Canadian unity and they expressed such sentiments. 3. Parental-Familial Pressure—if one member of the family whose address we picked met our requirements, the other members of the family would frequently try to persuade that person to be interviewed. 4. Historical-—while explaining the survey, we mentioned that we were comparing older styles of speech to younger styles and that we would be asking informants to talk about Ottawa in the olden days. Many older people showed a keen interest in this topic. Empathy for a door-to-door canvasser—many people had some time in their lives knocked on doors while working on school projects,'selling something, or canvassing for a political party or an organization such as the Red Cross. Some of these people expressed comradeship and agreed to be interviewed. Inability to say no—a few people obviously did not want to be interviewed but could not say 'no'. One woman asked half way through the interview whether she had to do this; I told her"yes,"and she continued. Curiosity—probably most people who agreed to be inter viewed had some degree of curiosity. Ennui—many people said they had nothing better to do. Pride in identity—while explaining the survey, we mentioned that we would be comparing people by age, class, sex, occupation, ethnic background, etc. Many people took pride in representing one or more of these groups. Religious—one woman had a magazine rack in the vestibule laden with religious pamphlets just waiting for someone to knock on her door. She and her husband would submit to almost any interview but certainly demanded equal time. Waste this salesman's time—many people believed that the survey was just another magazine sales promotion. Two men stated they began the interview in order to see how long the interviewer would continue to lie. 77 12. Helping a nice young man—a number of elderly women stated they wanted to help a nice young man. 13. Interest in topic—at least ten informants said they were very interested in the topic. Five people who refused to be interviewed showed unequivocally that they were not. This is a possible source of biassed results, as those who were interested in the topic readily agreed to be interviewed, while a few who were not interested, refused or may have stated that they did not meet our requirements. Size of Sample Although the number of informants is considerably smaller than we would have liked, the number is adequate for a sociological urban dialectology survey. The Norwich Survey by Trudgill had only sixty 19 informants, and the Lower East Side Manhattan Survey by Labov was com-20 prised of 81 New Yorkers. The Ottawa Survey has 89 Ottawans. With 21 reference to size of sample, Trudgill quotes Labov, 1966a, p.638; the full quotation follows: It may therefore be concluded that the 26 New York City television informants show the same linguistic behavior as the 81 New York City ALS informants. If the previous studies of New York City had followed a systematic method of selecting informants, the 25 or 30 cases described would have been sufficient to show the outlines of a systematic structure of stylis tic and social variation. We may conclude that the structure or social and stylistic variation of language can be studied through samples considerably smaller than those required for the study of other forms of social behavior. Further, while discussing size of sample and instances within a cell, Labov states: 78 In Chapter IV, we found that from 10 to 20 instances of a given variable were sufficient to assign a value that fits consistently into a complex matrix of stylistic variation, while at the level of three or four instances, fluctuation unrelated to the matrix was noted. Similarly, we will find that from ten to twenty individuals will give us a value for a social class which fits consistently into an overall pattern of stratification while groups of four or five show unrelated fluctuation. In the case of (r), it will be possible to divide a.group, of 81 informants into six strata which are clearly separated in the same order for five stylistic levels. Thus we see that numbers which might be totally inadequate for the study of attitudes, say, towards racial segregation, with the associated reluctance to give a straightfor ward personal response, are quite adequate for the study of the phonological variables.22 A few years later he revised his number of required informants downward to five stating: ...we find that the basic patterns of class strati fication, for example, emerge from samples as small as 25 speakers. Extremely regular arrays of stylis tic and social stratification emerge even when our individual cells contain as few as five speakers and we have no more than five or ten instances of the given variable for each speaker. With this regular and reproducible data, we are in a position to speci fy what we mean by the "stylistic" or "social" meaning which seems so elusive when language is studied out of context.23 We do, however, agree with Trudgill when he states: However, a rather large sample would have been useful for those cases where it is desirable to classify informants according to sex, age and social class simultaneously.24,25 Analysis of the Sample The sociological characteristics of the 89 Ottawa and 11 urban valley informants were placed into matrices with a view to evaluating the dis tribution. We chose to utilize those which were most similar to the Labov, 1966, Trudgill, 1974, and Scargill, 1974 , Surveys and which were configured so as to contain the largest possible number in each cell. The most interesting matrices are presented below; those which were utilized throughout either Chapter 5 or Chapter 6 are boxed. Table 4.3.1 AGE-GROUP >40 <40 Total Ottawa 37 52 89 Valley 7 4 11 Total 44 56 100 Table 4.3.2 SEX-GROUP M F Total Ottawa 43 46 89 Valley  7 11 Total 47 53 100 Table 4.3.3 GENERATION CANADIAN New Canadians 0 + 1st gen. Old Canadians 2 or more gens. Total Ottawa Valley Total 38 0 38 51 11 62 89 11 100 Table 4.3.4 RURAL BACKGROUND OTTAWA VALLEY BACKGROUND rural urban Total Ottawa Valley Total 23 11 34 66 0 66 89 11 100 Table 4.3.5 ETHNIC BACKGROUND Irish Scots English French Other Ottawa 15 20 27 6 19 Valley 5 5 0 0 1 Total 20 25 27 6 20 80 Table 4.3.6 SEX, AGE, CLASS (OTTAWA) Male >40 Female >40 Male <40 Female <40 Total Low-wrk Low-mid Mid-mid Up-mid Low-up Total 2 1 4 3 3 13 4 5 7 4 4 24 2 9 13 4 2 30 3 7 7 4 1 22 11 22 31 15 10 89 Table 4.3.7 SEX, AGE, CLASS (OTTAWA AND VALLEY) Male Female Male Female Total > 40 > 40 < 40 < 40 Low-wrk 3 5 2 4 14 Low-mid 1 8 10 7 26 Mid-mid 4 9 15 7 35 Up-mid 3 4 4 4 15 Low-up  2 1 10 Total 14 30 33 23 100 Table 4.3.8 SEX/AGE CLASS CONVERGED (OTTAWA) Male Female Male Female Total > 40 > 40 < 40 < 40 L-WK-LM 3 9 11 10 33 Mid 4 7 13 7 31 UM LU 6 8 6 5 25 Total 13 24 30 22 89 81 Table 4.3.10 ETHNIC BACKGROUND/CLASS (OTTAWA AND VALLEY) PQ US Eng Scot Ire Itl Ukr Pol Cz Eur Jwsh SAm Total Lower 0 0 1 .0 0 0 1 0 0 0 0 0 .2 Working 2 1 2 3 4 0 0 0 0 0 0 0 12 Low-Mid 2 1 4 4 7 3 0 0 1 3 0 1 26 Middle 1 2 11 11 7 2 0 1 0 0 0 0 35 Up-Mid 1 1 4 3 2 0 2 0 0 1 1 0 15 L-Up 0 0 5 4 1 0 0 0 0 0 0 0 10 Total 6 5 27 25 21 5 3 1 1 4 1 1 100 Table 4. 3.11 AGE CONVERGED CLASS; . SEX CONVERGED CLASS (OTTAWA) •>40 <40 F M L WK LM 12 21 17 16 Mid Mid 11 20 17 14 UM LU 14 11 12 13 Total 37 52 46 43 This matrix was utilized throughout Chapter 6. Table 4.3.12 SEX/AGE (OTTAWA AND VALLEY) Male >40 Female>40 Male<40 Female<40 Total Ottawa 13 24 22 30 89 Valley 1 6 1 3 11 Total 14 30 23 33 100 This matrix was utilized throughout Chapter 5. 82 Table 4.3.13 CLASS (OTTAWA AND VALLEY) Ottawa Valley Total Lower-Working 11 3 14 Lower Middle 22 4 26 Middle 31 4 35 Upper Middle 15 0 15 Lower Upper 10 0 10 Total 89 11 100 This matrix was utilized throughout Chapter 5. Linguistic Sample The linguistic items which we investigated in the five styles con sisted of 27 items. Below, these items are listed with their frequency of occurrence in the questionnaire for each style; the column headed 'Free Speech' contains numbers indicating the total number of utter ances of that item during free speech by all 100 informants. Table 4.3.14 LINGUISTIC SAMPLE MP WL P SR Reading Free Speech 1. VtV 9 21 9 10 41 1,513 2. ntV 3 10 4 16 12 554 3. -ing 3 5 7 3 20 1,404 4. nj,tj,dj 4 9 3 1 6 133 5. rV=rV 0 9 4 1 3 57 6. st 0 1 3 0 4 300 7. h 4 0 0:0 12 205 8 v##v 0 2 3 0 6. 79 83 MP WL P SR Reading Free Speech 9. 0 9 1 0 4 79 10. AU 4 5 3 2 8 493 11. Aut 1 •0 0: 0; 5 217 12. ai 3 1 3 3 10 431 13. ait 1 2 0 0 4 194 14. An 0 2 0 0 1 16 15. nd 1 4 4 1 5 723 16. vr-+er 6 5 0 0 7 77 17. vr->-ar 0 1 0 0 0 114. 18. SB 0 2 0 0 4 231 19. hw 4 2 1 0 6 328 20. kt, pt 1 1 1 0 5 89 21. ~D 3 4 0 0 5 221 22. 3, 9 3 5 6 2 15 1,780 23. or 0 1 1 0 4 0 24. going to 0 0 0 0 3 110 25. milk 0 1 0 0 1 0 26. good [a/ 0 1 0 0 3 16 27. tomato 0 1 1 0 1 0 A presentation of the frequencies for all the individual variable which go into the. phonological study, Chapter .5, and the grammatical, pronunciation, and vocabulary study, Chapter 6, can be found in Appendix B. SR stands for series; this style was dropped from the analysis mainly because of the inability of the informants to recite poems, lists, etc. 4. The Questionnaire The structure of the questionnaire was designed to put the informant at ease. The first section deals with the informant's background which allows the informant to talk about himself, his family, and times past. The next section asks the informant to identify common objects by means 84 of pictures; this task easily convinces every informant that he can indeed do the interview, and it demonstrates to the informant that we are interested in what he actually says rather than what he may think is 'correct', thereby setting the tone for the entire interview. The third section is a word list of 120 items which the informant is to read. Then comes the grammar section which requires only that the informant orally fill in the missing word while the interviewer says the rest of the sentence. In the second portion of this grammar section, the informant is asked to choose from two sentences spoken by the interviewer the one that he would most likely say. Following this, the informant is asked to supply or choose local words and phrases, then to read the reading passage, to talk about life in Ottawa, and to give a narrative. Next, the informant is asked to tell a story by following a sequence of pictures, then to read a list of minimal pairs and finally, to respond to questions of language attitude concerning the Canadian, American, and British varieties of English and the informant's own speech patterns. The entire questionnaire is reprinted in Appendix A. 5. Interviews Of the 100 interviews, 64 were conducted by myself, 15 by Margaret Murdoch and 21 by Steffi Ortiz. All three of the interviewers are graduate students in Linguistics Departments with experience in dialec tology and a special interest in Canadian English. All the interviews were conducted in a similar friendly and informal style, and there appears to be no difference in the three sets of data. The three inter viewers ensured that all items in the questionnaire were elicited and encouraged the Informants to speak freely and digress when they so 85 desired. Most interviews were conducted in the informants' homes. Their friends and family were encouraged to remain in the same room. This arrangement tended to relax the informant and enabled us to record casual conversations among friends during breaks in the interview. 6 . Transcription and Analysis From the questionnaire, a key was created which numbered all the variables, the anticipated values, and left numbered blanks for unexpected values. For the Free Speech style, we listed the variables, their values and provided space to tick off the occurrence of each value. We then played the recorded interview and ticked off the appropriate value of the variables as we progressed through the tape. Thus, rather than transcribing every bit of the interview, we selected the value of only those variables we had previously incorporated into the questionnaire. Using this technique, we saved man-months of transcription time. Our next task was to transfer the results from the 100 individual code books or keys to computer Fortran Coding Forms and from these sheets to computer punch cards. The punching and punch verification was done by the Computer Services Department of the University of British Columbia. The next stage, that of arranging and presenting the data, was per formed with the help of the Amdahl 470/V6, Model II computer under the 26 MTS operating system with a Tektronic 4012 Graphics Terminal. The Midas programme was employed throughout this stage of our survey by Lewis James, Computer Consultant for the Arts Faculty. The analysis of our data was carried ouC with computer assisted tests as well as with traditional means. For the data in Chapter 5, we con-86 sidered the association of each of our linguistic variables in relation to our sociological parameters using Somers' D (a measure of ordinal variation). The data in Chapter 6 was analysed similarly, except that instead of giving the percentage of the time the informant responded in a particular way, we have dichotimized the informants' scores as to whether they fell above or below the median. Chi-square tests and in some cases the Fisher test for probability were used to measure whether there was a-significant difference in the frequency of occurence of values in Ottawa versus the Ottawa Valley urban centres. The above tests were carried out by Virginia Green, Statistics Consultant to the Arts Faculty at U.B.C. 7. Limitations The limitations of this study are the logical consequences of the theoretical design of the survey, the equipment used, and the unavoidable biases of one transcriber. Limitations due to Survey Design Although dialectology as a science has existed for well over one hundred years and although it has developed established and proven, techniques, the linguist is ever aware that his presence, as part of a survey, affects the speech which he is trying to observe. Urban socio-dialectologists, who attempt to be more sensitive to stylistic and sociological factors than did workers in traditional dialectology, are even more keenly aware of this problem. The fact that we were present observing speech has no doubt altered that speech. Another effect on the speech we observed is a consequence of our having devised tasks 87 which the informants were to perform; some of these tasks were quite unnatural, and perhaps never before performed by the informants. A further limitation to a purely unbiased description of speech is that we then take these speech utterances while performing quite artificial tasks and equate them to specific speech styles. The design of this survey containing such tasks, though an improvement upon the older survey method which elicited answers in only one stylistic mode, nevertheless must be evaluated as a limitation. With the new methodology, one does observe a wide gamut of stylist variation, but one is left to speculate about the informant's full range of styles. Lack of American Data Throughout the entire time of the survey and especially while developing the questions and analysing the results, we were aware of a lack of sociological and stylistic surveys of main stream urban 27 Americans with which we could have made comparisons. We now know, for instance, the frequency percentages of values of -ing in Ottawa (and Vancouver) but we know very little about this same item in the United States. The Number of Informants Because of time restrictions, we could interview only one hundred informants. This small number of informants provides us with an even smaller number of informants in each cell when two or three of the indicators (sex, age, class) were combined. At no time was the number in a cell allowed to go below eleven. 88 Nature of Recruitment The fact that our informants were volunteers gives a definite skew to our data. It was my observation that one group of those who refused to be interviewed were not interested in English, and they probably spoke much like the two informants in the lower class. Others who appeared to be of the lower upper class declined, claiming lack of time. Probably in all classes those who liked English or who felt that they were good at English agreed to be interviewed more readily.than those who felt that their usage was not standard. We also did not have the facilities, time, or aim of conducting a fully random sampling which would meet the rigorous standards of an extensive sociological survey. Limitations due to Equipment Because we needed to have the recording equipment light, self-contained and unintimidating in appearance, we used a Sony 110-B cassette tape recorder utilizing the built-in microphone and battery power with Sony 28 Low-Noise C-90 cassettes. The frequency response is 50-10,000 Hz. The frequency response was adequate to make valid transcriptions of all speech recorded but was certainly not of a quality to merit further analysis on a sound spectrograph. Informants' Difficulties Four women informants did not wear their dentures while being inter viewed and one man complained about his dentures while being interviewed. Six people had some trouble reading with or without glasses, and one was somewhat hard of hearing. These factors may have influenced some linguistic items minimally. 89 Limitations of the Computer The Midas programme and the hardware configuration referred to above served our purposes well. We did find, however, that the computer printers could not easily be adjusted to use IPA (the International Phonetic Alphabet); this necessitated several coding transferrals by hand with the accompanying chance of errors with each transferral. Limitations of Transcriber To a certain degree one has a propensity to hear what one has been trained to hear and not hear what one has not been trained to hear. This transcriber, like all other transcribers, had such a bias which will inevitably skew the results to some small extent. The transcriptions of this Ottawa project and the Vancouver project will however be fairly similar because the transcribers have had similar training. 90 Chapter 4: Footnotes "'"Concerning the geographical parameter, it is planned that the results of this Ottawa study will be compared to the results of the Vancouver study. Moreover, it is hoped that comparable studies will be carried out in other urban communities across Canada. Where the temporal parameter is concerned, we will make com parisons of data from this study with data from previous studies, e.g. Avis, 1954, 1955, 1956; McDavid, 1951. 2 A working definition of class for this study is: 'a major social group, members of which are of approximately the same economic position, prestige, occupational rank, power, value orientations, and characterized by interaction and class consciousness," taken from H.K. Reading, A Glossary of Sociological Terms (London: Sociologia, 1976), p.27. 3 See our Socio-economic Class Index below. We do in fact take the average of the individual's socio-economic indicators. 4 Op.cit., pp.211-220; op.cit., pp.38-41. "'There are three additional points which may be assigned to an informant in the education indicator. Two points are given for private school attendance and one point for returning for further training. Bernard Barber, Social Stratification (New York: Harcourt,.Brace, and World, 1957), pp.102-104. We also consulted B. Blishen and Hugh McRoberts, "A Revised Socioeconomic Index for Occupations in Canada," Canadian Review of Sociology and Anthropology, vol.13, no.l (1976), pp.71-79. ^W. Labov when summarizing data on social mobility concludes: "1) Upwardly mobile persons adopt the norms of an exterior reference group - as a rule, the norms of the next higher group with which they are in contact. 2) A group which shows a past history of social stability tends to be governed more by its own linguistic norms - more precisely, to achieve a balance in which own and external norms are reflected in fairly consistent performance, without a wide range of style shifting. 3) A downward mobile category deviates in its nonacceptance of the normative patterns which other segments recognize. Here we are speaking of a set of individuals who deviate from the principal subgroup in which they were raised." W. Labov, "The Effect of Social Mobility on Linguistic Behavior," Sociological Inquiry, vol.36, no.2 (Spring 1966), pp.202, 203. We have made provision for social mobility data in this survey and hope to be able to utilize it in a future study. 91 g See the Dictionary of Canadianisms, op.cit., and the Dictionary of  Canadian English, op.cit. 9 For the purpose of simplicity, we will label both the task and the style with the same heading. We hypothesize the tasks elicit 'styles' which correlate closely to the range of styles found in natural human communication. See Labov, 1966, pp.90-131 for discussion of this point. 10See Noam Chomsky, Aspects of the Theory of Syntax, (Cambridge: M.I.T. Press, 1965), pp.3-15 for a clarification of the linguistic terms competence and performance. "'""'"This task, first used by me in the Kootenay Region Survey during the summer of 1970 (Gregg, 1973, pp.105-116), is not part of the tech niques displayed by Labov or Trudgill. Murray Kinloch in his article, "The Use of Pictures in Elicitation," American Speech, vol.46 (1971), pp.38-46, discusses the advantages of this technique. 12 Op.cit., pp.92-109; op.cit., p.47. 13 See Margaret Murdoch, "Reading Passages and Informal Speech," un published paper given at the Third International Conference on Methods in Dialectology (London, Ontario), 1978, pp.1-7. 14 P.R. Hawkins, Social Class, The Nominal Group arid Verbal Strategies, (London: Routledge and Kegan Paul, 1977), p.56. "^Canadian Urban Trends: Metropolitan Perspective, Vol.2, ed. D.M. Ray, (Toronto: Copp Clark, 1977), p.53. "^Census Tract Bulletin: Ottawa-Hull, (Ottawa: Statistics Canada, 1971), p.2. "^Ibid., pp.2-37. 18 The above comments on the motivation of the informants and the manner of approach ori the part of the interviewer can best fit into a discussion of field techniques as found in: Roger Shuy, Walter Wolfram, and William Riley, Field Techniques in an Urban Language Study, (Washington: Center for Applied Linguistics, 1968), pp.20-28; William Samarin, Field Linguistics: A Guide to Linguistic Field Work, (New York: Holt, Rinehart and Winston, 1967), pp.33, 34, and 108; David Sankoff and Gillian Sankoff, "Sample Survey Methods and Computer Assisted Analysis in the Study of Grammatical Variation," Canadian Languages in  their Social Context, ed. Regna Darnell, (Edmonton: Edmonton Linguistic Research, 1973), pp.12-23; and W. Labov, "The Study of Language in its Social Context," Studium Generale, vol.23 (1970), pp.45-49. 92 19Trudgill, 1974, op.cit., p.27. 20 Labov started with 122 interviews but found that he had to eliminate some because they were not native New Yorkers and others because they were non-native speakers. The Norwich Survey and the Ottawa Survey saved much time by setting strict criteria of place of birth and native language from the outset. 21Trudgill, 1974, op.cit., p.27. 22 Labov, 1966a, op.cit., p.181. 23 Labov, 1970, op.cit., p.43. 24 Trudgill, 1974, op.cit., p.27, footnote 2. 25 Labov's work, The Social Stratification of English in New York  City, has been criticized for using as few as five linguistic variables as the base for a sociolinguistic survey; see J. Pellowe et al., "A Dynamic Modelling of Linguistic Variation: The Urban (Tyneside) Linguistic Survey," Lingua, vol.30, no.l (1972), pp.1-30. 26 For further specifications, see J.L. Leigh and Tina Duke, UBC  Facilities, (Vancouver: U.B.C. Computing Centre, 1978), pp.1-31. 27 As previously stated in Chapter 1, most American surveys have been of rural or inner-city speech. 2 8 Sony TC-110B specification sheet. 93 CHAPTER 5 THE CO-VARIATION OF THE PHONOLOGICAL VARIABLES WITH SOCIOLOGICAL AND STYLISTIC PARAMETERS 1. Measurement of Co-variation A major aim of this study is to investigate the co-variation of phonological items with sociological and stylistic parameters. In order to measure this type of correlation the following procedure was under taken: 1) on Fortran coding forms, we assigned a coded value to each of the 752 questionnaire items according to each informant's response; 2) the questionnaire items which contained any of the 27 phonological variables were then selected from the rest of the items and grouped together according to variable and the task in which the informant was, at the time of utterance, performing; 3) each informant was placed in a socio-economic class,"'' an age group, a sex group, a rural or urban back ground group, an ethnic background group, and a several generation or new Canadian group all according to sociological information given during the interview and as a result of applying the socio-economic class index criteria. This procedure allowed us to obtain frequency scores for each variable with reference to each sociological group and each style. By means of these scores, we were able: 1) to analyse the nature and extent of the correlation between phonological variables and socio economic class, social context, age, sex and the other sociological factors; 2) to discover which variables are most subject to co-variation with the above listed sociological parameters; and consequently 3) to 94 ascertain whether our hypotheses, assumptions, indices, sampling pro cedures, age 'cuts', class 'cuts', task ordering, etc., were valid. We will now present the 27 phonological items, items which seemed likely to provide some sociological or stylistic differentiation, and the results of the analysis for each item in graph form. The five con textual styles are plotted along the abscissa; the styles range on a linear scale from the most formal, Minimal Pairs (MP), through Word  List (W), Pictures (P), and Reading (R), to Free Speech (FS), assumed the most informal. The index scores for one value of each phonological variable are displayed in percentages and plotted along the ordinate. These percentages are the mean average of the frequency of a value with reference to the utterances of a variable by all members of a sociological group. The symbols represent the sociological groupings, and the lines connect the scores obtained by each group in the five contextual styles. The classes represented in the graphs are: Working (W), Lower Middle (LM) , Middle (M) , Upper Middle (UM) and Lower Upper (LU). A clear case of ordered sociological stratification would be illustrated by a graph displaying ordered and non-intersecting lines spaced with good separation; contextual variation would be illustrated by the slope of the lines. 1. The Variable (VtV), medial /t/ Examples: city, Ottawa, little, out of, and 146 more words and phrases. This variable is the post tonic /t/, pronounced [t] or [d], which is found between vowel sounds or between the liquid sounds /l/ and /r/ and vowel sounds, e.g. shelter, delta, tilted, belted and dirty, party, thirty, quarter, and smarty. In addition, this 95 variable was found to occur frequently after most.voiceless frica tives; after /f/ in after, often, fifteen, and fifty; after /s/ in sister, sixteen, mister, twister, blister, etc.; and after /J/ in wished our, washed (h)is, fished it, etc. Furthermore, a medial /t/variation was noted to take place after In/, e.g. carpenter, seventy, seventeen, winter, centre, and pointed; occurrences of this last variation were always tabulated under the variable (ntv), how ever. The medial Id rule also applied after /k/ in picture and arctic. 96 Figure 5.1.a. VtV = VtV by Class •=WQRKING SLOWER MIDDLE X^MIDDLE •=UPPER MIDDLE SLOWER UPPER Figure 5.1.a. displays the average percentage scores for (VtV) = [VtV] for the five social classes established from our socio-economic class index (Table 4.1) in each of the five contextual styles. Figure 5.1.a. reveals that the phonological variable (VtV) is 1) involved in a good amount of ordered social 97 class differentation as illustrated by the distance between the lines and the almost perfect linear sequencing of the classes at each stylistic marker. In fact if we used only three class lines, as Labov frequently did, LU/UM, M, and LM/W, we would have no cross overs or intersections. Furthermore, a 20 percent separation is maintained between the lower upper class and.the two lowest classes. Stylistic variation is also strong with scores generally declining as we progress from left to right on the axis of formal to informal. The scores drop 45 percent for some classes. The steepest gradient is between W and P, both representing non-connected speech, rather than between P and R, for example, which would have represented a variation between non-connected and connected speech. The percentages range between 67 and 7 and a few individuals scored 100 percent in MP and 0 percent in FS. Some of the working class, especially those in the lower class, showed almost no stylistic variation at all, with the percentages being below 15 in all styles. There appears to be some stigma attached to having too low a per centage of [t] in medial frj position especially in more formal styles. Notice that for the lower upper class, the reading style score for this variable is lower than the free speech score. Many informants of the lower upper and upper middle classes demonstrate this pattern throughout the survey. 98 Figure 5.1.b. VtV = VtV by Sex/Age PJ=MHL£S GE 40 YEARS OF AGE A-=MALES LT 40 YEARS OF AGE X-FEMALES GE 40 YEARS OF AGE <t>=FEMALES LT 40 YEARS OF AGE Figure 5.1.b. displays the variable (VtV) when analysed according to sex/age groups. We can see that females over 40 years of age are the most formal through the entire range of styles while the young males were consistently the most informal; this 2 is a pattern which recurs frequently in our study. 99 Avis, 1956; Gregg, 1957; and CEU treat the usage of this 3 variable in Canadian English. There is no parallel study of Northern American usage with regard to this variable, and consequently no comparison can be made. The Webster's Third International Dictionary transcribes words with post-tonic medial /t/ as [d] as first choice. 2. The Variable (ntV) Examples: plenty, centre, twenty, winter, and 44 other words. This variable has three possible pronunciations, namely, /nt/, /n/, or /nd/. 100 Figure 5.2.a. ntV = nt by Class •^WORKING A=LQWER MIDDLE X=MIDDLE •=UPPER MIDDLE *=L0WER UPPER Figure 5.2.a. displays the average percentage scores for the value /nt/. This graph clearly reveals that for this phonological variable, the upper middle class has consistently higher scores than does the lower upper class; we will want to check the pattern throughout our survey as well. No other clear social class 101 differentiation pattern emerges from our statistics. On the other hand, the gradients of the lines demonstrate a strong stylistic variation, with the frequency of /nt/ decreasing in all classes as the formality of the context decreases. The scores range from 94 to 13 percent. This range of scores forcibly illustrates the inherent weakness of postal surveys which inevitably must ask questions such as, "Does the -tt- of butter sound like the -dd- of shudder?"4 The lexical item centre elicited pronunciations more frequently containing /nt/ than other words contains the variable (ntV); the high prestige of the National Arts Centre in Ottawa undoubtedly played a role in this anomaly. 102 Figure 5.2.b. ntV = nt by Sex/Age m=MHL£S GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMALE5 GE 40 YEARS OF AGE •^FEMALES LT 40 YEARS OF AGE Figure 5.2.b. represents the same variable and value according to sex/age groups. We can observe that females are much more formal than males in all styles, and that age does not appear to be signi ficant. 103 Figure 5.2,c, ntV = nd by Age/Sex •"=MALES GE 40 YEARS OF AGE A^MALES LT 40 YEARS OF AGE X=FEMALES GE 40 YEARS OF AGE <2>=FEMALES LT 40 YEARS OF AGE o a _J 00 I a _ I—co LU LJ -CK UJ STYLE Figure 5.2.c, displays the variable (ntV) pronounced as /nd/. Males over forty, and to a lesser extent males less than forty, predominate in this pronunciation. 104 3. The Variable (-ing) Examples: doing, fishing, morning, building, and 97 other items. This variable is the grammatical morpheme suffix which marks the progressive aspect and gerund. It is also a nominalizing suffix in words such as morning and building but is not part of mono-syllabic stems such as thing, sing, or wing. It never carries primary stress and is pronounced in three different manners, namely, Ao/, /in/, and /an/. Figure 5.3.a. -ing = iQ by Class •=WQRKING A=LQWER MIDDLE X=MIDDLE SUPPER MIDDLE *=LDVER UPPER Figure 5.3.a. displays the average percentage for each class for the variable (-ing) realised as /in/. This graph illustrates fairly consistent differentiation and ordering according to class, some intersections and cross-over of lines notwithstanding. The percentages range from 94 for the lower upper class to 34 for the working class in minimal pair style with a separation of about 20 points maintained between these classes in all other styles. A strong stylistic variation is present with the steepest gradients again occurring between W" and P. The scores range between 94 and 11 percent. Comparing parallel studies on this variable,"' we notice that the pronunciation /an/ is considered to be the only alternative to /in/. 106 Figure 5.3.b. -ing = an by Class •^WORKING A=L0WER MIDDLE X=MIDDLE SUPPER MIDDLE *=L0VER UPPER 2n si UJ (_j -STYLE Figure 5.3.b. displays the frequency of /en/. One immediately sees that the working class stands out predominantly for the pro nunciation in FS but that for all other classes and styles the frequency of this stigmatized form is consistently very low. How ever our study in Ottawa reveals, and preliminary investigations elsewhere in Canada substantiate, that both /in/ and /en/ are not 107 the most frequent realizations of variable (-ing), but that [in] is. This pronunciation, not yet discussed in linguistic journals, is 6 treated as usage to be avoided by CBC newscasters in You Don't Say. When analysing our corpus, we frequently discovered that our informants pronounced such words as being and bean, paying and pain, playing and plain, saying and sane, and especially beings and beans in a very similar manner, with the vowel in the word containing the -ing suffixzbeing somewhat longer. An example of the confusion which can arise from this pronunciation comes from Margaret Murdoch, chief interviewer for the Vancouver Survey, interviewer for our Ottawa Survey, and native of Ottawa-Rockcliffe Park; she reports that a frequently elicited opposite of playing is fancy.^ 108 Figure 5.3.c. -ing = in by Class •=WQRKING A=L0WER MIDDLE X=MIDDLE •=UPPER MIDDLE *=L0VER UPPER Figure 5.3.c. shows the average frequency of /in/ for each class in each style. No regular class differentiation emerges except to mention that as the lower upper class and the lower class have high occurrences of /inV and /en/ respectively, their scores for /in/ are lower than the three middle classes. There is a strong stylistic gradient between W and P. 109 Figure 5.3.d. -ing = in by Sex/Age m=MflLES GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMALES GE 40 YEARS OF AGE <!>=FEMALES LT 40 YEARS OF AGE Figure 5.3.d. displays the frequency of /in/ according to sex/ age groups. It is important to notice that females and males less than forty years of age employ this usage more than the older generation does. It is the opinion of the author that /in/ is an attempt towards the pronunciation /in/ and away from /en/. By far, most pronunciations of /in/ in our corpus were of tertiary stress 110 level or lower. Only when a speaker raises the stress level of /in/ to secondary or higher levels do some listeners react negatively to its usage; otherwise its usage was eyaluated as fully acceptable or g unnoticed in our informal surveys. If we compare the Manhattan statistics with the Ottawa statistics assuming that the Ottawa /in/ and /in/ are equivalent to Manhattan An/ and Ottawa /on/ is equi valent to Manhattan /©n/, we will readily see in Figures 5.3.e. and 9 5.3.f. that the two cities vary greatly in (-ing) usage. 4. The Variable (ty, dy, ny) palatal glide Examples: tune, due, new student, duke, stupid, avenue, nude, nuclear, mature, tube, tune, Tuesday, and a few others. This variable is the presence or absence of the palatal glide [j] after /t/, /d/, or /n/ and before /u/. In Canadian English, the palatal glide [j] is normally not heard in the word suit and always heard after /f/, /v/, /m/, and /b/; /k/ is followed by either value of (yu). Ill Figure 5.4.a. ty, dy, ny = ty, dy, ny by Class •^WORKING ASLOWER MIDDLE X=MIDDLE <J>=UPPER MIDDLE *=LQWER UPPER Figure 5,4.a. presents the averaged percentage scores for the [j] option of variable (ty, dy, ny) for all five classes in all five styles. The graph illustrates that there is a strong and ordered pattern of class differentiation with the exception of the working class. The non-linear ranking of the working class requires some comment. For a number of phonological items, the working class and 112 the lower upper and upper middle classes have similar patterns. An explanations for this lies in the fact that Ottawa and the surrounding Ottawa Valley was settled in the 1840s, 50s, and 60s, by Scots and Irish of mainly humble social standing. Many of the descendents of these settlers have now reached the upper middle and lower upper classes of Ottawa while a large number have remained in the working class. The results found for this item and others such as (ou) , Cai) , C.hw) , (ar) , (potato), (or) and (film) point to this common heritage. The remaining classes, LM and M, are to a large extent made up of new Canadians and newcomers, many of whom follow more closely General Canadian usage patterns. Females over forty years of age have much higher scores than the other sex/age groups; see Figure 5.4.b. 113 Figure 5.4.b. ty, dy, ny = ty, dy, ny by Sex/Age LTJ^MRLES GE 40 YEARS OF AGE A """MALES LT 40 YEARS OF AGE X"""FEMALES GE 40 YEARS OF AGE 0=FEMALES LT 40 YEARS OF AGE The stylistic variation is significant only between R and FS where the gradient is quite steep. For further information on Ontario and Canadian usage of this variable see Avis, 1956 and CEU."'"^ Speakers of Northern American generally do not appear to have this palatal glide as a goal in their most formal styles. 114 5. The Variable (rV),r metathesis Examples: apricot, professor, Africa, agriculture, presume, hundred, introduce, precisely, presented, promoted, provincial. Figure 5.5.a. rV = rV by Class •^WORKING A-LOWER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=LDVER UPPER o 2n STYLE 115 This variable is the presence or absence of r metathesis. The pattern of variation is illustrated in Figure 5.5.a. The percentages represent the frequency of metathesis for all classes in four styles. We can see that in spite of cross-overs, there is some class differ entiation for this variable as the upper classes have the higher scores and the lower classes have the lower scores. There is no clear pattern of stylistic variation for this variable. The pattern which we expected for all classes materialized for the working class only. An investigation of this variable according to sex revealed that females had consistently fewer occurrences of metathesis than did males. 6. The Variable (st) Examples: He just left, half past eight, last Saturday, fast car, almost two, youngest, must be, best part, West Virginia, mostly, oldest, and first thing. This variable is the presence of absence of the /t/ of the cluster /st/ in morpheme final position. 116 Figure 5.6.a. st = st by Class •=V0RKING A=LOWER MIDDLE X=MIDDLE •-UPPER MIDDLE *=LQVER UPPER a a. (—to LU CJ Cr: UJ Q_, P' STYLE Figure 5.6.a. displays the presence of /t/ for all classes for four styles; W and P are combined. We can see some class differ entiation, with the working class having the lowest scores through out and the lower upper class having the highest scores twice. The scores for the three middle classes reveal no regular pattern. Too H7 few /st/ values in morpheme final position would appear to be a stigmatized feature especially in formal contexts. The stylistic variation is quite regular as all scores decrease as we move from formal to more informal styles. We have no reports to tell us what the situation with regard to this and the previous variable in Northern American might be; however, we believe that it would be very similar. 7. The Variable (h) Examples: he, him, her, his, hers, them, and come here. This variable is the presence or absence of initial aspiration in words internal to phrases and in unstressed position. 118 Figure 5.7.a. h = h •=WORKING A=LQVER MIDDLE X-MIDDLE <!>=UPPER MIDDLE *=LOWER UPPER a a _ a CD STYLE Figure 5.7.a. illustrates the frequency of this initial aspira tion of the five classes in MP, R, and FS. No regular pattern appears evident for sociological co-variation of this variable. However, the stylistic gradient from formal to informal is quite steep for all groups. Notice that the lower upper class ranges 74 percent in MP 119 to 0 percent in FS. Certain environments appeared to trigger a higher frequency of initial aspiration than did others; for example, the phrase bike he bought elicited a low frequency of /h/ (39%) whereas, the phrase guess he'll be elicited a higher frequency of /h/ (76%). This may lead someone to investigate the liaison patterns of plosives versus sibilants regarding silent 'h'."""""" Furthermore, there was a marked tendency among our informants towards initial aspiration after medial Itlwhen pronounced as [t] and towards the absence of initial aspiration when medial t was pronounced as [d]. The phrase to invite her serves as an example for this interesting case of rule ordering and interdependence. 8. The Variable (V##V) Examples: the apple, *a apple, to eat, to invite, the egg, the answer, the owner, the air, the Arts Centre, the ice, the other, the old, the end, the area, the alley, the Ex, the apartment. Almost all grammar books prescribe a rule which states that a function word before a word beginning with a vowel sound must not end with a schwa. This variable (V^^V) measures the presence or absence of that schwa. 120 Figure 5.8.a. V#V = a V by Class LH=W0RKING A=L0WER MIDDLE X=MIDDLE SUPPER MIDDLE *=L0WER UPPER a oo STYLE Figure 5.8.a. indicates the frequency of the occurrence of the schwa before words that begin with vowel sounds. Here again the working class and the lower class is differentiated strongly from all other classes in styles R and FS, the styles of connected speech. This pronunciation across word boundaries would appear to be a 121 stigmatized feature. There is no other pattern of sociological or stylistic variation which is evident. It is important to observe that Northern American speakers, especially Michiganders are known for this schwa usage, and that this usage may be a contributing factor to the low assessment of Northern American English by our Canadian informants; see data on language attitudes, Appendix E. 9. The Variable (t#y, d#y) Examples: would you, did you, could you, but you, bet you, get you, beside you, that you, told you, what (are) you, thought you, at you, what you've done. This variable is realised either with a strong hiatus between the aveolar plosives /t/ or /d/ and the /ju/ that follows, or as the affricated palatal glides [tju] or [d^,u]. 122 Figure 5.9.a. t y, d y = t y, dy by Class •^WORKING A=LDWER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=L0VER UPPER o a _ I—CO LU CJ> -CK LU Figure 5.9.a. displays the average>frequencies of the full hiatus as spoken by the five classes in three styles. Although no clear sociological patterns or stylistic patterns are evident, the graph does demonstrate that the affricated palatal glide is the prevailing usage for all classes in all styles. 123 10. The Variable (ou): the Canadian Diphthong Examples: house, south, mouth, lout, shout, houses, without, about, trout, out, and mouse. This variable (-ou) is the major phonological feature which differentiates Canadian English from Northern American English. This diphthong is pronounced relatively more close [AU] before voiceless consonants and more open [ao] before voiced segments and 12 in final position; see Chapter 3, Section 2. Northern American speakers pronounce the more open variety before the voiceless consonants as well as before the voiced environments. ' 124 Figure 5.10.a. ou = AU by Class •^WORKING A=L0VER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=L0VER UPPER I—LD LU O CL. 3' o CM MP P STYLE R "?S Figure 5.10.a. displays the average frequency of [AU]before voiceless consonants for the five classes in the five styles. There is no significant stylistic co-variation for the five classes in the five styles. Further, there is only a partial pattern of socio economic co-variation, with the lower upper class having the highest 125 scores and the middle and lower middle classes having the lowest scores. However, the pattern becomes quite confused when we add the working class and the upper middle class. The important aspect of the data for Canadian English is the high frequencies of [AU] before voiceless consonants for all classes in all styles. We attempted to find other sociological factors which might influence the frequency of this value. First, we hypothesized that informants with rural backgrounds would have a higher instance of this value than would those informants with urban backgrounds. We contrasted these two groups and, as Figure 5.10.b. reveals, we found no evidence to substantiate our hypothesis. Figure 5.10.b. ou = AU, Rural/Urban •=RURRL A=URBflN 127 Figure 5.10.C. ou AU, Several Generation Canadians versus New Canadians •^SEVERAL GENERATIONS A=NEW CANADIANS a a _ r—iO LU CJ CLYL CL. MP STYLE ?S Secondly, we hypothesized that informants who were new Canadians or whose parents were new Canadians would have a lower instance of this diphthong [AU] than would Canadians of longer lineage. Figure 5.10.C. proved this hypothesis to be correct. Finally, we investi gated the sex/age groups and found that females who were less than 128 40 years of age had markedly lower scores than the other three groups; see Figure 5.10.d. This is a somewhat disquieting sign for this Canadian diphthong, as it is often young women in society who set the trends in language usage for the future. Figure 5.10.d. ou = AU by Sex/Age •=MALES GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMALES GE 40 YEARS OF AGE 0=FEMALES LT 40 YEARS OF AGE °KP ? ? R1 ?S STYLE 129, Further, this diphthong was noted to have a range of height of the initial vowel from [A] to [e] . The initial vowel was evaluated to be higher among male and female informants over 40 years of age and among young males than among young females. Three informants who were university students and who had travelled and lived in other countries stated separately that they had been singled out for this pronunciation and ^h and that they had tried to suppress both. Moreover, it was noted on several occasions that when a word containing the variable (ou) was pronounced slowly 13 or with hesitation, the diphthong was pronounced [ ao] , not [ AU] . 11. The Variable (outV): the Canadian diphthong plus medial It/ Examples: pouter, shouted, about it, about an, out in, out on, etc. The variable (ou) before medial/t/presents us with an interesting rule ordering problem and with scores much different from those for variable (ou) . Rule (1) ou -»- Au/—t V when medial /t/rule, t -> d/v-V, is not applied. Rule (2) ou -> ao/-dV when medial/t/rule, t ->• d/V-V, is applied. Figure 5.11.a displays a strong and ordered socio-economic stratification in MP and FS but not in R. 130 Figure 5.11.a. outV = Aut by Class •=W0RKING A-LOVER MIDDLE X=MIDDLE •=UPPER MIDDLE *=L0WER UPPER As illustrated in Figure 5.11.a., there are high scores in styles which generally elicit the medial /t/ as [t], and there are generally low scores when the styles elicit the medial /t/ as [d]. There was, however, no one to one relationship present here, for there was occasionally an [acw] before a [t] and a 17 percent instance of [AU ] before [d]. 131 12. The Variable (T) This Canadian diphthong is in many ways parallel to the var iable (ou), but far less known and consciously perceived. Examples: kite, night, right, bike, tonight, alright, invite, sliced, write, sjLte, etc. This variable (T)is a major phonological feature which differ entiates Canadian English from Northern American. This diphthong is pronounced relatively more close, [ai ] before voiceless consonants and more open Jai] before voiced environments and in final position; see illustrations 3.2 and 3.3. As is the case with (ou), speakers of Northern American pronounce the more open variety, [at] before voiceless as well as voiced environments. 132 Figure 5.12.a. T = ei by Class •^WORKING A=L0VER MIDDLE X=MIDDLE SUPPER MIDDLE *=LQWER UPPER LU CJ CK LU CD . CM MP "7 STYLE Figure 5.12.a. shows the average frequencies of the value [ei] before voiceless consonants for the five classes in all five styles. We can see that there is a fairly consistent pattern of socio-economic variation, with the exception again of the working class. The common background of the lower upper class and the working class of Ottawa 133 was discussed above when we presented the variable (ty, dy, ny). There is no clear pattern of stylistic variation. It is important to observe the high percentage of [ei] before voiceless consonants for all classes and styles. The lower middle and middle classes had the lowest scores for this typically Canadian item. Many informants told us that the Canadian diphthongs are stronger in the rural areas than in the cities. Following this lead, we cut our sample accordingly. 134 Figure 5.12.b. i = si, Urban versus Rural •=RURflL A=URBflN o UJ CJ UJ CM O MP P' STYLE "ft Figure 5.12.b. displays consistent separation at all stylistic markers with urbanites scoring lower. Among those with urban back grounds, however, was a significant minority of new Canadians, while all those with rural backgrounds were Canadians of several generations, 135 Figure 5.12.C. T = ei by Generations •-SEVERAL GENERATIONS A-NEW CANADIANS CD a _ LU CJ cn LU CM MP STYLE R ?S We.then examined the scores for new Canadians versus Canadians of several generations. Figure 5.12.c, not surprisingly, illustrates that Canadians of several generations have higher scores of this characteristically Canadian diphthong than do new Canadians. 136 Figure 5.12.d. T = ai by Sex/Age GAMBLES GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMALES GE 40 YEARS OF AGE <!>=FEMALES LT 40 YEARS OF AGE Looking for further sociological patterns in the variation of [ai ], we analysed the scores according to our four sex/age groups, Figure 5.12.d. As is the case for the value [AU], females under 40 years of age had scores much below the other groups. The frequency percentages for this group leads one to predict a general but slight 137 decline of this Canadian marker for all Canadians in the next few decades. As was also the case with [AU], we noticed a range of height for the initial and final vowels of this diphthong; females under 40 generally had a lower initial /a/ and glided /i/ than did the other sex/age groups. 13. The Variable (TtV) 14 Examples: writer, invite (h)er, invited, right up, right away. Because of the rule ordering options, the high diphthong [ai] immediately before a medial /t/ position had to be investigated separately. Rule (1) ai -> aiV—tV when medial t rule, "t -* d/V_V, is not applied. Rule (2) ai ->- ai/—dV when medial t rule, t -> d/V_V, is applied. 138 Figure 5.13.a. itV = aitV by Class CD=WORKING SLOWER MIDDLE X=MIDDLE •=UPPER MIDDLE *=LOVER UPPER Figure 5.13,a, displays the average percentages per class of the value [ai] when immediately before medial/t/. As we can see, the scores for (XtV) closely follow the pattern for medial/t/, i.e. [ai] before medial/t/as [t] and [at] before medial/t/as [d]. However, it should also be noted that this correlation is not a one to one relationship, e.g. we transcribed several instances of ;[ait] and, more interestingly, 139 we transcribed even more instances, approximately 10 percent, of [aid]. 14. The Variable (un-) Examples: untrue, unbelievable, unbearable, unforgetable, unreal. This variable, a negative prefix, can be pronounced in two dif ferent ways, [An] or [an]. Informants from Smith's Falls and Renfrew of the working and lower middle classes had very high scores of [an]. 140 Figure 5.14.a. un = an, Rural/Urban/Class •=RURflL WORKING I LOWER MIDDLE A^RURRL MIDDLE 8, UPPER X=URBRN WORKING I LOWER MIDDLE DURBAN MIDDLE I UPPER Figure 5.14.a. shows the frequency of the pronunciation:[an] in three styles W, R, and FS, and according to four sociological groups. Informants with rural backgrounds of the working and lower middle classes pronounced the negative prefix -un as:[an] much more fre quently than did any other group. Informants with rural backgrounds of the middle, upper middle and lower upper classes were the group 141 with the next highest scores, although the spread between these two groups was very large. Our usual five socio-economic groups and urban-rural background groups were replaced by rural-class versus urban-class groups to more forcibly illustrate the sociological forces at play as regards this variable, i.e. rural and class. It is interesting to note that the working and lower upper classes do not closely approximate one another in the pronunciation of this item although one might have suspected it. This pronunciation [an] is a stigmatized feature of Canadian English. Some speakers of Northern American English very definitely have this feature [an] in their speech though we do not yet know to what extent, and whether there exists a sociological correlation. This item requires further investigation with refer ence to stress. 15. The Variable (nd) Examples: sandwich, grandmother, husband, grandfather, second, almond, around, sound, and his, knives and forks, hundreds, friends, kinds, kind of, landscapes, and grounds. This variable is the presence or absence of the /d/ in the cluster /nd/. 142 Figure 5.15.a. nd = nd by Class •=WORKING A=LOWER MIDDLE X=MIDDLE <S>=UPPER MIDDLE *=LOWER UPPER Figure 5.15.a. shows the scores for the value /nd/ for the five classes in five styles. The stylistic variation follows a pattern similar to that for (VtV), (ntV), (-ing), (outV), (TtV), and (hw). No co-variation patterns related to socio-economic class are evident •in the first, three styles, but there is strong stratification in R and FS. It is, however, very interesting to observe that again the upper 143 middle class is more informal-in R than in FS; the lower upper class maintained equal scores for these two styles. We also see that P is a very informal style. This observation is very important as this is a new task/style not used in either Labov or Trudgill. The choice of the lexical items for each style was extremely important and a potential source of skewed results; some words such as sound had a very high score of /nd/ while and was usually pronounced /n/. For this item, we recorded a few instances of (nd) as [nt], i.e. final devoicing. This latter pronunciation was almost exclusively limited to the working class. 16. The Variable (aer), (eel) Examples: Marry, guarantee, caramel, Barry, wheelbarrow, Carp, (balcony) and (Aylmer). In Chapter 3, page 41 we discussed the convergence of vowel sounds before /r/. This variable is the pronunciation of (eer) as either [aer] or [er], the converged form. 144 Figure 5.16.a. ssr =. er by Class •=WORKING A=LOWER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=LOVER UPPER Figure 5.16.a. displays the frequency in percentage of (j&r) pronounced as [er], the converged form, for all classes in three styles, MP, W, and R. We can see that as the formality of the style decreases, there is a slight increase in the instances of this value, a moderate amount of stylistic variation. 145 We also investigated this value with reference to rural versus urban background, and sex/age groups. Figure 5.16.b. ser = er, Rural/Urban •=RURBL A-URBRN o_4 CM MP 7" P' STYLE FS 146 Figure 5.16.C. asr = er by Sex/Age •"""MALES >= 40 YEARS OF AGE A=MALES < 40 YEARS OF AGE X=FEMALE5 >= 40 YEARS OF AGE <> """FEMALES < 40 YEARS OF AGE o si fM MP P' STYLE ?S Figures 5.16.b. and c. display the scores for these two socio logical parameters, respectively; we can see that our informants with rural backgrounds converge /aa/ before /r/ less frequently than our urban informants and that women over 40 converge /eer/ to /er/ less frequently than all other sex/age groups. 147 Figure 5.16.d. eer = er by Age •= GE 40 YEARS OF AGE A- LT 40 YEARS OF AGE Figure 5.16.d. demonstrates that there was significant differ entiation with regard to age alone. Our Somers' D index number 16 shows the difference by age to be .24896. Two other sources which deal with this variable in Canadian English are Gregg, 1957, and SCE.17 148 Speakers of Northern American are known for merging marry, merry, and Mary to an [er] pronunciation. The retention of [aer], therefore, is a distinguishing marker for Canadian English. The pronunciation of Carp as [keerp] occurred only twice as true elici-tations, but possibly all informants knew and imitated it as Ottawa Valley "twang". • Examples: balcony, and Aylmer. A very limited statistical base indicates. the variable (ss I) would follow a pattern similar to that of (aer). 17. The Variable (Vr->sr) Examples: for, particular, forget, there's, or, and you're. This variable is the pronunciation of the vowel quality before /r/ or the reduction of the vowel before /r/ to schwa. 149 Figure 5.17.a. Vr = 3r by Class •^WORKING I A=L0VER MIDDLE X=MIDDLE SUPPER MIDDLE *=L0VER UPPER ! i o o _, s-1 oo 0.4 I—CD ZZL LU CJ H CK CL. MP STYLE ?5 Figure 5.17.a. displays the frequency of vowel quality reten tion before /r/ for all five classes in two styles only. We can see a good deal of stylistic variation with vowel quality retention increasing as the formality of the style decreases. This same figure shows that with regard to socio-economic parameters, the upper classes have less vowel reduction in formal style than do the other classes. 150 Figure 5.17.b. Vr = er by Sex/Age LTJ-MRLE5 GE 40 YERRS OF RGE A^MRLES LT 40 YERRS OF RGE X^FEMRLES GE 40 YERRS OF RGE <5>=FEMRLE5 LT 40 YERRS OF RGE o o _ CD o. h-uo ZZL UJ CJ • CK LU CKr-, CM HP i? p1 PT" STYLE ?S Figure 5.17.b. reveals that females both old and young have markedly less vowel reduction in non-connected speech than do males. The scores of all groups converge in Free Speech, however. As vowel reduction in English is very much dependent upon rhythm 18 and stress patterns, and as Canadian and Northern American English 151 rhythm and stress patterns are very similar, one would expect this variable (Vr->ar) to yield similar scores in surveys of both dialects. 18. The Variable (se) . Examples: that (in stressed position), glass, grass, and last. This variable (ae) has two phonological realizations; they are [ee] and [ae] . At the present time in Canadian English, the lower ae, symbolized . [ae], is found optionally only in a few words some of which are listed above. 152 Figure 5.18.a. ae = ae by Sex/Age •=MALES GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMflLES GE 40 YEARS OF AGE «>=FEMALES LT 40 YEARS OF AGE o o _, si STYLE Figure 5,18.a. illustrates that females who are less than forty years of age pronounce Ig] far more frequently than any other sex/age group, this leads us to believe that this pronunciation may increase in the future. Our statistical analysis of socio-economic variation and of stylistic variation did not yield any conclusive results. 153 Canadian English, like Northern American, has retained the . [as] in words such as grass, dance, can't, half, path, and pass, whereas 19 Standard Southern British has adopted [a:]. Only one woman in the Ottawa survey consistently pronounced the.[a:] in the British manner. She had lived in Britain for a short time, was married to a Britisher, and found her identity in British culture, a not uncommon Canadian phenomenon. 19. The Variable (hw) Examples: where, why, what, when, which, wheelbarrow, whipped, and whether. This variable is the presence or absence of the feature '~voice' represented by the /h/. before the /w/. The frequencies of /hw/ for all the classes and styles are displayed in Figure 5.19.a. Our graph demonstrates a definite and ordered social class stratification, with the lower upper class being followed in sequence by the upper middle, middle, lower middle, and working classes. The graph also reveals that there is a progressive increase in the range of stylistic variation as one moves up the social structure. Further, we can observe that the upper middle class was more informal in R than in FS. Figure 5.19.a. hw = hw by Class •=WORKING A=LQWER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=L0WER UPPER 155 Figure 5.19.b, hw = hw by Sex/Age m=MALES GE 40 YERRS OF AGE A=MALES LT 40 YEARS OF AGE X=FEMALE5 GE 40 YEARS OF AGE <J>=FEMALES LT 40 YEARS OF AGE A close look.at.the sex/age groups, Figure 5.19.b., reveals a strong generation gap. The stylistic variation for females and males over forty years of age is very similar to the stylistic patterns for (VtV), (ntV) , (-ing), and (nd). The lines representing stylistic variation for females and males under 40 are much more horizontal, 156 representing little change from style to style. It is important to notice that young women consistently employ this pronunciation less frequently than any other group. This tendency is again a possible sign of less frequent usage of (hw) in the future. Further, we notice again that for one group, the reading style was less formal than the free speech style. Figure 5.19.c. reveals that informants whose families have lived several generations in Canada have /hw/ as a goal in their speech (see scores for MP and W) much more so than do the new Canadians. c 157 Figure 5.19.C. hw = hw by Generations •=SEVERRL GENERATIONS A=NEV CANADIANS For further discussion and data regarding this variable in Canadian, British, and American English, see Avis, 1956; Gregg, 1957 and 1972; and CEU.20 158 20. The Variable (kt) , (pt) Examples: picture, perfectly, asked, exact, exactly except, perfect. This variable is the consonant cluster of two voiceless stops; one value is both consonants articulated, kt or pt, and the other value is only one of the consonants articulated, k,t, or p. Figure 5.20.a. kt = kt by Class •=V0RKING A=L0WER MIDDLE X=MIDDLE 0=UPPPR MIDDLE *=L0VER UPPER o STYLE 159 Figure 5.20.a. displays the frequency of the instances of both consonants being articulated for all classes in four styles. We can see that in the non-connected speech styles there is a large dif ference in scores among the socio-economic classes, but that the ordering is somewhat confused. There is a strong stylistic varia tion which takes place mainly between the non-connected speech styles, MP and P, and the connected speech styles, R and FS. Notice again that the upper middle class are more formal in their free speech style than in their reading style. Further, those items which had clusters of three consonants, e.g. perfectly and exactly, elicited much lower scores. 21. The Variable /»/ Examples: shone, collar, cot, daughter, ()ttawa, caught, hot, bought, lock, so>ccer, spot, not, Rochester, sock, rock, lot, jodd, fog, God, box, politics, pod, shock, holiday, stop, block, Boston. In Canadian English, there is no phonemic distinction between /r>/ and;/a/ as there is in Northern American. In Ottawa, therefore this variable is pronounced either [x>] or: [a]; the degree of lip rounding does not distinguish meaning at all. As a consequence, this variable is actually pronounced with a continuous reduction in the range of lip rounding from [n] to [a], but for the purposes of transcription, we placed each utterance of this variable into either the /D/ type or the /a/ type of the Northern American phonemic vocalic system. 160 Figure 5.21.a. v = v by Class LTJ=VQRKING A=L0WER MIDDLE X=MIDDLE SUPPER MIDDLE *=LQWER UPPER Figure 5,21.a. reveals no systematic socio-economic pattern of variation, but it does display the strongest stylistic variation of this study, ranging from 100 percent to 14 percent. The steepest gradient is between R and FS. 161 Figure 5.21.b. v = t» by Sex/Age •=MALES GE 40 YEARS OF AGE A=MALES LT 40 YEARS OF AGE XrFEMALES GE 40 YEARS OF AGE <5>=FEMALES LT 40 YEARS OF AGE Figure 5.21.b. reveals that women under forty years of age have consistently higher scores than any other sex/age group. 22. The Variable (th) Examples: whether, thirsty, Thursday, grandfather, that, some thing, South, father, grandfather, grandmother, three, 162 thirty, nothing, Dorothy, the, either, theatre, and many more. The variable (th) had the following values ordered by frequency from left to right: S, 6, n, z, t, and d. •^WORKING A=L0WER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=LDVER UPPER l—to LU CJ -Cr: LU Figure 5.22.a. th = 5, 9 by Class Q-J °f? J P1 R1 ?5 STYLE 163 As illustrated by Figure 5.22.a., there was no perceptible variation in the first four styles for this variable. However, in Free Speech we detected socio-economic variation which was con sistent with our socio-economic findings elsewhere. The upper middle class had the highest score followed by the lower upper, middle, lower middle and the working classes. The value /n/ was pronounced in phrases such as: and that, and this, and and those. The value /z/ was often detected in the phrase Is that right, (eh)? Both these pronunciations, /n/ and /z/, are cases of progressive contextual assimilation which are natural to native speakers of English. The /t/ and /d/ pronunciations were in a few cases mere performance errors, and in other cases they were instances of language interference from French and other mother tongues. We recorded no /t/ or /d/ sounds from informants with 21 Irish backgrounds. Because of the class ordering of the pronun ciation of this variable, we can assume some stigmatization for those who frequently substitute other sounds for the variable (th). Figure 5.22.b. Labov's th = 5, 9 by Class 164 (th) A comparison of statistics for Ottawa (Figure 5.22.a.) and 22 Lower East Side Manhattan (Figure 5.22.b) reveals the fact that people from the Lower East Side, some of whom are third and fourth generation Americans, have not been fully assimilated into English speech norms, whereas, generally speaking, all anglophone Ottawans have. This comparison provides further evidence that the Labov 165 Manhattan survey cannot claim to represent main stream American 23 speech. 23. The Variable (or) Examples: orange, sorry, Dorothy, and porridge. This variable had five allophonic values; they were: [o], [v], [QJJ to], and [r>:]. The value [D] was used most frequently by all groups in the three styles measured, with no systematic socio economic or stylistic pattern evident. The sociological findings for this variable were: 1) the working class and the lower upper class again had similar usage patterns, this time in the frequency of [a] in the word sorry in word list style (see Figure 5.23.a.); 2) informants whose families have been in Canada for several generations more often pronounced the over-rounded sound [ o] before /r/ than did newer Canadians (see Figure 5.23.b.); and 3) informants over forty pronounced this value more frequently than did those under forty (Figure 5.23.C.). Figure 5.23.a. or = a by Class •^WORKING A=LGVER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=L0WER UPPER o SH r- S| STYLE 167 Figure 5.23.b. or = o by Generation •=SEVERRL GENERATIONS A=NEW CflNRDIflNS CD CD o . CD CD . -CO LU CJ OC Lu Q_ 168 Figure 5.23.c. or = o by Age •= GE 40 YEARS OF RGE A= LT 40 YEARS OF AGE a a _, o. I—CD UJ O Words and Phrases The following variables are not phonological items but words or phrases. 169 24. The Variable (going to) This variable has many realisations including [goiQtu], [gointu], [g6inta] and [gomta] all of which we have designated goto 1 and [gene], [gana], and [gona] which we have designated goto 2. Figure 5.24.a. going to = goto 1 by Class CD=WORKING ASLOWER MIDDLE X=MIDDLE «>=UPPER MIDDLE *=L0WER UPPER a a. CD . I—to UJ CJ LU 0_, 3" O -MP STYLE 170 Figure 5.24.a. displays the frequency of goto 1 for all five classes in the Reading and Free Speech styles. The frequencies in the Reading style reveal no systematic variation according to class, with 66.8 percent of all occurrences being [gomta]. We do see, on the other hand, a great deal of socio-economic variation when we analyse the scores for FS. We notice, once more, much stylistic variation and the fact that the upper middle class are more formal in FS than in R style. Also worth special notice is the fact that the working class stands alone in their low frequency of goto 1 in FS. This high frequency of goto 2 (90 percent) by the working class causes it to be evaluated as a stigmatized form if used with great frequency; on the other hand, a fifty-fifty mixing seems acceptable. Out of 397 elicitations of the variable (going to) in the Ottawa survey, the following occurrences were obtained: Form Occurrences Percentage 1. go inta 230 57.9 2. goana 59 14.9 3. go i nt a 52 13.1 4. gana 34 8.6 5. gone 11 2.8 6. goanta 7 1.8 7. goirytu 4 1.1 100.2 American linguistic research has caused all ESL textbooks printed in the United States to teach [gana] as the goal for connected speech; our evidence strongly suggests that, in Canada, [gointa] would be more appropriate. 171 25. The Variable (milk) This variable is pronounced either [milk] or [melk], and has to do with the reduction of the number of vowel phonemes before /!/. Figure 5.25.a. milk = milk by Class •^WORKING A=L0VER MIDDLE X=MIDDLE <?>=UPPER MIDDLE SLOWER UPPER o o _ 7 P1 ? STYLE ?s ! 172 Figure 5.25.a. displays the frequency of [milk] for all classes in two styles only, namely W and R. The working class has the lowest scores for this value in both styles. The scores for the other classes show no consistent pattern of social differentiation. The stylistic variation is fairly large between the two styles. 26. The Variable (good) This variable has two pronunciations of the vowel, 1) with slight lip rounding and the [o ] in position as shown in Illustration 3.1, page 37 and 2) with no lip rounding and the [o] brought forward to a centralized [©']. 173 Figure 5.26.a. good = gc£d by Sex/Age LTJ^MALES GE 40 YEARS OF AGE A^MALES LT 40 YEARS OF AGE X=FEMALES GE 40 YEARS OF AGE 0=FEMALES LT 40 YEARS OF AGE STYLE Figure 5.26.a, displays the frequency of the latter version. We can see that males who are less than forty years of age generate this pronunciation in Free Speech far more than any other group. Females who are less than forty years of age appear to be a distant second in the production of this sound. This pronunciation is 7 174 found mostly among teenagers, and it seems to be a modern substitute for slang words such as neat, cool, and groovy. We note that the Somers' D analysis indicates a strong socio-economic differentiation for this variable. We would require many more data, however, before we could be certain of this trend. 27. The Variable (tomato) This variable has at least three values in Canadian English, and all were found in Ottawa: [tem&to], [-temffito], and [temeido]. The first two values have prestige value for women over forty years of age. The:/a/ value is associated with British English and has the highest prestige. Many informants who had pronounced /•&,/ later claimed that they always said /a/. The /ei/ pronunciation is associated with American English and has no stigma attached to it. Figure 5.27.a. tomato = tamlajro by Sex/Age 175 feel •=MALE5 GE 40 YEARS OF AGE A-MALES LT 40 YEARS OF AGE X-FEMALES GE 40 YEARS OF AGE O-FEMBLES LT 40 YEARS OF AGE Figure 5.27.a. demonstrates that women over forty stand out, predominately over all other sex/age groups in the pronunciation of [temdto] and 11: amart o j" combined. A further break-down of the statis tics of this variable for these women according to their socio economic class reveals interesting patterns in variation. Figure 5.27.b. tomato = tam(_a)to Women over 40 Upper Middle Class Figure 5.27.c Key: Figure 5.27.d. Figure 5.27.e. Figure 5.27.f. 1001 80 60 40 20 7 / / / / / / / W / / / / / / / i / / / I R = .33 Lower Upper Class 100 80 60 40 J 20 / / / / w 7 / / / / / / / JZL R 100 80 60 40 20 Middle W R Lower Middle Class 100 80 60 40 20 W 100 80 60 40 20 Working Class W R 177 Figures 5.27.bv, c. , d., e. , and f. , display the frequencies for the three values, /a/, /ss/, and /ei/ in the three styles W, P, and R for women over forty of the upper middle, lower upper, middle ' class, lower middle, and working classes respectively. We notice again that the upper middle class has higher scores than the lower upper class. Moreover, we can see that as we move down the socio-economic ladder, we move from a British pattern to an American pattern; this tendency was observed throughout the survey with other variables. Other sources which deal with the Canadian, British, and American pronunciation of (tomato) are Avis, 1956; Gregg, 1973; and CEU.24 178 2. Evaluation and Summary  Cross-analysis The data for the 27 linguistic variables presented above provide us with ample evidence of the co-variation of phonological features on the one hand and sociological and stylistic phenomena on the other. The Canadian English spoken by our sample of informants from Ottawa clearly reveals socio-economic differentiation and stylistic variation for the majority of variables. Thus far in this chapter, we have analysed each phonological variable individually by means of the sociological and stylistic parameters; we will now attempt to analyse the same data and graphs by focusing on the sociological and stylistic parameters as refer ence points. Stylistic Variation Stylistic variation was a major hypothesis of this study. It was hypothesised that a number of linguistic items would show phonological variation directly related to the degree of formality of the task to be performed by the informant. From our data, we can observe that of the 27 phonological items with which we dealt in this survey, 20 items underwent stylistic variation in a fairly regular manner. These items are: VtV = VtV; ntV = nt; -ing = -in; ty, dy, ny = ty, dy, ny; rV = rV; st = st; h = h; d#y = d#y; out = Aut; Tt = ait; nd = nd; Vr = er; Vr = er; hw = hw; kt = kt; v = r>; th = th; going to = goto 1; milk = milk; and tomato ?? tem\d)to. Only the following seven items did not show clear stylistic variation: V//V = e#V; , ou = AU; T = e i; un = an; a = as; or - or; and good - go<d. This record of, 20 out of 27 items displaying 179 stylistic variation demonstrates the weakness of the uni-stylistic dialect studies and provides a strong case for including stylistic parameters in future dialectology surveys. Socio-economic Variation Socio-economic class differentiation and linguistic co-variation was the other major hypothesis of this study. It was hypothesized that, for many items, the linguistic variation would be directly related to the socio-economic class of the speaker and that,.as one moved up the social structure one would observe more and more formal values conforming to prescribed standards. From our data we can see that ordered socio economic stratification occurred with reference to the following values: VtV = VtV, -ing = 'in, nj = nj, rV = rV, st = st, out = Aut, Tt = ait, nd = nd, hw ~.hw, and th = th. In addition to such socio-economic dif ferentiation and ordering, it is clear to see that informants higher in the social structure have a much broader range of styles at their command than do those informants lower in the social structure. See in partic ular the graph for hw = hw; it is probably the case that an informant would be able to communicate at any level of formality below that at which he performed during the interview but less likely that an informant would speak in register ranges much above that which was recorded during the interview. This range of delivery on the part of the upper classes substantiates the often heard working class claim that the upper classes are dishonest and two-faced, talking to different people in different ways, while they are themselves honest and truthful, talking the same way to all people. 180 Figure 5.28.a. hw = hw Register Control •"""WORKING A ""•LOVER MIDDLE X=MIDDLE <I>=UPPER MIDDLE SLOWER UPPER o o _ STYLE Hatch marks indicate area of control for our LU and Working classes. Our data shows that socio-economic differentiation is a major para meter related to phonological variation and demonstrates the weakness of dialect studies which do not include socio-economic data. The graph for hw, for example, displays a 70 percent spread between the lower upper and working classes in Minimal Pairs. 181 Data with Reference to Working Class From the data and graphs available for our phonological items, we are able to see that the working class has recorded lower scores than any other class for the following preferred values: VtV = VtV; ntV = nt; -ing = in; tj, dj, nj = tj, dj, nj; rV = rV; st = st; V#V = V#V; out = Aut; Tt = ait; un- = An; Vr = Vr not or; hw = hw; th = th; or = or, going to = is goto!, milk = milk; and tomato = tam(a>to. This set of data builds a strong case to refute any claim that there is no or only very little social differentiation in Canadian English. Stigmatized Forms In addition to attaining the lowest scores for the above listed items, the working class set itself apart from all other classes in its pronun ciation of -ing = on, rV = Vr", st = s#, V#V = e#V, un- = an, and th = n, z, t, and d, going to = goto I, and milk = me Ik. The scores for these latter eight items were so far removed from the scores of the other classes that they could not be considered simply the lowest, but they had to be cate gorized as drastically different. These forms are judged to be stigmatized values when used in formal situations or when used too frequently in the more informal styles. The graphs for these stigmatized form are presented below. Figure 5.29.a ing = an •=VORKING A=LOWER MIDDLE X=M1DDLE e^UPPER MIDDLE *=LOWER UPPER o STYLE Figure 5.29.b. rV = rV •=V0RKING A=L0WER MIDDLE X=MIDDLE OrUPPER MIDDLE *=L0WER UPPER o °* i? ? S ?s STYLE 183 Figure 5.29.C. st = st [^WORKING A=L0VER MIDDLE X=M1DDLE <J>=UPPER MIDDLE *=L0VER UPPER Q Figure 5.29.d. V*V = sh CD=WORKING A=L0WER MIDDLE X=MIDDLE 0=UPPER MIDDLE *=L0VER UPPER Figure 5.29.e. un- =F an •=VORKING A=LBVER MIDDLE X^MIDDLE •=UPPER MIDDLE *=LOVER UPPER Figure 5.29.f. th = 5, 9 •=VORKING A=LOVER MIDDLE X=MIDDLE <»=UPPER MIDDLE *=LOVER UPPER LU a: LU MP P STYLE 185 Figure 5.29.g. going to = goto 1 O=U0RKING A=L0WER MIDDLE X^MIDDLE 0=UPPER MIDDLE *=L0VER UPPER UJ U or LU 0-, MP STYLE 7s Figure 5.29.h. milk = mi Ik •^WORKING A-LOWER MIDDLE X=MIDDLE SUPPER MIDDLE *=L0WER UPPER I—CD LU (_> -or LU MP P STYLE 186 Reading Style In Chapter 4, when introducing the styles which would be elicited and analysed in this study, we mentioned that it appeared as though many infor mants read our reading passage in a more informal style than their own Free Speech style. An analysis of the data and graphs reveals that the lower upper class read the variables VtV = VtV, d^y = d#y, and rV = rV more informally in the reading passage than they pronounced them in their own Free Speech style. Similarly, the upper middle class read the vari able rV = rV, nd = nd, Tt = eit, kt = kt, hw = hw, and going to = goto! more informally than they pronounced them in their own Free Speech. It would appear that the upper middle class and the lower upper class infor mants undertook role playing while reading the passage. The reading pas sage was purposely written with a view to eliciting quite informal speech. The two above mentioned classes read the passage more slowly than the middle class but more rapidly than all the other classes. As we measured reading skills with reference to our social classes we saw a progressive increase in ease and speed of reading as we moved up the classes from lower, to working, lower middle, and middle. However, when we came to the upper middle and lower upper classes we noticed a marked increase in role playing and in a concern for timing and hesitation. This reading style was accompanied with a decrease in speed. The graphs for the eight variables are presented below along with a table which presents the duration of reading by class. Figure 5.30.a. 187 VtV = VtV •=V0RKING A=L0VER MIDDLE X=MIDDLE SUPPER MIDDLE *=L0VER UPPER Figure 5.30.b. •^WORKING A=L0VER MIDDLE X-MIDDLE •=UPPER MIDDLE *=LOVER UPPER STYLE Figure 5.30.C. rV = rV •^WORKING A^LOVER MIDDLE X=MIDDLE «=UPPER MIDDLE *=LOVER UPPER LU O LU MP P STYLE Figure 5.3CLd. nd = nd •^WORKING A^LOVER MIDDLE X=MIDDLE «>=UPPER MIDDLE *=L0VER UPPER Figure 5.30.e. Tt = alt O=W0RKING A-LOWER MIDDLE X=MIDDLE «=UPPER MIDDLE *=10VER UPPER Figure 5.30.f. kt = kt •^WORKING A=LOVER MIDDLE X^MIODLE 0=UPP-"R MIDDLE *=LOVER UPPER Figure 5.30.g. hw = hw •FORKING A=L0WER MIDDLE X=MIDDLE <J>=UPPER MIDDLE *=L0VER UPPER STYLE Figure 5.30.h. going to = gotol CQ=WORKING A=L0WER MIDDLE X=MIDDLE •=UPPER MIDDLE *=i_0WER UPPER v1 ? r? ft STYLE 191 Table 5.30.i. DURATION OF READING BY CLASS Class Seconds Lower 280 Working 234 Low Mid 217 Middle 205 Up Mid 211 Lower Up 212 It would be remiss here not to discuss Labov's findings concerning Reading style: A few upper middle class speakers seemed to have the degree of control and self-awareness needed to modify their reading style in the direction of con versational style, but this is a rare effect and not a very large one.25 The fact that our data demonstrate a definite and recurring pattern of this phenomenon while the Lower East Side Manhattan informants produced only rare instances would tend to substantiate our claim that the East Side Manhattan Survey is a survey of a truncated portion of American society which did not include the upper-upper, the lower upper, or upper middle classes. Our middle class informants appeared to modify their reading about as often as Labov indicates his designated upper middle class did. This comparison would lead one to conclude that Labov's sample generally performed as did the bottom three classes in the Ottawa Survey. A further comparison of data from both surveys on the variables (th) and (-ing) also reveals very different patterns and provides evidence that the Manhattan Survey investigated and analysed inner-core urban speech which contained features of foreign language interference, not the speech of a broad sample of American society. 192 Pictures The task and style labelled Pictures was, as stated in Chapter 4, an addition to the method devised by Labov in 1966 and used by Trudgill in 1973. There was some doubt as to the proper placement of Pictures in the sequence of styles along the abscissa from formal to informal. We decided to designate and place it as the least formal of the non-connected speech styles. Let us now analyse the data available for this style. Our data reveal that for four variables, namely VtV, -ing, rV, and nd, Pictures would have been more appropriately placed to the right of Reading, i.e. more informal than Reading. The printed word in the reading passage, it would appear,.caused many informants to produce reading pronunciations and more careful speech for some variables than that which was elicited while informants identified pictures. Conversely, Pictures was more formal than Word List for the variable V#V. For the remainder of vari ables for which Pictures was a style, Pictures was correctly placed. One could evaluate from the above analysis that Pictures, and indeed all the other styles, were adequately placed and that the sequence of tasks/styles 26 established above could well be used in future surveys. Sex/ Age Groups When analysing the data with a view to comparing the performance of the four Sex/Age groups, we see evidence for the following general state ments : 1) Females over forty years of age stand out from the other sex/age groups by their tendency to pronounce more frequently the formal values of our variables. Evidence of this tendency can be seen in the graphs for VtV = VtV, ntV = nt, -ing = -in, ny = ny, Tt = ait, Vr = Vr, hw = hw, 193 and tomato = tomidjto. This tendency of females over forty to be more formal than the other sex/age groups is seen to be the case very fre quently in Chapter 6 as well. 2) The opposite tendency is observable among males less than forty years of age. They, more than any other sex/age group, pronounced the informal values of the variables. Evidence of this tendency can be seen in the graphs for VtV = VtV, ntV = nt, ny = ny, hw = hw, and good = go^d. 3) Males over forty years of age usually had scores between the extremes of the-, two previously mentioned sex/age groups. They did, however, standout in their pronunciation of ntV = nd and Vr = or. 4) Females under forty years of age also generally had scores between the extremes of the older females and younger males, but they did stand out for their lower frequency of the pronunciation of both Canadian diphthongs, (ou) and (T) and their higher frequencies of the lowered [ae] and the lip rounded /n/. The data regarding all four sex/age groups con-27 form generally to the results of the Survey of Canadian English. 194 3. Somers' D Statistical Analysis In addition to the summary in the previous Section, we have applied the Somers' D statistical description in order to further substantiate our claim of linguistic and sociological co-variation. The Somers' D analysis is best suited to the analysis of data which can be divided into simple groups of two. In order to conform to this format, we divided our four sex/age groups into two age groups, those over 40 and those under 40, and into two sex groups, female and male. Further, we dichotomized all our informants into two socio-economic groups rather than our previous five. This division was accomplished by finding the median score of the Socio-economic Class Index (27.5 points) and placing the 45 informants who were below that score into a group labelled Below. The remaining 44 informants were placed in a group labelled Above. It was then possible to run two by two grids while holding one parameter con stant. A positive sign (+) in front of the Somers' D value is here arbitrarily associated with a positive correlation to the top division of class, the older division in age, and females. The magnitude of the Somers' D value is equal to the difference between the percentage of people who had the chosen usage within one sociological group and the percentage of people who had that same usage in the opposite group. The 28 Somers' D statistical description follows. (Immediately preceding the Somers' D value for Class, Age, and Sex, we present the linguistic mean for each respective group. The value attained by the 'Above' group, Over 40, and Females is placed above the 'Below' group, Under 40, and Males.) SOMERS' D DESCRIPTION Phonological Items Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 1. VtV = VtV 2. ntV = nt MP W R FS MP W R .63275 .46002 .60117 .40846 .23837 .09198 .23388 .18972 .19710 .07555 .94574 .90698 .74109 .63877 .71124 .59074 .5000 .44574 23256 .22687 ,29147 .13953 .30334 .07000 ,22687 ,16899 ,18605 .64631 .47444 .61986 .42147 .26114 .09593 .27148 .16883 . 21013 .08030 .95370 .90667 .73241 .65855 ,65278 .64744 ,52315 ,43667 ,29111 .25427 .38889 .44222 .44906 .11700 .06624 .10470 .18000 .62908 .45562 .60685 .39356 .23960 .08389 .27403 .14350 .18164 .08360 .98519 .86179 .78815 .58475 .75926 .53488 .57160 .36450 .37832 .27338 .34522 .42276 .28160 .22407 .27338 .14935 .42927 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M ,3. -ing = in 4. ty dy ny = ty dy ny FS MP W R FS MP W FS .27993 .22804 .67442 .48062 .63721 .56889 .30255 .20910 .37497 .26800 .25679 .15432 .63953 .44574 .58430 .35675 .60465 .48148 .70930 .51550 .49138 .11795 .14206 ,23256 ,19948 .20155 ,23256 ,32204 .25581 ,29147 .17003 ,37209 ,35942 .26407 .24532 .62963 .54000 .63889 .57692 .30899 .21722 .36352 .29122 .25482 .16755 .63657 ,47500 .60268 .37466 .62963 ,48077 .71296 .54000 ,50333 ,15778 .03638 .12444 .05128 ,16239 .02000 .26767 , 35 889 34188 .25427 ,26000 ,40000 .30456 .19808 .62222 .52846 .61333 .59070 .25868 .25066 .34334 .29750 .21414 .19280 .61481 .46341 .56270 .36877 .57037 .51163 .70926 .50610 .44872 .19483 ,08291 ,10515 .01757 ,02584 .09431 .08291 ,12304 .16331 ,10284 ,20054 ,29708 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 5. rV = rV W st = st 7. h = h R FS W-P R FS MP FS .23611 .17603 .46111 .36822 .37209 .26357 .42391 .20930 .61302 .55667 .48256 .39922 .34363 .26646 .62760 .58065 . 30749 .37712 .06199 ,06212 .21499 .14005 .18605 .19919 .05271 .02326 .01220 .04940 , 13953 -.04389 .24279 .15471 .48878 .31019 .34506 .28704 .36538 .25676 .61111 .55769 .46528 .42333 .38693 .24191 .60897 .60135 , 41165 ,29238 .11538 .02355 ,13248 .14744 ,13667 .09875 ,10684 -.06667 .13665 .02911 32333 ,11966 .23837 .17654 .42054 .41111 .34959 .28889 .37209 .27174 .56667 .59303 .45370 .42683 .30183 .30783 .63426 .56481 .42610 .25034 ,11326 .01406 ,14884 .04186 ,15501 .16077 .05271 .01138 ,03624 .06481 .33604 .20417 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 8. V#V = V#V W 9. d#y, t#y 10. OU = AU R FS W R FS MP W .91860 .91111 1.0000 1.0000 .87829 .80775 .88889 .77193 .22222 .33333 .18605 .16667 .20833 .07331 .89147 .75194 .73837 .72407 .72093 .76667 .01500 NA .11628 , 15200 -.11111 .01938 ,09211 .18600 .04400 -.01520 .95833 .88462 1.0000 1.0000 .87500 .82000 .85000 .83333 .30000 .26087 .28571 .10000 .09841 .16250 .89815 .76667 .73611 .72756 .75000 .74631 .14800 NA ,20778 .04200 .03913 .18571 .00000 .14600 .03700 .01300 .90000 .93023 1.0000 1.0000 .88815 .79350 .86957 .81159 .23529 .31250 .22222 .12500 .02506 .26563 .76296 .88618 .65556 .81008 .64444 .84884 -.06000 NA ,21843 .08700 -.07721 .09722 -.20724 ,11300 -.09200 -.21100 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 11. outV = Aut 12. T = ai FS MP R FS MP W FS .79070 .73643 .72073 .61858 .81429 .65667 .36163 . 38140 .23073 .13858 .75581 .66279 .80952 .68889 .77519 .71852 .74806 .68217 .66892 .62070 .06900 .11630 .16478 -.04651 ,11630 ,04600 ,12100 .05400 .02300 -.05662 ,79630 ,74038 ,74170 ,74000 .85778 ,62449 .40833 ,34500 .24170 .15544 .80556 .64000 .86111 .66667 .86111 .66667 .81019 .64667 .72318 .58766 .01000 .02178 ,05329 .15600 ,11778 .17000 .19400 ,19100 ,24700 ,26403 .72593 .80488 .66860 .66722 .85000 .84390 .35889 .38537 .66860 .66722 .72222 .69512 .73333 .76190 .73333 .75969 .72593 .70325 .67567 .61012 .03700 .05662 -.00610 -.07480 -.05662 .01500 -.02900 -.03700 .06800 ,14661 Item Style Linguistic Somers' D Linguistic Mean Class Mean Above/Below Above/Below >40/<40 13. Tt = ait MP W 14. un = An 15. nd = nd R FS W R FS MP W R .93023 .79070 .62791 .51111 .27907 .27907 .24892 .18070 .81395 .82222 .88372 .83721 1.0000 1.0000 .09302 .06976 .56395 .50926 .18217 .14074 .30930 .20349 ,13900 .08837 .02326 .07662 -.03700 .04700 NA .02326 -.02687 .11059 .27907 .91667 .82000 .73611 .45192 .40278 .19000 .32318 .18766 .81944 .81731 .88889 .84000 1.0000 1.0000 .11111 .06000 .52778 .54167 .20833 .12821 .30972 .21800 Linguistic Somers' Mean S ex F/M F/M Somers' D Age >40/<40 .09700 .28419 .28000 .26403 .04900 .04700 NA .05111 .01282' .20940 .35556 .95556 .75610 .65556 .47674 .36667 .18293 .27567 .17012 .80000 .83721 .88889 .82927 1:0000 1.0000 .04444 .12195 .58704 .48256 .17778 .14341 .28667 .22317 .20000 .18450 .31816 .14661 -.09000 .06000 NA -.07751 .20879 .11680 .15393 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 16. aer = asr 17. Vr = Vr 18. 8B = 33 FS MP W R W FS W 19. hw = hw FS MP W .28505 .15906 .47287 .42287 .52326 .48074 .62857 .58804 . M444 . 51163 .27778 .19638 .81395 .83333 .95349 ,94767 .44657 .52043 ,55233 ,20349 ,39535 ,16667 ,26427 .06977 ,00672 .11473 ,13200 ,19616 ,01300 .02300 -.11765 .39535 .26615 .26614 .18818 .49400 .38380 .58269 .38426 .70028 .48263 .63889 .53846 .26727 .21795 .87500 .78846 .98611 .92500 ,55227 ,45061 ,65972 , 17500 ,48611 ,13462 .03351 ,10778 ,36111 ,35718 .10100 -.10187 ,13000 16400 ,11067 ,46222 ,43803 .23181 .20958 .48667 .40528 .51938 .48444 .67687 .54658 .75556 .39535 .25604 .21964 .93023 ,72222 ,99390 ,91111 .58952 ,43932 ,41667 ,33537 ,30000 ,25581 .03393 ,11870 ,03876 31677 ,36100 .06117 .28600 ,24300 ,32500 ,12087 .05220 Style Linguistic Somers' D Mean Class Above/Below Above/Beli R FS MP W-P R FS MP W R FS .17073 .11111 .17442 .08217 .10474 .03322 .62791 .62791 .47675 .47929 .16744 .20465 .13095 .10526 .90698 .91473 .81589 .90926 .62558 .59225 .15864 .10021 .05962 .09302 ,17308 .00000 -.02643 -.02326 -.02005 -.04600 -.15000 .02326 .05058 Linguistic Somers' D Mean Age >40/<40 >40/<40 Linguistic Somers' Mean S ex F/M F/M .20588 .09615 .21296 .06733 .12364 .02803 .69444 .58000 .48849 .47115 .21111 .16800 .13725 .10507 .87963 .93333 .81713 .89583 .56111 .64333 .12783 .13241 .10973 .22778 .27713 .11400 .03681 .05889 .06138 -.15000 -.11111 -.15889 -.02262 .15909 .11905 .14444 .11057 .07761 .05665 .64444 .60976 .47929 .47625 .19556 .17561 .08772 .14683 .90370 .91870 .89815 .82752 .63481 .58049 .18696 .06359 .04004 .00976 .09402 .03400 -.02643 .01355 -.02005 -.03200 .10500 .12087 .16725 Item Style 22. th = 96 MP W R FS 23. or = or W R 24. going to = goto 1 R FS 25. milk = mi Ik W Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' Mean Sex F/M F/M 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 .97442 .90543 .0000 .0000 1.0000 .97778 .64922 .72674 .83740 .75000 .61765 .35933 .93023 .88889 NA NA NA NA .13900 NA .02200 -.09302 .06900 .28706 .04100 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 .91892 .95288 .0000 .0000 .97222 1.0000 .70602 .67500 .87879 .73551 .60185 .36042 .94444 .88462 NA NA NA NA -.20000 NA .02800 .04667 .17400 .23611 .05900 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 .94891 .91183 .0000 .0000 .97778 1.0000 .68148 .69512 .88372 .68981 .55167 .38409 .93333 .88372 NA NA NA NA .09700 NA -.02200 -.05962 .14900 .14091 .04900 Item Style Linguistic Somers' D Linguistic Somers' D Mean Class Mean Age Above/Below Above/Below >40/<40 >40/<40 Linguistic Somers' D Mean Sex F/M F/M 26. good = god 27. tomato = tamvvto R W R FS W R Positive Correlation .76744 .58140 .95349 .95556 .75969 .77519 .75000 .00000 .37209 .11111 .39535 .13333 .39535 .16279 77 of 101 ,18600 -.00300 -.04600 7500 .26098 .26202 ,23256 75 of 101 .66667 .68000 1.0000 .92308 .76852 .76667 .00000 .50000 .41667 .11538 .50000 .09615 .44444 .16000 84 of 101 .01300 .07700 .00400 ,50000 30128 ,40385 ,28444 87 of 101 .60000 .75610 .97778 .93023 .73333 .80488 1.00000 .25000 .48409 .11628 .49031 .13953 .50252 .09756 74 of 101 -.15600 .04800 -.19200 75000 ,23928 ,23874 ,34688 74 of 101 NJ O 205 Chapter 5: Footnotes "'"This was done with the aid of our Socio-economic Class Index. Of course, the cut off points used to determine the placement of informants were arbitrary, but the socio-economic continuum is very real. We realize that some people will be upset with the mentioning of socio economic class. The working class and the lower class have been merged for all figures in this chapter, because there.were too few informants in the lower class to give an adequate indication of usage patterns. The two major matrices used in this chapter are presented below. Table 4.3.12.p. SEX/AGE (OTTAWA) Male > 40 Female > 40 Male < 40 Female < 40 Total Ottawa 13 24 22 30 89 Table 4.3.12.b. CLASS (OTTAWA) Ottawa Lower-Working 11 Lower Middle 22 Middle 31 Upper Middle 15 Lower Upper 10 Total 89 206 "Trudgill, when describing the sex differentiation of the variable (ng) noticed a similar pattern and offered the following explanation: This is a fact which is not, on the face of it, particularly surprising, but one that is at the same time in need of some explanation. There would appear to be two inter-connected explanatory factors: 1. Women in our society are:more status-conscious than men, generally speaking, and are therefore more aware of the social significance of linguistic variables. There are probably two main reasons for this: (i) The social position of women in our society is less secure than that of men, and generally speaking, subord inate to that of men. It is therefore more necessary for women to secure and signal their social status linguis tically and in other ways, and they are more aware of the importance of this type of signal. (ii) Men in our society can be rated socially by their occupation, their earning power, and perhaps by their other abilities: in other words, by what they do. For the most part, however, this is not possible for women, who have generally to be rated on how they appear. Since they cannot be rated socially by their occupation, by what other people know about what they do in life, other sig nals of status, including speech, are correspondingly more important. This last point is perhaps the most important. 2. The second, related, factor is the WC speech, like many other aspects of WC culture, has, in our society, connotation of masculinity, since it is associated with -the roughness and toughness supposedly characteristic of WC life, which are, to a certain extent, considered to be desirable masculine attributes. They are not, on the other hand, considered to be desirable feminine charac teristics. On the contrary, refinement and sophistication are much preferred. 3 Avis investigated this variable (VtV) in 1955 among upper middle class Ontarians. He asked the informants to analyse their own pronun ciation of minimal pairs containing intervocalic It/;all the words in the list were similar or identical to those in the present study. Of 102 informants answering the question, 52 claimed that they pronounced medial/t/only as [t]. W.A. Avis, "Speech Differences Along the Ontario-United States Border: III Pronunciation," JCLA, Vol.2, No.2, (1956), pp. 54, 55. Gregg analysed the casual speech of Vancouverites and states: "The distinction between post-tonic, intervocalic [t] and [d] has been lost in natural Van. speech. The [t] in this position has been voiced, so that matter,and madder are both pronounced [ 'maedsr] hit it and hid it, both [hid it]V(R.J. Gregg, "Notes on the Pronunciation of Canadian English as Spoken in Vancouver, B.C.," JCLA, Vol.3, No.l, [October 1957], p.25). CEU reveals that 51 to 72 percent of the parents claimed to pronounce medial/t/as [t]. M.H. Scargill, Modern Canadian English Usage: Linguistic  Change and Reconstruction, (Toronto: McClelland and Stewart, 1974), pp.66, 67, henceforth cited as CEU. 207 Our data reveals that the three sources above were probably not contradictory but merely measuring different styles. 4CEU, pp.66-67. 5Labov, pp.394-399, and Trudgill, pp.91-95. You Don't Say, vol.2, issue 4, editors Lamont Tilden and George Rich, (Toronto: CBC, 1978), p.2. ^Personal communication. Mention of informal surveys gives me the opportunity to report that this survey has caused a great deal of interest among a large number of people with whom I had contact during the two year period of this project. These people formed an unorganized network of advisors and evaluators of the survey. They offered 1) suggestions regarding which linguistic items should be investigated, 2) suggestions as to the difference between Canadian and Northern American English, 3) opinions on the results of the survey, and 4) their own subjective attitudes toward Canadian English usage. These informal surveys had individual contributors from the following groups of people: 1. My assistants on the survey, Margaret Murdoch and Steffi Ortiz. 2. Informants, their family, friends, and neighbours who would discuss ,Canadian English usage at great length after the interview. 3. My colleagues at the Federal Language Bureau: Vera McLay, Cornelius von Baeyer, Michael Sutton, Edith Pahlke, and Christine Deeble. 4. Members of the Arts Computer Consultancy Services Unit of the Univer sity of British Columbia: Lewis James, Virginia Green, and Olga Elias. 5. Many British-Canadians, American-Canadians, and Canadians who had experienced the different usage patterns in various English speaking lands. 208 The statistics for Manhattan are taken from Labov, p.398, Figure 3. Notice that Labov's graphs need to be turned 180° to be compared with Trudgill's and ours. In Labov's graph, A = Casual speech, B = Careful Speech, C = Reading style, and SEC = socio-economic class with 0 = Lower Class, 1 = Mixed Lower Class and Working Class, 2-5 = Working Class, 6-8 = Lower Middle Class, and 9 = Upper Middle Class. Figure 5.3.e. -ing = in + in by Class •=VuRKING A=L0WER MIDDLE X=MIDDLE <!>=UPPER MIDDLE *=L0VER UPPER o. r—CO LU "91 a KP P STYLE 209 Figure 5.3.f. Labov's -ing = LQ by Class Contextual style Avis observes that in Ontario, the palatal glide enjoys prestige, but that there is a remarkable degree of variation for individual words. He presents the following data: Tuesday news dew duke tune due student lute suit [juj/liu] 97 93 64 58 56 52 73 19 18 [u] 56 51 41 47 49 53 75 86 87 W.S. Avis, op.cit., pp.48, 49. CEU presents a short historical sketch of the once fashionable [duk tuzdi] and data according to sex, age, and province. CEU, pp.52, 53. "'""'"Stress patterns, too, should be investigated more thoroughly. 12 Avis calls this Canadian diphthong "fast" with a relatively high beginning. Avis, 1956, op.cit., p.42. 13 Before we leave this variable, we would like to mention that we recorded a few cases of the word how pronounced as [hAu]. 210 14 This phrase has two steps of optional rules which can ultimately affect the pronunciation of the diphthong: Rule (1) h -»- h in function words in sentence internal position. -> <f> in function words in sentence internal position. Rule (2) ai -> ei/_tV when medial t rule, 't-*dA/-V is not applied. ai/-dV when medial t rule, ' t-Kl/V-V is applied. "'""'Our choice of the phrase, Did he find them, has obviously distorted our graph in the style MP. We would have expected a 70 to 90 reading for nd in Minimal Pairs. 16 Our Somers' D analysis of the data is presented in Section 3 of this chapter. "^Gregg, op.cit., p.22, when describing the vowel [e] states: "Some speakers use this vowel in words like Barry, parry, etc., which are thus homophones of berry, Perry ['beri] ['peri]. The same speakers also treat as homophones Harry and hairy ['heri], marry, Mary, and merry ['meri]. CEU presents a table representing the pronunciation of guarantee according to sex, age and provinces, pp.95-99. 18 For an explanation of low stress levels on function words see my booklet Rhythm and Unstress (Ottawa: Government of Canada, 1978), pp.1-7. 19 For further discussion of /ee/, see Ruth McConnell, Our Own Voice, (Toronto: Gage Educational Publishing Limited, 1979), pp.22-23. 20 Avis, op.cit., p.53, also investigated this variable (hw) in Ontario among the upper middle class by means of a "written questionnaire and a minimal pair task. Of 159 informants, 68 claimed to pronounce [hw], 50 claimed just [w], and 41 acknowledged inconsistent usage. Gregg, op.cit., p.26, again analysing free speech states: "17. The younger folk in Van. seem largely to have lost the [hw] sound, the unvoiced counterpart of [w]. With them witch and which have fallen together as [wits], wile and while, as [wael], weal and wheel, as [wil]." CEU presents an historical sketch of [hw] and [w] and proceeds to confuse the question and the answer choices so that the data must be rejected, pp.94, 95, and the questionnaire insert. Young people in Ottawa appear to preserve the [hw] somewhat more than do their counterparts in Vancouver, the Kootenays, and the Northern United States. 21 There is a number of Public Servants of Irish ancestry from Prince Edward Island and Newfoundland in Ottawa who frequently substitute [t] and [d] for /6/ and /5/ respectively. (Personal observation.) 211 22 This graph of Manhattan's usage of (th) is taken from Labov, 1966, op.cit., p.260. Notice that Labov's graphs must be turned 180 degrees to be compared with ours. 23 Labov when describing the social distribution of his survey area admits: "The absence of a steady segment of the upper middle class is a reflection of the particular social history of the Lower East Side. It is a port of entry for immigrants and a place of nurture for those on the way up, but normally not a permanent home for children of upper middle class parents" (Juanita Williamson and Virginia Burke, eds., "The Effect of Social Mobility on Linguistic Behavior," A Various Language: Perspec  tives on American Dialects, [New York: Holt, Rinehart and Winston, 1971], pp.640-659). 24 Avis, op.cit., p.53, records that of 152 informants, 108 preferred /temeto/, 32 /temaeto/, and 12 /tamdto/. CEU describes British and North American regional pronunciations and presents data broken down according to sex, age, and province, pp. 65-66. Gregg reports that the teenagers in the Kootenays use only /temeto/. R.J. Gregg,."The Linguistic Survey of British Columbia: The Kootenay Region," Canadian Languages in their Social Context, ed. Regna Darnell (Edmonton: Edmonton Linguistic Research, 1973), pp.109-113. For further information on B.C. usage, see Roberta Stevenson's, The Pronunciation of English in  British Columbia, unpublished M.A. thesis, (Vancouver: University of British Columbia, 1976), pp.1-153. 25 » Labov, 1966, op.cit., p.96, and footnote 6, p.132. 26 A sixth task/style, Series, was attempted and abandoned because of very irregular performance and the inability of informants to recite or list. 27 Compare data with SCE or CEU. Also see H.J. Warkentyne, "Contem porary Canadian English: A Report of the Survey of Canadian English," American Speech, Vol.46, (1971), pp.193-199. This latter article contains a two page section entitled "Effects of Education on Usage" which sum marizes the SCE data on a sociological basis. 28 For a deeper understanding of the Somers' D statistical description, see G. David Garson, Handbook of Political Science Methods, (Boston: Holbrook Press, Inc., 1971), pp.161-162. 212 CHAPTER 6 THE CO-VARIATION OF GRAMMATICAL, PRONUNCIATION AND VOCABULARY VARIABLES WITH SOCIOLOGICAL PARAMETERS Measurement of Co-variation In the previous chapter, we investigated phonological items and the extent and degree to which they varied according to sociological and stylistic parameters. We demonstrated that for many items there was systematic variation according to socio-economic class, sex, and age. Furthermore, our data revealed even stronger evidence of stylistic varia tion according to the tasks the informants were asked to perform. In this chapter, we will investigate 71 items which are important in Canadian English because: 1) the items are believed to be in a state of change, 2) the items are expected to show linguistic and sociological co-variation, or 3) the items are peculiarly Canadian or characteristically Canadian, especially when compared to Northern American. In the design of the questionnaire, no systematic attempt was made at achieving a stylistic analysis for each item. Although there no doubt would be stylistic variation for each item, we believe that this would be much less than for the phonological items of the previous chapter. As stated in Chapter 1. of this dissertation, one of the major motivating forces for this study was the Survey of Canadian English"*"and 2 its derivative book Canadian English Usage; the SCE reveals the linguis tic and sociological co-variation of 103 variables according to age, sex and province. The SCE, however, excluded socio-economic class. 213 It is our belief that one will find more linguistic variation within a city,but across class boundaries, than one will find in travelling 3,142 kilometers from Ottawa to Vancouver, but staying within the same class. Only when the Vancouver Survey and other Canadian urban surveys have been completed will we be able to combine our. results and see whether our hypothesis is correct. In order to partially substantiate our hypothesis, we will present data for linguistic items which we believe will show variation with respect to socio-economic parameters as well as the other sociological parameters of age, sex, ethnic background, rural/urban background, and new Canadian/several generation Canadian background. The 71 variables will be presented in alphabetical order in the following three sections: 1) Grammar and Syntax, 2) Pronunciation, and 3) Vocabulary, thereby 4 following closely the format of SCE and Avis' three studies. 1. Grammar and Syntax The first section of this chapter deals with the correlation between the syntactic variation of an item on the one hand and the sociological variation on the other. All the variables in this section were elicited orally in a style very similar to the style of response to written questionnaires. Our data, therefore, are directly comparable to SCE, Avis 1954, 1955, and 1956, and Gregg, 1973.5 Avis (1954, p.14) when introducing his grammar and syntax data stated: "In matters of grammar especially, cleavages are more commonly social than regional,..." Our intuition would seem to agree with this statement. Let us now look at the sociological differentiation of our items with a view to evaluating 214 whether our socio-economic differentiation is more enlightening than the provincial difference as presented in SCE. 1. Between John and Me #263 Question: John, Mary, and I are sitting in a row. Mary is sitting between John and A. me B. I C. you Table 6.1.1 BETWEEN JOHN AND ME Age > 40 Age < 40 Female' Male All A B C A B C A B C A B C A L WK LM 27 73 0 14 86 0 22 78 0 14 86 0 19 MID 73 27 0 35 60 5 57 43 0 41 53 6 48 UM LU 79 21 0 73 27 0 85 15 0 67 33 0 76 OTTAWA 61 39 0 35 63 2 51 49 0 40 58 2 All data for this variable reveal a systematic socio-economic progres sion. The lower classes tend to use the hyper-correction between John and I_, while the upper middle and lower upper classes mainly employ between  John and 'me. Middle class informants over 40 years of age closely approx imate the upper classes while younger middle class informants are midway between the upper and lower classes. For form A, the informants over the age of 40 had markedly higher scores, average 61, than those under 40, average 35, and females consistently had higher scores than males. The generation gap which is present in our data is also evident in the CEU, p .26. 215 2. Just Between You and Me #264 Question: Just between you and , I think that they're not telling the truth. A. me B. I Table 6.1.2 JUST BETWEEN YOU AND ME Age > 40 Age < 40 Female Male All A B A B A B A B A L WK M 64 36 95 5 89 11 79 21 84 MID 73 27 58 42 69 31 59 41 63 UM LU 100 0 100 0 100 0 100 0 100 OTTAWA 81 19 82 18 86 14 77 23 This collocation was included to demonstrate that the usage of a linguistic item can vary greatly depending on its;environment. The evidence suggests that the lower classes are quite confident in their usage, that the middle class is uncertain and frequently utters, hyper-urbanisms, and that the, usage of the upper classes is in agreement with (prescriptive standards. 3. Eh Wh-interrogative #312 Question: Do you ever say something like, What are they trying to  do, eh? A. yes B. no C. abhorrence Table 6.1.3a EH WH-INTERROGATIVE Age > 40 Age 40 Female Male All A B C A B C A B C A B C B,C L WK LM 10 80 10 25 75 0 22 72 6 17 83 0 80 MID 9 64 27 16 68 16 14 79 7 13 56 31 87 UM LU 14 71 14 9 82 10 8 85 8 13 67 17 88 OTTAWA 11 71 17 18 74 8 16 78 7 15 67 17 216 Most informants felt that the eh was inappropriate after a question and even more so after a long question. An ordered socio-economic pattern is evident but quite weak. Eh is one of the most characteristic markers of Canadian English. In this survey, we investigated 8 different types of eh based on the structure of the preceding sentence and the full question tag which eh replaces.^ Eh types six and seven are displayed here because their usage varies drastically from that of the others. Table 6.1.3b. presents data Cor. all 8 varieties of eh. Table 6.1.3b. Type of Eh Uses (%) Does not (%) Abhorrence (%) 1. Reversed polarity, agreement 72 28 2. Reversed polarity, confirma-: . tion 58 41 3. Constant polarity 64 36 -4. Imperative 2 485. Exclamation 73 27 6. Wh-interrogative 15 73 12 7. Narrative 6 47 47 8. Pardon 43 42 16 A sample sentence for each type of eh was recited by the interviewer, see Questionnaire numbers 306-314, and the informant was asked to answer whether he used that type or not. In addition we noted a ninth variety of eh which did not fit into our structured categories; examples include: No,eh; Thanks, eh; and Good luck, eh. See CEU, pp.75-76 for further discussion and data on this variable. CEU's question 24 is identical with our eh type 3; the percentages are almost identical. 217 4. Eh Narrative #313 Question: Do you ever say something like. This guy is up on the 27th  floor, eh, then gets out on the ledge, eh, then the police come, eh... A. yes B. no C. abhorrence Table 6.1.4 EH. NARRATIVE Age > 40 Age < 40 Female Male All A B C A B C A B C A B C B,C L WK- LM 0 40 60 20 45 35 11 39 50 17 50 33 86 MIDDLE 0 73 27 5 63 32 7 71 21 0 63 38 97 UM LU 0 14 86 0 45 55 0 15 85 0 42 58 100 OTTAWA 0 40 60 10 52 38 7 42 51 5 52 42 This variety of eh, usually called the 'narrative eh', has been des-g ignated a stigmatized form by Avis, 1972, types 6 and 8 obviously are too Table 6.1.4 reveals ordered ranking; the lower class informants have this usage more than the other classes, middle class informants use this form less frequently and the upper classes claim never to use it. 5. Fewer #296 Question: What is the opposite of this sentence? There are more people here tonight than last night. A. fewer B. less Table 6.1.5 FEWER Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 56 44 30 70 47 53 25 75 38 MIDDLE 55 36 37 63 43 57 50 50 43 UM LU 71 29 56 44 77 23 50 50 65 OTTAWA 62 35 38 63 52 45 42 58 218 Table 6.1.5 reveals that the upper middle and lower upper classes consistently had the highest scores for fewer, the prescribed form. The table also demonstrates a large sociological differentiation with refer ence to the two age groups, with those over 40 scoring high for fewer. Further, females averaged higher scores for fewer than did males. The ordered socio-economic correlation outlined above recurs consistently for the items of this section. 6. Have You Got #277 Question: If you needed a match, what would you ask your friend: a match? A. Have you got (Canadian) C. Have you (British) B. Do you have (American) D. Give me, Could I have, etc. Table 6.1.6 HAVE YOU GOT Age > 40 Age < 40 Female Male All A B C D A B C D A B C D A B C D A L WK LM 0 14 29 57 29 62 0 10 13 50 13 25 33 50 0 17 18 MIDDLE 45 9 36 9 25 45 5 25 29 36 21 14 35 29 12 24 33 UM LU 43 29 7 21 18 55 9 18 23 38 8 31 42 42 8 8 29 OLD 35 20 25 20 24 45 3 28 20 36 16 28 38 33 8 21 NEW 33 17 17 33 26 65 4 4 22 50 11 17 35 47 6 12 OTTAWA 34 19 22 25 25 54 4 17 21 42 14 23 37 39 7 17 Avis, 1954, p.16 states: According to my survey, do you have has surprisingly little currency in Ontario; of 85 persons questioned, only nine responded with the American form and three would use either. It is significant that several of the 12 persons who use the form have lived in the States at one time or another. 219 Table 6.1.6 reveals a strong generation gap, with the older infor mants of the middle and upper classes preferring have you got and the majority of the younger informants employing do you have. The British form have you has moderate usage among the older informants of the lower and middle classes but less usage among the younger informants. Young informants of new Canadian backgrounds had the highest frequency scores for the American form; this is a pattern which we first observed in the 97 Kootenay survey and which we will evaluate throughout this chapter. 7. If It Were #288 Question: They would go for a walk if it warmer. A. were B. was Table 6.1.7 IF IT WERE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 50 50 14 86 29 71 25 75 27 MIDDLE 33 67 14 86 25 75 17 83 20 UM LU 73 27 25 75 70 30 33 67 53 OTTAWA 61 39 18 82 48 52 26 74 Table 6.1.7 reveals a strong differentiation with reference to age and sex; the informants over 40 and females had much higher scores for A, 61 and 48 percent respectively, than did the younger informants and males, 18 and 26 percent respectively. In addition, the upper classes, i.e. the upper middle and the lower upper classes, differentiate themselves sharply from the other classes. The data for the middle class are not in linear progression. CEU, pp.40-41 reveals similar but weaker socio logical patterns for age and sex, little difference according to province, and of course no reference to socio-economic class. 220 8. If You Had #287 Question: We would've helped you if you asked us. A. would've C. had've B. had D. 0, i.e. nothing in the blank Table 6.1.8 IF YOU HAD Age > 40 Age < 40 Female Male All A : B . C D A B C D A B C D A B C D B L WK LM 22 56 11 11 24 52 24 0 29 59 6 6 15 46 38 0 53 MIDDLE 0 91 9 0 10 75 10 5 7 86 7 0 6 76 12 6 81 UM LU 0 100 0 0 18 82 0 0 8 92 0 0 8 92 0 0 92 RURAL 0 95 5 0 25 58 17 0 11 89 0 0 7 71 21 0 URBAN 10 76 10 5 16 70 12 2 18 73 6 3 10 71 16 3 OTTAWA 6 85 6 3 17 67 13 2 16 77 5 2 10 71 17 2 ALL TOT 5 85 7 2 18 67 13 2 16 78 4 2 9 71 18 2 These data demonstrate strong and regular socio-economic differentiation with the upper classes consistently having the highest scores for form B. We also observe minor differentiation with reference to age and sex. The informants over 40 years of age and the female informants had higher average scores for form B than their counterparts did. Of special interest is form C, had've; the lower, working, lower middle, and middle classes all had incidences of this value , but the upper middle and lower upper classes did not have a single incidence of this form. Had've was employed by 38 percent of the male informants of the lower classes. 221 9,10. Lie, Lay, Has Lain //s 275 and 276 Question: He lies in the sun every day. Yesterday, he there for • 3 hours • (A. laid B. lay C. lied D. lie) So far tc he has there for 5 hours. (A. laid B . lain C . layen D. lied) Table 6. 1.9 LIE PAST Age > 40 Age < 40 Female Male All .'A B C D A • B C D . A B C D A B C D B L WK LM 40 60 0 0 43 43 10 5 35 53 6 6 50 43 7 0 48 MIDDLE 9 91 0 0 20 65 15 0 14 79 7 0 18 71 12 0 74 UM LU .7 93 0 0 '..9 82 19 0 18 92 0 0 8 83 8 0 88 OTTAWA 17 83 0 0 27 60 12 2 20 73 5 2 26 65 9 0 Table 6. 1.10 LIE PRESENT PERFECT Age > 40 Age < 40 Female Male All A B c C A B C D ABC D A B C D B L WK LM 38 25 25 0 48 29 0 19 50 19 13 19 38 38 0 8 28 MIDDLE 9 73 18 0 45 40 15 0 21 64 14 0 41 41 18 0 52 UM LU 0 93 7 0 18 73 0 0 8 85 0 0 8 83 8 0 84 OTTAWA 12 70 15 0 40 42 6 8 28 53 9 7 31 52 10 2 Both tables reveal strong and regular socio-economic differentiation with reference to the Past and Present Perfect tenses of the verb lie. Strong differentiation according to age is also evident. CEU, pp.34-35 investigates the Present Participle of these two verbs, lie and lay, and finds the confusion percentage to be 39 percent, which is similar to our data. 222 11. ? Not ? #744 A count was taken of the occurrences of questions generated by the informants in free speech which contained the non-contracted form not. Table 6.1.11 ? NOT ? Age > 40 Age < 40 Female Male Total A A AAA L WK LM 1 1 2 0 2 MIDDLE 6 0 0 6 6 UM LU 4 7 7 4 11 OTTAWA 11 8 9 10 In the section of the questionnaire which deals with Subjective Attitude and Language Awareness, the question was asked: "What is the difference between Canadian and American speech?" A small number of informants responded that Canadians ask some questions with the full form not while the Americans do not. Nineteen such questions were recorded during the survey; Table 6.1.11 presents those occurrences with reference to our sociological parameters. Although only a few occur rences are available to make a generalization, we can see that an ordered correlation by class is evident, i;e. our upper classes have the highest frequency followed by the middle class, and finally the lower classes have the lowest frequency of occurrence. Further investigation should be conducted concerning this variable. 223 12. Past Perfect #315 Question: Do you ever use verb forms like these; had given, had gone? Give an example in a sentence showing that the simple Past couldn't be used. (An example might be: it had been found when the police arrived.) A. uses Past Perfect demonstrating the difference B. uses Past Perfect but doesn't demonstrate the difference C. gives only Conditional III construction. D. no Table 6.1.12 PAST PERFECT Age >40 Age < 40 Female Male All A' B C-J D ABCD ABCD ABCD A L WK LM 0 88 13 0 15 60 10 15 0 71 12 18 27 64 9 0 11 MIDDLE 45.27 18 9 26 63 5 5 29 57 7 7 38 44 13 6 33 UM LU 46 31 8 15 64 27 0 9 67 33 0 0 42 25 8 25 54 OTTAWA 34 44 13 9 30 54 6 10 28 56 7 9 36 44 10 10 The ability to generate a sentence which demonstrates the use of the Past Perfect was strongly related to socio-economic class in each age and sex group. More than two-thirds of all informants were unable to generate a sentence which distinguished the Past Perfect from the Past, and it would appear that the majority of those informants do not use the Past Perfect in their speech or writing. 224 13. Snuck #280 Question: They sneak into the movie theatre. Yesterday, they into the movie theatre. A. sneaked B. snuck Table 6.1.13 SNUCK Age > 40 Age < 40 Female Male All AB AB AB AB A L WK LM 30 70 0 100 12 88 7 93 10 MIDDLE 91 9 10 90 43 57 35 65 39 UM LU 64 36 18 82 46 54 42 58 44 OTTAWA 63 37 8 92 32 68 28 72 Table 6.1.13 shows the strongest generation gap of the survey. Some socio-economic differentiation is also evident. A few elderly ladies indicated that they would not generate either sneaked or snuck, and over 20 informants laughed or chuckled while answering this question. The CEU, pp.43, 44 reveals a similar split between the age groups. 14. Subject-Verb Non-agreement A count was taken of the occurrences of non-agreement in number between subject and verb. All occurrences of subject-verb non-agreement were of the following construction: There is + plural. Table 6.1.14 SUBJECT-VERB NON-AGREEMENT Age > 40 Age < 40 Female Male Total A A A A A L WK LM 16 22 19 19 38 MIDDLE 4 17 7 14 21 UM LU 3 1 2 2 4 OTTAWA 23 40 28 35 ALL TOT 0 0 0 0 225 The above data show that there is a direct and ordered relationship between subject-verb agreement and socio-economic class; the lower the class, the higher the frequency of there is + plural. Informants over 40 and females had the lowest scores, 23 and 28 respectively. We notice that for this variable,the middle class informants over 40 years of age conform to the upper classes usage pattern while the middle class informants less than 40 side with this lower class usage. Data of this nature alarm many educators and parents. 15. Take #318 Question: You and Sally are on the third floor, Mrs. Fraser is on the sixth floor. Ask Sally to carry a letter up to Mrs. Fraser. A. take B. bring Table 6.1.15 TAKE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 78 22 85 15 88 12 75 25 83 MIDDLE 90 10 95 5 86 14 100 0 93 UM LU 100 0 91 9 92 8 100 0 96 RURAL 89 11 83 17 88 12 85 15 URBAN 90 10 88 12 88 12 90 10 IRISH 86 14 86 14 89 11 80 20 SCOTS 100 0 100 0 100 0 100 0 ENGLISH 100 0 93 7 94 6 100 0 FRENCH 0 0 100 0 100 0 100 0 OTHER 60 40 77 23 70 30 75 25 OTTAWA 91 9 90 10 88 12 92 8 ALL TOT 90 10 87 13 88 12 88 12 It has long been my observation that people from the Ottawa Valley and from the'city of Ottawa itself will frequently employ bring when referring to the act of someone carrying something away from themselves or their spatial reference point. Our data reveal that this pattern 226 usage does in fact exist, and that it is partially related to class but much more definitely„to Irish background. We are looking forward to results from the Ottawa Valley Survey concerning this variable. 2. Pronunciation The speaker of Canadian English has a phonological system as out lined and analysed in Chapter 4 and 5 respectively. This phonological system allows the speaker to generate an infinite set of possible words by means of placing the component parts of the system into various com binations. In addition to analysing this sound system in Chapter 5, we undertook in this study to elicit and analyse the pronunciation of individual words. These individual words consist of the phonological segments which we have outlined and analysed above. We, therefore, are analysing the choice of phonological segments, e.g. either may be pro nounced /aiSer/ or /foer/. We will see to what extent sociological parameters correlate to variation in word pronunciation. The pronunciation choices are frequently associated in the linguistic literature with a British form, an American form, and in some cases a uniquely Canadian form. This is done in spite of the fact that all linguists are aware of a great deal of variety in usage within these countries. 227 1. Africa #53 Task; Pictures : A. [aef reko ] B. [ aef er ke] Table 6.2. 1 AFRICA Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 64 36 38 62! 61 39: ' 29 71 49 MIDDLE 91 0'9 75' 25 ' 7:9' 21 82 18 82 UM LU 64 36 • 45 55.-• •54 46, . .'58 42 55 OTTAWA 72 28 52 46 64 3.61 58 42 Table 6.2.1 reveals there is no linear correlation between the word pronunciation and socio-economic class. We can see, however, that the middle class consistently has 30 percent higher scores for form A than do the other two class groups. Further, the older informants and the female informants have higher scores, 7 2 and 64 percent respectively, than do the younger informants and the males, 54 and 58 percent respectively. 2. Again #146 Task; Word list: A. [egei n] B. [ eg£n ] Table 6.2. 2 AGAIN Age > 40 Age < : 40 Female Male All A B A B A B A B A L WK LM 27 73 19 81 17 83 29 71 22 MIDDLE 36 64 30 70 43 57 24 76 32 UM LU 14 86 36 64 23 77 25 75 24 OLD 18 82 17 83 19 81 16 84 NEW 36 64 39 61 37 63 39 61 OTTAWA 25 75 27 73 27 73 26 7.4 Our data demonstrate that [egein] is still a significant marker of Canadian English. Our data on the other hand show no regular co-varia tion of pronunciation and sociological parameters except that Ottawans 228 with new Canadian background have consistently higher scores for [agein]; Avis, 1956, p.45 tells us that [egein] reflects British practice and that it occurs most frequently in stressed position. Many informants read the word list in breath groups; if the word again was at the beginning of a word group or breath group it would frequently be stressed and more likely be pronounced [egein]. If the word was in the middle of the breath group,it would normally have less stress and be pronounced [agen]. We also noticed more stress on the first few words of a column or page and less stress on words within the list. CEU, p.72 presents a self analysis on the part of the informants of the variable (again). The data were selected by means of a written questionnaire; the item was elicited in complete isolation in the most formal style attainable. The results, showing a preference for [ogein] in all sociological groups, are in marked contrast to our data and demon strate the repercussion of different survey methodologies. 3. Anti-(pollution) #195 Task; Word list: A. [aent i ] B. [sentat. ] C. [ffinte ] Table 6.2. 3 ANTI-Age > 40 Age < 40 Female Male All A B C A B C A B C A B C A L WK LM 91 0 9 100- 0 0 94 0 6 100 0 0 97 MID 91 0 9 95 5 0 100 0 0 88 6 6 94 UM LU 93 7 0 91 9 0 100 0 0 83 17 0 92 OLD 86 5 9 '93 7 0 96 0 4 84 12 4 NEW 100 0 0 100 0 0 100 0 0 100 0 0 OTTAWA 92 3 6 96 4 0 98 0 2 91 7 2 The form [ssnti] is a very strong Canadianism and it is gaining fur ther strength among the younger informants. We can see that informants 229 with new Canadian backgrounds all had scores of 100 percent for this value. This suggests that new Canadians accept the prevailing form around them. Answer C. [asnta] was elicited from only a few informants all of whom were over forty. Avis, 1956, p.47 and CEU, pp.60-61 present similar percentages and reveal a strengthening trend for [jjeirti ]. Avis, 1956, gives us a quarter of a century's historical perspective of this trend. 4. Apricots #65 Task; Pictures: A. [aepreknts] B".' [elprakbfs] Table 6.2.4 APRICOTS Age > 40 Age < 40 Female Male All A B A B A B A B B L WK LM 73 27 85 15 78 22 85 . 15 19 MIDDLE 91 9 89 11 100 0 82 18 10 UM LU 57 43 73 27 69 31 58 42 36 OLD 73 27 93 7 84 16 83 17 NEW 71 29 74 26 79 21 67 33 OTTAWA 72 28 84 16 82 18 76 24 Table 6.2.4 reveals a tendency toward [aeprekct], the American form, among the younger informants. The trend away from British forms toward Northern American forms is substantiated in this study, noteworthy exceptions notwithstanding. CEU, pp.54-55 presents similar percentages for Ontario and data which demonstrate the strengthening of [sprakxits] among younger informants, while Gregg, 1973, pp.111-113 and CEU reveal that British Columbia is 30 to 40 percentage points removed from the rest of Canada in the direction of LeIprakbtsJ, the normal British form. 230 5. Asphalt #123 Task; Word list: A. [aesfalt] B. [aejfalt] C. [eszfalt] Table 6.2.5 L WK LM MIDDLE UM LU OTTAWA Age > 40 ABC 30 40 30 64 9 27 57 29 14 51 26 23 ASPHALT Age < 40 ABC Female ABC Male ABC 5 90 15 80 18 82 12 84 5 5 0 4 6 81 13 21 64 14 36 50 14 29 59 12 31 62 8 50 42 8 23 65 12 33 56 12 All A 13 32 40 The most striking sociological differentiation for this variable is found by comparing the age groups. We can see that the informants over forty had high scores for [aesfalt] and fairly high scores for [sezfalt], while the younger informants have very high scores for [aejfalt ]• The lower classes consistently had the highest scores for [aejfalt]. No comparative data are available for this variable. 6. Aunt #79 Task; Pictures: A. [aent] B. [tint] C. [ant] Table 6.2.6 AUNT Age > 40 Age < 40 Female Male All A B C A B C A B C A :.B C B,C L WK LM 91 0 9 95 5 0 89 6 6 100 0 0 6 MIDDLE 91 0 9 95 0 5 93 0 7 94 0 6 7 UM LU. 86 0 14 91 9 0 77 8 15 100 0 0 12 OTTAWA 89 0 11 94 4 2 87 4 9 98 0 2 The data for this variable deomonstrate that form A, [aent ] is pre dominant for all sociological groups. Females of the upper middle and lower upper classes scored 23 percent for [nnt] and [ant] combined. We notice here a similar usage pattern to the variable (tomato). Avis, 1956, 231 p.52 states: "Aunt and drama, probably because of the social environment in which they are used, appear to have a higher incidence of the 'broad a' than most other words in the group;...." See Gregg, 1973, p.112 and CEU, pp.84-85 for further discussion and almost identical data on this variable. 7. Balcony #234 Task; Word list: A. [be'lkeni] B. [baelkani] C. [balkeni] D. [bnlkeni] Table 6.2.7 BALCONY Age > 40 Age < 40 Female Male All A BCD A B C D ABC D ABC D B L WK LM 36 45 18 0 48 43 0 10 44 50 6 .0 43 36 7 14 44 MIDDLE 18 36 27 18 20 45 20 15 14 50 21 14 24 35 24 18 42 UM LU 21 71 7 0 18 82 0 0 23 77 0 0 17 75 8 0 76 OTTAWA 25 53 17 6 31 52 8 10 29 58 9 4 28 47 14 12 Table 6.2.7 reveals that the upper middle and lower upper classes have predominantly high scores for /balkeni/, while the lower, working, arid lower middle classes have the highest scores for /be* I ken i/. For further reference see variable (Vr->£r) of Chapter 5. 232 8. Been #181 Task; Word list: A. [bin] B. [bin] C. [ ben ] Table 6.2. 8 BEEN Age > 40 Age < 40 Female Male All A B C A B C A B C A B C A L WK LM 30 60 10 81 19 0 65 29 6 64 36 0 64 MIDDLE 55 36 9 70 30 0 57 43 0 71 24 6 64 UM LU 43 57 0 91 9 0 54 46 0 75 25 0 64 OLD 55 36 9 83 17 0 65 31 4 76 20 4 NEW 23 77 0 74 26 0 50 50 0 61 39 0 RURAL 70 20 10 85 8 8 78 17 6 73 13 13 URBAN 27 68 5 74 26 0 52 45 3 66 34 0 OTTAWA 43 51 6 79 21 0 59 39 2 70 28 2 ALL TOT 48 45 7 77 21 2 61 35 4 68 28 4 12 Our data would seem to indicate that the Canadianism [bin] is growing in frequency of usage. Table 6.2.8 reveals that there is a large genera tion gap between the age groups>with the younger informants having con sistently higher percentages for form A [bin], than did the older infor mants. The table further reveals that both [bin] and [ben] are uttered more frequently from informants with rural backgrounds than from urban informants. Avis, 1956, p.45 points out that [bin] is most likely to be uttered in stressed position. Form B [ b in ] is the normal one in unstressed position in both Northern American and Canadian English. 233 9. Blouse #88 Task; Picture: A. [b1aoz] B. [blAus] or [b 1 aos ] Table 6.2.9 BLOUSE Age > 40 Age < 40 Female Male All A B A B A _ B A B A L WK LM 91 9 86 14 100 0 71 29 88 MIDDLE 91 9 100 0 100 0 94 6 97 UM LU 100 0 91 9 100 0 92 8 96 OLD 100 0 97 3 100 0 96 4 96 NEW 86 14 87 13 100 0 72 48 OTTAWA 94 6 92 8 100 0 86 14 All female informants pronounced this variable [blaoz]. Lower class males and informants with new Canadian backgrounds had the highest scores for [blAus] or [ bIaos}. Value A. [blaoz], is the only pronunciation given by the Concise Oxford Dictionary. 10. Caramel #152, 153, 154 Task; Word list: A. [keYamel] B. [ kaereme I ] C. [ kdreme I ] D. [ kdrmeI] Table 6.2.10 CARAMEL Age > 40 Age < 40 Female Male All A B C D A B C D A B C D A B C D B L WK LM 44 22 22 11 94 0 0 6 75 13 6 6 78 0 11 11 6 MIDDLE 89 11 0 0 80 5 5 10 86 7 7 0 80 7 0 13 6 UM LU 31 62 8 0 90 10 0 0 67 33 0 0 45 45. 9 0 38 OLD 58 17 5 0 77 8 4 12 70 22 9 0 68 18 ' 0 14 NEW 42 33 17 8 100 0 0 0 84 11 0 5 69 15 15 0 OTTAWA 52 35 10 3 87 4 2 7 76 17 5 2 69 17 6 9 Form D [kdrmeI], a popular form in the States and Ireland, was not once uttered by any informant of the upper middle or lower upper classes in Ottawa. Also evident from this table is the fact that the younger 234 informants collapse [asr] to [er] more than any other group. Form B [ ksereme I ] consistently received its highest scores from the upper middle and lower upper classes. CEU, pp.67, 68 provides divergent data on this item. It focuses on whether the variable has two or three syllables, and informants claim a 48 percent frequency for form D. This question in the SCE would have been difficult for the informants to answer. 11. Catching #530 Task; Reading: A. [ksetf in] B. [ke'tjin] C. [k£tj IQ] Table 6.2. 11 CATCHING Age > 40 Age < 40 Female Male All A B C A B C A B C" A B C A L WK LM 91 9 0 70 25 5 83 11 6 69 31 0 77 MIDDLE 100 0 0 89 0 .6 86 0 14 100 0 0 93 UM LU 86 14 0 91 9 0 92 8 0 83 17 0 88 OTTAWA 92 9 0 82 12 6 87 7 6 85 15 0 Table 6.2.11 demonstrates that among the younger informants this variable varies directly with socio-economic status. Further,the table shows that only young female informants lowered the /ee/ to [as]. The lower classes scored higher than the other two classes with reference to /e/. There appears to be a common phonological background among the lower and upper classes which is not shared by the middle class, see Chapter 5, p. 112 for some historical background to this point. 235 12. Congratulate #191 Task; Word list: A. [ kangraet Je I ei t ] B. [kengrsd^aleit] Table 6.2.12 CONGRATULATE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 36 64 14 86 28 72 14 86 22 MIDDLE 73 27 30 70 36 64 53 47 45 UM LU ' 79 21 45 55 69 31 58 42 64 OTTAWA 64 36 27 73 42 58 42 58 This variable has very strong and ordered socio-economic differen tiation for all four age and sex groups. The higher one moves up the social ladder, the more frequently one will hear form A. In addition, one can see a very strong shift in usage between the two age groups. One could expect that the younger informants might spell the word with a 'd'. The data in CEU, page 91 substantiates our claim that this variable is undergoing a change with reference to age group. 13. Decal #222 Task; Word list: D. [de*kal] A. [de5k9l] B. [de*kael ] C. [dfkal ] Table 6.2.13 DECAL Age > 40 Age < 40 Female Male All A B C D A B C D ABC D A B C D A L WK LM 55 9 9 0 67 0 14 0 50 6 22 0 79 0 0 0 62 MIDDLE 27 27 0 9 75 5 5 5 43 14 7 0 71 12 0 12 58 UM LU 43 21 7 0 100 0 0 0 69 8 8 0 67 17 0 0 68 OLD 41 23 9 5 86 3 3 0 54 8 12 0 80 16 0 4 NEW 43 14 0 0 65 0 13 4 53 11 16 0 61 0 0 6 OTTAWA 42 19 6 3 77 2 8 2 53 9 13 0 72 9 0 5 236 Values A, B, and D are typically Canadian; value C and [dakasl] are American. The major sociological differentiation for this variable is with reference to age group. The younger informants have much higher scores for [de*kel ] than do the older informants. Again we have evidence of a Canadianism gaining in strength. 14. Egg #55 Task; Pictures: A. [eig ] B. [eg] C. [eig] D. [ii! Table 6. 2.14 EGG Age > 40 Age < 40 Female Male All ABCD A B C D A B C D A B C D B L WK LM 91 0 9 0 86 10 0 5 83 6 6 6 93 7 0 0 6 MIDDLE 82 18 0 0 90 10 0 0 79 21 0 0 94 6 0 0 13 UM LU 71 14 14 0 82 18 0 0 77 23 0 0 75 8 17 0 16 OTTAWA 81 11 8 0 87 12 0 2 80 16 2 2 88 7 5 0 No strong pattern of linguistic and sociological co-variation is evident in this data. Form A. [eig] is strongly preferred by all socio logical groups. The upper classes had the highest incidences of forms B and C. The scores for value D, /iig/ may be a few occurrences of per formance errors. 13 See Gregg, 1957, p.23 for further discussion on this variable. 237 15. Either #179 Task; Word list: A. [aiSar] B. [.f5ar] C. [aroar] Table 6.2.15 EITHER Age > 40 Age < 40 Female Male AH A B C A B C A B C A B C A L WK LM 36 55 9 29 71 0 39 61 0 21 71 7 31 MIDDLE 27 73 0 35 65 0 43 57 0 24 76 0 32 UM LU 57 43 0 36 64 0 62 38 0 33 67 0 48 OTTAWA 42 56 3 33 67 0 47 53 0 26 72 2 CEU, pp.78 and Avis, 1956, pp. 51-52 associate [ai 5a r ] and TH 5ar] with British and American English respectively. It is typical of Ontario and Ottawa that both forms are used. The upper classes, however, strive for the prestigious pronunciation /aiSar/ more than the other classes do. Notice that socio-economic class is strongly differentiated among females, especially among older women; this is a recurring phenomenon in our survey. Comparing the two age groups and the two sex groups in UM LU we can see that the older informants and the.female informants consistently -chose, the more prestigious, form. 16. February #92, 93 Task; Pictures: A. [f£bjueri] ' B. [f^bueri] C. [februeri] D. [febari] Table 6.2.16 FEBRUARY Age >. 40 Age < 40 Female Male All , A B C D A B C D A BCD A B C D C L WK LM 50 0 33 17 67 33 0 0 80 0 20 0 38 50 0 13 10 MIDDLE 17 33 50 0 50 17 17 17 0 60 20 20 57 0 43 0 25 UM LU 0 55 36 9 22 22 33 22 8 25 50 17 13 63 13 13 32 OTTAWA 17 35 39 9 48 26 15 11 33 22 33 11 35 39 17 9 238 Amomg all young informants and female informants the incidence of form C .[ftfbrueri] increased as we moved up the social class structure. We can also see that [ftfbjueri], the typically American form, is very popular among the lower classes, but that this popularity diminishes rapidly as we move up in class. Form D [fefberi] is a form which does not have the secondary stress on the penultimate syllable; other examples of this are the British forms secretary [s£*kratri], library [ laibri ], military [rrulatri], etc. Form B [fcfbueri] enjoys a fairly high frequency. 17. Fertile #196 Task; Word list: A. [ f e rt a 11 ] B. [fartal ] C. [f arde1 1] Table 6.2. 17 FERTILE Age > 40 Age < 40 Female Male Al] A B C A B: C. A B C. A B C A L WK LM 82 9 . 9 100 0 0 100 0 0 86 7 7 94 MIDDLE 36 64 0 90 10 0 71 29 0 71 29 0 71 UM LU 50 29 21 91 9 0 62 31 8 75 8 17 68 OLD 59 27 14 93 7 0. 85 12 4 72 20 8 NEW 50 43 7 96 4 0 74 26 0 83 11 6 OTTAWA 56 33 11 94 6 0 80 18 2 77 16 7 Our data demonstrate again that younger informants use the typically Canadian form far more frequently than.do the older informants. Further, informants of the lower classes use the typically Canadian form more frequently than do the other classes. Both of these trends are the con trary to our initial expectations. Avis, 1956, p.46 provides further discussion and data on this var iable plus senile, virile, docile, futile, missile, projectile, and pro file. The 25 year perspective which his study now allows us shows us that -ile [ail] is maintaining strength for most words and has gained in fertile and futile. 239 18. Futile #215 Task; Word list: A. | : fjut ail ] B . [fjudel ] C. [fjut el] Table 6.2. 18 FUTILE Age > 40 Age < 40 Female Male All A . B C A B C A B C A B C A L WE LM 80 10 10 95 •5 0 94 6 0 86 7 7 90 MIDDLE 82 0 18 90 0 10 86 0 14 88 0 12 87 UM LU 86 7 7 91 9 0 92 0 8 83 17 0 88 OLD 86 0 14 90 7 3 92 4 4 84 4 12 NEW 79 14 7 96 0 4 89 0 11 89 11 0 OTTAWA 83 6 11 92 4 4 91 2 7 86 7 7 Table 6.2.18 shows a high incidence of the typically Canadian form /fjutail/ for all sociological groups. As was the case with (fertile), the new Canadians over forty years of age have lower scores than the several generation Canadians in their use of typically Canadian forms, but younger new Canadians surpass several generation Canadians in the younger generation; this pattern frequently recurs. 19. Film #237 Task; Word list: A. [film] B. [filom] Table 6.2.19 FILM Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 45 55 76 24 61 39 71 29 66 MIDDLE 55 45 75 25 79 21 59 41 68 UM LU 86 14 82 18 92 8 75 25 84 OLD 55 45 72 28 69 31 60 40 NEW 79 21 83 17 84 16 78 22 RURAL 57 43 54 46 58 42 53 47 URBAN 59 41 81 19 76 24 72 28 OTTAWA 64 36 77 23 76 24 67 33 ALL TOT 58 42 75 25 69 31 66 34 240 Those concerned about the "purity" of the English language may be pleased to see that the younger informants conform much more frequently to the prescribed standard than do the older informants and that new Canadians have consistently higher scores for [film] than do old Canadians. Females show the strongest socio-economic differentiation. Urban informants have higher scores for [film] than do informants with rural backgrounds. It is of interest to note that all informants on Elm Street in a working class district of Ottawa pronounced their street Ii lam/. No where else was elm recorded with two syllables. 20. Garage #71, 72, 73, 74 Task; Pictures: A. [garcb,] B. [ga~rd;$] C. [garse^] D. [grd^] Table 6.2.20 GARAGE Age > 40 Age < 40 Female Male All A B C D ABCD A BCD A BCD A L WK LM 60 0 20 20 33 47 13 7 36 36 27 0 44 33 0 22 25 MIDDLE 57 0 43 0 20 20 20 40 30 20 30 20 33 8 25 33 23 UM LU 33 25 33 8 50 50 0 0 42 42 17 0 38 25 25 13 32 OTTAWA 46 13 33 8 32 37 13 18 36 33 24 6 38 21 17 24 No clear sociological pattern emerges from our data. It is inter esting, however, that the popular Northern America form [grad^] and the popular British Columbia form [greedy] were not among the four most fre quently elicited forms. 241 21. Generally #233 Task; Word List: A. [ d^nara I i ] B. [d^nrali] C. [denarii] D. [d^enar lli] Table 6.2.21 GENERALLY Age •>• 40 Age < 40 Female Male All A B C D A B C D A BCD A B C D D L WK LM 18 9 36 36 10 29 24 29 11 17 28 39 14 29 29 21 31 MIDDLE 45 18 0 27 15 10 15 50 21 29 0 36 29 0 18 47 42 UM LU 8 15 0 62 9 18 9 55 8 31 0 46 9 0 9 73 58 OTTAWA 23 14 11 43 12 19 17 42 13 24 11 40 19 10 19 45 The upper middle and lower upper classes preferred value D, [d^narlli], over all other forms. The lower classes had higher incidence of form C, [d^enarli], than did any other class; the frequency of this form C decreased sharply as one moved up the social classes. It appears that excessive vowel and syllable reduction is a characteristic of stig matized speech, cf.anti- lieutenant, naturally, potato, recognize, regular, and temperature. 22. Genuine #157 Task; Word List: A. [d^njuan] B. [d^nj uai n ] Table 6.2. 22 GENUINE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 36 64 29 71 17 83 50 50 31 MIDDLE 73 27 50 50 71 29 47 53 58 UM LU 79 21 55 45 69 31 67 33 68 RURAL 43 57 8 92 42 58 13 87 URBAN 73 27 49 51 48 52 66 34 OTTAWA 64 36 42 58 49 51 53 47 ALL TOT 58 42 39 61 46 54 49 51 242 Avis, 1956, page 47, states that form B is heard widely in spite of its being proscribed for more than a century. He has also stated to Professor R.J. Gregg that he suspects it of having higher currency among rural people. Our data proves Avis correct on both counts. In addition to the above comments we can see that this variable displays a great deal of socio-economic differentiation,with the lower classes preferring form B. 23. Hundred #576 Task; Series: A. [hXndred] B. [hXnCd)er"t ] C. [hXnrad] D. [hAnderd] Table 6.2.23 HUNDRED Age > 40 Age < 40 Female Male All A B C D A B C D A BCD A B C D A L WK LM 56 11 11 22 63 11 16 11 72 6 17 6 40 20 10 30 60 MIDDLE 45 9 18 27 67 0 16 16 64 7 7 14 53 0 13 13 58 UM LU 67 8 17 0 63 0 25 13 83 8 8 0 38 0 38 13 65 OTTAWA 56 9 16 16 64 4 13 9 73 7 11 7 45 6 18 18 Generally, the higher one goes up the socio-economic class system, the more one will hear form A, [hXndred]; the lower one goes in the socio-economic classes, the more one will hear form B, [hAn(d)art ], 14 with possible [d] deletion, r metathesis and final devoicing. Females had a much higher percentage for the preferred form, form A, than did the males. 243 24. Khaki #52 Task; Pictures: A. E-kdrki] B. Ekaki] C. Ekdki] D. unknown Table 6.2.24 KHAKI Age > 40 Age .< 40 Female Male All ABCD ABCD ABCD ABCD A L WK LM 64 27 9 0 17 44 0 11 38 19 6 13 31 62 0 0 35 MIDDLE 82 18 0 0 20 40 10 10 50 21 14 0 35 41 0 12 42 UM LU 93 7 0 0 55 18 9 18 85' 0 8 8 67 25 0 8 76 OLD 82 14 5 0 30 37 4 15 56 12 8 12 50 42 0 4 NEW 79 21 0 0 23 36 9 9 56 17 11 0 33 44 0 11 OTTAWA 81 17 3 0 27 37 6 12 56 14 9 7 43 43 0 7 Khaki pronounced [kdrki] is an established Canadianism current in all social levels in Canada. In Table 6.2.24, we can see that this item displays a great deal of socio-economic differentiation with the lower classes having closer ties to American usage. The younger generation is definitely moving away from this Canadianism; there is a 54 percent spread between the younger and older informants for this value. The history of this item may shed some light on this shift. Khaki, which is derived from Urdu khaki meaning 'dusty', was borrowed into English in the nine teenth century by the British Army in India. The Canadians adapted it from the British; they assumed that what they heard, Ekdki ], was a reflec tion of the r-less speech of most Britains, so they inserted an _r. In Britain and Canada the khaki uniform is of two types: for winter it is a heavy wool, which is dark greenish-brown in colour; for the summer the uniform is a light cotton cloth and beige in colour. In the States, how ever, the winter khaki is not known and the beige or tan coloured tropical military cloth is called [kaeki]. For the past two or three decades at least, pants and jackets for teenagers have been made of this chino fabric in several colours. During one interview a teenage informant got up from 244 her chair, went to her closet and showed me her latest blue khakis [kaekiz]. This is a striking example of linguistic change. The meaning of khaki had changed from a dark greenish brown colour to a light shiny cloth and style. Avis, 1956, pp.43-44 states that only 4 of 109 informants said [ksski]; this constitutes the greatest linguistic change we are able to see by comparing these two studies. 25. Library #84 Tasks; Pictures: A. [laibreri] B. [laiberi] C. [laibri] D. [ I a i bar i 3 Table 6.2.25 LIBRARY Age > 40 Age < 40 Female Male All A B C D A B C D A BCD A B C D A L WK LM 18 27 36 18 76 24 0 0 50 28 11 11 64 21 14 0 56 MIDDLE 64 27 9 0 60 30 10 0 64 14 21 0 59 41 0 0 61 UM LU 29 21 43 7 82 9 0 9 46 15 23 15 58 17 25 0 52 OTTAWA 36 25 31 8 71 23 4 2 53 20 18 9 60 28 12 0 Table 6.2.25 demonstrates that the informants under forty conform to the prescribed form A much more than the older informants do. The table also reveals that the British form [laibri] is fairly current among the informants over forty but not among the younger informants. The form [laiberi] is found less frequently in the upper classes. The middle class over forty years of age conforms much more to the present norm than the other two classes do. 245 26. Lieutenant #147 Task; Word list: A. [Ieftenant] B. [lutenant] C. [late'nant Table 6.2.26 LIEUTENANT Ag ;e > 40 Age < 40 Female Male All A B C A B C A B C A B C A L WK LM 45 45 9 14 86 0 28 72 0 21 71 7 25 MIDDLE 91 9 0 25 75 0 57 43 0 41 59 0 48 UM LU 71 29 0 40 60 0 69 31 0 45 55 0 58 OLD 68 32 0 21 79 0 46 54 0 36 64 0 NEW 71 21 7 27 73 0 53 47 0 35 59 6 OTTAWA 69 28 3 24 76 0 49 51 0 36 62 2 Our data demonstrate, that the British form [Ieftenant] is a pres tigious form, for one sees that generally the higher the class, the higher is the percentage of its occurrence. Further, from our data we-see a strong shift towards the American form among the younger informants The trend toward American forms is very strong in military terminol ogy, see (khaki) and (missile). The older informants had contact with the British military in World War II, while the younger informants know the military through American movies, news, and protests. Only lower class males uttered form C. Form C, [late'riant] is British naval usage. CEU, p.73, also displays a strong shift among younger informants to the American form. 246 27. Luxury #129 Task; Word list: A. [lAg^eri] B. [iXkJeri] C. [IXkjr i] Table 6.2.27 LUXURY Age > 40 Age < 40 Female Male All A B C A B C A B C A B C B L WK LM 9 82 9 5 95 0 6 94 0 7 86 7 91 MIDDLE 0 100 0 10 90 0 0 100 0 12 88 0 94 UM LU 0 100 0 9 91 0 0 100 0 8 92 0 96 OTTAWA 3 94 3 8 92 0 2 98 0 9 88 3 No pattern of linguistic and sociological co-variation is evident in this table. Form B enjoys over 82 percent frequency by all groups and sub-groups. Form C, [lAkJYi], with deleted second syllable was elicited only from males of the lower classes. Our data is very similar to that of CEU, pp.77-78. 28. Mirror #41 Task; Picture: A. [mirer] B. [nur] C. [mtral] D. [nurou] Table 6.2.28 MIRROR Age > 40 Age < 40 Female Male All A B C D A B C D A B C D A B C D A L WK LM 64 27 0 9 71 24 5 0 61 28 6 6 79 21 0 0 69 MIDDLE 100 0 0 0 75 25 0 0 86 14 0 0 82 18 0 0 84 UM LU 100 0 0 0 80 20 0 0 92 8 0 0 91 9 0 0 92 OTTAWA 89 8 0 3 75 24 2 0 78 18 2 2 83 17 0 0 Here we have a clear case of socio-economic differentiation. The prescribed form [mtrer] is more frequently elicited as we moved up the class structure. The form[mir] increases in frequency as we moved down the class structure . The forms [mirel] and [rrurou] may result from confusion with the word 'mural'. 247 29. Missile #87 Task; Picture: A. [nusail] B. [misal] Table 6.2.24 MISSILE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 27 73 33 67 17 83 50 50 31 MIDDLE 30 70 26 74 17 83 35 65 28 UM LU 29 71 64 36 38 62 50 50 44 OLD 24 76 36 64 17 83 44 56 NEW 36 64 39 61 32 68 44 56 OTTAWA 29 71 37 63 23 77 44 56 The data for this variable are markedly different from the data for the variables (fertile) and (futile). This fact reinforces our claim that the trend toward American forms is very strong in military termin ology, see (khaki) and (lieutenant). The British form, [nusail], never theless, ^-is growing in prestige among the upper class informants who are less than forty years old. Notice that most women opt for the American form. Males who have been in the Canadian Forces overwhelmingly choose the British form. See Avis, 1956, p.46 and CEU, pp.80-81, for further information and similar data on this variable. 248 30. Morning #440 Task; Reading: A. [mornt-n] B. [mornan] C. [mornin] Table 6.2.30 MORNING Ag e > 40 Age < 40 Female Male All A B C A B C A B C A B : C A L WK LM 45 0 55 45 0 55 56 0 44 31 0 69 45 MIDDLE 64 0 36 42 0 58 50 0 50 50 0 50 50 UM LU 71 0 29 45 0 55 69 0 31 50 0 50 60 OTTAWA 61 0 39 44 0 56 58 0 42 44 0 56 Table 6.2.30 demonstrates that for the older informants, the pro nunciation of the nominalizing morpheme -ing is socio-economically dif ferentiated. For the younger informants there appears to be no correla tion to socio-economic class, as all classes range between 55 to 58 percent for [-in]. As is often the case in our survey, the female informants conform in higher numbers to prescribed and careful speech. Not one occurrence of [an] was recorded. See variable (-ing) Chapter 5, for further reference. 31. Multi-(national) #223 Task; Word list: A. [mXl tl-] B . [m Altai -] C. [ mAl Its-] Table 6. 2.31 MULTI-Age > 40 Age < <40 Female Male All A B C A B C A B c A B C A L WK LM 91 0 9 100 0 0 94 0 6 100 0 0 97 MIDDLE 91 0 9 95 5 0 100 0 0 88 6 6 94 UM LU 79 7 14 91 9 0 85 0 15 83 17 0 84 OLD 77 5 18 93 7 0 88 0 12 84 12 4 NEW 100 0 0 100 0 0 100 0 0 100 0 0 OTTAWA 86 3 11 96 4 0 93 0 7 91 7 2 249 All sociological groups have high scores for value A. The upper classes, however, consistently have the lowest. Informants whose families have been in Canada for more than three generations have lower scores than new Canadians. Only people over forty uttered [mXlta-]. In addition there were three utterances of [mAldi-]; these are included in the [mA'lti-] statistics. See variables (anti-) and (semi-) for further reference. 32. Natv-rally #190 Task; Word list: A. [nstjarsli] B. [nsetjrali] C. [nstjerli] D. [ nastjor I I i ] Table 6.2.32 NATURALLY Age > 40 Age < 40 Female Male All A B C. D A B C D A BCD A B C D D L WK LM 36 0 18 36 5 20 60 15 24 18 24 29 7 7 71 14 23 MIDDLE 45 9 27 18 15 20 50 5 14 36 36 14 35 0 47 6 10 UM LU Q 43 21 29 0 9 45 36 0 23 38 38 0 33 25 25 32 OTTAWA 25 19 22 28 8 18 53 16 14 25 32 27 16 12 49 14 We placed this item in the questionnaire hypothesizing that forms D and C would show socio-economic differentiation, i.e. that the upper classes would have the highest incidences of form D and that the lower classes would have the highest incidences of form C. Our data only partially supports our hypothesis. Compare variable (generally) Table 6.2.21; there our same hypothesis faired better. 250 33. Ottawa #258, 259, and 260 Task; Word list: A. [ndewb] B. [ntewib] C. [ndawa] D. [ddawe] Table 6.2.33 OTTAWA Age > 40 Age < 40 Female Male All A B C D A B C D A BCD A B C D B L WK LM 73 9 0 18 74 16 0 11 76 12 0 12 69 15 0 15 12 MIDDLE 57 29 14 0 56 13 25 6 50 30 20 0 62 8 23 8 13 UM LU 50 30 10 10 67 33 0 0 63 38 0 0 50 25 13 13 20 OTTAWA 61 21 7 11 66 17 10 7 66 23 6 6 62 15 12 12 How one pronounces the name of one's home town is often an indicator of the standing one has within that community. We can see from Table 6.2.33 that to begin Ottawa with an [a] is to risk being categorized into the lower classes, see especially the data for females and all informants under forty. On the other hand to begin Ottawa with an [n] followed by a [t] and ended with an [x>] or a schwa is to risk being taken for someone from Rockcliffe or at least Clemow; [ndewb] is neutral and non-descriptive. Form C, [ndewe], is employed only by upper and middle class informants whose families had been in Ottawa for several generations. 251 34. Pictures #36 Task; Pictures: A. [pikjarz] B. [pitjarz] C. [pikd^arz] D. [piktjarz] Table 6.2.34 PICTURES Age > 40 Age < 40 Female Male All A B C D A B C D A BCD A B C D C L WK LM 10 0 70 20 14 0 43 33 0 0 71 24 29 0 29 36 52 MIDDLE 27 0 45 27 15 0 80 5 29 0 57 14 12 0 76 12 68 UM LU 0 14 79 7 18 0 82 0 8 8 77 8 8 8 83 0 80 OTTAWA 11 6 66 17 15 0 65 15 11 2 68 16 16 2 63 16 Table 6.2.34 reveals that all groups preferred value C, [pikd^arz], and surprisingly that only the upper middle and lower upper classes said [pitjarz]. They also had the lowest scores for the prescribed form, [piktjarz]; The highest score for [piktjarz] was elicited from males of the lower, working, and lower middle classes. The upper classes con sistently scored highest for value.C. Form C, [pikd5arz], represents an extension of the medial/t/rule, i.e. post tonic t is pronounced [d] after [k] and when part of the affricate [d^]. See variable VtV in Chapter 5 for further discussion of the medial /t/ rule. i 252 35. Potato #369 Task; Reading: A. [peteido] B. [pedeido] Table 6.2.35 POTATO Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 36 64 75 25 61 39 62 38 61 MIDDLE 64 36 74 26 93 7 50 50 70 UM LU 79 21 73 27 92 8 58 42 76 RURAL 50 50 42 58 53 47 38 62 URBAN 59 41 76 24 82 18 58 42 IRISH 29 71 57 43 56 44 20 80 SCOTS 60 40 67 33 80 20 57 43 ENGLISH 75 25 73 27 88 13 55 45 FRENCH 0 100 75 25 33 67 100 0 OTHER 63 17 85 15 100 0 63 38 OTTAWA 61 39 74 26 80 20 56 44 ALL TOT 55 45 69 31 71 29 52 48 This variable occurred three times in the questionnaire; numbers 59-62 in Pictures, numbers 244-247 in Word list and number 369-371 in Reading. We have purposely chosen numbers 369-371 in order to display the highest incidence of value B. Our data reveal that among informants over forty there is very strong and ordered socio-economic co-variation, and that the pronunciation is decreasing in frequency among the younger informants. We can see that form B is most prevalent among informants of Irish descent and that it is more common among those with rural back ground. Males had much higher scores for form B than did females, and males with Irish background recorded the top score of 80 percent. We elicited only two occurrences of the soft, non-aspirated initial [ b ] which is characteristic of Irish Gaelic and Irish English. The Ottawa Valley Survey should give us some interesting data for this variable. 253 36. Recognize #140 Task; Word list: A. [rekegnaiz] B. [n £kenai z] Table 6.2. 36 RECOGNIZE Age : > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 73 27 95 5 83 17 93 7 87 MIDDLE 100 0 100 0 100 0 100 0 100 UM LU 93 7 82 18 92 8 93 7 88 OTTAWA 89 11 94 6 91 9 93 7 The middle class conforms perfectly to the prescribed form whili upper and lower classes stray occasionally from the mark. 37. Route #156 Task; Word list: A. [ ru t ] B. . [rAui ] Table 6.2. 37 ROUTE Age > 40 Age < 40 Female Male All A B A- B A B A B A L WK LM 73 27 90 10 89 11 79 21 84 MIDDLE 82 18 80 20 86 14 76 24 81 UM LU 93 7 100 0 100 0 92 8 96 OTTAWA 83 17 88 12 91 9 81 19 Table 6.2.37 reveals a fair amount of socio-economic variation with this pronunciation item. Value B which is frequently heard in the United States is little used by the upper classes in Ottawa, but it enjoys 24 to 27 percent frequency among the lower and middle classes. Gregg, 1973, pp.109-113 and our data were similar while the CEU, pp.87-88, is consistently 20 to 30 percentage points lower for [rut]. 254 38. Sandwiches #89, 90 and 91 Task; Pictures: A. [ saendw i t'Jez ] B. [ seenw itjez] C. [seem (w)itjaz] Table 6.2.38 SANDWICHES -* Age > 40 Age < 40 Female Male All A B C A B C A B . C A B C A L WK LM 60 10 30 19 52 29 47 18 35 14 64 21 31 MIDDLE 75 25 0 32 53 16 36 55 9 50 38 13 43 UM LU 46 38 15 55 36 9 62 38 0 36 36 27 48 IRISH 80 20 0 25 50 25 57 14 29 33 67 0 SCOTS 63 25 13 33 44 22 75 25 0 38 38 23 ENGLISH 50 25 25 40 40 20 56 31 13 27 36 36 FRENCH 100 0 0 50 50 0 67 33 0 50 50 0 OTHER S 40 40 20 15 69 15 20 60 20 25 63 13 OTTAWA 58 26 16 31 49 20 49 34 17 34 46 20 Informants under forty years of age and female informants display ordered socio-economic stratification with reference to form C. The sociological pattern which we hypothesized for the retention or deletion of the [d] is fully realized only as applied to informants under forty. Two individuals pronounced this variable as [ saenwi t Jez ], which is a Scottish form and which enjoys fairly high frequencies among older people in the Ottawa Valley. Compare this variable with (hundred) for d- deletion statistics. 255 39. Schedule #125 Task; Word list: A. [ Jffd^u 1 ] B. [skcfd^ul ] Table 6.2. 39 SCHEDULE Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 36 64 24 76 17 83 43 57 28 MIDDLE 36 64 15 85 36 64 12 88 23 UM LU 57 43 18 82 54 46 25 75 40 OLD 45 55 17 83 38 62 20 80 NEW 43 57 22 78 26 74 33 67 OTTAWA 44 56 19 81 33 67 26 74 Value A [ Je^u I ], the current British variant has been consciously promoted by the CBC as a prestigious form. This seems to have had some effect on older informants of the upper classes but very little or a negative effect on younger people. In 1955 when Avis did his survey, about 33 percent of his students opted for form A; today in Ottawa the percentage for that same age group is about half that number. Avis, 1956, op.cit., pp.53,54, Gregg, 1973, pp.109-113 and CEU, pp.55-56 show even higher-frequencies for[sk£d^ul ]. 40. Semi-(circle) #67 Task; Word list: A. [ s£mi -] B. [s6 ma i- ] C.[s e'ma- ] Table 6. 2.40 SEMI-Age > 40 Age < 40 Female Male All A B C A B C A B C A B c A L WK LM 82 9 9 100 0 0 94 0 6 93 7 0 94 MIDDLE 82 0 18 95 5 0 92 0 8 88 6 6 90 UM LU 93 0 7 82 18 0 100 0 0 75 17 8 88 OLD 82 5 14 96 4 0 96 0 4 84 8 8 NEW 93 0 7 91 0 0 95 0 5 89 11 0 OTTAWA 86 3 11 94 6 0 95 0 5 86 9 5 256 Table 6.2.40. demonstrates that form A [s£mi] is the prevailing form for all sociological groups. The two age groups and the two sex groups, however, display opposing socio-economic patterns. Avis notes that in the fifties, Americans were referred to as /se'maiz/ by Canadian teenagers; today, that usage is unknown. See variables (anti-) and (multi-) for further information and references. 41. Sentence #119 Task; Word list: A. [s£ntans] B. [st5?ns] C. [se'ntns] Table 6.2.41 SENTENCE Age > 40 Age < 40 Female Male All ABC ABC L WK LM 18 64 18 48 43 10 MIDDLE 100 0 0 60 20 20 UM LU 50 43 7 82 18 0 OTTAWA 56 36 8 60 29 12 ABC ABC A 33 56 11 43 43 14 37 93 7 0 59 18 24 74 85 15 0 42 50 8 64 67 29 4 49 35 16 No clear sociological pattern is evident. It is interesting to see, however, that form B has a consistently high score among groups of infor mants of the lower classes. 257 42. Something #143 and 144 Task; Word list: A. [sXmGir)] B. [sXmGin] C. [sAm?en] Table 6.2.42 SOMETHING Age > 40 Age < 40 Female Male All A B c A B c A B C A B c A L WK LM 36 45 18 84 16 0 67 22 11 67 33 0 63 MIDDLE 82 18 0 75 25 0 79 21 0 76 24 0 77 UM LU 71 29 0 70 30 0 69 31 0 73 27 0 71 OTTAWA 64 31 6 78 22 0 71 24 4 72 27 0 Form C, [sXm?an] was elicited only from women over 40 of the lower classes. The majority of all groups read form A, [sXmGin], this fact adds further evidence to the pattern that [-in#] is preferred only as a verbal, i.e. progressive aspect and gerund morpheme. See item 30, morning for comparable data. 43. Toronto #s 254, 255, 256, and 257 Task; Word list: A. [tenoirto]- B. [terdrrto] C. [tenondo] D. [ tanono] Table 6.2.43 TORONTO Age > 40 Age < 40 Female Male All A B C D A B C D A B C D A B C D A L WK LM 50 40 Q 10 41 24 6 29 44 38 6 13 45 18 0 36 39 MIDDLE 36 55 9 0 33 28 28 11 57 36 0 7 13 40 40 7 32 UM LU 71 7. 21 0 44 44 11 0 69 31 0 0 50 10 40 0 56 OTTAWA 54 31 11 3 39 30 16 16 56 35 2 7 33 25 28 14 Form D, [ten$no] displays a fully ordered socio-economic correlation; it received its highest scores from the informants of the lower, working, and lower middle classes; the middle class had fewer instances of this form and the upper classes did not have any occurrences at all. Form C , 258 received its highest scores from older males. Forms A and B show no clear sociological pattern. The form [trdna] heard frequently among the working class in the city of Toronto was not once read by our native Ottawans. 44. Temperature #158 Task; Word list: A. [ temparat Jar ] B. [ te'mprat Jar ] C. [ te'mpart Jar ] D. [ te'mpat Jar ] Table 6.2.44 Age > 40 ABCD L WK LM 0 64 0 36 MIDDLE 27 36 0 36 UM LU 7 50 14 14 OTTAWA 11 50 6 28 TEMPERATURE Age < 40 Female ABCD 0 40 5 55 0 55 20 25 9 64 0 27 2 51 10 37 ABCD 0 59 0 41 14 43 7 36 8 69 8 8 7 57 5 30 Male ABCD 0 36 7 57 6 53 18 24 8 42 8 33 5 44 12 37 All B 48 48 Form A appears to be a reading pronunciation which is infrequently pronounced even in this task. The prevailing form is B for all groups. Form D is favoured by the lower classes more than by any other group; the lower classes tend to reduce (and eliminate) syllables more than other classes. Excessive use of this reduction results in stigmatized speech. 259 45. Vase #37 Task; Picture: A. [vnz ] B. [vaz] C. [veiz] D. [veis] Table 6.2.45 VASE Age > i 40 Age < t 40 Female Male All A B C D A B C D A B C D A B C D C L WK LM 64 36 0 0 14 33 48 5 44 33 22 0 14 36 43 7 31 MIDDLE 38 64 18 0 15 45 40 0 21 71 7 0 12 35 53 0 32 UM LU 36 29 36 0 9 55 36 0 38 31 31 0 8 50 42 0 36 OLD 36 36 27 0 14 48 38 0 35 46 19 0 12 40 48 0 NEW 43 50 7 0 13 35 48 4 37 42 21 0 11 39 44 6 OTTAWA 39 42 19 0 13 42 42 2 36 44 20 0 12 40 47 2 Table 6.2.45 reveals that the American form, form D, is rarely used in Ottawa. The Canadian form, formC, enjoys widespread use in Ottawa with younger informants and male informants having the highest scores. Form B [vaz] with an unrounded back open vowel is more frequent than the form [VDZ]. Forms A and B were more frequently elicited from females and older informants than from their counterparts. The pronunciation [veez] was not recorded. Avis, 1956, p.43, Gregg, 1974, pp.108-113 and CEU, p.58, and we agree on about 30 percent for [veiz], Gregg and we agree on 60 to 70 percent for forms A and B combined while Avis and CEU had 40 and 30 percent respectively. This was undoubtedly a difficult variable to deal with through a postal questionnaire. 260 46. Weren't #167 Task; Word list: A. [warnt] B . [ wernt ] Table 6.2. 46 WEREN' T Age > 40 Age < 40 Female Male All A B A B A B A B B L WK LM 91 9 95 5 94 6 93 7 6 MIDDLE 91 9 84 11 86 7 88 13 10 UM LU 86 14 91 9 77 23 100 0 12 RURAL 85 15 77 23 74 26 93 7 URBAN 82 18 93 5 85 12 94 6 OTTAWA 89 11 90 8 87 11 93 7 ALL TOT 83 17 89 9 81 17 93 7 Table 6.2.46. established that there is a moderate currency of form B, [weYnt ], throughout society in Ottawa. This form has an average over-all percentage of about 13 percent with higher percentages among females of the upper classes. Form B is also more popular among those informants with rural background. The word 'were', item #600, was noted to be pro nounced [wer] about 6 percent of the time but only when in a stressed position. The forms [-weYnt] and [wer] enjoyed , significantly higher frequencies, 40 and 46 percent respectively, in our Ottawa Valley urban centres of Renfrew and Smith's Falls. The rural area of the Ottawa Valley undoubtedly would yield even higher frequencies. The high frequencies of these forms in the rural portion of the Ottawa Valley, a unique dialect pocket area, may have caused Ian Pringle, co-worker of the Ottawa Valley Survey, to assume this usage for the majority of Canadians.^ In fact, [weYnt] and [wer] are minority usage forms. R.J. Gregg estimates 18 a less than 5 percent frequency for [weY] and [weYnt] in Vancouver. No other data are available for comparison concerning this variable. 261 47. Wilson #339 Task; Reading: A. [wilsen] B. [wiltsen] Table 6.2.47 WILSON Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 82 18 100 0 94 6 92 8 94 MIDDLE 100 0 100 0 100 0 100 0 100 UM LU 86 14 100 0 85 15 100 0 92 RURAL 80 20 92 8 74 26 100 0 URBAN 86 14 100 0 . 94 6 97 3 OTTAWA 89 11 100 0 93 7 98 2 ALL TOT 83 17 98 2 87 13 98 2 We can see from this table that the intrusive 't' at least in this position is dying out among the younger generation. It has fairly high currency among older females and especially among rural females. The intrusive 't' does not appear to be losing ground in final position in words such as false, sense, and fence. Further we recorded four instances of across pronounced as [akrnst^a form very common in eastern Michigan. 48. Zebra #48 Task; Pictures: A. [zfbre] B. [ z£bre] Table 6.2.48 ZEBRA Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 80 20 100 0 94 6 92 8 94 MIDDLE 82 18 95 5 93 7 88 12 90 UM LU 86 14 100 0 92 8 92 8 92 OLD 76 24 97 3 88 12 88 13 NEW 93 7 100 0 100 0 94 6 OTTAWA 83 17 98 2 93 7 90 10 262 Avis' surveys in 1949-50 and 1954-5 reveal that the majority of his 19 informants chose form B [z£bre]; this is definitely not the case today in Ottawa. Our informants over forty years of age averaged only 17 per cent for [ze'bre] and our informants under forty averaged only 2 percent. The study in the Kootenays reveals a similar decline in percentages for [ze'bra] from the older generation's, 35 percent, to the teenagers' 6 per-20 cent. For comparative data see: Avis, 1956, pp.44-45 and Gregg, 1973, pp. 112-113. 3. Vocabulary The variables in this category were selected as representative samples of lexical items which have long been considered typically Canadian in usage, meaning, or frequency. The terms of reference remain Canadian English versus Northern American English. We hypothesize that the data will reveal sociological differentiation with reference to age group and possibly generation group more than to any socio-economic criteria. 1. Chesterfield #42 Task; Picture: Identify by name a large piece of furniture which can seat three to four people. A. sofa B. chesterfield C. couch Table 6.3.1 CHESTERFIELD Age > 40 Age < 40 Female Male All A B C A B C A B C A B C B L WK LM 18 55 27 24 19 57 11 44 44 36 14 50 31 MIDDLE 27 73 0 35 40 25 36 50 14 29 53 18 52 UM LU 14 79 7 27 36 36 15 62 23 25 58 17 60 OLD 23 59 18 28 34 38 19 50 31 32 40 28 NEW 14 86 0 30 26 43 21 53 26 28 44 28 OTTAWA 19 69 11 29 31 40 20 51 29 30 42 28 263 Our data reveal a strong drift away from chesterfield towards sofa and couch among those informants less than 40 years of age. Avis in 1955 recorded that up to 88.8 percent of his informants chose chesterfield. In general, young people seem to have far fewer lexical Canadianisms in their speech than older people. The three choices have somewhat different meanings according to a sizable minority of those sampled. Chesterfield is a prestige piece of furniture for the living room; couch is a piece of furniture for relax ation in'the recreation room and sofa is the neutral term. We can see that there is some ordered socio-economic differentiation with regard to this linguistic variable. Also interesting to note is the fact that 86 percent of the informants with new Canadian background who were over 40 said chesterfield while young informants with new Canadian backgrounds chose couch and sofa. Females conformed in higher percentages to the Canadian usage than did males. Davenport, a very frequent form in Michigan,was mentioned only two times and in both cases it was a third or fourth choice. See Avis, 1955, pp.13-18, Gregg, 1973, pp.110-116, and CEU, pp.106-107 for further discussion and slightly divergent data concerning this variable. 264 2. Blinds #51 Task; Picture: Identify by name an opaque cloth on a roller which when rolled down in a window prevents light from entering or leaving. A. Blinds B. Shades Table 6.3.2. BLINDS Age > 40 Age • < 40 Female Male All A B A B A B A B A L WK LM 82 18 60 40 67 33 69 31 68 MIDDLE 80 20 60 40 85 15 53 47 67 UM LU 93 7 45 55 77 23 67 33 72 OLD 86 14 54 46 76 24 58 42 NEW 86 14 61 39 74 26 67 33 OTTAWA 86 14 57 43 75 25 62 38 Again , we see that the two age : groups display a 30% difference : their overall age scores. Avis' data in 1955 provide us with a quarter century's time perception ;< he recorded that then 94 .5 percent of his informants chose blinds, . Our data reveal that our informants over 41 years of age averaged 86 percent, but younger informants averaged only 57 percent. We notice further from our data that we more frequently elicited the Canadianism from females than from males. See Avis, 1955, pp.15-16 and Gregg, 1973, pp.108-115 for further discussion and generally similar data on this variable. 265 3. Brush (for chalk board) #63 Task; Picture: Identify by name the felt instrument used to cle chalk boards. A. brush B. eraser Table 6.3. 3 BRUSH Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 64 36 57 43 72 28 43 57 59 MIDDLE 64 36 85 15 86 14 71 29 74 UM LU 50 50 70 30 54 46 64 36 58 OLD 64 36 83 17 73 27 76 24 NEW 50 50 55 45 68 32 35 65 OTTAWA 58 42 71 29 71 29 59 40 This variable and zed are the only two' vocabulary ft ems classified as Canadianisms which received higher scores from informants under 40 than from those over 40. "Informants whose families have been in Canada several generations had consist/gntly higher scores than did the-newer Canadians. No other studies of this variable are extant. Americans do not understand-the Canadian* us age of form A. 266 4. Orange #322 Task; Local words and usage: Question: What colours are the lights in a traffic light? A. orange B. yellow C. amber (or caution) Table 6.3.4 ORANGE Age > 40 Age < 40 Female Male All A B C A B C A B C A B C A L WK LM 27 55 18 15 80 5 22 61 17 15 85 0 19 MIDDLE 36 36 27 32 42 26 29 43 29 38 38 25 33 UM LU 38 54 8 27 64 9 33 67 0 33 50 17 33 RURAL 35 55 10 8 75 17 22 61 17 29 64 7 URBAN 32 45 23 26 57 17 30 52 18 26 55 19 IRISH 29 57 14 14 86 0 22 67 11 20 80 0 SCOTS 30 50 20 44 56 0 20 80 0 43 43 14 ENGLISH 33 50 17 20 53 27 38 44 19 9 64 27 FRENCH 0 100 0 25 50 25 0 67 33 50 50 .0 OTHER 60 20. 20 23 62 15 30 50 20 38 50 13 OTTAWA 34 49 17 24 62 14 27 57 16 29 56 15 ALLTOT 33 50 17 22 61 17 27 55 18 27 58 16 Throughout the Ottawa Valley,, including the cities of Ottawa and" Montreal,one frequently hears talk of "running an orange light". Our 2' data demonstrates that the usage of 'orange' is indeed considerable. We can see, however, that this usage is waning among the younger infor mants. No other sociological patterns are evident. It is noteworthy that several informants offered 'caution' as a colour. This item has not been investigated in previous studies dealing with Canadian English. 267 5. Taps #80 Task; Picture: Identify by name a combined hot and cold water outlet in a bathroom sink. A. tap( s) B. faucet Table 6.3. 5 TAPS Age > 40 Age < 40 Female Male All A B A B A B A B B L WK LM 73 27 52 48 72 28 43 57 41 MIDDLE 64 36 65 35 71 29 59 41 36 UM LU 71 29 36 64 54 46 58 42 44 OLD 77 23 69 31 81 19 64 36 NEW 57 43 35 65 47 53 39 61 OTTAWA 69 31 54 46 67 33 53 47 Table 6.3.5. reveals no systematic differentiation according to socio-economic class. We notice, however, that the older informants and females had much higher scores for tap(s) than did the younger informants and males, respectively. New Canadians consistently had 20 to 40 points higher for the American value faucet than did informants whose'family had been in Canada for several generations. Valve and spigot - also classified as Americanisms - were never mentioned as possible answers. Almost all informants who were asked supplementary questions indicated that the outside outlet for the garden hose was called a tap and that a separate individual outlet for hot or cold water in a kitchen or bath room sink was called a tap. When comparing other data for this variable, Avis, 1956, p.18, Gregg, 1973, it is important to realize that these technical distinctions were not held constant in the various surveys on Canadian English. 268 6. Past versus After #s 101,; 102, 103 Task; Picture: What time is it on this clock? (11:15) A. quarter past B. quarter after Table 6. ,3.6. PAST VERSUS AFTER Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 11 89 5 90 6 88 7 93 7 MIDDLE 27 73 20 75 36 57 12 88 23 UM LU 33 67 63 38 58 42 25 75 45 OTTAWA 25 75 20 76 31 64 13 87 All columns indicate a strong and ordered socio-economic differen tiation for this variable. The higher the class the higher the score for past, the typically British form. Generally, however, our Ottawans preferred after. Again, females and those over 40 had the highest scores for the British form. Comparative data are not available for this variable. 7. Railway Crossing #68 Task; Picture: Identify by name the junction of a road and train tracks. A. railway crossing B. railroad crossing C. level crossing Table 6.3,7. RAILWAY CROSSING Age > 40 Age < 40 Female Male All A B C A B: C A B C A B C A L WK LM 45 55 0 62 38 0 50 50 0 64 36 0 56 MIDDLE 36 64 0 45 45 10 43 57 0 41 47 12 42 UM LU 79 21 0 55 45 0 62 38 0 75 25 0 68 OTTAWA 56 44 0 54 42 4 51 49 0 58 37 5 269 No ordered, linear sociological correlation is evident for this variable. Railway is generally considered British and railroad American; our informants seemed to share these two variants almost equally. Level  crossing is the normal British term in current usage. 8. Z #108 Task; Pictures: Please read these letters out, H, W, X, Y, Z. A. zed B. zee Table 6.3.8. ZED Age > 40 Age < 40 Female Male All A B A B A B A B A L WK LM 64 36 86 14 78 22 79 21 78 MIDDLE 82 18 95 5 93 7 88 12 90 UM LU 93 7 82 18 92 8 83 17 88 OLD 73 27 97 3 85 15 88 12 NEW 93 7 78 22 89 11 78 22 OTTAWA 81 19 88 12 87 13 84 16 Our data reveals that a higher percentage of our informants aged 16 to 39 said zed than did those over 40. Almost all who commented on this variable felt that American children's programs such as Sesame Street would change the usage pattern for the next generation. For informants over 40, this variable had regular socio-economic differentiation, with the American value zee receiving its highest scores from the lower classes and the British value zed obtaining its highest scores from the upper classes. See Avis, 1956, p.50 and CEU, pp.59-60 for further information and quite similar data concerning this variable; it appears as though zed has lost 10 percentage points over a quarter of a century. 270 4 • Summary Cross-Analysis The data for the 71 linguistic variables presented in this chapter provide us with evidence of the co-variation of grammatical, pronunciation, and vocabulary items on the one hand and sociological parameters on the other. Although Canadian English is considered to be quite uniform from the Ottawa River to the Pacific, the variation within a given city, in this case Ottawa, is substantial. This variation is correlated closely to the sociological factors of socio-economic class, age, and sex. So far in this chapter, we have analysed each linguistic item individually with reference to the sociological factors mentioned and occasionally augmented by other factors such as ethnic background, rural versus urban background, and new Canadian versus several generation Canadian background. We will now conduct a cross-analysis of the same data by focusing on the sociological factors.as reference points, thus grouping together the linguistic items. Grammar The grammar section of this chapter contains data which most strongly displays ordered socio-economic differentiation. These data, therefore, support our hypothesis that, for many items, the linguistic variation is directly related to the socio-economic status of the speaker. From our data in Table 6.4.1, we can see that clear socio-economic differentiation and linear ordering occurred in 13 of the 15 variables. This stratifica tion occurred with reference to the following values: 6 7 1. Between John and me; 3. Eh ; 4. Eh ; 5. Fewer; 6. Have you got; 271 8. If you had; 9. Lie, Lay; 10. Lie, Lay, Lain; 11. ?not? 12. Past Perfect; 13. Sneaked; 14. Subject/verb agreement; and 15. Take. We rank ordered the socio-economic groupings by giving a 1 to the group with the highest frequency, a 2 to the group with the second highest frequency, and a _3 to the group with the lowest frequency of the preferred value. Table 6.4.1. GRAMMAR BY SOCIAL CLASS Highest Frequency of Prescribed Value (rank ordered) Class/Item 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 T L WK LM 3 2 3 3 3 3 2 3 3 3 3 3 3 3 3 43 MIDDLE 2 3 2 2 2 2 3 2 2 2 2 2 2 2 2 32 UM LU 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 15 Further, Table 6.4.1. reveals partial ordering for values 2. Just  between you and me, and 7. If it were. When we add the rank scores, we find UM LU in first place with 15 points, MIDDLE in second place with 32 points, and L WK LM in last place with 43 points. The data in Table 6.4.1. present a strong case for socio-economic stratification in Canadian English, and they demonstrate the need for including socio-economic parameters in future dialect studies of Canadian English. In addition to this analysis on socio-economic class, we undertook an analysis of the Grammar items according to the four sex and age groups. 272 Table 6.4.2. GRAMMAR BY SEX AND AGE GROUP Highest Frequency of Preferred Value (rank ordered) Item 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 T 0 1 3 1 1 1 2 1 1 1 1 1 2 1 1 3 21 F 2 1 2 3 2 4 2 2 2 2 3 4 2 2 4 37 M 3 4 3 2 3 1 3 3 3 3 2 1 2 2 1 38 Y 4 2 4 4 4 3 4 4 4 . 4 4 3 4 4 2 54 F = Female 0 Over 40 years old M = Male Y = Under 40 years old These data reveal that the informants over 40 years of age had the highest frequencies of the preferred values of the fifteen variables. When we add the rank score for all the items, we see that the informants over 40 years of age were first with 21 points, females were second with 37 points, males next with 38 points, and informants under 40 years of age were last with 54 points. The fact that the younger informants placed fourth is no doubt disconcerting to parents and educators, but this may only be the result of an age role which changes as one grows older. We note the closeness in scores between the males and females; only within this section of grammar items do males score close to females. It is interesting to note that males had the highest scores for the time and space relationships for items 12. Past Perfect and 15. Take, while 22 females had the lowest ranking for these two items. Generation Gap. • A large difference in scores, at least 20%, for informants over 40 versus informants under 40, i.e. pre W W II and post W W II respectively, is referred to as a generation gap in this study'. A generation gap was revealed for: 1. Between John and me; 5. Fewer; 6. Have you got; 7. If it were; 8. If you had; 9. Lie, Lay; 10. Lie, Lay, Lain; and 13. Sneaked. 273 For all items, we notice that the younger informants were moving away from the prescribed values. Differentiation by Sex Females had a 21 percent higher frequency for form A of 7. If it were. Word Pronunciation In Chapter 5 we analysed twenty-seven phonological segments in order to determine the co-variation of these items with sociological parameters. In this portion of this chapter, we will analyse' the choice of phono logical segments associated with particular words. Most of these words come from the traditional Canadian English studies such as the SCE, Avis 23 1955, Gregg and Poison. These words were chosen for. the above x mentioned studies mainly for their ability to reveal differences between American and Canadian speech, differences between British and Canadian speech, and to reveal trends in linguistic change for comparing the speech patterns of the different generations, sexes, and provinces' within Canada. We will now conduct an analysis of our data, applying our socio logical parameters as reference points. Socio-economic Stratification Table 6.4.3 displays the rank ordering of the three class groupings as determined by the frequency of pronunciation of the preferred value for each pronunciation variable. 274 Table 6.4.3. PRONUNCIATION BY CLASS Highest Frequency of Preferred Value (rank ordered) Item 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 "L. WK LM 331233213 2 3 3 2 3 3 3 1 1 3 2 3 3 2 3 MIDDLE 112322311 :'l 1 2 3 2 2 2 2 3 2 3 2 2 3 2 UMLU 223111112121111132111111 Table 6.4.3 continued Item 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 T LWKLM 233323123333232133223321 116 MIDDLE 1222322322213232 .1 1322213 97 UMLU 311111311 112111322111132 69 Our data reveal the following pronunciation values had ordered co variation with socio-economic class: 5. [sasfalt], 6. [nnt] and [ant], 12. [ kengrat jaleit ], 14. [eg], 15. [atoar], 16. [fe'brueri], 19. [film]-, •21. [d^enarl I i ], 22. [d^njuan], "24. [ korki ], .26,1 I ef tenant ],. 27..[ I Xkferi ], 28. \[miYar ], 30. [morninj, 33.. [ pt.?WT> ],. 34...~fe&k<d;3ar z.], 35. [pateido], 38. [ seandw it Jaz ], 45. [veiz] , and 46. [weYnt]. This analysis indicates that 20 of the 48 pronunciation items demonstrated ordered socio-economic stratification while 24 other items revealed partial socio-economic stratification. Four items^ namely 3. [aenti-] , 17. [fartail] , 31. [mXlti-] , and 40. [s£mi-], demonstrated ordered socio-economic dif ferentiation contrary to that hypothesized for them. The sum totals of the rank ordering for each item adds up to 69 points for UM LU, 97 points for MIDDLE, and 116 points for L WK LM. Although our data demonstrated the general trend of socio-economic stratification which was hypothesized, our data also suggest that our list of vocabulary items, taken from older and rural surveys, needs some revision. 275 Stigmatized Forms The following values had frequency scores which stood out in the four sex and age groups of the lower, working, and lower middle classes and have been categorized as probable stigmatized forms: 5. [aezfalt], 8. [ben], 12. [ kengresd^aleit ], 16. [fe'bjueri], 19. [filam] , 20. [greedy], and [grad^] , 21. [denarii] , 22. [ d^cfnjuain ] , 23. [ hAn(d)art ], -25. [laiberi] , 28. [mir] , [miral] and [mi*rou], 30. [mornan], 35. [padeido], 38. [ s'sern(w).i tj ]-arid •[ saeQwit J,], 41. [se?ns], 42. [s/\m?an] and 43. [t(a)r&no]. In addition, the data for the forms 36. [rdkanaiz] and 37. [raot] showed a partial pattern of socio-economic differentiation which would lead one to categorize these values as possible stigmatized forms. Contrary to expectations, the value 34. [pitjar] for picture was 24 pronounced only by the upper middle and lower upper classes. Prestige Forms The following values stand out as prestige pronunciations because of their higher frequencies among the upper middle and lower upper classes: 6. [rmt ] and[ant], 7. [beelkani], 15. [aiSar], 21. [ d^narl I i ], 24. [kdrki], and [kdki], 26. [ I ef te'nant ], 32. [ nset Jar I I i ], 33. [ntawc], 39. [Je^ul], 43. [tarcnto] and 45. [VDZ] and [vaz]. Generation Gap A large number of the variables, 29 of 48, were especially well suited to the task of demonstrating the differences in usage between the age groups. The values which were distinguished by their higher frequencies among the older informants are: 1. [aefraka], 4. '[ ei prakcis ], 5. [aesfalt] 276 and [aezfalt], 8. [ben], 10. [kaeramel], 15. [aioar], 19. [fi'lam], 20. [gsra^], 24. [korki] , 25. [laibri], 26. [ I efferent ], 30. [mornin], 31. [mAlta-], 35. [padeido], 38. [ seendw it faz ], 39. [J^d^ul], 40. [se'ma-], 45. [vnz], 47. [wi Itsan]and . 48. [ z£bra]. Conversely, the values which had usage patterns peculiar to the informants under 40 years of age are: 5. [aeJfDlt ], 8. [bin], 12. [kangrad^aleit], 13. [de"kal], 16. [feYjueri ], 17. [faYtail], 22. [d^njuam], 28. [mir], 32. [natjarli], 43. [t(a)rdno], and 45. [veiz]. Table 6.4.4. GENERATION GAP 1 = conforms more to prescribed or prestigious forms 0'= conforms less to prescribed or prestigious forms Item 1 4 5 8 10 12 13 15 16 17 19 20 22 24 25 26 28 30 31 32 35 38 39 40 42 43 45 47 48 T Old 11101111100111011101011001100 19 Young 00010000011000100010100110011 10 Table 6.4.4. reveals that the informants over 40 years of age conformed more to the prescribed and/or prestigious forms than did the informants under 40 years of age, in 19 out of the 29 cases. It is interesting to observe that our younger informants had the higher frequencies for 8. [bin] 13. [de'kal], 17. [fartail] and 45. [veiz], all four of which are considered Canadian markers. One would have expected the older informants to have higher frequencies here, too. Differentiation by Sex Females had at least a 10 percent higher frequency than did males for the following values: 6. [ant], [nnt], 7. [baelkeni], 13. [blaoz], 15. [aiSar], 17. [ f e*brueY i ], 23. [hAndrad], 24. [kdrki] , 26. [ I ef tenant ], 277 30. [morninj, 32. [ naetjerl I i. ], 36. [peteido] , 38. [seendwit Jez], 41. [sentans] , 43. [tanSnto] , and 44. [te'mpretJar]. Sex and Age Groups We can compare the patterns and roles of the sex and age groups most effectively when we analyse the data for: 1) the 20 variables which reveal ordered socio-economic differentiation, 2) the 17 stigmatized forms, and 3) the 14 prestige forms. Regarding the first set of variables, we will be observing to what extent each of our four sex and age groups maintained ordered and regular class distinctions within itself. The second and third sets of variables will be examined in order to measure the relative frequency of stigmatized and prestige forms uttered by each sex and age group. Our cross-analysis of the 20 variables revealing clear socio-economic stratification is illustrated in Table 6.4.5. Table 6.4.5. SOCIO-ECONOMIC STRATIFICATION MAINTAINED Item 5 6 12 14 15 16 19 21 22 24 26 27 28 30 33 34 35 38 45 46 T F 00111111111110100101 14 0 011 0 0 01111011110 1011 13 " • Y 1 Q 1 0 1 1 0 1 1 1 1 0 1 0 0 1 0 1 0 0 11 . M 101010011 1 1110010000 10 1 = maintained stratification F = female 0 = over 40 0 = did not maintain M = male Y = under 40 The above data reveals that females and older informants maintained the strongest internal class hierarchy of linguistic/socio-economic co variation. They maintained a class stratification, for 14 and 13 Items respectively.. Our informants under 40 years of age, were next with 278 11 items. Finally, our male informants showed the least amount of socio economic class differentiation by maintaining a regular ordering for only 10 variables. The data for our four sex and age groups in this chapter appears consistent with the data for the previous chapter, wherein the ordering from formal to informal for the four sex-age groups was: 1) females over 40, 2) males over 40, 3) females under 40, and 4) males under 40. When we analyse the performance of our four sex and age groups with regard to utterances of stigmatized forms we find equally interesting results. Table 6.4.6. FEWEST STIGMATIZED FORMS (RANK ORDERED) Item 5 8 12 16 19 20 21 22 23 25 28 30 35 38 41 42 43 T F 22 2 12013313011122 21 Y 1142103412402 3214 35 0 441340214310 3 2441 43 M 23243042242044333 45 F = female 0 = over 40 M = male Y = under 40 Table 6.4.6. reveals that females uttered the fewest stigmatized forms. Informants under 40 years of age were the next group with an index score of 35. Both these groups are subject to a great deal of pressure to conform to prescribed norms and to not stand out from others around them. The SCE revealed that young females consistently conformed to prescribed norms more than any other group; this was, in part, a reflection of the role of teenage girls in the Canadian school system. The highest instances of stigmatized forms occurred among the older informants, 43 index points, and male informants, 45 index points. The older informants, and the older males especially, had high occurrences 279 of what are often considered rural forms ,such as [ aesfal t ],.[ ben ], [filam], [hAnert], [ padeido],.C rAut ], [se*?ns] and [sAm?an]. Our male informants had high occurrences of 16. [febjueri], 19. [filam], 21. [d^narl i ],25. [ la'iberi ], 35. [ padeido], 38. [saemwit J ],41. [se^ns], and 43. [tardno]. These pronunciations tend to be classified as 'sloppy' by English teachers, which in turn probably encourages teenage boys to retain or adopt them in order to differentiate themselves from the girls. The third perspective for evaluating the performance of our four sex and age groups is by means of our prestige forms. Table 6.4.7. MOST PRESTIGE FORMS BY AGE AND SEX (RANK ORDERED) Item 6b 6c 7 15 21 24a 24c 26 32 33 39 43 45a 45b T F 12114 2 122121 2 123 0 31222 1 311212 1 3 25 Y 14333 4 243343 3 2 42 M 33441343433444 48 Table 6.4.7. reveals that females are ranked number one again, this time for having the highest instances of prestige forms. Our female informants had the highest occurrences of 6b. [nnt], 7. [ball kan I], 15. [ai'Sar], 24c. [kdki], 33. [ntawb], 43. [ tarnnto] and 45b. [vaz]. The informants over 40 years of age were ranked very close in second position. They had the highest frequencies for- 6c. [ant], 24. [kdrki], 26. [leftenant], 32. [ naetJWI I i ], 39. [Je^ul], and 45a. [ vnz ]. All these prestige forms except for [naetjarlli] find their source in British usage. In contrast, the younger generation of informants has lost most of these British prestige forms and have shifted to North American forms. As is the case with most of our other perspectives, the male informants were ranked last. 280 For all perspectives, we see that the females are ranked first and that the males are consistently ranked last. This ranking provides strong evidence for sex and age roles in language usage, and concurs with the 25 results in Chapter 5 and the SCE. It is interesting to speculate whether these language roles can change; our discussions with informants point to the fact that informants changed their speech patterns as they grow older and when they moved up or down the socio-economic structure. Vocabulary All items in this category, with the exception of 4 and 6, are Canadian markers which differentiate Canadian speech from Northern American. Our first analysis will be to see if there is a linear linguis tic and socio-economic co-variation for the following preferred values of our variables: 1. chesterfield, 2. blinds, 3. brush, 4. orange, 5. taps, 6 past, 7. railway crossing, and 8. zed. Table 6.4.8. VOCABULARY ITEMS BY CLASS Highest Frequency of Preferred Value (rank ordered) Item 1 2 3 4 5 6 7 8 T L WK LM 3 2 2 2 2 3 2 3 19 MIDDLE 2 3 1 1 3 2 3 1 16 UM LU 1 1 3 1 1 1 1 2 11 Table 6.4.8. reveals that only items 1. 'chesterfield' and 6. 'quarter past' have clear and ordered socio-economic stratification. The most interesting analysis of the data for these vocabulary items will be to determine to what extent our sociological groups have retained these Canadian vocabulary items. We will compare the three socio-economic 281 groups first, followed by the four sex and age groups, and finally the new Canadians will be compared to Canadians whose families have been in Canada for at least three generations. Table 6.4.8. displays the rank ordering of the three socio-economic groupings for each lexical item. The Table demonstrates that the upper middle and lower upper classes retained the highest frequency of Canadian lexical markers, followed by the middle class. The lower, working, and lower middle classes had the lowest frequencies of these Canadian markers. Table 6.4.9. VOCABULARY ITEMS BY SEX AND AGE Highest Frequency of .Preferred Value (rank ordered) Item 1 2 3 4 5 6 7 8 T 0 1 1 3 1 1 2 2 4 15 F 2 2 1 3 2 1 4 2 17 M 3 3 2 2 4 4 1. 3 22 Y 4 4 1 4 3 3 3 1 23 Table 6.4.9. displays the rank ordering of the four sex and age groups for each lexical item. One immediately sees, when observing the data, that the informants over 40 years of age retained the highest frequencies of the Canadian vocabulary items while the informants under 40 years of age had the lowest frequencies for the same Canadian markers. This trend toward a more general English among younger informants is not unexpected when one considers all the language levelling that has taken place since World War II in radio, television, popular music, movies, travel, migration, and the education system. Females had higher frequencies of these markers than did males. 282 Table 6.4.10. VOCABULARY ITEMS/NEW AND OLD CANADIANS Highest Frequency of Preferred Value (rank ordered) Item 1 2 3 4 5 6 7 8 T OC 2 2 1 0 1 0 0 1 7 NC 1 1 2 0 2 0 0 2 8 OC = Several Generation Canadian NC = New Canadian The figures in Table 6.4.10. suggest that we should not assume that informants whose families have been in Canada several generations will have more Canadian markers than will new Canadians. Generation Gap Our older informants chose 1. chesterfield, 2. blinds, 4. orange, and 5. taps at least 10 percent more frequently than did the younger informants. Differentiation by Sex Our female informants chose 2. blinds, 3. brush, 5. taps, and 6. past at least 10 percent more frequently than did males. 5. Somers' D. Analysis We applied the Somers' D test (a measure of ordinal variation) in order to obtain numerical evidence to substantiate our claim of linguistic and sociological co-variation. Our two age groups and two sex groups fitted easily into the Somers' D format, but we had to dichotomize our linguistic choices into 1) a chosen value and 2) other(s). Further, we had to abandon our socio-economic groupings and dichotomize all informants 283 into two groups, one group consisting of 45 people below the median score of 27.5 points on our Socio-economic Class Index scale and the other group consisting of those 44 people above the median score. The positive sign (+) was arbitrarily designated to the top division of class, to the older division in age, and to females. The magnitude of the number is equal to the difference between the percentage of the people who had a particular usage within one sociological group and the percentage of people who had that same usage in the opposite group. The following table is a presentation of the Somers' D analysis which also serves as a summary of this grammar, pronunciation, and vocabulary study. 284 Table 6.5.1. SOMERS' D RESULTS Grammar Item Question naire Value Social Class Age Sex 1. Between John and Me 263 A +.52093 +.26496 +.11576 2. Just between and me you 264 A +.13372 -.01797 +.09619 3. Eh6 312 B,C +.07420 +.06571 -.00556 4. Eh7 313 B,C +.07197 +.10000 -.01667 5. Fewer 296 A +.17073 +.24265 +.10167 6. Have you got 277 A +.22462 +.09375 -.15655 7. If it were 288 A +.14066 +.42929 +.21303 8.. If you had 287 B +.32558 +.17986 +.05844 9. Lie (Past) 275 B +.29175 +.23242 +.07611 10. Lie (Past Participle) 276 B +.39860 +.27389 +.01107 11. ?not? 744 A +.14257 +.03898 +.08241 12. Past Perfect 315 A +.42381 +.04375 -.07990 13. Sneaked 280 A +.28277 +.55165 +.03911 14. Subject verb non-agree s 704 A +.33418 -.05249 -.07078 15. Take 318 A +.09756 +.00625 +.03936 Total = 15(+) i of 15 13(+) of 15 10(+) of Pronunciation Question- Social Item naire Value Class Age Sex 1, Africa 53 A +.08140 +.15220 f .06032 2. Again 146 A -.01085 -.01923 +.01085 3. Anti- 195 A -.02532 -.04487 +.07080 4. Apricots 65 B +.05074 +.12091 -.06032 5. Asphalt 123 A +.23256 +.39664 -.09302 6. Aunt 79 B,C -.01912 +.05342 +.11008 7. Balcony 234 B +.11473 +.00855 +.11266 8. Been 181 A +.01480 -.35989 -.10677 285 Pronunciation Item Question naire Value Social Class Age Sex 9. Blous e 88 A +.04238 +.02137 +.13953 10. Caramel 905 B +.12338 +.26556 +.01623 11. Catching 530 A +.09302 +.09667 +.01301 12. Congratulate 191 A +.22377 +.36966 +.00362 13. Decal 222 A -.08527 -.35256 -.18760 14. Egg 55 B +.00517 -.00427 +.08579 15. Either 179 A +.15297 +.08974 +.21085 16. February 903 c +.19712 +.20430 +.17241 17. Fertile 196 A -.21447 -.38675 +.03256 18. Futile 215 A -.00264 -.09451 +.04863 19. Film 237 A +.10078 -.13034 +.08114 20. Garage 901 A +.17106 +.07479 +.01085 21. Generally 233 D +.14444 +.00544 -.05238 22. Genuine 157 A +.18243 +.21581 +.04599 23. Hundred 576 A -.01012 -.08194 +.27273 24. Khaki 52 A +.27076 +.54025 +.12957 25. Library 84 A +.07132 -.35043 -.07132 26. Lieutenant 147 A +.23651 +.45915 +.13175 27. Luxury 129 B +.08786 +.02137 +.09406 28. Mirror 41 A +.19365 +.14379 -.05556 29. Missile 87 A +.03896 -.08683 -.20930 30. Morning 440 A +.09302 +.17111 +.13875 31. Multi- 223 A -.07183 -.10043 +.02636 32. Naturally 190 D +.00476 +.12091 +.13319 33. Ottawa 907 B +.12041 +.03205 +.06150 34. Pictures 36 C +.03805 +.00330 +.05391 35. Potato 369 A +.06977 -.12889 +.23902 36. Recognize 140 A +.01912 -.05342 -.01912 286 Item 37. Route 38. Sandwiches 39. Schedule 40. Semi-41. Sentence 42. Something 43. Toronto 44. Temperature 45. Vase 46. Weren't 47. Wilson 48. Zebra Vocabulary 1. Chesterfield 2. Blinds 3. Brush 4. Orange 5. Taps 6. Quart Past 7. Railway Crossing 8. Z Question- Social naire Value Class 156 A +.08475 902 A +.14167 125 A +.05891 67 A -.05238 119 A +.27649 904 A +.11746 906 A +.14746 158 B +.05761 37 C -.00775 167 B +.00635 339 A -.00000 48 A -.07082 Total: 36(+) of 48 Age Sex -.05128 +.09716 +.21569 +.13178 +.25214 +.07752 -.08007 +.09408 -.04060 +.17829 -.10621 +.02063 +.21593 +.24762 -.00980 +.12632 -.22863 -.26512 +.03268 +.03968 -.11111 -.04228 -.15220 +.02857 26(+) of 48 35(+) of 48 42 B +.13488 +.38675 +.09251 51 A +.00866 +.28852 +.13095 63 A +.04444 -.10294 +.13968 322 A +.05371 +.10286 -.01996 80 A -.04083 +.15598 +.13178 103 A +.32497 +.04592 +.18132 68 A +.07028 +.01709 -.07028 108 A +.06150 -.07906 +.02946 Total: 7(+) of 8 6(+) of 8 6(+) of Grand Total: 58(+) of 71 45(+) of 71 50(+) of 71 287 6. Comparison with SCE Every survey design has inherent advantages and disadvantages. The major advantage with a written questionnaire and postal survey such as the Survey of Canadian English is that it can reach thousands of people rather than the few informants we were able to recruit and interview. The major disadvantage with a written questionnaire is that the informant is asked to make sophisticated linguistic judgements and to evaluate much of his own subconscious speech. The SCE had an additional disadvantage in that it was directly associated with the school system and teachers of English and consequently many students and their parents undoubtedly reacted to it as a test, choosing what they believed to be "correct" rather than indicating their natural usage.. These two facts, that it was written and that it was from the schools, would lead one to expect that the answers would only be of one style and that that style would be roughly equivalent to our Minimal Pairs style. The Ottawa Survey, despite its relatively few informants, has a much wider range of linguistic usage than the SCE because it has an almost full range of socio-economic classes and frequently a full range of linguistic styles. This difference and the differences between the sex and age groups make it difficult to compare the two surveys. If, however, we were to assume that the SCE data represented the average of all socio-economic classes and that the style of answer were somewhere near our style of Minimal Pairs and Word List, we would see that our data are quite similar. The difficult aspect in comparing the two sur veys is in trying to compare the conclusions. For although there is an excellent comment after each individual linguistic item in the SCE, there are almost no summarizing remarks about major trends and patterns in Canadian English usage. One of the reasons for the lack of such 288 summaries may be that the SCE ehose early tc exclude socio-economic parameters. The survey was analysed according to provincial usage, and except for Atlantic Canada, very little was forthcoming as to what was interesting and different in Canadian English usage by region. English Canada is after all one of the largest homogeneous linguistic areas of the world. The conclusions of the SCE regarding age.group differen tiation also suggest problems in summarizing the data; for example the statement, "...youngsters are preserving forms of English which their parents either are allowing to disappear or, in some instances are not using at all," is based mainly on the use of riz instead of raised, to  home instead of at home, yelk instead of yolk and deaf to rhyme with leaf. The fact that these choices are listed and present on the questionnaire form encourages some bookkeeping errors, some prankish activity, and some multiple-choice random answering. Male students unsurprisingly always had the highest frequencies for these choices. In Ottawa, we found little evidence to substantiate these claims of the SCE, either for those specific items, which we randomly checked, or for such a general trend. We do, however, agree with the SCE remarks about girl students preferring more traditional or conservative forms, and that American English is exerting an influence on Canadian English. 289 Chapter 6: Footnotes "''M.H. Scargill and H.J. Warkentyne, "The Survey of Canadian English: A Report," English Quarterly, Vol.5, No.3 (1972), pp.47-104. 2 M.H. Scargill, Modern Canadian English Usage: Linguistic Change  and Reconstruction, (Toronto: McClelland and Stewart, 1974), pp.1-143. When referring to this book, we will cite CEU. 3 Because we desired to analyse the age and sex co-variations simul taneously with socio-economic co-variation and because we wanted no fewer than eleven persons in each cell, we found it necessary to converge our five arbitrary classes used in Chapter 5 to three and to separate age and sex. Below is Table 4.3.11, the matrix which is employed throughout this grammar, pronunciation, and vocabulary study of Chapter 6. Table 4.3.11 AGE/CONVERGED CLASS: SEX/ CONVERGED CLASS (OTTAWA) >40 <40 F M L WK LM 12 21 17 16 Mid Mid 11 20 17 14 UM LU 14 11 12 13 Total 37 52 46 43 This matrix is utilized throughout Chapter 6. Op.cit., SCE and CEU and: W.S. Avis, "Speech Differences Along the Ontario-United States Border: I Vocabulary," JCLA, Vol.1, No.l, (October 1954), pp.13-18. , "Speech Differences Along the Ontario-United States Border: II Grammar and Syntax," JCLA, Vol.1, No.l, (March 1955), pp. 14-19. , "Speech Differences Along the Ontario-United States Border: III Pronunciation," JCLA, Vol.2, No.2, (October 1956), pp. 41-59. 290 R.J. Gregg, "The Linguistic Survey of British Columbia: The Kootenay Region," in Canadian Languages in Their Social Context, ed. Regna Darnell (Edmonton: Linguistic Research, 1973), pp.105-116. The linguistic items will be listed alphabetically within their groups; the questionnaire number is given after the variable title. The question or task is always given as are the values. 7These eight varieties of eh were taken from Deborah J. Gibson's A Thesis on Eh, unpublished Master's thesis, (Vancouver: University of British Columbia, 1976), pp.1-77. Another reference work consulted for this variable was W.S. Avis, "So Eh? is Canadian, Eh?" CJL, Vol.17 (1972), pp.89-104. 8Avis, 1972, Ibid., p.93. 9 Gregg, op .cit, pp.105-110. "^By "Conditional III construction" we mean If + Past Perfect = Past Conditional (would have). "'""'"Ian Pringle and Enoch Padolsky are presently working on a dialect survey of the Ottawa Valley. Both are professors at Carleton University, Ottawa. 12 By Canadianism, we mean any usage which can be used as a factor in distinguishing a speaker of Canadian English from a speaker of Northern American. See Appendix D for a list of such Canadianisms. 13 R.J. Gregg, "Notes on the Pronunciation of Canadian English as Spoken in Vancouver, B.C.," JCLA, Vol.3, No.l, (October 1957), p.23. 14 Other words subject to devoicing and heard during the survey were valued [veeljut ], valid, [valet], and Howard [haowert]. "^Avis, 1956, op .cit., p.44. 16 Avis, 1956, op.cit., p.47. 17"Topnotch book on Canadian English," The Citizen, April 14, 1979, Ottawa, p.40. The quote is: "On p.29, a comment asks 'How many of these words do you differentiate in speech: where, wear, were?' My own answer is two: the first two sound the same. Canadians have the second and third sounding the same; others have all three distinct." "^Personal discussion. 291 19 Avis, 1956, pp.44-45. 20 Gregg, 1973, pp.112-113. \ly son's day-care centre teaches the words red, orange, and green on a safety poster. Orange construction paper is used for the middle light. Data from the Ottawa Valley Survey on this variable should prove to be interesting. 22 Of course, the age and sex categories presented here and below are not mutually exclusive, e.g. the 0 group consists of both M and F, M is made up of both 0 and Y, and so on. We present the four age and sex groups in rank order so that we can display the relative importance of each sociological indicator. 23 Gregg, 1973, op.cit. and J. Poison, A Linguistic Questionnaire  for British Columbia, unpublished M.A. thesis, (Vancouver: University of British Columbia, 1969). 24 We would wish to see many more data before we would consider categorizing this form as prestigious. Op.cit., SCE and CEU. 292 CHAPTER 7 CONCLUSION 1. Methodology In this dissertation, we have attempted to show to what extent there is a co-variation of linguistic items on the one hand and socio logical and stylistic parameters on the other in the English spoken by native Ottawans. In order to be able to demonstrate such correlations of linguistic variation with sociological and stylistic variation, we wrote a questionnaire and conducted a survey of 100 informants, 89 native anglophone Ottawans and eleven native anglophones from urban centres in the Ottawa Valley. The informants were chosen at random according to their residence in particular census tracts with a view to filling the various cells within our sociological matrix. The questionnaire, Appendix A of this dissertation, contained 752 variables including items which focussed on phonological segments, mor phology and syntax, word pronunciation, and vocabulary. Additional items focussed on language attitudes and para-linguistic items such as stuttering, swearing, ingressive speech, and hesitation phenomena. The first section of the questionnaire dealt with the informant's background and allowed the informant to talk about himself and to become partially at ease with the interview situation. The answers to these background questions later allowed us to determine the sociological status of the informant. The socio-economic status of an informant was determined by 293 the equal representation of seven indicators: occupation, father's occupation, income, education, spouse's education, house/apartment value, and residence location. Informants were judged to have new Canadian backgrounds if they or their parents came from another country. Informants were categorized as having rural backgrounds if they or their parents came from the Ottawa Valley or other rural areas. All infor mants were assigned an ethnic origin; these ethnic origins included England, Scotland, Ireland, the United States of America, French Canada, and any* other nation claimed by the informants. Tasks and Styles Included in the questionnaire were different tasks to perform. These tasks included reading minimal pairs of words, reading word lists, naming pictures, and reading a passage. It was hypothesized that we would elicit different styles of speech from the informants while they were performing these tasks. It was further hypothesized that we would elicit the most formal and careful speech from an informant while he was reading minimal pairs; that the second most formal and careful style would be elicited while the informant was reading word lists; and that the least formal style among non-continuous speech would be elicited while the informant was identifying pictures. The style elicited from the informant while reading the reading passage about adolescents in a family setting was hypothesized to be less formal than that for the non-continuous speech styles. In addition to the four closely controlled tasks mentioned above, the questionnaire was designed so as to contain opportunities for the 294 informant to speak freely about topics which he knew intimately, e.g. his youth, his opinions of Ottawa, his closest encounter with death and danger, his trips and vacations, his favourite jokes and anecdotes, a short story through pictures, and conversations during breaks in the interview. This task, entitled Free Speech, was hypothesized to elicit a less formal style than would be elicited from the four previously mentioned tasks. 2. The Co-variation of Phonological Variables  with Sociological arid Stylistic Parameters The linguistic items which were investigated and the data therefrom were grouped into two major parts. The first part, Chapter Five of this dissertation, deals with the correlation of phonological variation on the one hand and stylistic and sociological variation on the other. The second part is concerned with the co-variation of sociological factors and grammatical, pronunciation, and vocabulary items. The first part derives its format and methodology essentially from William Labov's The Social Stratification of English in New York City (1966)1 and Peter Trudgill's The Social Differentiation of English in 2 Norwich (1974). In order to make our study relevant to Canadian English, we made modifications to the above works on the basis of the findings from the Pilot Project of the Vancouver Urban Dialectology Survey.^ Our study investigated 27 phonological variables and endeavored to elicit from each informant at least three realizations of each variable in each of our five tasks/styles. 295 The Phonological Variables The phonological variables in this part of the study were chosen because they were suspected of having stylistic or sociological varia tion—the major purpose and hypothesis of this survey—or because they were characteristic Canadian phonological items which we wished to investigate further. The items were: 1. VtV (medial t), 2. ntV, 3. -ing, 4. tj, dj, nj (palatal glide), 5. rV (r metathesis), 6. st, 7. h, 8. V##V, 9. d##y, t##y, 10. ou (the Canadian diphthong [AU]), 11. outV (the Cana dian diphthong [AU] followed by medial t), 12. T (the Canadian diphthong [oi], 13. TtV (the Canadian diphthong [ai] followed by medial t), 14. un-(negative prefix), 15. nd, 16. aer,.eel, 17. Vr->ar, 18. ae, 19. hw, 20. -kt, -pt, 21. x>, 22. th, 23. or, 24. going to, 25. milk, 26. good, 27. tomato. Some important arid heretofore unreported phonological findings made during the course of the study are outlined below: 1. The medial _t rule which recognizes that /t/ when in intervocalic and post-tonic position can be realized as [d] is expanded by speakers of English in Ottawa and elsewhere in Canada, allowing the It/ to be preceded by the frictionless continuant [r] (as in party) , the lateral [i] (as in filter), the voiceless fricatives [f] (as in ' after), [s] (as in sister), and [J] (as in washed our), and occasionally by the nasal [n] (as in ninety) and the plosive [k] (as in picture). Our last example demonstrates that the expanded medial t rule applies to the affricate tj as well. 2. The suffix -ing was most often pronounced [in] and not ILQ] or [an]. To date, linguistic literature has assumed only the last two pos sibilities. Nouns such as building, morning, and evening have a 296 lower frequency of [in] than dp verb forms such as doing, playing, and running. 3. The Canadian diphthongs ou and T are pronounced [AU] and [ai] in positions other than immediately before voiceless consonants, namely before voiced consonants and occasionally in word final position. 4. The auxiliary verb form going to was most often pronounced [gointa] while the second most frequent was [gene]. The pronunciation [gana] was very infrequent. This fact has profound implications for ESL as most American texts teach [gana].. 5. An open variant of /se/, transcribed [a],was frequently elicited from females, especially when they pronounced a limited set of words consisting of that (when stressed), grass, glass, pass and class. Sociological Findings From our data, we observed that 20 items demonstrated phonological variation directly related to the degree of formality of the task per formed by the informants. These items were: VtV = VtV, ntV = nt, -ing = in, tj, dj, nj = tj, dj , nj , rV = rV, st = st, h = h, d#y = d#y, out = Aut, it = ait, nd = nd, ssr = er, Vr = ar, hw = hw, kt = kt, v = x>, th = th, going to = goto 1, milk = milk, and tomato = tam[ajto. These data forcefully prove our hypothesis of linguistic and stylistic co-variation. Our data also revealed ordered socio-economic stratification for the following values: VtV = VtV, -ing = -i<j, t j , d j , nj = t j, d j , nj., rV = rV, st = st, out = Aut, Tt = ait, nd = nd, hw = hw and th = th. 297 Although- there were not as many values demonstrating socio-economic stratification as there were demonstrating stylistic variation, our data presents strong and indisputable evidence of linguistic and socio-economic co-variation. In addition to such socio-economic stratification, it was clear to see that informants higher in the social structure use a much wider range of styles than do those informants lower in the social structure. Our data also provided forceful illustrations of those values which patterned as stigmatized forms. The following values were designated stigmatized forms: -ing = an, rV = Vr, st = s, V#V = a#V, un = an, and th = n, z, t and d, going to = goto 2 and milk = melk. Further, our data allowed us to compare the usage patterns of the four sex/age groups. We found that females over forty years of age stood out from the other sex/age groups by their tendency to pronounce more frequently the formal values of our variables and by their main taining the strongest class ordering. At the opposite end of the scale were the males under forty years of age. They pronounced the informal values of the variables more frequently and the prestigious forms less frequently than did any other group. Males over forty years of age and females under forty were between the two extremes. The older males had more rural usages and the younger females conformed more to present norms than did the other groups. Although we are somewhat disconcerted that our data confirm several stereotypes, we are pleased that at last we have hard data from which we can talk about Canadian usage. 298 The Manhattan, Norwich, and Ottawa Surveys Compared Included in the Ottawa Survey but absent from both the Manhattan and Norwich studies was the task of naming objects from pictures. The Ottawa Survey elicited 83 variables by means of pictures. This method was found to be fast and effective, with little need for supplementary questions. These pictures drew attention away from the linguistic interview situation and helped to convince the informant that we were interested in what he called things. Some pictures in isolation were somewhat difficult to identify; in these instances we were able to record quite unguarded speech. The data demonstrate that our task/style labelled Pictures was the most informal of the non-continuous speech styles, and in some cases this task was more informal than our Reading Style. A second, new technique was employed involving pictures; we asked each informant to tell a story following a sequence of pictures.""" Whili the informant was telling the story, he would be asked to speak the roles of the scolding man and the retorting boy. The channelling cues observed during this activity indicate that we were eliciting very unguarded speech.^ The major advantage of this technique was that it afforded us some structure in the Free Speech style, i.e. it allowed us to record many of the same lexical items from informant to informant while they were generating their own continuous speech. We found the methodology pioneered by Labov and Trudgill fully adaptable to the Canadian situation. 299 Socio-economic Scope In both the Manhattan and the Norwich Surveys, a full range of ; society was not represented. For example, the Norwich Survey ranged in seven groups from less than $1,200.00 annual income to more than $5,000.00 annual income,7 a very small range; our Ottawa Survey, in contrast, differentiated people on a range from basic welfare ($2,000.00.. annual income) to people of the lower upper class (more than $35,000.00 annual salary). The Manhattan Survey informants were also narrowly restricted in annual income, ranging from $2,000.00 to g $5,000.00 income per year. This range did not include the upper middle, lower upper, or the upper upper classes when the survey was conducted in 1964. The Manhattan Survey was also restricted to a very small and unique linguistic area, the Lower East Side, of a linguistic ally unique city. Labov, a few years after his survey when writing about social mobility and the Lower East Side, states: It [the Lower East Side and New York City in general] is a port of entry for immigrants and a place o,f nurture for those on the way up, but normally not a permanent home for children of upper middle class parents.^ This statement makes it clear that one should not expect to hear much upper middle class speech in this area. Furthermore, the Manhattan Survey selected its informants from a previous survey, the Mobilization for Youth project, which investigated juvenile delinquency in the inner-core of the city. This is not a full range of society. May I hasten to add that these clarifications are not intended as adverse criticism of either survey sociologically or linguisticallyj they are meant to indicate a difference in aim and scope and to dissuade anyone from believing that either the Norwich or Manhattan Surveys are 300 surveys of full representative samples of mainstream society in their respective countries. The direct comparison of data among the three surveys is not possible because the linguistic items were not identical. However, the Manhattan and Ottawa Surveys had two phonological variables, _th and -ing, which covered largely the same ground. The Ottawa Survey revealed that all informants pronounced th as [0] or [5] over 98 percent of the time in the first four tasks/styles and over 85 percent of the time in Free Speech, and that the most frequent substitutions were [n] and [z], both cases of progressive contextual assimilation, e.g. and then [en nen] and is that [iz zaet]. Both these cases are natural to native speakers of English. The Manhattan data, on the other hand, reveal an average of 93 percent for [0] or [3] intthe most formal style and an average of 53 percent for [0] and [ 0.Q in the most informal style. These data demonstrate that most informants from the Lower East Side have not been fully assimilated into English speech norms, and that, consequently, the data for the Manhattan Survey cannot claim to be representative of main stream American speech on a linguistic basis, either. The second phonological variable which we can compare in both the Ottawa and Manhattan Surveys is -ing. Our Ottawa informants averaged over 95 percent for [in] in the first four styles and 84 percent in Free Speech, while the Manhattan informants averaged 91 percent in the most formal style but only 55 percent in the most informal style. This, too, is a large difference in usage patterns, but it.seems plausible and representative of Ottawan and New York City speech. Our informal comparative surveys of radio programs and public transportation speech in both cities tend to confirm the difference. 301 The final difference in the Ottawa and Manhattan data which we wish to bring to the fore here is with reference to the Reading and Free Speech tasks/styles. Labov writes: A few upper middle class speakers seemed to have the degree of control and self-awareness needed to modi fy their reading style in the direction of conver sational style, but this is a rare effect and not a very large one.^-0 Our data, on the other hand, demonstrate a definite and recurring pattern of the phenomenon. Our lower upper class informants read the variables VtV = VtV, d#y = d#y, and rV = rV more informally in the reading passage than they pronounced them in their own Free Speech style. Similarly, the upper middle class read the variable rV = rV, Tt = ait, nd = nd, hw = hw, kt = kt, and going to = gotoj more infor mally than they pronounced them in their own Free Speech style. These two classes demonstrated a feeling for the roles they were reading, for timing and hesitation phenomena and even linguistic and phonological style. As a consequence of this, they read more slowly than did the middle class. The fact that the Lower East Side informants produced only rare instances of this phenomenon would tend to substantiate our claim that the Lower East Side Manhattan Survey is a survey of a truncated portion of American society which did not include the upper middle, lower upper, or upper upper classes. 3. The Co-Variation of Grammatical, Pronunciation, and  Vocabulary Variables with Sociological Parameters The. others maj.or'-part -of ~ this .'thesis,, •• found in Chapter 6, deals with the correlation of grammatical, pronunciation, and vocabulary variation 302 on the one hand and sociological variation on the other. This part derives much of its format and many of its 71 linguistic items from M.H. Scargill's and H.J. Warkentyne's The Survey of Canadian English"'""'" and W. Avis' three articles "Speech Differences along the Ontario-12 United States Border." We made a number of adjustments in order to focus on urban socio-dialectology, and we added a few items in order to investigate local Ottawa and Ottawa Valley usage. All the sociological parameters made use of in the phonological study were also utilized in this study; these included: socio-economic class, age, sex, urban/rural background, ethnic background, and new Canadian/several generation Canadian background. The inclusion of the parameter of socio-economic class makes this study a major innovation in the field of Canadian English dialectology. There was no systematic attempt to repeat the 71 items in the five tasks, and consequently no analysis of stylistic co-variation was possible. Chapter 6 is divided into three sections which deal uniquely with grammar and syntax, word pronunciation, and vocabulary items respectively. Grammar and Syntax The grammatical and syntactic variables in this section are subject to the greatest degree of socio-economic co-variation of any of our items in this entire survey. Thirteen of the 15 demonstrate clear differentiation and progressive ordering; these are: 1. between John 6 7 and me; 3. Eh ; 4. Eh ; 5. Fewer; 6. Have you got; 8. If you had; 9. Lie, Lay; 10. Lie, Lay, Lain; 11. ?not?; 12. Past Perfect; 13 Sneaked; 14. Subject/verb agreement; and 15. Take. The two other items demon strated only partial co-variation. We analysed our data according to 303 age and sex groups and found that the informants over 40 years of age had the highest frequency of prescribed values. Females, males, and informants under 40 years of age followed in rank order. A separate analysis was run in order to determine which items were treated dramat ically differently by the two age groups. Word Pronunciation In our phonological study in Chapter 5, we analysed which allophones of given phonemes our informants pronounced with reference to their social class, age, and sex and to the style of speech. In this section we analysed which phonemes and which sequence of phonemes were uttered within words and phrases in correlation to the full gamut of sociological parameters. Most of the words in this section were taken from traditional Canadian English dialect studies and some were found not to be fully suitable for an urban socio-dialect study of Canadian English. Our data, nevertheless, demonstrate that the following pronunciation values have ordered co-variation with socio-economic class: 5. [aesfnlt], 6. [tint], and[ant], 12. [ kengrat fa I e i t ], 14. [eg], 15.•[ai5ar], 16. [februeri], 19. [film], 21. [d^enarlli], 22. '• [ d^njuari*}, 24. [kdrki], 26. [ I ef tenant ], 27. £ I Xkjari ], 28. [ iru'rar ], 30. [mornin], 33. [ctawn], 34. [pikd^arz], 35. [pateido], 38. [saendwitJaz], 45. [veiz], and 46. [weYnt]. Three other values typical of American usage demonstrate socio-economic stratification but in a reverse order to that which was hypothesized; they are: 3. [eentai-], 31.''[:m'X I tai-] and 40. [s£mai-]; In addition to the above listed values which varied according to the socio-economic status of the informant, we were able to find 17 values 304 which patterned as stigmatized forms, 14 values which patterned as prestige forms, 21 values which had usage patterns peculiar to infor mants over 40 years of age, and 11 values which had usage patterns peculiar to informants under 40 years of age. These forms are pre sented in the lists below: Stigmatized Forms The following values have been categorized as probable stigmatized forms: 5. [33zfa.lt], 8. [ben], 12. [kangrad^aIeit], 16. [febjueri], 19. [filam], 20. [greedy], and [grad^], .21. [d^enerli], 22. [ d^nj ua in ], 23. [hAn(d)art], 25. [laiberi], 28. [mir], [miral] and [mirou], 30. [mdrnan], 35. [pedeido], 38. [ sasm(w) it J ] and [ saenwit J ], 41. [se7ns], 42. [sAm?n], and 43. [t(a)rdno]. In addition, forms 36. [re*kanaiz], and 37. [ rAut ] have been categorized as possible stigmatized forms. Prestige Forms The following values stand out as prestige pronunciations: 6. [x>nt] and [ant ], 7. [ bael kani ], 15. [aiSar], 21. [ d^nar I I i ], 24. [kdrki]and: . [kdki], 26. [leftenant], 32. [ nart Jer lli], 33. [ntewn], 39. [Je^ul], 43. [terunto], and 45. [VDZ] and [vaz]. Forms Used Chiefly by Older Informants 1. [aefraka], 4. [eiprakMs], 5. [abfT>11 ] and [azfclt], 8. [ben], 10. [kasramel], 15. [ai6ar], 19. [ fLI am],.20. [garee^], 24. [kdrki], 25. [laibri], 26. [ I ef tenant ], 30. [mornin], 31. [mAlta-.], 35. [pedeido], 305 38. [saendwitjaz], 39. [jSd^ul], 40. [sdma-], 45. [.VDZ], 47. [wiltsan], and 48. [zifbra]. Forms Used Chiefly by Younger Informants 5. [sJfDlt], 8. [bin], 12. [ kangreed^al eit ], 13. [dekal ], 16. [fe'bjueri ], 17. [fartail], 22. [d^njuain], 28. [nur], 32. [naet Jar I i ], 43. [t(a)rdno], and 45. [veiz]. Sex and Age Groups In order to determine the performance characteristics of the sex and age groups, we cross-analysed the 20 variables which reveal clear socio-economic differentiation, the 17 stigmatized forms, and the 14 prestige forms. Our data revealed that the female informants main tained the strongest class hierarchy among themselves, had the lowest frequency of stigmatized forms, and had the highest frequency of pres tige forms. Males, on the other hand, placed last in maintaining a class hierarchy through linguistic means, had the highest frequency of stigmatized forms, and had the lowest frequency of prestige forms. Our informants under 40 years of age placed second for the least amount of class hierarchy, had the second lowest frequency of stigmatized forms, and placed third for most prestige forms. The informants over 40 years of age were second in socio-economic stratification, second in frequency of prestige forms and third in fewest stigmatized forms. See Tables 6.4.5. to 6.4.7. 306 Vocabulary The goal in this section of the study was to see what particular word the informant would choose when asked to identify an object. Each variable contained a traditional Canadian marker, a Northern American variant, and frequently a British variant. Only the items 1. chesterfield and 6. quarter past had clear and ordered socio-economic stratification. The sociolinguistically interesting aspect of this section was to determine to what extent each sociological group retained the Canadian vocabulary items. Our data revealed that the use of these Canadian vocabulary items was very much related to class. The two upper classes, upper middle and lower upper, had the highest frequency of the Canadian lexical markers, followed by the middle class, and then followed by the lower classes. The retention of Canadian vocabulary items was also investigated by age and sex group. Here, our data demonstrated that the informants over 40 years of age had the highest frequency of Canadian words followed by females, males, and finally the younger informants. This suggests to us that Canadian items are losing ground as we see that the American options are gaining in popularity. A com-13 parison of Avis' items and frequencies over a quarter of a century ago compared with our data gives us a time perspective and reinforces the evidence that Canadianisms are losing ground to American usage and to usage patterns which are gaining in the rest of the English speaking world. Comparison with the Survey of Canadian English There are two major differences between this survey and the SCE. The first difference is in the mode of eliciting the linguistic material. 307 The SCE used a written questionnaire which required that the informant make sophisticated linguistic judgements about his own usage. This self-14 evaluative procedure led to problematic results for several items. Our survey, on the other hand, elicited the linguistic material by means of a recorded interview which required the informant to perform several different tasks. These tasks often distracted the informant from the linguistic interview situation, and informants frequently produced unguarded speech. We were able to observe, transcribe, and analyse this conscious and subconscious speech. The disadvantages of our method is, of course, the low number of interviews which were possible. The other major difference between the SCE and our survey is that we incorporated socio-economic and stylistic parameters in our study. We have demonstrated that a great deal of linguistic variation is directly correlated to these two parameters, and that, consequently, future and past dialect studies which do not include these parameters suffer a considerable handicap. In summation, we did that which we set out to do; we demonstrated that there was a great deal of variation in Canadian English within one city; we demonstrated that that variation was directly correlated to the speakers sociological status and to the immediate social con text, I.e. the linguistic task. We gained a greater knowledge of the usage patterns and variants of linguistic items. And, finally, we were able to compare the linguistic roles that our various sociological groups were socialized to play. We look forward to the availability of comparable data from other cities and regions from within Canada and without, so that we may more consciously understand the language which we speak and manipulate and which undoubtedly manipulates us. 308 Chapter 7: Footnotes William Labov, The Social Stratification of English in New York  City, (Washington, D.C.: Center for Applied Linguistics, 1966), pp.l-655. Peter Trudgill, The Social Differentiation of English in Norwich, (Cambridge: Cambridge University Press, 1974), pp.1-211. 3 R.J. Gregg, "Urban Dialectology: a pilot survey of the English spoken in the city of Vancouver, B.C. (1976-77)," unpublished paper given at the 1977 Learned Societies Conference in Fredericton, N.B. (Vancouver: University of British Columbia, 1977), pp.1-9; another source from this pilot project is Margaret. Murdoch, "Reading Passages and Informal Speech," unpublished paper given at the Third International Dialectology Methods Conference in London, Ontario in 1978 (Vancouver: University of British Columbia, 1978), pp.1-7. 4 See A.M. Kinloch, "The Use of Pictures in Elicitation," American  Speech, No.46 (T971), pp.38-46, for further discussion on this topic. "The sequence of pictures was taken from P.R. Hawkins, Social Class, the Nominal Group and Verbal Strategies, (London: Routledge and Kegan Paul, 1977), pp.56-57. For a description of five channel cues related to casual speech, namely a change in tempo, a change in pitch range, a change in volume or rate of breathing, and laughter, see Labov, 1966, op.cit., pp.109-112. 7Trudgill, 1974, op.cit., p.39. O Labov, 1977, op.cit., pp.214-219. 9 William Labov, "The Effect of Social Mobility on Linguistic Behavior," in A Various Language, ed. J.V. Williamson and V.M. Burke, (New York: Holt, Rinehart and Winston, 1971), p.649 and footnote 18. Another source for what classes are to be found in the Lower East Side is Labov, 1966, op.cit., p.219, where in Labov writes: "As far as the upper class is concerned, we would not expect to find representatives of this group living on the Lower East Side,..." Labov's upper class may include any informant above low middle class status in the Ottawa Survey. 309 Labov, 1966, op.cit., pp.96 and 132. '''"'"M.H. Scargill and H.J. Warkentyne, "The Survey of Canadian English: a Report," English Quarterly, Vol.5, No.3, (Toronto: Canadian Council of Teachers of English, 1972), pp.47-104. See also M.H.Scargill, Modern Canadian English Usage: Linguistic Change and  Reconstruction, (Toronto: McClelland and Stewart, 1974), pp.1-143. "'"^W.S. Avis, "Speech Differences along the Ontario-United States Border: I Vocabulary," JCLA, Vol.1, No.l, (October 1954), pp.13-18; W.S. Avis, "Speech Differences along the Ontario-United States Border: II Grammar and Syntax," JCLA, Vol.1, No.2, (March 1955), pp.14-19; W.S. Avis, "Speech Differences along the Ontario-United States Border: III Pronunciation," JCLA, Vol.2, No.2, (October 1956), pp.41-59. 13 Avis, 1954, 55, 56, Ibid. (Speech Differences). 14 The items most affected by this limitation in the SCE were: 24. eh, 27. vase, 38. eh, 48. tomato, 49. butter, 50. caramel, 51. father, 52. calm, 54. squirrel, 59. again, 78. cot and caught, 83. film, 101. congratulate, 106. whine and wine, and 110. guarantee. "'""'The SCE had the intention of evaluating linguistic variation with reference to level of education attained, one of the indicators in our Socio-economic Class Index, but did not follow through when it was found that this parameter was very sensitive. See H.J. Warkentyne, "Contemporary Canadian English: A Report of the Survey of Canadian English," American  Speech, Vol.46 (1971), p.94. 310 BIBLIOGRAPHY Avis, W.S. 1954. "Speech Differences Along the Ontario-United States Border, I Vocabulary." The Journal of the Canadian Linguistics  Society 1:1, pp.13-18. . 1955. "Speech Differences Along the Ontario-United States Border, II Grammar and Syntax." The Journal of the Canadian  Linguistics Society 1:1, pp.14-19. . 1955. "Speech Differences Along the Ontario-United States Border, III Pronunciation." The Journal of the Canadian Linguistics  Association 2:2, pp.41-59. . 1967. A Dictionary of Canadianisms on Historical Principles. Toronto: W.G. Gage. . 1972. "So eh? is Canadian, eh?" Canadian Journal of Linguistics 17:2, pp.89-104. . 1973. "The English Language in Canada: A Report." Current Trends in Linguistics 10:1, pp.40-74. , Drysdale, P.D., Gregg, R.J., and Scargill, M.H., eds. 1967. The Senior Dictionary. Toronto: Gage Educational Publishing. Revised and updated, 1973. , and Kinloch, A.M. 1979. Writings on Canadian English 17 29-1975: An Annotated Bibliography. Toronto: Fitzhenry and Whiteside. Backhouse, J.K. 1967. Statistics: an Introduction to Tests of Significance. London: Lowe and Brydone. Barber, Bernard. 1957. Social Stratification. New York: Harcourt, Brace, and World. Bernstein, Basil. 1973. Class, Codes and Control: Volume I Theoretical  Studies Towards a Sociology of Language. 2nd revised edition. London: Routledge and Kegan Paul. . 1975. Class, Codes and Control: Volume 3 Towards a Theory of Education Transmissions. London: Routledge and Kegan Paul. 311 Blishen, B., and McRoberts, Hugh. 1976. "A Revised Socioeconomic Index for Occupations in Canada." Canadian Review of Sociology and  Anthropology 13:1, pp.71-79. Canada Year Book 1976-77. Ottawa: Statistics Canada. Canadian Urban Trends: Metropolitan Perspective. 1977. Toronto: Copp Clark/Government of Canada co-publication. Census Tract Bulletin: Population and Housing Characteristics Ottawa- Hull. 1974. Ottawa: Statistics Canada. Census Tracts, Ottawa-Hull. 1971. Ottawa: Statistics Canada. Chambers, J.K. 1973. "Canadian Raising." Journal of the Canadian  Linguistic Association 18, pp.113-35. . 1974. "The Ottawa Valley 'twang'." Canadian English: Origins and Structures, ed. J.K. Chambers. Toronto: Methuen. - , ed. 1975. Canadian English: Origins and Structures. Toronto: Methuen. Cowan, H.I. 1961. British Emigration to British North America. Toronto: University of Toronto Press. Darnell, Regna, ed. 1971. Linguistic Diversity in Canadian Society. Edmonton: Linguistic Research. , ed. 1973. Canadian Languages in their Social Context. Edmonton: Linguistic Research. Davis, Alva L., McDavid Jr, Raven J., and McDavid, Virginia G., eds. 1969. A Compilation of the Worksheets of the Linguistic Atlas  of the United States and Canada and Associated Projects. 2nd ed. Chicago: University of Chicago Press. Gardner, Robert C, and Lambert, Wallace E. 1972. Attitudes and Motivation in Second-Language Learning. Rowley Newbury House Publishers. Gibson, Deborah Jean. 1976. A Thesis on Eh. Unpublished M.A. Thesis. Vancouver: University of British Columbia. Gregg, Robert J. 1957. "Notes on the Pronunciation of Canadian English as Spoken in Vancouver, B.C." Journal of the Canadian  Linguistics Association 3, pp.20-26. . 1957. "Neutralization and Fusion of Vocalic Phonemes in Canadian English (Vancouver)." The Journal of  the Canadian Linguistic Association 3:2, pp.78-83. 312 Gregg, Robert J. 1964. "Scotch-Irish urban speech in Ulster." Ulster dialects: an introductory symposium. Holywood: Ulster Folk Museum. . 1968. "Notes on the Phonology of a Country Antrim Scotch-Irish Dialect." Orbis 7, pp.392-406. 1973. "The Linguistic Survey of British Columbia: The Kootenay Region." Canadian Languages in their Social Context, ed. Regna Darnell. Edmonton: Edmonton Linguistic Research. . 1973. "The diphthongs ai and ai in Scottish, Scotch-Irish and Canadian English." Journal of the Canadian Linguistic  Association 18, pp.136-45. . 1975. "The Phonology of Canadian English as Spoken in the Area of Vancouver, British Columbia." Canadian English: Origins  and Structures, ed. J.K. Chambers. Toronto: Methuen, pp.133-144. , forthcoming. "Canadian English." Varieties of English: Commonwealth English Series, ed. Y. Matsumura. Japan: Manuscript. Hamilton, D.E. 1958. The English Spoken in Montreal: a Pilot Study. Unpublished M.A. thesis. Montreal: University of Montreal. Hawkins, P.R. 1977. Social Class, the Nominal Group and Verbal Strategies. London: Routledge and Kegan Paul. Joos, Martin. 1942. "A Phonological Dilemma in Canadian English." Language 18, pp.141-44. Kinloch, M.A. 1971. "The Use of Pictures in Elicitation." American  Speech 46, pp.38-46. . 1973. "The Survey of Canadian English: A First Look at New Brunswick Results." English Quarterly 5:4, pp.41-51. Kurath, Hans. 1949. A Word Geography of the Eastern United States. Ann Arbor: University of Michigan Press. Labov, William. 1966. The Social Stratification of English in New York  City. Washington, D.C: Center for Applied Linguistics. . 1966. "The Effect of Social Mobility on Linguistic Behavior." Sociological Inquiry 36:2, pp.186-203. . 1969. The Study of Nonstandard English. Washington, D.C: Center for Applied Linguistics. 1970. "The Study of Language in its Social Context." Studium Generale 23, pp.30-87. ••• • . 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. 313 Lambert, W.E., Hodgson, R.C., Gardner, R.G., and Fillenbaum, S. 1960. "Evaluational Reactions to Spoken Languages." Journal of Abnormal  and Social Psychology 60, pp.44-51. , Frankel, H., and Tucker, G.R. 1966. "Judging Personality through Speech: A French-Canadian Example." The Journal of  Communication 16, pp.305-321. Macaulay, R.K.S. 1977. Language, Social Class, and Education: A Glasgow  Study. Edinburgh: Edinburgh University Press. Marckwardt, A.H. 1971. "Principal and Subsidiary Dialect Areas in the North Central States." Readings in American Dialectology, ed. H.B. Allen and G.N. Underwood. New York: Appleton-Century-Crofts, pp.74-82. McConnell, R.E. 1979. Our Own Voice: Canadian English and how it is  Studied. Toronto: Gage Educational Publishing. Newman, Peter C. 1975. The Canadian Establishment. Toronto: McClelland and Stewart. Nie, Norman H., and Hull, Hadlan, C. 1975. Statistical Package for the  Social Sciences. Toronto: McGraw-Hill. Orkin, Mark M. 1970. Speaking Canadian." English. Don Mills: General Publishing. Padolsky, E., and Pringle, I. 1977. "Reflexes of M.E. Vowels before /r/ in Ottawa Valley Dialects of Hiberno-English Types." Unpublished paper read at the Learned Societies Conference at Fredericton, N.B. Ottawa: Carleton University. Pellowe, John, Nixon, Graham, Strong, Barbara, and McNeany, Vincent. 1972. "A Dynamic Modelling of Linguistic Variation. The Urban (Tyneside) Linguistic Survey." Lingua 30, pp.1-30. Poison, J. 1969. A Linguistic Questionnaire for British Columbia. Unpublished M.A. thesis. Vancouver: University of British Columbia. Reed, Carroll, E. 1967. Dialects of American English. Cleveland: The World Publishing Company. Rodman, L. Winter 1974/1975. "Characteristics of B.C. English." English  Quarterly 7:4, pp.49-82. Samarin, William J. 1967. Field Linguistics: A Guide to Linguistic Field  Work. New York: Holt, Rinehart and Winston. Sankoff, David, and Sankoff, Gillian. 1973. "Sample Survey Methods and Computer Assisted Analysis in the Study of Grammatical Variation." Canadian Languages in their Social Context, ed. Regna Darnell. Edmonton: Edmonton Linguistic Research. 314 Sankoff, Gillian and Cedergren, Henrietta. 1971. "Some Results of a Sociolinguistic Study of Montreal French." Linguistic Diversity  in Canadian Society, ed. Regna Darnell. Edmonton: Edmonton Linguistic Research. Scargill, M.H. 1974. Modern Canadian English Usage: Linguistic Change  and Reconstruction. Toronto: McClelland and Stewart. . 1977. A Short History of Canadian English. Victoria: Sono Nis Press. , and Warkentyne, H.J. 1972. "The Survey of Canadian English: A Report." The English Quarterly 5:3, pp.47-104. Shuy, Roger W., Wolfram, Walter A., and Riley, William K. 1968. Field  Techniques in an Urban Language Study. Washington, D.C: Center for Applied Linguistics. Stevenson, Roberta C. 1976. The Pronunciation of English in British  Columbia. Unpublished M.A. thesis. Vancouver: University of British Columbia. Trudgill, Peter. 1974. The Social Differentiation of English in Norwich Cambridge: Cambridge University Press. . 1974. Sociolinguistics: An Introduction. Aylesbury, Bucks: Hazell Watson and Winey. . 1975. Accent, Dialect and The School. London: Butler and Tanner. • 1978. Sociolinguistic Patterns in British English. London: Edward Arnold. von Baeyer, Cornelius, 1976. Talking about Canadian English: and  Canadian French as well. Ottawa: Supply and Services Canada. . 1977. The Ancestry of Canadian English. Ottawa: Supply and Services Canada. Warkentyne, H.J. 1973. "Contemporary Canadian English." American Speech 46, pp.193-99. This article was printed in 1974 for the 1971 issue. Williamson, Juanita V., and Burke, Virginia.M., eds. 1971. A Various  Language: Perspective on American Dialects. New York: Holt, Rinehart and Winston. Wolfram, Walter A. 1969. A Sociolinguistic Description of Detroit Negro  Speech. Washington, D.C: Center for Applied Linguistics. '__ , and Fasold, Ralph. 1974. The Study of Social Dialects in American English. Englewood Cliffs: Prentice Hall. 315 Woods, Howard. 1977. Sound Production : h arid the elimination of the  intrusive h. Ottawa: Government of Canada. . 1979. Rhythm and Unstress. 3rd edition. Ottawa: Government of Canada. . 1979. Syllable Stress and Unstress. Ottawa: Government of Canada. APPENDIX A OTTAWA ENGLISH QUESTIONNAIRE APPENDIX A OTTAWA ENGLISH QUESTIONNAIRE  Background Information (I'd like to ask you to tell me some information about your own background and that of your family). 2. Sex 3. What year were you born? 4. Where were you born? 5. (If not born here) At what age did you come to Ottawa? 6. Where else have you lived? 7. For how many years? 8. Do you speak any other languages fluently besides English? (Specify) 9. Where did you father come from? 10. What was his native language? 11. How long has he lived in Canada? 12. How long has he lived in Ottawa? 13. Where did your mother come from? 14. What was her native language? 15. How long has she lived in Canada? 16. How long has she lived in Ottawa? 17. What generation Canadian are you? 18. What ethnic background do you have? 19. (If applicable) Where does you present spouse come from? 20. What was his/her native language? 21. How long has he/she lived in Canada? 22. How long has he/she lived in Ottawa? 23. What is your spouse's occupation? 24. How much education has your spouse had Now think back to when you were twelve years old: 25. Where did you live? 26. What was your father's main occupation 27. What kind of schools did you go to? 28. What grade did you complete? 29. Have you any further education? (Describe) 30. What job do you do now? (Ascertain exactly. If retired or unemployed, ask for last job held). Picture = P (What is this?) 31. (P) orange 32. (P) tomatoes 33. 34. 35. (P) pumpkins 36. (P) pictures 37. (P) vase 38. (P) kite 39. (P) winter 40. (P) night 41. (P) mirror 42. (P) sofa with 43. (P) house 44. (P) 45. 46. 47. 48. (P) 49. (P) 50. (P) 51. (P) 52. (P) 53. (P) 54. (P) 55. (P) 56. (P) 57. (P) 58. (P) 59. (P) 60. (P) 61. (P) 62. (P) 63. (P) 64. (P) 65. (P) 66. 67. (P) 68. CP] 69. 70. (P: p_o_ta_toes zebra film lion blinds khaki cloth Africa tuba an egg South right an apple professor whipped cream centre chalk board brushes pointer apricots semi-circle semi-truck (short f semi-trailer railway crossing lever 320 71. 72. 73. 74. (P) garage roof 75. 76. 77. 78. 79. (P) father grandfather grandmother daughter aunt 80. (P) tap in kitchen sink 81. (P) plural 82. (P) cutlery 83. (P) window 84. (P) library 85. 86. (P) dog on floor. How would you describe what this dog is doing? 87. (P) missile 88. (P) blouse 89. 90. 91. (P) sandwiches 92. 93. (P) February 94. 95. (P) frying pan 96. What is cooking in it? (bacon and eggs) 97. (P) What is the second work day of the week? 98. Is there a difference between a dinner and a supper? 99. (Clock-face A - 2:45) What time is it on this clock? 100. Would you say that time (2:45)in another way? 321 101. (Clock-face B - 11:15) What time is it on this clock? 102. Would you say that time (11:15) 103. in another way? Quarter Past 104. (Clock-face C - 8:30) What time is it on this clock? 105. Would you say that time (8:30) in another way? Please read these letters out. 106. (H) 107. (W) 108. (Z) 109. (P) Group singing 110. 111. What do you call the living room 112. in your house? Living 113. (P) 27 114. (P) newspaper Word List = W (Please read out these words as naturally as you can.) 115. marry 116. writer 117. matter 118. new 119. sentence 120. absurd 121. It shot across a clear sky. 122. calm 123. asphalt 322 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. provincial schedule regular nuclear tube luxury grandfather singing sandwiches that grass greasy milk agriculture recognize tune caught something mouth again lieutenant ration leg berry That's untrue. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. caramel library route genuine temperature columnist cot government merry guarantee quantity February weren't almond lecture They didn't go. vocabulary valued There's lots to eat. International Harvester always shouldn't ei ther favourite 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. been known English nothing Africa student lengthen bike doing naturally congratulate introduce good also anti-pollution fertile Could you stop smoking? resource centre beautiful anyone particular without wheelbarrow 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. dual secretary mature united water futile attempt theatre perfectly He pointed at me. daughter dirty decal multi-national prestige He just left. newspapers which often tomatoes sorry generally balcony Mary Barry film 238. fishing 239. leisure 240. the egg 241. 242. about 243. around 244. j3o_ta_toes 245. 246. 247. 248. garage 249. 250. Grammar = G Fill in the blanks of these sentences: 263. John, Mary and I are sitting in a row. Mary is sitting between John and . 264. Just between you and I think that they're not telling the truth. 265. We really missed a great game last night; we should've 266. He sees some money in the box. How would you make that negative? He 267. I often bring them home with me. Last year I often them home with me. 268. I have often them home with me. 251. pumpkin 252. south 253. unbelievable 254. Toronto 255. 256. 257. 258. Ottawa 259. 260. 261. Alberta 262. Winnipeg 325 269. He's always short of money, 270. isn't he? Liaison. How would you change this sentence, so that you're talking about yourself? I'm always short of money, . 271. There's an egg in the fridge. Please say the same thing about ten eggs. ten eggs in the fridge. 272. I don't bother him. If it's the other way around, starting with "he" and ending with "me", how would you say it? He bother me. 273. We often run into old friends there. Yesterday, we into Jerry. 274. He did it alone; he did it all by 275. He lies in the sun every day. Yesterday, he there for 3 hours. 276. So far today, he has there for 5 hours. 277. If you needed a match what would you ask your friend? a match? 278. Sam was at the Air Show when the plane crashed; he must've it happen. 279. She dives into the pool a lot. Yesteday, she from the 12 M board. 280. They sneak into the movie theatre. Yesteday, they into the movie theatre. 326 281. The team gives blood to the Red Cross. Yesterday, they 27 pints. 282. Sometimes we see bears along this highway. Yesterday, we several bears. 283. The police often come down here. Yesterday, four of them all at once. 284. You're looking for your gloves; you ask your friend, " my gloves?" 285. He drinks 3 glasses of milk. Yesterday, he 3 glasses of milk. 286. So far today, he 3 glasses of milk. 287. We would've helped you if you asked us. 288. They would go for a walk if it warmer. 289. I would vote against it if I you. 290. Can guys tell me where the post office is? 291. You don't have to prove that; it's already been . 292. Don't help him; him fix it himself. Make the following sentences negative: 293. We used to go there. 294. Let's take the bus. 295. He likes everything. What is the opposite of this sentence? 296. There are more people here tonight than last night. Choose the sentence you would say. 297. 1) It's real hot in here. 2) It's really hot in here. 298. 1) It was an historical event. 2) It was a historical event. 299. 1) Those trees are dying. 2) Them trees are dying. 300. 1) They've gotten out of control. 2) They've got out of control. 301. 1) There used to be a fight between the two groups, but anymore that has stopped. 2) There used to a fight between the two groups, but now that has stopped. 302. 1) Thank you anyways. 2) Thank you anyway. 303. 1) He gave it to Jason and me. 2) He gave it to Jason and I. 304. 1) To whom did you give the book? 2) Who did you give the book to? Larry gave Tom 5 dollars and Tom has agreed to pay Larry back. 305. What did Larry do? 306. What did Tom do? Here are a number of sentences containing 'eh' Read them aloud and tell me which ones you say 307. Nice day, eh? 308. It goes over here, eh? 309. Oh, you're still here, eh? 310. Think about it, eh? 311. What a game, eh? 312. What are they trying to do, eh? 313. This guy is up on the 27th floor, eh? then he gets out on the ledge, eh... 328 314. Eh, what did you say? Which ones do you consider ungrammatical or in bad taste? 315. Do you ever use verb forms like these; had given, had gone? Give an example in a sentence showing that the simple past couldn't be used. 316. Do you ever use the word shall? Give an example. Local Words and Usage = L (I'm going to ask you about some local words and expressions.) 317. What does take it over to the cash mean? 318. (Use bring or take) You and Sally are on the third floor, Mrs. Fraser is on the sixth floor. Ask Sally to carry the letter up to Mrs. Fraser. 319. Now pretend you are Mrs. Fraser on the phone. Ask Sally to carry the letter to you. 320. Do you know what snye means? 321. How do you pronounce Iroquois? 322. What colours are the lights in a traffic light? Can you think of any other local Ottawa words? What do they mean? What are other words for a high school? 323. high school 324. 1) suspenders 2) braces 325. veranda 326. shadow 327. regardless 329 Reading Passage = R (I'd like to read this short story aloud. Please don't read it as if you were in school, but as naturally as you can. Pretend you are reading it out to some friends.) (Story on separate sheet) 328. Barry 356. in her 329. shouted 357. leisurely 330. shouted 358. invited 331. as he 359. Aunt 332. house 360. also 333. what's 361. going to 334. tonight 362. 335. library 363. trout 336. used_ to 364. father 337. high 365. caught 338. Mary 366. last. 367. last Sa_turday 339. Wilson 368. Saturday 340. going to 369. potato 341. 370. 371. 342. call her 372. vegetables 343. like 373. sliced 344. to invite her 345. 374. toma_toes 346. 375. 347. 376. 348. about it 377. hot 349. 378. that 350. f ixing 351. fixing her 379. sound 352. humming 380. nothing 353. _tune 381. matter 354. probably 382. that 355. alright 383. when 384. going to. 385. 386. 387. Mary 388. want to 389. Dorothy 390. 391. film 392. the Arts 393. Centre 394. Peter 395. supposed to 396. 397. good 398. better 399. quarter 400. eating 401. 402. sorry 403. about an 404. 405. asked 406. plenty 407. anyway 408. been 409. out in 410. 411. garage 412. 413. 414. working 415. moJ:orbike_he 416. 417., bike he 418. bought 419. Tuesday 420. 421. making 422. progress 423. but I 424. guess he'll 425. greasy 426. right up 427. 428. to his 429. 430. something 431. tp_eat 432. right away 433. 434. the egg 435. egg 436. sandwiches 437. 438. 439. got up 440. morning 441. just had 442. porridge 443. glass of 444. orange juice 445. 446. 447. 448. 449. 450. 451. 452. 453. 454. 455. 456. 457. 458. 459. 460. 461. 462. 463. 464. 465. 466. 467. 468. 469. 470. 471. 472. 473. right starving what's just help Barry butter i syrup milk would you the empty bottles garage company Tell him plenty wa_ter new tube little later when were sitting around wasn't it 474. 475. 476. 477. 478. 479. 480. 481. 482. 483. 484. 485. 486. 487. 488. 489. 490. 491. 492. 493. 494. 495. 496. 497. 498. 499. 500. 501. 502. interesting meeting again always pictures were last double-u_ but I didn't recognize him recognize im unreal What've doing Mary asked Barry's grandmother visiting her visiting her daughter winter las_t business 503. February 504. 505. been 506. secretary 507. doing 508. proj ect 509. _0ttawa 510. 511. 512. South 513. inter esti: 514. 515. buildings 516. beautiful 517. enjoying 518. being 519. student 520. again 521. writer 522. 523. whether Questions about Ottawa = FS A. B. 524. either 525. provincial 526. government 527. congratulated 528. Barry's 529. fa.ther 530. catching 531. and his 532. 533. sat out on 534. 535. 536. good 537. cook 538. pa_tio 539. talking 540. thirty 541. theatre about Ottawa). place to live? What particular things do (don't) you like about it? 333 C. What changes have you noticed in the years you've lived here? D. What improvements do you think could be made? E. Is there much to do in this city? F. If you knew someone who was trying to decide whether or not to move to Ottawa, what would you tell them about it? 542. (P) eleven 543. SR round 544. BG census tract now (BG = Background) v.. Spontaneous Narrative = FS Either A. Have you ever been in a situation where you thought you were in serious danger of being killed—where you thought to yourself, "This is it"? Could you tell me about it? OR B. Have you been in a situation, recently or some time ago, where you had a good laugh, or something funny or strange happened to you, or you saw it happen to someone else? Could you tell me about it? OR C. Would you tell me about your last trip in some detail? Series = SR Would you please say for me the days of the week? 545. Sunday 546. Monday 547. Tuesday 548. 549. Wednesday 550. Thursday 551. Friday 552. Saturday 553. Would you please .count from 15-30? 554 555 556 557. 558 559. 560 561 562 563 564 565 566 567 568 569 Would count now by tens from 40-100? 570 571 572 573 574 575 576 Would you make up pairs of opposites. Example: long and short 577. AU # 578. ao // 579. ei // 580. ei # 581. ai # Now make pairs like this. Example: jump. jump—jumping Use: run, eat, swim, hunt, fish. 582. in // 583. en # 584. in # Now make pairs like this. Example: 33 - 33, 34. 585. 25 586. 27 587. 78 588. 95. Minimal Pairs = MP (There's one last set of words I'd please read these out in pairs?) 589. white—wide 590. whether—weather 591. aunt—ant 592. been—bean 593. Mai—Mel 594. bury—berry 595. lout—loud 596. do—dew 597. house—houses 598. 599. futile—feudal 600. were-where 601. 602. matter—madder 603. site—side 604. knife—knives 605. Mary—marry 606. 607. inter-city—inner-city 608. 609. 610. powder—pouter 611. winter—winner 612. merry—Mary 613. writer—r id er 614. 615. latter—ladder 335 like you to read aloud. Would you 616. shone—shown 617. picture—pitcher 618. 619. thirsty—Thursday 620. 621. daughter—dodder 622. baking—bacon 623. Allan—Ellen 624. vary—very 625. dew—due 626. 627. granted—granite 628. falling—fallen 629. collar—caller 630. caught—cot 631. Barry—berry 632. why—Y 633. taking—taken 634. he hit it—he hid it 635. news—noose 636. merry—marry 637. Did he find them? 638. Did he fine them? 639. 640. 641. 336 Subjective Attitudes and Language Awareness = A A. What do you think of the English spoken in Ottawa? B. Other than the Maritimes and Newfoundland, can you tell where an English Canadian comes from? C. Can Americans tell from your speech whether you're Canadian or American? If so, can you give me some examples of the difference? 642. eh 644. ai 643. AU 645. TO"-Versus a D. Do you pronounce some words differently from either of your parents? Or do you use any different words from those your parents used? Or from your children? (Examples) E. Have you changed your pronunciation of any words over the years? Do you now use a different term to refer to something than you used to? Have you changed any of your grammar? F. How do you feel about your own speech? (Put an [X] on the rating scale.) very quite somewhat neutral somewhat quite very 646. dissatisfied : : : : : : :satisfied G. Please indicate how you rate these varieties of English. (A for American) Place (B for British ) on each row (C for Canadian) of this rating scale very quite somewhat neutral somewhat quite very 647 . (A) slangy : : : : : : : formal 648. (B) 649. (C) 650. (A) clipped : : : : : : :drawled 651. (B) 652. (C) 653. (A) superior : : : : _: : :inferior 654. (B) 655. (C) 656. (A) friendly 657. (B) 658. (C) :authoritative H. Now rate the English of the news readers for these networks: A for NBC (American) B for BBC (British) C for CBC (Canadian) P for CKGO (Popular) 659. (A) worst : : : = : : _:best 660. (B) 661. (C) 662. (P) Free Speech = FS 663. medial t = ytV 682. ao = Au 664. medial t = _ dy 683. ao = ao 665. medial nt = ntV 684. a i = e i 666. medial nt = nV 685. a i = e i 667. medial nt = ndV 686. a i = a i 668. - ing = in 687. un- = An 669. - ing = en 688. un- = an 670. - i ng = in 689. 00 = 0 671. tus dus nu = tjus djus nju 690. oo = Qc 672. tu 1 du 1 nu = tu4 dus nu 691. nd# = nd 673. rV er 692. nd# = n 674. st* = st liaison 693. Vr er 675. st* = s liaison 694. 33 = 32 676. = h liaison 695. 33 = 33 677 . *h = 0 liaison 696. wh = hw 678. V# + h = V# + #V liaison 697. wh = w 679. V* + #V = e# + #V liaison 698. kt, pt = 680. d# + #ys t# + #y = same 699. kts pt = 681. d# + #Yi t# + #y = t[ liaison 700. O = V 338 701. o = a 702. th = n 703. th = z 704. subject verb disagreement 705. you know tag 706. yes 707. yeah 708. yeap 709. mh hmm 710. uh huh 711. mm 712. ah hesitation 713. ahm hesitation 714. um hesitation 715. sure 716. right 717. eh 718. okay 719. like hesitation 720. well hesitation 721. I mean hesitation 722. I think 723. I don't know 724. I guess 725. you 726. your 727. they 728. swear words 729. gnne 730. gotnta 731. income (background) 732. house value (background) 733. pulmonic ingressive speech 734. lad 735. swear word substitutes 736. th = t 737. th = d 738. really 739. not really 740. right = re it 741. oh yeah 742. the +C = 5i#C 743. great = greit 744. interrogatives formed with not 745. I & we 746. uncommon words 747. embarassed laughter -giggling 748. w writer 749. w united 750. opinion introducing gambit to set social contract 751. stuttering 752. reading passage time 339 READING PASSAGE "Hi, Mum!" Barry shouted as he ran into the house. "What's for supper tonight? Today at the library I met this girl I used to know back in high school, her name's Mary Wilson. I'm going to call her now and I'd like to invite her to supper. How about it, eh?" His mother was in the bedroom fixing her hair and humming a tune. "That's probably alright, dear," she answered in her usual leisurely way. "I've invited your Aunt Beryl for supper also. We're going to have trout (you know, the fish your father caught last Saturday), and potato salad, mixed vegetables, and sliced tomatoes with it. Pumpkin pie for dessert. A real hot weather meal. How's that sound?" "Fantastic, Mum! Nothing the matter with that! But when are we going to eat? Mary and I want to meet Dorothy and Al to see a film at the Arts Centre. It's The Pink Panther with Peter Sellers. It's supposed to be very good. We'd better be there by a quarter after eight in case there's a line-up. Any chance of eating early?" "I'm sorry son, we can't possibly eat for about an hour and a half. We could've eaten sooner if you'd asked me, but don't worry, you'll have plenty of time. Anyway, Murray's been out in the garage for hours working on that old motorbike he bought on Tuesday. He doesn't seem to be making much progress, but I guess he'll be greasy right up to his ears and will need lots of time to wash up." "Okay, Mum. Say, can I have something to eat right away? I only had the egg sandwiches you gave me for lunch, and I got up so late this 340 morning I just had time for a bowl of porridge and a glass of orange juice. Right now I'm starving, what's in the fridge?" "Just help yourself, Barry. There's bread and butter and maple syrup if you like, and a glass of milk. By the way, when you're finished, would you carry the empty pop bottles out to the garage for me, please? While you're there you can tell your brother to hurry things up, as we're having company for supper. Tell him there's plenty of hot water, and a new tube of grease cutter." A little later when they all were sitting around the table, Mary said "Wasn't it interesting our meeting again today. I always remember Barry's picture and mine were both on the back page of the school annual. I guess that's because both our last names begin with a 'W'. But I almost didn't recognize him—it was unreal." "What've you been doing since graduation, Mary?" asked Barry's grandmother, who was visiting her son and daughter-in-law. "The winter before last I went to business college, and since February I've been working as a secretary with a firm of architects. We were doing a project in Ottawa South, which was quite interesting, and our office was in one of those new buildings with a beautiful view of the river. It was good experience, but now I'm really enjoying being a student again. Some day, I'd like to be a writer. But at the moment, I am deciding whether or not to become a Canadian citizen. If I do, maybe I can get a job with either the federal or provincial government." After dinner, everybody congratulated Barry's father for catching 341 such great fish, and his mother for being such a good cook. Then they all sat out on the patio talking until 7:30, when it was time for Barry and Mary to go to the theatre. 342 white—wide whether—weather aunt—ant been—bean Mai—Mel bury—berry lout—loud do—dew house—houses futile—feudal were—where matter—madder site—side knife—knives Mary—marry inter-city—inner city powder—pouter winter—winner merry—Mary writer—rider WORD PAIRS latter—ladder shone—shown picture—pitcher thirs ty—Thursday daughter—dodder baking—bacon Allan—Ellen vary—very dew—due granted—granite falling—fallen collar—caller caught—cot Barry—berry why—Y taking—taken he hit it—he hid it news—noose merry—marry Did he find them?—Did he fine them? 343 WORD LIST marry leg nothing multi-national writer berry Africa prestige matter (That's untrue) student (He just left.) new caramel lengthen newspapers sentence library bike which absurd route doing often (it shot across genuine naturally tomatoes a clear sky) temperature congratulate sorry calm columnist introduce generally asphalt cot good balcony provincial government also Mary schedule merry anti-pollution Barry regular guarantee fertile film nuclear quantity resource centre fishing tube February beautiful leisure luxury weren't anyone the egg grandfather almond particular about singing lecture without around sandwiches (They didn't go) wheelbarrow potatoes that vocabulary dual garage grass valued secretary pumpkin greasy (There's lots mature south milk to eat.) united unbelievable agriculture International water Toronto recognize Harvester futile attempt Ottawa tune always theatre Alberta caught shouldn't perfectly Winnipeg something either (He pointed mouth favourite at me.) again been daughter lieutenant provincial dirty ration English decal PICTURES an orange lever tomatoes garage roof pumpkins family pictures father vase grandmother kite grandfather night daughter mirror aunt chesterfield tap house cutlery potatoes window zebra library film dog lying lion missile blinds blouse khaki sandwiches Africa frying pan tuba bacon and eggs an egg Tuesday south dinner vs. supp right 2:45 an apple 11:15 professor 8:30 whipped cream WXYZ centre group singing chalk board with brushes living room and pointer newspaper apricots Tuesday semi-circle 27 semi-truck newspaper railway crossing 30 APPENDIX B ITEM BY ITEM PRINT OUT < ONEWAY V=ALL OP=MARG$ CASES=ABEA:OTTAWA > OK'EfiflY T ABU1ATICNS CASES=AREA:OTTAW A 2. SEX H F N= 89 4 3 46 MARG% 4 8,3 51 .7 3, BOBN (3) (7) (8) (9) (10) (11) N = 89 1 1 4 1 2 2 MAEGK 1, 1 1, 1 4. 5 1 ,1 2. 2 2.2 (12) (13) 2 1 2. 2 1. 1 (21) (23) (24) (25) (27) (28) 1 3 5 4 1 1 M A B G % 1. 1 3.4 5.6 4.5 •1. 1 1. 1 {29) (33) 2 1 2, 2 1. 1 (34) (35) (38) (3*9) (41) (42) 2 1 2 3 2 1 «ABG% 2, 2 1, 1 2, 2 3.4 . 2. 2 1.1 (43) (44) 1 2 1. 1 2, 2 (45) (46) (47) (48) (49) (50) 1 1 3 1 1 5 KABG% 1. 1 1. 1 3.4 1.1 1. 1 5.6 (51) (52) 1 4 1. 1 4.5 (53) (54) (55) (56) (57) (58) 4 4 2 3 4 3 M A B G % 4,5 4.5 2. 2 3.4 4. 5 3.4 (59) (61) 2 1 2, 2 1, 1 (62) (91) (99) 1 1 1 MARG% 1„ 1 1.1 1,1 *4. WKEHEB OT PEBTH OTVAL EG NT CNT MONT H= 89 68 1 .3 1 6 1 MAEG5I 76. 4 1. 1 3.4 1.1 6. 7 1. 1 WCDA PQFBC 1 1 •1. 1 1.1 DPS THY USA ENG SCOT SAM 3 1 1 1 1 MAEGf 3. 4 1 . 1 •1.1 1 • 1 1.1 5.AGEHEBE MISS (D <2) (3) (4) (5) N = 21 68 8 5 3 4 1 HABGX 38. 1 23. 8 14.3 19. 0 4.8 6,WfiEBELS MISS OTVAL EONT OKI MONT WCDA M= 28 61 1 1 9 2 4 M AEG? 3.6 3.6 32. 1 7, 1 14. 3 AT LCD A NYCM DC 3 2 10. 7 7. 1 USA ENG EUE AFB 2 1 2 1 MAEG% 7. 1 3. 6 7. 1 3.6 7. YEABSTH MISS (1) (2) {3) <4) (5) N = 28 61 8 8 5 2 3 MABG% 28. 6 28. 6 17.9 7. 1 10.7 (6) 2 7,1 8.OLANG EN G FB ITL . GEE DKR IYOTH N= 89 73 9 4 1 1 1 HABG* 82,0 10, J 4.5 1.1 1. J 1.1 9.FATHFRM MISS CT EENFB SMFLS PEETH OTVAL »= 88 1 29 1 1 1 8 MABG% 33, 0 1. 1 1.1 •1. 1 9. 1 ECNT GNT 1 9 1. 1 10. 2 MONT WCDA ATLCDA PQFBC UPSTNY USA 2 4 3 2 2 1 HABGX 2, 3 4.5 3.4 2.3 2. 3 1. 1 F.NG 10 11, 4 SCOT 4 4. 5 348 ITL OSSB OKR SAM CZ 5 1 2 1 1 MAEG% 5.7 1. 1 2. 3 1.1 1.1 10.FATHLAN MISS ENG FE ITL GEE OKB N= 88 1 72 5 5 1 2 MARG% 81. 8 5.7 5.7 1.1 2.3 POL 1 1. 1 IYOTH 2 2. 3 11. HOHLONGC MISS LIFE CHILD B= 88 1 61 6 MARG% 6 9,3 6,8 12. HOWLONGH MISS LIFE CHILD N= 88 1 29 8 BABGX 33. 0 9.1 TEEN YADOLT 8 13 9.1 14.8 TEEN YADOLT 7 8.0 44 50. 0 13.HOTHFRK M= 89 MARG3& OT PERTH C1VAL EC NT CNT KC NT 26 1 15 3 12 2 29.2 1. 1 16. 9 3.4 13. 5 2.2 WCDA ATLCEA 3 3 3. 4 3.4 MARG% MABGS 14.HOTHLAM N= 89 PQ PRC NXCW.DC 4 a. 5 ITL 1 1. 1 EUR 1 1. 1 ENG 72 80. 9 1 . UKR 1. SAM 1 FR 7 7, 9 USA 1 1.1 CZ 1 1.1 ITL 5 5. 6 E.KG 9 10.1 JPN 1 1.1 SP 1 1. 1 SCOT 2 2. 2 GER 1 •1. 1 IRE 1 •1. 1 UKR 1 1. 1 POL 1 1. 1 15. H01LCHGC LIFE N= 89 70 fl.ARG.SS 78,7 16. HOILONGH LIFE H= 89 28 M A R G % .31.5 IYOTH 1 1 . 1 CHILD TEEN YADULT 1 3 15 1.1 3.4 16.9 CHILD TEEN YADOLT 6 6 49 6.7 6. 7 55.1 349 17.MBATGEBR (1) (2) (3) N= 89 25 12 9 M ARG55 28,1 13,5 10.1 (4) 15) (6) 17 9 6 19.1 10.1 6.7 (7) 4 4, 5 (8) 3.4 KABGl ZERO 4 4, 5 18.ETBSICBG MISS PQFRC USA ENG SCO! IRE B= 87 2 6 4 27 20 15 HJLBGX 6.9