__ _”—-—‘~v m.‘ i ABSTRACT TOWARD A THEORY OF SOCIAL CONFLICT: A STUDY OF FRENCH-ENGLISH RELATIONS IN AN ONTARIO COMMUNITY By John D. Jackson The principal objective of this thesis was to amplify certain hypotheses concerning social conflict through an empirical study of inter-ethnic relations in a particular community. 0f secondary interest was the relationship between social conflict and ethnic assimilation. The community selected for this study, located in southwestern Ontario, Canada, was composed of both French- and English-speaking residents. As such, the study provides an instance of French-English relations in the Province of Ontario as distinct from the more studied situation in the Province of Québec. Social conflict was defined as consisting of any relationship between two or more sets of individuals exhibiting some degree of organi- zation and incorporating one or more of the following conditions: incom- ,patible values, position scarcity, and resource scarcity. The position nassumed was that a state of conflict is a structural arrangement. Observed phenomena which is often referred to in the literature as repre- senting overt or covert conflict, manifest or latent conflict, or potential or actual conflict were taken as relating to variations in the intensity and militangz of a conflict situation. John D. Jackson A conflict exhibiting a low level of intensity and militancy would correspond to a covert, latent, or potential conflict in more common terminology. Intensity was defined as the magnitude of party involvement, militancy as the use of coercive and violent strategies in the pursuit of party interests. Institutionalization and community identity were posited as independent variables influencing the degree of militancy. Nhether decision-making power relevant to the issues in dis- pute is located in the vertical system (i.e., beyond the local community) or in the horizontal system (i.e., within the local community) and the extent to which ethnic, religious, and class groupings are superimposed were taken as independent variables influencing the degree of intensity. The research upon which this thesis was based proceeded in three phases. During the first phase, frequent trips were made to the selected community to delineate the major components of its social structure. Unstructured interviewing of community leaders was the primary method of data collection used during this phase. An eight month period of inten- sive interviewing and observation followed. An analysis of these data led to the development of the hypotheses referred to above. These hypo- theses were tested during the final phase using structured interviewing, documentary analysis, and a questionnaire administered to a residential sample. The hypotheses relating institutionalization and community identity to militancy and the strength of the community horizontal pattern and ll“IPBI'iuposition to intensity were found, with some exceptions, to corre- 813011:! to the observed direction of the indicators. The thesis concludes John D. Jackson with a series of multivariate hypotheses to account for these exceptions. The data also pointed to a relationship between the intensity of conflict and structural assimilation, questionning the sequential and inevitable move toward assimilation in the race relations cycle model. TOWARD A THEORY OF SOCIAL CONFLICT: A STUDY OF FRENCH-ENGLISH RELATIONS IN AN ONTARIO COMMUNITY BY 4 John waJackson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1967 @ Copyri ght by JOHN DAVID JACKSON December 1967 ACKNOWLEDGMEN'E As with any piece of work of this nature, assistance in a variety of forms has come from several sources. The research problem itself had, no doubt, its origins in my own past, in my early life in a French- speaking continuity in the Pravince of Québec. But these experiences must be translated into ideas and the ideas again related to experience. I am indebted to Professors Walter E. Freeman and James B. McKee for their encouragement, permissiveness, and intellectual stimulation which enabled me, freely and openly, to grapple with the issues to which this thesis is addressed. As the research took shape, others entered the picture-- Professors Hillimn A. Paunce, Duane 1.. Gibson, and Seymour Parker--each with a contribution, but each permitting me to make the decisions which, for better or for worse, are represented in the following dissertation. A project of this type requires material support. Such support was provided by the Institute for Comnity Development, Michigan State University, and a Canada Council Award. Without this support, it would have been iupossible to carry out a project requiring such a lengthy period of field work. Considerable assistance in the form of advice, encouragement, and access to office machinery and supplies was provided by Professor Rudolf Belling, Chairman of the Department of Sociology, University of Windsor, and Father Edward'soyce, Director of the General Mission Centre of Essex County. It was through the efforts of the latter organization that volunteer interviewers were located. Professor Helling's assistance in -ii- paving my way into the study community was most appreciated. The people of Normanville, to whom I owe a deep debt of grati- tude, must remain unnamed. Their willingness to accept a stranger, a prying stranger, in their midst with such openness and hospitality was indeed appreciated. To those who read this thesis, and some will, I hope only that it will serve to lead each to a deeper appreciation and understanding of their town and their people. It is most difficult to acknowledge the contributions of those closest to one's self. My student colleagues--Max Flood, John Pease, Lee Sloan, Joseph Smucker, and Wbrth Summers--will recognize many of the ideas in this thesis as having emerged from many hours of discussion, debate, and appraisal. It is customary, at this point, to acknowledge one's wife for her patience and understanding. It is more difficult for me to express my feelings in this respect when my wife not only showed considerable patience and understanding, but also acted as critic, editor, secretary, and research assistant. I can only say, "Thank you.” ~111- ACKNOWLEDGMENTS . . . . . . LIST OF TABLES . . . . . . LIST OF FIGURES . . . . . . Chapter I. INTRODUCTION . . . The Problem . . Contributions . Organization of II. REVIEW OF THE LITERATURE AND DEVELOPMENT TABLE OF the CONTENTS Thesis 0F HYPOTHESES O O O O O O O O O O O O 0 soc181 conflict 0 O O O O O O O O I O O O O The Psychological Versus the Sociological Perspective . . . . . . Conflict Versus Consensus Social Conflict Defined . The Variables of Conflict Theoretical Hypotheses The Community and Conflict Community Variables . . . ,Assimilation and Conflict . Assimilation: Ideology or Social Process C O O O O O O O O O O O O O Assimilation: Unidimensional or Multidimensional. The Variables of Assimilation . . Assimilation and Conflict . . . . Summary . . . . . . . . . . . . . . III. THE RESEARCH DESIGN . . . . . . . . . The Testable Hypotheses . . . . . . . . . . . . . . Hypothesis I: Institutionalization and Militancy. O O O O O O C O O O O O O Page ii ix xiii ll 12 12 19 24““ 3O 37 40 42 47 51 52 59 63 65 65 66 Chapter Hypothesis II: Superimposition and Intensity. . Hypothesis III: Vertical-Horizontal Patterns of Organization and Intensity. . . . . . . . . . Hypothesis IV: Community Identification and Militancy O I O O O O O O O O O O The Empirical Conditions for Conflict . Methods of Data Collection. Observation . . . . . . . Interviewing . . . . . . Documents . . . . . . . . thations . . . . . . . . The Questionnaire . . me Smle O O O O O O O Validity and Reliability: of Limitations . . . . 8‘1er 0 O O O O O O O O O A Discussion IV. HISTORICAL BACKGROUND: THE DEVELOPMENT OF THE CONFLICT PARTIES . . . New France: 1760. . . . . . A Commercial Society . . An Agricultural Society . Internal Conflicts . . . British North America: Assimilation Versus Survive O o o o The New Power Structure . . . . . . . . . Assimilation Versus Survival: 1763-1840 . Assimilation Fails: 1840-1867 . . . . . . . Political Realities and Institutionalized conf11Cteooossoososooo English Separatism . . . . . . . . . . Confederation: An.Alternate Solution . French Minorities and the School Question French and English Minority Rights . . Ontario: Bilingual and Catholic Schools The De Jure Basis of Catholic Schools in Ontario . . . . . . . . . . . . . The De Facto Basis of French Schools in Ontario . . . . . . . . . . . . . . Assimdlation, Biculturalism, and Separatism The English Canadian Position . . . . . . -v- 0 O O O O O O O O O O O O O O O 0 O O O O O O O O O O O O 0 0 O O O O 0 coop-loose O O 0 O O O O O Page 69 72 74 76 79 8O 92 95 97 97 102 106 113 114 115 116 117 118 120 120 123 132 132 135 137 141 142 144 145 146 153 154 Chapter The French Canadian Position . . . . . . . . . The Franco-Ontarian Position . . . . . . . . . sumry . O O O O O O O O O O O O O O O O O O O O V. NORMANVILLE, PART I: THE TONN AND ITS PEOPLE . . . Early Settlement and Development . . . . . . . Early Development: 1800-1900 . . . . . . . . Urbanization and Industrialization: 1900-1966 The People of Normanville . . . . . . . . . . The Metropolitan Region: A Comparison . . . The Significance of Language . . . . . . . Knowledge and Use of French in Normanville 01d and New Residents . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . VI. NORMANVILLE, PART II: THE TOWN AND ITS INSTITUTIONS Communal Institutions . . The bbrld of Work. . . Government and Politics Non-Communal Institutions Church and Parish . . . . The Schools: Protectors of the Faith Voluntary Associations . . . . . . . Sumaty....o........... VII. THE EMPIRICAL CONDITIONS FOR CONFLICT . . . . . . . The Parties: Quasi-Groups or Statistical Categories? . Interpersonal Interaction . . . . . . . . . . . Institutional Separation . . . . . . . . . . . Subjective Awareness . . . . . . . . . . . . . Contact and Interaction: Conflict Agencies andAgentSoooooooeoseoeo Conflict Agencies . . . . . . . . . . . Conflict Agents . . . . . . . . . . . . A Note on Interpersonal Contact . . . . Position Scarcity, Resource Scarcity, and Value Incompatibility . . . . . . . . . . . . . .v1- 0 O O O O O O O O O O O O O O O O O O O O Page 157 161 163 166 168 169 172 176 176 180 182 193 194 197 197 197 204 214 214 219 231 236 238 “‘ 238 239 248 249 253 255\ 270““ 281 286 Chapter Page Position and Resource Scarcity . . . . . . . . . . 287 Incompatible Values . . . . . . . . . . . . . . . . 290 sumary O O O O O I O O O O O O O O O O O O 0 O O O 0 2 9 5 VIII. THE VARIABLES OP CONFLICT, PART I: INDEPENDENTVARIABLES................ 298 Institutionalization . . . . . . . . . . . . . . . . 299 Period I: 1900 to 1930 . . . . . . . . . . . . . . 300 Period II: The Thirties . . . . . . . . . . . . . . 307 Period III: The Sixties . . . . . . . . . . . . . . 313 Summary . . . . . . . . . . . . . . . . . . . . . . 323 Superimposition . . . . . . . . . . . . . . . . . . . 324 Periods I and II: 1900 to 193 . . . . . . . . . . 325 Period III: The Sixties . . . . . . . . . . . . . . 330 Summary . . . . . . . . . . . . . . . . . . . . . . 334 The Strength of the Horizontal Pattern . . . . . . . 335 Period I: 1900 to 1930 . . . . . . . . . . . . . . 335 Period II: The Thirties . . . . . . . . . . . . . . 337 Period III: The Sixties . . . . . . . . . . . . . . 340 Summary . . . . . . . . . . . . . . . . . . . . . . 341 mnity Identity 0 O O O O O O O O 0 O O 0 O O 0 O 342 Period III: The Sixties . . . . . . . . . . . . . . 343 SWYOOOOOOOOO0000000009000 346 Smry O O O O O O O O O O O O O O O O O O O O O O O 347 IX. 131E VARIABLES OF CONFLICT, PART II: THE DEPENDENT VARIABLES AND TEST OF HYPOTHESES . . . . . . . . . . 350 Intensity . . . . . . . . . . . . . . . . . . . . . . 350 Period I: 1900 to 1930 . . . . . . . . . . . . . . 350 Period II: The Thirties . . . . . . . . . . . . . . 356 Period III: The Sixties . . . . . . . . . . . . . . 361 sum 0 O O O O O O O O O O O O O O O O O O O O O 371 Militancy . . . . . . . . . . . . . . . . . . . . . . 372 Period I: 1900 to 1930 . . . . . . . . . . . . . . 373 Period II: The Thirties . . . . . . . . . . . . . . 376 Period III: The Sixties . . . . . . . . . . . . . . 377 Summary . . . . . . . . . . . . . . . . . . . . . . 378 Test of Hypotheses . . . . . . . . . . . . . . . . . 379 Hypothesis I: Institutionalization and Militancy . 379 -vii- Chapter Sumary O O O O O O O O O O O O O O O O O O O 0 Hypothesis IV: Community Identity and Militancy mm ry O O O O O O O O O O I O O O O O O O O O Hypothesis II: Superimposition and Intensity . sumry O O O O 0 O O O O O O O O O O O O O O O Hypothesis III: The Strength of Pattern and Intensity . . . . Summary . . . . . . . . . . . . Summary . . . . . . . . . . . . . X. CONCLUSIONS AND SUMMARY . . . . . . Conclusions . . . . . . . . . . . Reassessment . . . . . . . . . Assimilation and Conflict . . . Proposals for Phrther Research smary O O O O O O O O O O O O I APPEme O O O O O O O O O O O O O O O O O B IBLIOMHY 0 O O O O O O O O O O O O O 0 -viii- the Horizontal Table II. III. IV. V. VI. VII. VIII. IX. XI. XII. XIII. LIST OF TABLES The Relationship Between Militancy and Intensity . . Proportion of Time French Spoken by Frequence of Use of French Scores. . . . . . . . . . . . . . A Comparison of the Distribution of Four Characteristics in the Sample with the POPUIation O O O O O O O O O O O O O O O O O 0 Summary of Variables and Indicators 0 C O O O O O 0 Summary of Indicators and Related Instruments . . . Religious Affiliations of Normanville Population and Rank in Relation to Forty-five Metropolitan Regional Census Tracts Ethnic Origins of the Normanville Population and Rank in Relation to Forty-five Metropolitan Regional Census Tracts Extent of Education Received by Persons not Attending School in Normanville and Rank in Relation to Forty-five Metropolitan Regional Census Tracts . . Normanville wage and Salary Income per Family and Rank in Relation to Forty-five Metropolitan Census Tracts O O O C O O O O O I O O O O O O O 0 Conflict Parties by Language Spoken Use of French by Those Who Speak French EXtent to which French is Used by Those who Speak FrenCh O O O O O O O O O O O O O O O O O O O O O 0 Age by Extent of Use of French . . . . . . . . . . . nix- Page 32 101 105 108 111 177 178 178 179 183 184 187 190 XVII. XVIII. XIX. XXI. XXII. XXIII. XXIV. XXVI. XXVII. XXVIII. XXIX. Length of Residence in Normanville by conf11Ct Parties 0 O O O O O O O O O O O C O O 0 Parents' Residence in Town by Conflict Parties . . Conflict Parties by Place of Employment of Heads Of stehOIds O O O 0 O O O O O O O O O 0 Occupational Status Distribution for Normanville Regidential smle O O O O O O O O O I O O O O 0 Conflict Parties by Occupational Status Levels . . Conflict Parties by Annual Income of Family Units Conflict Parties by Age in Years of Heads Of Housemlds O O O O O O O O O O O O O O O O 0 Conflict Parties by Family Income of Employed Heads of Households . . . . . . . . . . . . . . Normanville Church Membership . . . . . . . . . . Extent to which French is Used by Attendance at FreHCh Mass 0 O O O O O O O O O O O O O O 0 0 School Enrollment by Extent of Use of French . . . School Enrollment by Family Income of Employed Heads of Households . . . . . . . . . . . . . . Conflict Parties by Association Membership . . . . Residential Distribution of Conflict Party Members . . . . . . . . . . . . . . . . . Catholic-Protestant Association Scores by conflict Part1 as 0 O O O O O O O O O O O O O O O Catholic-Protestant Association Scores by Family Income Levels . . . . . . . . . . . . . . -x- Page 193 194 199 199 200 201 202 203 215 218 228 231 236 240 244 245 Table XXXI. XXXII. XXXIII. XXXIV. XXXVI. XXXVII. XXXVIII. XLI. French-English Association Scores . . . . . . . . . French-English Association Scores by Family Incom Lave 1 s O O O O O O O O O O O O O O O O O 0 Distribution of Votes Received by Three Incumbents of Positions on the Town Council in December 1965 According to French and English Residential sectors 0 O O O O O 0 O O O O O O O O O O O O O 0 Comparison of Percentage of Votes Received by Candidates for Councillor in December 1965 in Polling Divisions 10 and 4 . . . . . . . . . . . . Conparison of Percentage of Votes Received by Candidates for Deputy Reeve in December 1965 in Polling Divisions 10 and 4 . . . . . . . . . . Ethnic Origin and Occupations, Male Labor Force, Ontario 1931, Percentage of Over- and Under- Representation in Occupation by Ethnic Group . . . Ethnic Origin and Occupations, Male Labor Force, Ontario 1961, Percentage of Over- and Under- Representation in Occupation by Ethnic Group . . . Conflict Parties by Annual Famdly Income . . . . . . Religious Affiliation by Ethnicity and Occupational Status Levels of Employed Heads of Households . . Summary of Measurements of Association between Religious and Ethnic Categories and Occupational Status Levels of Euployed Heads of Households . . The Annexation Vbte by French and English P0111“ Divisions 0 O O C O O O O O O O O O O O I A comparison of "Ranking-Ratios" on Ethnic-Religious Issues According to Party and Sub-Community Membership . . . . . . . . . . . . . . . . . . . . Page 246 247 279 280 280 327 328 331 332 332 344 346 {table )CLII. XLIII. XLV. JCLVI. JXHUVII. XLVIII. Summary of State of Independent Variables with Reference to the Bilingual School Issue by Time Period and Regional Foci . . . . . . . . . Summary of State of Independent Variables with Reference to the Catholic School Issue by Time Period and Regional Foci . . . . . . . . . French and English Catholic Ranking of the Issue "Bilingual Classes for French-speaking Children to Grade 13" in Relation to Five Other Issues French and English Catholic Ranking of the Issue "Equal Grants for Public and Separate Schools" In Relation to Five Other Issues 0 o e e o o a French and English Catholic Ranking of the Issue "Extension of Government Support to Catholic Schools up to Grade 13" in Relation to Five OtherIssues................. Amount of French Language Copy by Volume in ”The Centurion" . . . . . . . . . . . . . . Summary of Dependent Variables with Reference to Bilingual School Issue According to Time Period and Regional Fbci . . . . . . . . . . . Summary of Dependent Variables with Reference to Catholic School Issue According to Time Period and Regional Foe! O O O O O O O O O O O O O O ~xii- Page 348 349 363 363 364 368 380 380 liigmzre 1.. LIST OF FIGURES Page Normanville Separate Schools' Enrollment, Bilingual and English Classes, 1955-1965 . . . . . 227 Ethnic and Religious Composition of Voluntary Associations in Nbrmanville, Ontario . . . . . . . 234 The Conflict System: Bilingual School Issue local Level, PeI‘iOd I, 1900-1930 e s s e o e o o o 404 The Conflict System: Bilingual School Issue Local Level, Period II, The Thirties . . . . . . . 405 The Conflict System: Bilingual School Issue Local Level, Period III, The Sixties . . . . . . . 406 The Conflict System: Bilingual School Issue Provincial Level, Period I, 1900-1930 . . . . . . 407 The Conflict System: Bilingual School Issue Provincial Level, Period II, The Thirties . . . . 408 The Conflict System: Bilingual School Issue Provincial Level, Period III, The Sixties . . . . 409 -xiii- CHAPTER I INTRODUCTION This thesis is a study of French-English relations in an Ontario continuity. The primary objective was to develop a set of hypotheses con- cerning social conflict. Much has been written on the Canadian problem since the two groups came together two centuries ago, but according to one study of the problem, "French Canada has been, on the whole, very 1“perfectly studied."1 The same may be said about race relations in the United States.2 In addition to the expected criticisms concerning an oVer-enphasis on description and a paucity of analytical studies, both 81titlations have suffered from the highly emotional polemics that are an 1nevitable consequence of intergroup conflict. Such dialogues have their affects on social science, ofttimes making it difficult to sort ideolog- 1i381 admonitions from systematic treatments. In the case of the American Negro, systematic research has exhibited sttong psychological orientations based, no doubt, on the American faith in individualism and a correSponding commitment to the hypothesis that the attitudes and beliefs of individuals directly determine the kinds of 3°¢ia1 organization which emerge from human interaction. Some students ¥ lPhilippe Garigue, A Bi lio ra hic I troduction to the Stud of L‘rench Cans mg . (Montréa1:McGi11 University Press, 19565, p. 5. 28cc Everett C. Hughes, "Race Relations and the Sociological Sination, " American Sociological Review, Vol. 28, December 1963, pP- 879-891; and E. F. Frazier, "Race Contacts and Social Structure, " ‘4“ 9!} Sociological Review, Vol. 14, 1949, pp. 1-11. -1- -2- of race relations hold this emphasis reaponsible for the inability of American social science to predict the current Negro revolt. If research on the American problem has suffered from a nominal- istic bias, the Canadian problem has suffered from an over-abundance of i deo logical polemics, 3 and a geographically narrow perspective. In a recent review of the literature on Hench Canada, only 170 out of 2949 entries were directly concerned with French Canadians outside of the Province of Québec.4 True enough, 76.6 per cent of those Canadians of IFrench origin reside within the borders of Québec. However, the impli- cations that the French minorities outside of the province have had and Will have on the totallity of French-English relations in Canada far exceed their numerical strength. A recent report of the Royal Commission 0!: Bilingualism and Biculturalism discusses the itrportance of the "French minorities” to the national picture.5 For the sake of clarification, the following résumé of English- French population ratios is presented. Utilizing the 1961 census, the t'-°ta1 population of Canada was slightly over eighteen and a quarter mil- lion. Of these, 30.4 per cent were of French origin, 43.8 per cent of BTitzish origin, the balance representing a mélange of other ethnic groups. In terms of the English-French dichotomy, around which much of the history and the institutional life of the country has grown, the residual category ~ 3Ph111ppe Garigue, op. cit., p. 5. “ma. P 5Canada, The Royal Comission on Bilingualism and Biculturalism, Luminagy Regort. (Ottawa: The Queen's Printer, 1965), p. 81. is often classified with the "English." Recently, this category has been referred to as the "third element, " an element now viewed as relevant to the current debate on biculturalism.6 Over the last six decades, the pro- portion of French Canadians to the total papulation has remained remark- ably stable, indicating a continuous growth in absolute numbers. The proportion of British Canadians, however, has declined, the over-all popu- lation growth being influenced by immigration. It was noted that 76.6 per cent of those of French origin were con- Centrated in the Province of Québec. Close to one-half of the remaining 23.4 per cent live in the Province of Ontario,7 the balance being scat- tered throughout the country, with a sizeable number in the Maritime Prov- 1aces. Historically, Québec and Ontario, both originating from the uPPer-Lower Canada partition of 1791, symbolize the dualism of the nation. And, it might be added, the French of Ontario have often been the brunt 95 the same kinds of discrimination and prejudice which are usually directed t("ward immigrant groups and racial minorities. Little has been written, from a sociological perspective, on the F1'anco-Ontariansuthe French Canadians of Ontario. The one bibliography at Our disposal lists only fourteen studies directly pertaining to 0“t.ario.8 It is doubtful that this has increased significantly during the ‘_ 512“” p. 51. F 7Of the total population of Ontario in 1961, 10.4 per cent were of tench origin, 59.5 per cent of British origin. 8Philippe Garigue, 22. 31 . last decade. In Spite of this lack of systematic studies of the problem outside of the Province of Québec, there has been a tendency to general- ize from the Québec situation to "French Canada" as a whole. The dangers inherent in this type of generalization are apparent in an observation made in the Royal Commission report: The divergent preoccupations of French Quebec and the 'French minor- ities' of the other provinces were very evident . . . regional dif- ferences seemed, during this first sampling of Opinion, to be almost as marked in French Canada as they were in English Canada.9 For this, and a number of other reasons soon to be stated, this thesis deals with French-English relations in Ontario, focusing on one community historically identified as a "French town." Students of minority groups have frequently taken into their u‘leoretical considerations two major types of contact situations; contact involving subordina- tion of an indigenous pOpulation by a migrant group . . . and con- tacts involving subordination of a migrant population by an indige- nous racial or ethnic group.1 Given the peculiarities of the American situation, empirical research was 998 tined to focus on the latter type of contact. Thus, we have a situa- tion in which a good deal of our understanding of cultural and racial 9Canada, cg, cit., p. 28. 10Stanley Lieberson, "A Societal Theory of Race and Ethnic Rela- tions," American Sociological Review, Vol. 26, 1961, p. 903. See also, Oscar I. Janowsky, Nationalities and National Minorities. (New York: The Macmillan Company, 1945); Everett C. Hughes and Helen M. Hughes, Liars Peoples Meet: Racial and Ethnic Frontiers. (Glencoe, Illinois: e Free Press, 1952); C. Wagley and M. Harris, Minorities in the New Lbrld: §ix Case Studies. (New York: Columbia University Press, 1958); and E. F. Frazier, Race and Culture: Contacts in the Modern World. (New York: Alfred A. KnOpf, Inc., 1957). minorities is derived from cases where incoming groups have been sub- jected to indigenous groups. With the exception of the Negro, some groups of Spanish Americans, and the American Indian, most of the contact situa- tions studied have resulted from the "voluntary" migration of subordi- nate peoples. 0n the other hand, the French Canadian case is a classic example 0f the first type of contact. To use Hughes' terminology, the French Canadian is a ”charter member" minority.“ Given the perspective that Contact between ethnic groups is a meeting of social organizations, then 81Snificant differences between the types of contact discussed above would be expected. The American emphasis on the "melting pot" ideology in con- trast to the Canadian emphasis on a pluralistic ideology is perhaps evi- dence of this hypothesis. Given this state of affairs, it was felt that reSearch into the Canadian situation was desirable in terms of the develop- ment of a sociology of intergroup relations. Moreover, the peculiarities 95 the Ontario situation require special attention in this context because 95 the decidedly numerical and social minority position held by the Franco- O“tarian in contrast to his Québec cousin. The community selected for this study, Normanville, Ontario, had a P0pu1ation of approximately 4500, 56 per cent of the citizens being of F1'ench origin, 30 per cent of British origin, the remaining 14 per cent b91118 composed of citizens of various ethnic origins.12 In no way was ¥ 11Everett C. Hughes and Helen M. Hughes, op, cit., p. 23. 12See Appendix II, Tables 1 to 3. this conmnity selected as being representative of some larger region. At best, it is representative of a small portion of southwestern Ontario. Normanville was studied as a particular case, the purpose being to test a model relating variations in comunity social structure to social con- flict. The criteria for selecting Normanville were as follows: 1. The town is located in Essex County, one of the four regions in Ontario with a sizeable French Canadian population. 2. Of the various communities in the urban portions of the county, Normanville reported the largest proportion of persons of French origin. It was assumed that this feature would provide a higher probability of locating a viable French Canadian sub-community. :3. Normanville is an urban area. Most studies of French Canada have been conducted in rural communities, so much so that the rural community has been presented as representative of French Canadian social organization. One only need pursue the history of the French in North America to discover the fallacy of this assertion.13 There is the problem of the limitations imposed on research by in"‘estigating a single case. Just how serious such limitations are depends on the general objectives of the study. Lipset and others have Cu~8t:inguished between two types of case studies. In the first type, referred to as a "particularizing analysis," emphasis is placed on a . . . description and explanation of the single case, to provide information concerning its present state, and the dynamics through Which it continues as it does.“ 13See Philippe Garigue, "The French Canadian Family," in M. wade (Editor), Canagigg Dualism. (Toronto: University of Toronto Press, 1960), pp. 181-201. 14Seymour M. Lipset et al., Union Democragy. (Garden City, New York: Anchor Books, Doubleday and Company, Inc., 1962), p. 471. In the second type, a "generalizing analysis," emphasis is placed on the development of enpirical generalizations or theory through the analysis of the single case, using it not to discover anything about it as a system but as an empirical base either for generalization or theory construction. This study includes both types of analyses, but the over-all research objectives were directed toward the latter type. That is, the primary goal of this thesis was to develop a model or theory from empiri- cal materials, a model which will then be available to test through com- parative analyses. Tests of hypotheses were based primarily on variations “hid! had occurred over time within the case or system under study.16 __'I'he_ PrOblem Apart from the substantive material itself, the primary focus of this thesis is social conflict. The relationship between conflict and assimilation is of secondary interest. Conflict is taken for granted; it is “finned that "social conflict is ubiquitous."17 The task then was one of 8Pecifying the eupirical conditions necessary for a state of conflict to eXist, certain variables of conflict states, and their interrelations. These objectives were met during the research process itself, a pr"Gees which proceeded in three distinct phases. During the first “'9 Phases, the conflict model tested was developed from an ongoing inter- action between the literature and the data. The resulting 83"- 0f \ 15M. 16%., p. 479. (S 17Ralf Dahrendorf, Class and Class Conflict in Industrial Society. tanford: Stanford University Press, 1959), p. 162. hypotheses were then tested during the final phase. These hypotheses were taken as guides for the collection and ordering of data. The major objective was to arrive at a more adequate delineation of the variables of conflict and to refine the hypotheses. The interest in assimilation resulted from a consideration of the race relations cycle model. This model suggests an inevitable sequence of the processes of contact, competition, conflict, accommodation, and aslsitllil.m';ion.18 The question was: Are conflict and assimilation neces- sarily sequential or are they interdependent? This question was approached through an attempt at "translating" variables of assimilation into the variables of conflict. This is a coumunity study though not in the classical sense of Presenting a picture of a comunity as a functional whole. Rather, given the conflict system studied, the interest is in the possible effects of Variations in community social structure upon an over-all conflict situa- tion encompassing a wider region. In effect, how and why do local com- munities vary in their responses to national or regional conflicts? The following four questions sumarize the research problem under- taken in this thesis: \ c 18The race relations cycle is based on Robert Park's theoretical onSideration of race relations: "They [race relations] involve racial ROWOtition, conflict, accoumdation, and eventually assimilation.” See Fobert Ezra Park, Race and Culture. (Paperback edition; New York: The 1'99 Press of Glencoe, 1950), p. 104. See also Robert E. Park and “nest H. Burgess, Introduction to the Science of Sociolggy. (Chicago: 0 e University of Chicago Press, 1921), p. 507; "Of the four great types :2 interactionucoupetition, conflict, accommodation, and assimilation-- etition is the elementary. . . ." 1. Phat are the necessary empirical conditions for a state of conflict to exist? Phat particular configurations of sets of individuals, values, and resources result in conflict relations? 2. Given a state of conflict, what attributes of this state vary in relation to others? What variables contribute to the intensity or strength of a conflict system? 3. that are the attributes of community social structure which condition conflict situations? 4. Is the race relations cycle, implying an irreversible pro- cess inevitably concluding in assimilation, an adequate model for the analysis of ethnic relations? What insights might be forthcoming by viewing ethnic relations as a special case of social conflict? Co t i utions The contributions of this thesis lie in two directions. First, there is the substantive area under study--the problem of French-English relations in Canada. This thesis recognizes the crucial role played by the FI'ench minorities outside of the Province of Québec. It is evident that there is a lack of studies in this area. Moreover, a sociology of ethnic relations can only benefit from materials based on situations “huh differ in many important aSpects from the situation in the United states, Secondly, empirical work focusing on social conflict is relatively rm1'9. Several attenpts at developing theories of social conflict have been made. Though most remain at a highly abstract level, they have pro- V idea the sources for the proposed model. The hypotheses proposed by -10- such students of the problem as Lewis A. Coser,19 Ralf Dahrendorf,20 and Raymond W. Mack21 have yet to be put to the test. This requires a pro- cess of testing, refining, and retesting: khat we need are statements of the direction and degree of associa- tion between sets of those variables [e.g., the intensity of con- flict, the nature of issues, etc.]; then we would be on the way to refining some of these hypotheses into prepositions.22 It is toward this end that this thesis is directed. Organization of the Thesis In the next chapter, the relevant literature is reviewed; the theoretical framework is set, and a series of hypotheses are advanced. Following this, in Chapter III, the research design is presented. Here, consideration is given to the deve10pment of empirical indicators and a set 0f testable hypotheses, and the methods of data collection are des- cribed. Chapter IV presents an analysis of the historical situation from which the conflict situation under study emerged. Chapters V and VI des- cribe the study coumunityuits peOple and its institutions. The empiri- Cal Conditions for conflict are identified in Chapter VII. In Chapters VIII and IX, the hypotheses are tested. A revision of the model in view of the findings is presented in the final chapter, Chapter X. \ 19 Th Lewis A. Coser, The Functions of Social Conflict. (New York: e Free Press of Glencoe, 1956). 20Ralf Dahrendorf, 02, cit, Pr 2l'Rrflymond H. Mack, "The Components of Social Conflict," Social \Och-mé, Vol. 12, Spring 1965, p. 396. zzlbid. CHAPTER II REVIEW OF THE LITERATURE AND DEVELOPMENT OF HYPOTHES ES The theoretical framework for this thesis is presented here in the form of a set of theoretical hypotheses. The terminology used by letter- t>erg is followed. "Theoretical hypotheses" are posited as propositions of high informative value requiring empirical support. "Ordinary (or testable) hypotheses" are prOpositions of low informative value also requiring empirical support.1 These latter hypotheses will be presented in Chapter III. It should be noted that the stated hypotheses were devel- Oped frOm two sources--the literature, as reviewed here, and data col- lected during the first phases of the research. The lack of empirical thunmrials in the literature necessitated a continuous interaction between theory and data throughout the research process. This was especially rlecessary in the formulation of the key variables and their indices. The thesis problem directed attention to three areas of sociolog- 1Gal theory: social conflict, comminity social structure, and intergroup Irelations with special reference to assimilation. In the first section c’f this chapter, the variables of conflict and their interrelations are e"eleped. Following this, a second set of hypotheses are constructed \ (.1, l'Hans L. Zetterberg, 0 Theor and Verification in Sociolo . hird enlarged edition; Totowa, New Jersey: The Bedminster Press, 1965), er. 79-86. The notion of informative value refers to the level of gen- ality of a proposition, "the higher the informative value . . . the ea'zer is the variety of events for which it can account." -11- -12- relating sapects of community social structure to the conflict variables. Finally, assimilation is viewed in relation to changes in conflict systems . I. SOCIAL CONFLICT No major areas of controversy are readily apparent in any exami- nation of the literature on social conflict. One opposes conflict to consensus and, depending on the views of a given polemist, attempts are made to demonstrate that either one or the other state is more basic to 809181 integration. Another involves those who would explain conflict from an individual or psychological perSpective as Opposed to those who work frOm a collective or sociological perSpective. A solution to or a 8Yntliesis of these positions is not an objective of this thesis. How- ever, the controversies cannot be avoided. In fact, they present a use- ful way in which to divide the literature into two broad categories. The 3 halo ical Versus the Sociolo ical Pers active In 1957. UNESCO published a comprehensive review of the litera- ture dealing with conflict.2 In this review, Professor Bernard refers I t° the social psychological approach as being ". . . through the individ- 3 “a1, his attitudes, opinions, and behavior pattern." This approach 19 \ (P 2The International Sociological Association, The Nature of Conflict. aria, France: United Nations Educational, Scientific and Cultural ganization, 1957). lb 3Jessie Bernard, "The Sociological Study of Conflict," in \1§., p. 56. readily illustrated by the following statements made by Professors Krech and Crutchfield in their Theory and Problems of Social Psychology: The problem of international tensions is ultimately a problem of the psychology of the individual. War is declared by and fought by individuals; peace and international unity are the products of the actions of individuals . . . the manner in which . . . economic, political, and other such influences work in international tensions is psychological; i.e., they become effective through . . . the needs, perceptions, beliefs, and attitudes of the individual. A similar view is cited in T. H. Pear's review of the psychological lit- erature pertaining to conflict. Referring to the work of H. J. Eysenck, Pear states that, he holds that ultimately all causative [italics mine] influences must find themselves reflected in the individual attitudes toward war and personal aggressiveness, built up by social pressure, by teaching and propaganda, by personal concept, by childhood experi~ ence, by parental emotional conditioning and the thousand-and-one agencies which determine our outlook. The key concepts in the psychological analysis of conflict appear to be "tension" and the "frustration-aggression" hypothesis. Thus, as Bernard notes, The [conflict] phenomena studied are sometimes viewed as tensions within the individual which from time to time, eventuate into 'open' conflict. That is resentments and frustrations, from whatever source, pile up within the individual until, in effect, they explode in overt aggression of some kind--in quarrelling, in fighting, in rioting, or what-have-you, as a means of reducing tension. 4David Krech and Richard Crutchfield, Theor and Problems of Social Psychology. (New York: McGraw-Hill Book Company, Inc., 1948), p. 586. 51. H. Pear, "The Psychological Study of Tensions and Conflict," in The International Sociological Association, 02, git., p. 144. 6Jessie Bernard, 22, git., p. 36. -14- The clash between the sociological and psychological perspectives arise when the tension concept is used to explain intergroup conflict. This position is attributed to Kurt Lewin whose work has had considerable influence on the analysis of social conflict. Whether or not Lewin did generalize his notion of "tensions" to the intergroup level is not too clear in his own work. His empirical referents remained at the level of infra-individual and intra-group conflict, although one may easily extract statements like the following from his work: . . . one of the most important factors in the frequency of conflict and in the building up of an emotional outbreak is the general level of tension at which the person or group lives. ihether or not a particular event will lead to a conflict depends largely on the tension level or on the social atmosphere in the groups.8 Any conflict creates tension, which leads to restlessness, unbalanced behavior, and over-emphasis in one or the other direction.9 The first two statements quoted above appeared in a treatment of the prob- lem of conflict within marriage; the third referred to individual "rest- lessness" and "unbalanced behavior" resulting from minority group member- Ship. With respect to intergroup conflict, Lewin stated in a 1943 arti- Cle dealing with Germany that ~. . . a fairly strong tendency was discernible to consider psycho- logical factors such as frustration or 'destructive traits' the basic causes for war. Accordingly, the avoidance of frustration was con- sidered the main road to peace. Since then a more realistic view of the importance of political and economic aspects seems to prevail.10 7Kurt Lewin, Resolvigg Social Conflict . (New York: Harper and Brothers, 1948), p. 89. 3mg. 9Ibld., p. 156. 1°Ibld., p. 43. -15- Nevertheless, a good deal of the research and amelioration pro- grams influenced by Lewin proceeded to generalize the individual tension concept to a concept of group tensions.11 Programs directed at improv- ing intergroup relations attempted to break down collective tensions through changing attitudes and values. Williams noted some of the assumptions iuplicit in these programs: 1. One of the most obvious of these premises guiding strategy is, in its least sophisticated formulation, 'Give the people the facts and prejudice will disappear.' 2. Action should be directed toward a direct change in values or attitudes. 3. Contact brings friendliness. .4, . . . experience changes behavior and there is a transfer of the changed behavior to other, more usual, types of situations.12 A similar generalization of the tension and frustration-aggression concepts appears in some of the literature dealing with community and etI'lxiic conflict. B. E. Mercer, for example, explains community conflict \ L llJessie Bernard, op, cit., pp. 59-60. ‘: 12Robin M. Williams, Jr., The Reduction of Intergroup Tensions. FNew York: Social Science Research Council, Bulletin 57, 1947), pp. 13-16. 1301‘ an example of a program based on these assumptions, see Hedley G. Eiulock, "Intercultural Relations Training: An Experiment in Citizenship Sducation." (Montreal: Centre for Human Relations and Coumunity Studies, 11:- George Williams University, 1964). (Mimeographed.) Although the a‘llflior duly noted that "conflict and hostility between groups is an - Iitergroup phenomena and cannot be explained in terms of hostile or pre- J‘Jdiced individuals. . ‘. . ," he went on to state that "citizenship or character is a set of attitudes and values . . . [and that] the goals of the project . . . [were directed toward] developing a better understanding among Indians and non-Indians, and increasing the citizenship and leadership capacity and responsibilities of all." -16- in terms of ”aggressiveness on the part of individuals." Accordingly, conflict may be controlled or reduced "through a reduction in hostility, the rechanneling of hostility to other subjects, or through the repression of overt conflict.“3 CorreSponding to this treatment of conflict, Shibutani and Kwan in a consideration of ethnic stratification state that "conflict . . . is an incidental part of the process of social Izliaange." Conflict is viewed as developing from "collective tensions [emerging] as frustrated men communicate with each other, reinforcing one i”izxother's apprehensions and anger."14 In a critique of this position, Bernard is of the opinion that "1 1 ttle definitive evidence has . . . been adduced to show that 'group temions,’ as either siuple or as weighted additive functions of individual tetisions, can explain inter-group relations."15 The problem, then, lies tn(>11 in the generalization of prOpositions from the individual level to the level of intergroup conflict than in any intrinsic fallacies in the cc>l'1cept of individual tensions. Certainly, psychologists and social psy- ct‘Ologists whose work touches on both individual and social conflict are a"were of this difficulty. Sherif, in his treatment of Intergroug Rela- t10 s and Leadershi draws attention to the issue when he points out that \ 1 133. g, Mercer, The American Communit . (New York: Random House, 956), pp. 178-179. 14T. Shibutani and K. M. Kwan, Ethnic Stratification: A Compara- Liive Approach. (New York: The MacMillan Company, 1965), p. 401. 15Jessie Bernard, 02. cit., p. 116. -17- we cannot extrapolate from the properties of individuals to the characteristics of group situations. It is equally erroneous to extrapolate from the properties of relations within a group to explain relations between groups. . . . Intergroup relations are patently determined by the process of interaction between the groups.16 .Likewise, T. H. Pear notes that "'psychological' and 'sociological' sig- ni fy two different levels of approach, necessitating their apprOpriate Iarrits of treatment and consequently appropriate conceptual tools."17 What then is the sociological view of conflict? According to 13<317nard, "the sociological conceptualization of conflict is in terms of 1:1113 relationship between or among systems."18 A relationship between CJJE' among groups defined as conflictive emerges from incompatible posi- 20 tilczns,19 mutually exclusive values, resource scarcity in terms of status, Power, and other means to achieving desired ends,21 22 and boundary main- t enance activity. \ 16Muzafer Sherif (ed.), Intergroup Relations ang_Leadership. (New York: John Wiley and Sons, Inc., 1962), pp. 8-9. 17T. H. Pear, Op, cit., p. 164. 18Jessie Bernard, op, cit., p. 37. 19Jessie Bernard, "Parties and Issues in Conflict," The Journal \Qf Conflict Resolution, Vol. I, No. 2, June 1957, pp. 111-112. See also Ralf Dahrendorf, pp, cit., p. 135; Kenneth E. Boulding, Copfllcc \ahg Defense. (New York: Harper Torchbooks Edition, Harper and Row, 1963), p. 5. 20Jessie Bernard, "The Sociological Study of Conflict," gp, cit., p- 38; Jessie Bernard, op. cit., p. 112; Robin M. Williams, Jr., ~SE£L‘_31§., p. 43; and Lewis A. Coser, Op, cit., p. 8. 21Lewis A. Coser, Ibid. ”Raymond 11. Mack, op, cit., p. 394. -l8- Implicit in this conceptualization is the notion of cost:23 parties to a conflict situation "cannot have 511 they want of anything."24 Whatever move one or another party makes implies gains or loses for each. Assuming rational behavior, an analogy between conflict so defined and games of strategy is drawn. In this manner, the theory of games of strat- egy is used as a model for the analysis of social conflict,25 the empha- sis being placed on the processes of bargaining and accommodation.26 Accordingly, social conflict is perceived as a consequence of structural processes rather than as a result of frustration produced tension and aggressive behavior, repressed hostilities, and misunderstand- ings resulting from poor communication. An illustration used by Lewis Coser is relevant here: A worker engaged in strike activity in order to increase his wages, his status or the power of his union, and one who releases aggres- sion against the boss because he perceives him as an oedipal figure, are dissimilar social types. Displaced father hatred may attach itself to any suitable object-~boss, policeman or staff sergeant. The economic fight of workers against the boss, on the other hand, is based on their articularppositions and roles in thg_economic and poligical system. [italics mine.]727 -— 23Jessie Bernard, "The Sociological Study of Conflict," OE, cit., p. 38. 24Raymond W. Mack and Richard C. Snyder, "The Analysis of Social Conflict-~waard an Overview and Synthesis," The Journal of Conflict Resolution, vol. I, No. 2, June 1957, p. 218. 25Jessie Bernard, "The Sociological Study of Conflict," op. cit., p. 100. 26Thomas C. Schelling, The Strategy of Conflict. (New York: Oxford University Press, 1963), pp. 4-5. 27Lewis A. Coser, pp, cit., p. 50. -19- Notwithstanding the polemics, the psychological and sociological views are not necessarily contradictory nor should one be Opposed to the other. Intra-individual, inter-individual, intra-group, and intergroup conflict are different types of conflict. Whether a general theory of conflict encompassing these and other types is possible or not is another issue. The psychological conceptualization is more apprOpriate for the first two, the sociological for the last type. The subject material of this thesis is intergroup or ppplpl conflict, and as such adopts a socio- logical in contrast to a psychological conceptual framework. Although the emphasis in this thesis is on structural variables, a complete view would require that the effects of personalities on a con- flict system be taken into consideration. It would be necessary, for example, "to make a distinction between what is due to individuals fill- ing the various roles in society and what is due to these roles them- selves."28 This distinction will be made through the concept of conflict ,gggp§_. In any case, one should be cognizant of the fact that it is only analytically that a separation of the psychological and sociological com- ponents of social conflict can be made. Capflipt Versus Consensus It is difficult to enter into the sociological literature on con- flict without finding oneself immersed in opposing ideological positions 28Raymond Aron, "Conflict and war from the Viewpoint of Historical Sociology," The International Sociological Association, 0p, cit,, p. 178. presented as answers to the question: "What holds society together?" In contemporary sociology these contrasting positions are treated under several labels--consensus versus conflict,29 Hegelian-Marxian dialectic versus functionalism,30 order versus conflict theories,31 integration 32 versus coercion theories, etc. The assumptions implicit in each posi- tion are summarized by Dahrendorf.33 29Irving L. Horowitz, "Consensus, Conflict and C00peration: A Sociological Inventory," Social Forces, Vol. 41, 1963, pp. 177-188. 3OPierre L. van den Berghe, "Dialectic and Functionalism," Amerigan Sociological Review, Vol. 28, October 1963, pp. 695-705. 31John Horton, "Order and Conflict Theories of Social Problems as Competing Ideologies," The American Journal of Sociology, Vol. 71, May 1966, pp. 701-713. 32Ralf Dahrendorf, op, cit., pp. 157-165. 33Ibid., pp. 161-162. "[Integration theory] is founded on a number of assumptions of the following type: 1. Every society is a relatively persistent, stable structure of elements. 2. Every society is a well-integrated structure of elements. 3. Every element in society has a function, i.e., renders a con! tribution to its maintenance as a system. 4. Every functioning social structure is based on a consensus of values among its members. . . . the coercion theory of society [assumes that], 1. Every society is at every point subject to processes of change; social change is ubiquitous. 2. Every society displays at every point dissensus and conflict; social conflict is ubiquitous. 3. Every element in a society renders a contribution to its dis- integration and change. 4. Every society is based on the coercion of some of its members by others." -21- A danger inherent in opposing the two perspectives is a tendency to adopt one or the other as an exclusive model for the analysis of social situations. In so doing, one moves closer toward formulating an ide010gica1 position and further from sociological theory.34 Dahrendorf points out that neither of these models can be conceived as exclusively valid or applicable. They constitute complementary, rather than alternative, aspects of the structure of total societies as well as every ele- ment of this structure. Indeed, as noted by the same author, it is next to impossible to deal with one perspective without taking the other into consideration. Conflict implies contact between parties--parties ". . . 'integrated into' a common frame of reference."36 In this research a conflict perspective was adopted, but not in opposition to an integration or consensus perSpective. Theoretically, the question, ". . . under what conditions [do] various kinds and degrees of conflict emerge?" was addressed to the data.37 It requires no stretch of the imagination to see that this question may be equally well phrased 34See Raymond w. Mack, op, cit., p. 390. "To phrase a theoretical issue in sociology as functionalism versus Marxism on the grounds that the functionalists assume social inte- gration whereas Marxists assume social conflict, is to ignore the fact that both are invalid as scientific theories. . . . There is nothing inhibiting to the study of conflict in functional analysis if by functional analysis we mean the attempt to describe and under- stand social structures and their consequences." 35 361bid., p. 164. Ralf Dahrendorf, pp, cit., p. 163. 37Amitai Etzioni, Stpdies in Social Change. (New York: Holt, Rinehart, and Winston, 1966), p. 112. as ". . . under what conditions [do] various kinds and degrees of conflict- 1imitation occur?"38 This latter phrasing implies integration. The second question is the obverse of the first. Substantively, an inter- ethnic situation is viewed in conflict terms rather than in terms of the more commonly employed "race relations cycle" model. Why? Simply because it is the conflict aspects of the situation under study which have here- tofore been neglected. The primary purpose of this research is to develop an alternative model for the analysis of intergroup relations, a model stressing separation as opposed to assimilation. In 1950, Jessie Bernard could ask, with some trepidation, "Where is the modern sociology of conflict?"39 Since that date several major works40 and a journal41 devoted to social conflict have appeared. This material is rich in perspectives, models, and hypotheses, but lacking in empirical work capecially in relation to social structure and social organization. This thesis is an attempt to partially remedy this situa- tion. Before proceeding with the task of defining concepts and develop- ing hypotheses, it would be well to state the basic assumptions upon which 38lpid. 39Jessie Bernard, "Where is the Modern Sociology of Conflict?" Americap Journal of §ppiology, Vol. 56, July 1950, pp. 11-16. 40Lewis A. Coser, op, cit.; The International Sociological Assoc- iation, 0p, cit.; Ralf Dahrendorf, 0p, cit.; Kenneth E. Boulding, op, cit.; Max.Gluckman, Custom and Conflict in Africa. (Oxford: Basil Blackwell, 1963); Pierre L. van den Berghe, South Africa: A Study in Conflict. (Middletown, Connecticut: hbsleyan University Press, 1965). 41The JOppnal of Copflict Resolution. Published quarterly by The Department of Journalism, University of Michigan, Ann Arbor, Michigan. -23.. the preposed model is based. First, the position which holds that social conflict is a result of individual maladjustment, misunderstandings resulting from poor com- munication, and frustration produced aggression is rejected. These factors may enter into a conflict system as variables, but are not to be taken as the source of social conflict. Social conflict emerges from boundary maintaining activity and the unequal distribution of resources. Second, the notion that conflict is necessarily dysfunctional or "patho- logical" is rejected. Social conflict may be either functional or dys- functional. Third, the notion that conflict necessarily entails an absence of integration, consensus, or c00peration is rejected. Conflict implies integration and vice-versa. Tb these assumptions may be added five additional ones posited as "empirical prOpositions" by Robin M. Williams, Jr., in answer to Horton's treatment of order and conflict theories as competing ideologies: 1. A11 interacting human papulations exhibit . . . social conflicts. 2. All interacting human populations that remain in interaction over time develop normative regularities. 3. All interacting human papulations manifest some continuity of social patterns. 4. All interacting human pepulations manifest change in social patterns, over time. 5. All interacting human pOpulations show both coerced and volun- tary conformity.42 42Robin M. Williams, Jr., "Some Further Comments on Chronic Coruxoversies," American Journal of Sociology, Vol. 71, No. 6, May 1966, p. 718. -24- Social Conflict Defined43 Social conflict is defined as consisting of any relationship between two or more sets of individuals exhibiting some level of organi- zation and incorporating one or more of the following conditions: incom- patible values, position scarcity, or resource scarcity. This defini- tion is similar in content to those prOposed by Coser,“4 Bernard,45 Dahrendorf,46 and Boulding.47 It suggests three empirical conditions which must be met if a relationship between two or more sets of indi- viduals is to be designated as conflictive. There must be parties, that is, sets of individuals exhibiting some level of organization. There must be some level of interactipp between these parties; without contact there can be no conflict. Finally, a condition of inceppatible values, and/or position anp resource scarcity must be present.48 Conflict parties. As obvious as these conditions may be, they are basic to a description of a conflict system. First, with reference 43Portions of the next three sections of this chapter are taken from the following paper by this student: John D. Jackson, "A Study of Ftench-English Relations in an Ontario Community," The Canadian Review 2! SopiOIOgy and Anthr0pology, Vol. 3, No. 3, August 1966, pp. 117-132. 44Lewis A. Coser, op, cit., p. 8. aSJessie Bernard, "Parties and Issues in Conflict," pp, cit., p.112. 46Ralf Dahrendorf, op, cit., p. 135. 47Kenneth E. Boulding, pp, cit., p. 5. 48Though not as elaborate, these conditions follow closely those listed by Mack and Snyder in a recent synthesis of materials on social conflict. Raymond W. Mack and Richard C. Snyder, op. cit., pp. 212-248. -25- to parties: a statistical category cannot be a party to a conflict situa- tion--a minimum level of organization is required. The category of ,-’49 Quasi-group, as advanced by Ginsberg and used by Dahrendor would appear to meet the minimum requirement. Ginsberg defines quasi-groups as: . . . aggregates or portions of the community which have no recog- nizable structure, but whose members have certain interests or modes of behavior in common, which may at any time lead them to form them- selves into definite groups. To this category of quasi-groups belong such entities as social classes [and, I would add ethnic groups] which, without being groups, are recruiting fields for groups.50 The degree of organization beyond this minimum requirement is an empiri- cal problem. Party contact and interaction. To Robert Park, conflict was "one of the four great types of interaction."51 Social conflict exists only insofar as the parties come into contact with each other in the pursuit of positions, resources, or value states. Given contact, interaction may occur at several levels of organization. In the case of inter-ethnic conflict, interaction may take place: (1) at the interpersonal level where individuals react to one another in terms of their ethnic membership; (2) at the level of conflict agents where persons designated as representa- tives of quasi-group interests meet to articulate the interests of their respective groups; and (3) at the associationsl level where conflipt agepcies, associations or organizations Specifically act to articulate 49Ralf Dahrendorf, pp, cit., p. 180. ”Morris Ginsberg, Sociology. (london, England: 1953), p. 40. 51Robert E. Park and Ernest W. Burgess, loc, cit. -26.. the interests of their respective groups. It is the latter two levels which are of primary interest in a consideration of intergroup relations. Generally, conflict relations are carried on by conflict agents and agencies. Drake, for example, specifies the necessity of "an organ- ized structure . . . for carrying on conflict."52 Williams refers to such agents and agencies as the necessary mediating links between parties involved in a conflict situation.53 Conflict agencies and agents are not, however, a requirement for a state of conflict to exist. It is through an identification of agencies and agents that the type of contact and forms of interaction between the parties may be specified. A conflict agency may be defined as an interest group; a group whose actions or policies indicate the presence of a conscious desire to move public policy--the allocation of values, or the allocation of resources in a particular direction.54 Within the limits of this thesis, any voluntary association which is ethnically enclosed is a potential conflict agency. Such associations may exhibit one or more of the follow~ ing functions : l. A supportive function consisting of actions or policies limited to increasing internal solidarity. 528t. Clair Drake, "Some Observations on Interethnic Conflict as One Type of Intergroup Conflict," The Journal of Conflict Resolution, Vol. 1, July 1957, p. 162. 53Robin M. Williams, Jr., "Racial and Cultural Relations," J. B. Gittler (ed.), Review'of Sociology: Analysis of a Decade. (New York: John Wiley and Sons, Inc., 1957), p. 451. 54Joseph La Palombara, Interest Grogps in Italian Politics. (Princeton, New Jersey: Princeton University Press, 1964), p. 16. -27- 2. An interest-articulating function consisting of actions which place party interests before the Opponent or the general public.55 3. An interest aggregating function wherein the interests of several other associations are drawn together to be articulated at a higher level.56 Not all ethnic associations exhibit all three functions. The supportive function can exist alone, but the interest aggregation and/Or articulation functions must be present if an association is to be designated as a con- flict agency. A labor union is an example of a conflict agency repre- senting worker interests. In a similar manner, SNCC, CORE, and the NAACP are conflict agencies within the context Of American race relations. Conflipt agepts serve a similar function, but refer to status- role complexes within institutional structures which function, either latently or manifestly, to articulate party interests.S7 As will be seen, these phenomena are clearly illustrated in the community under study where an English-speaking and a French-speaking inspector are appointed by the Provincial Department of Education to "represent" the interests of each group in the local school district. The incumbents Of these positions are not representatives of particular conflict agencies, but function to represent quasi-group interests in general. 55%., pp. 78-84. 56Ibigo. pp. 84-85. 7Merton's conceptualization of status-role is used here. Status 18 viewed as "a position in a social system occupied by designated indi- V1 duals." Role refers to the behavioral enacting of the patterned expec- tations attributed to that position. Robert K. Merton, Social Theory ppd Social Structure. (Revised edition; New York: The Free Press of Glencoe, 1957), pp. 284-286. -28- Ippogpatible values. position and resource scarcity. These con- ditions represent the foci Of a conflict situation: the answer to the question, "What is the conflict about?" A condition of position scar- plpy_arises when, for example, desirable positions in the social struc- ture are distributed according to party membership. It is very difficult in Canada for a person to be both a French Canadian and a corporation executive. The under-representation of French in relation to English- speaking Canadians in higher status occupations is well documented in John Porter's The Vertical Mosaic.58 Resource scarcity is a question of supply and demand; both parties cannot have all they want of desirable objects or "states of affairs."59 In the case under study, the school issue is relevant here. Roman Catho- lics and Protestants operate their own school systems, each drawing upon public resources-~a limited pool of monies, materials and personnel. These resources have been unequally distributed between the Protestant or "public" schools and the Catholic or "separate" schools, the former receiv- ing the greater share. This situation would be designated as a condition 0f resource scarcity. Empirically, it is difficult to conceive of a "pure" value conflict. .ipppppatiple values are usually compounded with position and resource Scarcity.6o The separation is, of course, analytically useful. It enables 58John Porter, The Vertipal Mosai . (Toronto: University of Toronto Press, 1965), pp. 91-98. 59Rayunnd W. Mack and Richard C. Snyder, Op, cit., p. 218. 60Ibid. -29- one to examine the manner in which contradictory value positions are related to position and resource scarcity. Kenneth Underwood, for example, approaches a religious conflict in a New England community from the per- spective of fundamental theOlOgical differences between the parties involved, but proceeds to relate these to an unequal distribution of posi- tions and resources.61 Gusfield analyzed the value positions of the "wets" and "drys" in the prohibition movement in terms of status groups.62 In the situation under study, the two main parties--French and English Canadians--assume contradictory positions with regard to assimi- lation. Complete assimilatiOn and complete separation are incompatible; one is the contradiction Of the other. Each party tends to move toward these extremes, English Canadians toward the former, French Canadians toward the latter. The actual state is one Of p1ura1ism--just how much separation is to be allowed (from an English-Speaking view) or how much is to be demanded (from a French-speaking view) is the issue. Thus, the empirical conditions or prerequisites for a conflict situation are these: if two or more quasi-groups (i.e., parties) are in contact and interacting under conditions Of position and resource scarcity or value incappatibility, then a state Of conflict exists among them. Social conflict, therefore, is defined as a structural state rather than 61Kenneth Underwood, Protestant and Catholic. (Boston: The Beacon Press, 1961) . 62J. R. Gusfield, Symbolic Crusade: Status Politics and The Arnerican Tepperance Movement. (Urbana, Illinois: The University Of Illinois Press, 1963). -30.. as a variable. This point Of departure demands that the variables Of conflict be defined and their interrelations specified. Inventories of variables appear in Coser,63 Dahrendorf,64 Mack and Snyder,65 and Williams.66 The selection of variables for this research was based partly on their importance in the literature and partly on their applica- bility to the situation under study. The Variaples of Conflict The four selected variables are members of what may well deveIOp into three classes of variables as empirical studies concerning social conflict increase. The first class has to do with the strength of the conflict and involves the variables Of intensity and militan y. The second class, involving institutionalization, is concerned with varia- tions in the regulation and formalization of the conflict relationship. The third class refers to the broader social context within which a particular conflict is located. The superimposition of several poten- tial conflict fronts (e.g., class, ethnicity, religion, urban-rural, etc.) is a member of this class. Iptensity and militancy. The strength of a conflict appears to 63Lewis A. Coser, pp, ci . 64Ralf Dahrendorf, Op, cit. 65Raymond W. Mack and Richard C. Snyder, Op. cit. 66 Robin M. Williams, Jr., Strapgers Next Door: Ethnic_Re1ations in American Conrnu ities. (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1964). vary along at least two dimensions which, though closely related, exhibit some degree of independence. In the community under study, several hours were spent attending meetings and conferences Of associations devoted to the articulation of local French Canadian interests. The numbers of peOple involved and the frequency of such meetings indicated a level of involvement at least beyond that which might be termed "low." At the same time, the sessions were extremely calm; there was an absence of threatening or intimidating statements directed toward Opponents; few press releases were made, and there had been no efforts to circulate petitions or organize demonstrations. In other words, the extent tO which members of one of the parties were involved in conflict agencies appeared to be fairly extensive. However, in terms Of tactics utilized, the actions of these agencies were extremely tranquil. These two aspects or dimensions of conflict behavior suggest the variables of intensity andmilitancy.67 ,Ipppppipy refers to the magnitude of party involvement.68 If all nmmbers of a quasi-group in a state of conflict with some other such group '67This distinction is based on Dahrendorf's conceptualization Of intensity and violence which appeared unsatisfactory to the data in Question, op, cit., pp. 212-213. His notion of violence, for example, included all "manifestations" Of conflict from peaceful discussion to the actual use of physical force. In other words, non-violent demon- strations would be included under the category Of violence! 68The meaning of the French word intensité is a closer approxima- timnlto the meaning intended here than the general meaning of "intensity" 111 English. The Larousse Dictionary defines intensité as "Expression de la valeur numerique d'une grandeur." This suggests the concept Of size orlmagnitude. Intensite' is readilytranslated into the English word vo 1mm. -32- were participants through conflict agencies, the level of intensity would be at a maximum. This maximum is seldom reached, but the degree of intensity does seem to vary according to certain conditions. These conditions will be specified later. In contrast, militancy refers to the nature of Specific acts directed against the Opponent rather than the extent to which quasi- group members are involved in these acts. Militancy is the use of coercive measures in pursuing party interests. A threat, the presenta- tion of petitions, the calling of a strike or demonstration are coercive measures. These measures or acts would be designated as indicators of a low to moderate militancy. The use of violence, or physical force with intent to harm or injure, would constitute a high level of militancy. Certainly the two variables are closely related. A certain degree of intensity would appear to be a prerequisite for militant behavior. However, intensity does not necessarily imply militancy. A high level Of intensity is possible without militant acts as a consequence. Referring to Table I, cells one, two, and four are empirical possibilities. The state suggested in cell three is assumed to be empirically rare. The TABLE I THE RELATIONSHIP BETWEEN MILITANCY AND INTENSITY ====================================:==================================== Militapgy ___Intensity High Low High 00000000000000 1 -- i 2 Low OOOIOOOOOOOOI. 3 a conditions represented in cells one, two, and four arise according to the influence Of other variables, both external and internal to a conflict system.69 The institutionalization of conflict. The regularization of con- flict situations is a continuous theme in the literature. According tO Horowitz, some "conflicts [are] programmed for continuation."70 Robert Dubin in a consideration of industrial conflict suggests that institution- alization is an inevitable outcome of continuous conflict situations.71 To Dahrendorf, "conflict regulation" is "in many ways [a] most crucial factor affecting the empirical patterns Of class conflict. . . ."72 The institutionalization Of conflict entails the regularization Of the relationship between the participating parties. An institutionalized conflict is one in which a set of regulations (i.e., rules or norms) and a set of positions and roles emerge to define and stabilize the relation- ship between the parties, permitting each to Operate in relation to its Opponent with some predictability.73 Accepting Parson's statement that 69Cells one and four, reapectively, correspond to a commonly made distinction between overt and covert conflict. In some cases, conflict per se is defined in terms Of cell one, cell four being referred tO as potential conflict or latent conflict. 70Irving L. Horowitz, op. cit., p. 183. 71Robert Dubin, "Industrial Conflict and Social Welfare," The .Journal of Confligt Resolution, Vol. 1, June 1957, p. 187. 72Ra1f Dahrendorf, cg, cit., p. 223. 73The reference to both rules and positions is consistent with the “rare general treatment Of institutionalization in the literature. See "institutionalization . . . is clearly a matter Of degree,"74 it follows that some conflict situations may be more institutionalized than others and, over a period of time, any one conflict may exhibit variations in institutionalization. In this sense institutionalization is viewed as a conflict variable. Labor-management relations provide a ready example of institution- alized conflict. Both parties are recognized and sanctioned; their interests are recognized and formal channels Of communication are avail- able for the articulation Of these interests. Once institutionalized, a conflict is more likely to continue as the parties develop a mutual inter- est in each other's survival.75 Thus, institutionalization does not resolve, in the sense of eliminating, social conflict. On the contrary, institutionalization implies a continuation of conflictive relations, regulated but not resolved. Indeed, according to the definition of con- flict advanced in this thesis, only the elimination of one Of the parties, through its withdrawal or destruction, resolves the conflict. If the parties remain intact, resolution may be thought Of in relation tO issues, but not in relation to the dissolution of the conflict system itself.76 L;- Talcott Parsons, The Social System. (New York: The Free Press Of Glencoe, 1951), pp. 36-39; Harry M. Johnson, Sociology: A Systematic Introduction. (New York: Harcourt, Brace and Nbrld, Inc., 1960), pp. 19-21. 74Talcott Parsons, Op, cit., p. 39. 750. Ram-Freund, "Intergroup Conflict and Their Settlement," finhe British Journal of Sogiology, Vol. 5, 1954, p. 201. 76Ralf Dahrendorf, op, 91c., pp. 223-224. -35- The theoretical significance of institutionalization as a conflict variable is that it provides a connecting link between conflict and inte- gration models. Through the institutionalization of social conflicts, conflict is contained within an integrated framework, or from the alter- nate parapective, social conflict insofar as it becomes institutionalized contributes to integration. Integration does not necessarily imply the absence Of conflict nor does conflict imply the absence Of integration. However, not all social conflicts become institutionalized nor, once institutionalized, will they necessarily remain so. The immediate inter- est in this thesis is not so much in conditions contributing to institu- tionalization, but in the effects Of variations in institutionalization upon a conflict system. Institutionalization is taken as an independent variable. The superimposition of ponflict categories. Several theoretical studies of social conflict77 have noted the coincidence Of significant social categories as a highly influential variable. That is, to what extent are such potential conflict fronts as class, religion, ethnicity, and political affiliation superimposed, creating a single, dominant clea- vage? To use a concrete exauple, if all and only French Canadians were Catholic, lower class, and voted socialist, and were confronted with an 77James C. Coleman, Community Conflic . (New York: The Free Press of Glencoe, 1957), p. 22; Lewis A. Coser, op, cit., p. 76; RObin M. Hilliams, Jr., Strangers Next Door. OpI cit., p. 369; Ralf Dahrendorf, op. pit., p. 213; Max Gluckman, Op, cit., pp. 1-4. -35- English Canadian population, all the members of which were Protestant, upper class, and voted conservative, there would be a high level of superimposition. A complete superimposition Of potential conflict fronts is perhaps empirically rare in industrial societies, but it is not diffi- cult to accept the prOposition that the degree of superimposition will influence the intensity of a conflict situation. Certainly, the energies involved or expended by conflict groups with overlapping memberships would reduce the potential for a highly intense and militant confronta- tion between any two. Coser suggests that such segmental participation . . . can result in a kind of balancing mechanism, preventing deep cleavages along one axis . . . the multi- plicity of non-cumulative conflicts provide one . . . check against basic consensual breakdown in an Open society. The empirical focus is, of course, individuals and their member- ships in various relevant categories. For example, an English Catholic in the community under study is subjected to at least one set Of cross- pressures. As an English-Speaker, he is open to the interests and pre- sures of English Canadians vis-a-vis French Canadians. As a Catholic, he is Open to the interests and pressures of Catholics, most of whom are French, vis-a-vis Protestants, most of whom are English. The presence of English Catholics suggests a lower level of superimposition between religion and ethnicity than if all English-Speakers were Protestant. The fact, however, that few French-Speakers are Protestant maintains a cer- tain level of superimposition. 78Lewis A. Coser, Op, cit., p. 79. The notion of cross-pressures and their effects on an individual's tendency toward reduced participation in relevant issues is a common explanatory device in voting studies. These studies have demonstrated that as "cross pressures increase, the individual tends to delay his voting decision and to become less interested in the election."79 Lipset suggests that such cross-pressures are basic for the maintenance of stable political systems.80 The concept of "status consistency" is also relevant.81 If the degree of superimposition of major conflict cate- gories in a given situation is high, it would follow that the proportion of individuals in the pOpulation exhibiting status consistency would also be high. Theoretical Hypotheses Having delineated four variables of social conflict, a set of hypotheses will now be constructed to deal with their interrelations. Institupiogalization and superimposition are taken as determinants; inten- sity and militancy are taken as resultants. The procedure adopted is to state in this chapter a set Of theoretical hypotheses relating the basic concepts. In the next chapter, a sub-set of testable hypotheses will be 79Seymour M. Lipset et al, "The Psychology of Voting: An Analysis Of Political Behavior," Gardner Lindzey (ed.) Hapdbook of Social Psyghology, Vol. II. (Reading, Massachusetts: Addison-Wesley Publishing Campany, Inc., 1954), p. 1133. 80Seymour M. Lipset, Political Man. (Garden City, New York: Doubleday and Company, Inc., ) p. 31. 81Leonard Broom, "Social Differentiation and Stratification," RObert K. Merton, Leonard Broom, and Leonard S. Cottrell, Jr., Sociology 232331. (New York: Basic Books, Inc., 1959), p. 431. ". . . a person's -33- stated utilizing empirical indicators of the basic concepts. Institutionalization and militan_y. A conflict which is institu- tionalized to one degree or another implies that the parties and their claims are recognized as legitimate and that the means are provided for the articulation of their claims or interests. Since militant behavior is generally undertaken as a means of demonstrating power and demanding recognition, it would seem to follow that conflicts, once institutional- ized, will exhibit less militancy. Militant tactics have been used by American Negro organizations at the local community level as a means of attaining recognition. Once negotiations between representatives Of these organizations and "the white power structure" are achieved, the militant behavior is usually reduced. This is not to say that militant tactics completely disappear from the scene, but there is a reduction, eSpecially Of violent forms.82 Lieberson and Silverman in a recent article on racial riots note that "a more responsive government [i.e., city government] makes riots less likely because it provides regular instigptional channels for expressing grievances." [italics mine.]83 Thus, the following hypothesis: statuses are considered inconsistent if he is 'high' on ethnicity and education but 'low' on occupation and income." 82$ee Ralf Dahrendorf, op, cit., p. 228; and L. Killian and C. Grigg, Rapial Crisis in America. (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1964), p. 136. 83Stanley Lieberson and A. R. Silverman, "The Precipitants and Underlying Conditions of Race Riots," American Sociological Review, Vol. 30, December 1965, p. 896. -39- Theoretical Hypothesis I The institutionalization of the relations between conflicting parties reduces the level of militancy exhibited by both parties. Superimposition and intensity. The problem of a possible rela- tionship between institutionalization and intensity is left Open. The extent to which potential conflict categories are superimposed in any given conflict situation would seem to have more effect on intensity than the degree of institutionalization. Involvement may be just as high in an institutionalized as in a non-institutionalized conflict. However, following Dahrendorf, the level Of superimpositiOn is taken as having the "utmost importance for variations in the intensity [italics mine] of a . . . conflict."84 . PrOpositions and findings from recent studies on voting behavior, the consequences of status inconsistency, and social conflict suggest that those subjected to cross-pressures resulting from membership in groups with conflicting interests-tend to withdraw from participation or exhibit less than a complete commitment to any one group.85 Robin Williams, Jr., expresses the relationship suggested here in the following manner: . . . a major factor in reducing conflict between groupings is the maximizing of cross pressures within individuals through multiple overlapping memberships.86 84Ralf Dahrendorf, pp= cit., p. zlS. 85Supra, p. 37. 86Robin M. williams, Jr., Strangers Next Door . . . , op. cit., p. 370. -40- Max Gluckman refers to the same phenomena in an analysis of the effects Of cross-pressures resulting from an overlap Of community and kinship loyalties upon conflict intensity among the Nuer.87 In line with these treatments of the problem, the following hypothesis is advanced: lhggretical Hypothesis II The greater the extent to which potential conflict categories are superimposed, the higher the degree Of intensity exhibited by any single conflict front. Summapy. Social conflict has been defined as a structural state-- as a type of relationship between identifiable quasi-groups or social systems. Two hypotheses were prOposed to deal with variations in conflict states. Though the model is sufficiently general to suggest applicabil- ity to any intergroup conflict situation, its use here is limited to inter-ethnic conflict. An additional problem with which this thesis is concerned is the extent to which certain local or community conditions influence the course of ethnic conflict. For example, the hypothesized connections between institutionalization and militancy may vary accord- ing to structural features exhibited by various local communities. II. THE COMMUNITY AND CONFLICT One need only cite Hillary's work of a decade ago to illustrate the range of materials which have been subsumed under the notion of 87Max Gluckman, pp, cit., pp. 1-26. .o..\ 1“ -41- community. After reviewing ninety-four definitions, Hillary reaches the conclusion that "beyond the cOncept that people are involved in community, there is no agreement as to the nature of the community."88 Apart from the elusiveness of the concept itself, many social scientists today hold that the community, however it may be conceived, is no longer relevant in "the mass society." The develOpment of the megalOpolis, regional decision- making agencies, and the increasing centralization in state and federal governments are pointed to as evidence of a decreasing relevance of local institutions. That is, all the important decisions affecting the lives of peOple are no longer made at the community level. That the autonomy, independence, and solidarity of local communi- ties are declining, there can be no doubt. This trend was demonstrated in community studies done as long ago as Warner's Yankee City series.89 More recently, the trend was taken as the very basis Of Vidich and Bensman's analysis in Small Town 1p Mass Society.90 Stein91 and warren92 attempt , 88G. A. Hillery, "Definitions of Community: Areas of Agreement," W. Vol. 20, June 1955, p. 119. 89W. Lloyd Warner et al., Yapkee City. (New Haven, Connecticut: Yale University Press, one volume abridged edition, 1963). The field work upon which the series was based was started in 1930. The initial volumes were published between 1941 and 1947. 90A. J. Vidich and J. Bensman, Small Town in Mass Society. (Garden City, New York: Anchor Books, Doubleday and Company, Inc., 1958). 91Maurice Stein, The Eclipse of Community. (Princeton, New Jersey: Princeton University Press, 1960). . 92Roland L. warren, The COmmunity in America. (Chicago: Rand McNally and Company, 1963). -42.. to deal with the problem at a more abstract level. Yet in Spite of this trend, pOpulations do cluster. Population clusters are observable pheno- mena, ranging from small isolated towns to large metrOpolitan centers. This is the geographic base of community, the Spatial anchoring point. Combined with this are patterns of relationships into which people enter by virtue of their clustering together in some location. This is not to suggest a geographic or ecological determinism. It is simply noted that population clusters are observable and that they have social and cultural relevancy. People do send their children to school, attend churches, pursue occupations, join associations, visit friends and relatives, strike, riot, live and die within some observable locality. Thus, following warren: Peoples' lives and their behavior are significantly influenced by their propinquity. . . . The intertwining of their lives on a locality basis, even in these days Of specialized interests, urban anonymity, and depersonalization, provides an important social reality and an important focus of study.93 This is the basis of the conception of community as used in this research. The relevance of community and the delineation of a community "is revealed in the functioning of the institutions which peOple have created: family, school, church, associations, etc."94 Commupity Variables DemOgraphic factors and superimpositiop. Population composition in 931b1d.. p. 9. 941b1d., p. 126, citing John Seeley et al., Crestwood _Heis,h__t8.. (Toronto: University Of Tbronto Press, 1956), pp. 3-4. -43- terms of ethnicity, class, and religious affiliations vary from community to community and within any single community over time. Events-~such as the introduction of new industries, farm to city, and central city to suburbia population movements which may have no direct bearing on a con- flict situation-~definitely effect the composition of the parties involved. There are at least two significant ways in which population changes of this type influence ethnic conflicts. First, there is the simple fact that a change in the ethnic com- position Of a given community may bring a regional or national conflict into local focus. If a French-Speaking community does not possess an English-speaking pOpulation, then one would hardly expect French-English conflict to have internal significance. External conflict is a possibil- ity. Such a community may be involved in an attempt to preserve local institutions in the face of a dominant English-Speaking population at the provincial or federal level. Second, and of more direct importance, pOpulation changes influ- ence the level of superimposition Of conflict fronts at the local level. The degree of superimposition in an ethnic conflict of regional or national scape may differ considerably from one community to another and any given community may exhibit quite a different pattern from the national one. Given the hypothesized effect Of the degree Of superimposition on the intensity of a conflict, population changes may well introduce new levels Of intensity to established conflict situations, reducing or increasing intensity according to the composition of new arrivals. Using only two categories, ethnicity and religion, data from the -44- community under study can be used as an illustration. During the four decades between 1920 and 1960, the community was transformed from an almost exclusively French Roman Catholic village to a town composed Of French Catholics, English Catholics, and English Protestants.95 A com- plete superimposition of ethnicity and religion was prevented by the presence Of an English-speaking Catholic group. It is in this manner that population shifts are important modifiers or catalysts of local conflict situations. Superimposition is, of course, the key variable. Population shifts are only noted here as a class of phenomena which must be taken into consideration in a description Of local conditions. Vertical and horizontal patterns. Ethnic conflicts focus on concrete issues at the local level. There is the question, therefore, of where and by whom are decisions effecting such issues made. With the growing participation of regional, provincial, and federal decision- making machinery in local matters, the likelihood of issues relating to education, employment, housing, etc., remaining strictly local is extremely low. Warren introduces the concepts of vertical and horizontal patterns of organization as a means of handling the relationships between commun- ity and extra-community systems. A community's vertical pattern is defined as "the structural and functional relations Of its various social "96 units and sub-systems to extra-community systems. Horizontal patterns 95See Appendix II. 96Roland L. Warren, op, cit., p. 161. -b. u a N 'l t.‘ H are defined as "the structural and functional relations of [the commun- ity's] various social units and sub-systems to each other."97 It is suggested that the strength of the horizontal system with reSpect to local issues may condition any relations posited between the variables of conflict. If local decision-making machinery is not involved in crucial issues, whatever they may be, the interest-articulating activity of con- flict agencies and agents will take place beyond the local community. Thus, the vertical system involving persons and associations on a regional basis will contain the actual confrontation, the horizontal system remain- ing relatively free of direct and extensive involvement. In this way, conflict agents and agencies may never confront each other locally. Agents and agencies located in provincial and federal capitals carry on the conflict, with little requirement for involvement on the part of the citizens of local communities. Following Warren, rather than dealing with this phenomena in terms of a dichotomy (i.e., horizontal versus vertical systems), the strength of a community's horizontal patgern is posited as the variable.98 Thus, the stronger the horizontal pattern with respect to decision-making regarding the issues in diSpute, the greater is the possibility of an intense conflict at the local level. The following hypothesis expresses this possibility: 97Ibid., p. 162. 98Roland L. Warren, op, cit., p. 13. ‘I -46- Theoretical Hypothesis III The weaker the horizontal pattern with reSpect to decision-making regarding the issues pertaining to a local conflict situation, the more likely will the intensity of a conflict be depressed at the local level. Community identification. The final community variable posited as having significance for this research was suggested by Coleman in his monograph on Community_Conflict. Pointing out that conflict depends on contact and interaction, he suggests: . . . when community members are highly involved with the community per se, identifying their own future with that of the community, that identification carries its own consequences as disagreement proceeds. Particularly, it appears to modify and constrain the disagreement.99 The presence of one or more ethnic groups in a community, though provid- ing a major prerequisite for conflict, also tends to establish a set of communal issues exclusive of ethnic interests. Roads must be built, sewers provided, recreational facilities made available, taxes levied, and so on. If members of a given community are strongly identified with these kinds of communal interests, there will be a tendency to avoid issues which might damage this level of COOperation. Hughes, in his study of French-English relations in a Quebec community, deals with this prob- lem in a chapter on "communal institutions."100 Killian deals with the 99James C. Coleman, Op, cit., p. 21. 100Everett C. Hughes, French Canada in Transition. (Chicago: University of Chicago Press, 1943), p. 84. -47- same notion from an Opposite point of view, where a Negro sub-community is prevented from identifying with over-all communal interests.101 A high degree of identification on the part Of the parties con- cerned with such communal interests would appear to have a more direct effect on the militancy of a conflict than on the intensity. A high level of involvement may remain, but the presence of a strong identity with the community per se would tend to reduce the probability of con- flict agencies and agents using coercive action at the local level. The following hypothesis expresses the prOposed relationship between communal identification and militancy: Theoretical Hypothesis IV The more that members of locally based conflict parties identify with communal interests, the less will be the degree of militancy exhib- ited by these parties. The purpose of this discussion on community was to illustrate how structural variations in particular communities may influence a local conflict situation. If advances can be made in this area of inquiry, the results should begin to provide explanations as to why conflicts with national implications vary considerably in their intensity and militancy from community to community. III. ASSIMILATION AND CONFLICT The model developed in this chapter is submitted as an alternative 101L. Killian, "Community Structure and the Role of the Negro Leader Agent." Paper presented at the Annual Meeting of the American Sociological Association, Montréal, August 1964. ‘o '0. tr '0 u. -48- to the "race relations cycle" for the analysis of ethnic relations. In this section, the relationship between the variables of conflict and the concept of assimilation will be considered. Park and Burgess include conflict and assimilation with competition and accommodation as basic processes of social interaction.102 It is from this conceptualization that the notion of the race relations cycle was develOped. The statement by Park that this cycle "of contacts, competition, accommodation, and eventual assimilation, is apparently progressive and irreversible"103 has formed the basic orientation of several studies on race and ethnic relations.104 The parallel development of studies based on the notion of "even- tual assimilation" and the existence of a prevailing ideological orien- tation toward the desirability and necessity of assimilation in the United States was perhaps no mere coincidence. Park himself alluded to the desirability of assimilation in his search for "a satisfactory "105 scheme of dealing with the problem of Americanization by education. [italics mine.] A conception of the inevitability of assimilation was 102Robert E. Park and Ernest W. Burgess, Op, cit., p. 507. 103Robert E. Park, Race and Culture. (New York: The Free Press of Glencoe, Paperback edition, 1964), p. 150. 10“For example, W. Lloyd Warner et a1, Op. cit.; and Raymond Breton, "Institutional Completeness of Ethnic Communities and the Personal Relations of Immigrants," American Journal of Sociology, Vol. 70, September 1964, pp. 193-205. 105Robert E. Park, Race and Culture, Op, cit., p. 283. .2“ .. Io -49- also in accord with the functionalist's emphasis on consensus as a pre- requisite for a viable social system. A low level of assimilation implies the absence of consensus between people or groups in contact and thus the impossibility of creating a functioning system containing both groups (e.g., a nation). Needless to say, this scheme fits in with those who propound the "one nation--one flag, one culture, one language" ideology. In studies of inter-ethnic relations cast in this framework, ques- tions asked are generally of the type, "How much assimilation has occurred?"; "Why are some individuals more assimilated than others?"; "Hhat detrimental effects does a lack of assimilation have on the society receiving immigrants?"; and "How can the assimilation of immigrants be facilitated?". The perSpective all but excludes the possibility of ask- ing such questions as "What prevents assimilation from occurring?"; "With a minimum of assimilation, what kinds of structures enable Opposing collectivities to coexist?"; and "In an integrated economic system can minority groups achieve equal participation in that economic structure without losing their identity?". It is questions of the latter type that direct attention not to consensus, but to conflict; not to assimilation but to separation. The perSpective is changed. Assimilation is no longer accepted as inevit- able, nor indeed as necessarily desirable. It is perhaps true that in the long run "the story of mankind consists of a succession of contacts of ethnic groups, their conflicts, their accommodations to one another, and their eventual fusion into new ethnic groups."106 But the "long run" 106T. Shibutani and K. M. Kwan, 02. _<=__it.. p. 116- consists of several generations, perhaps centuries, of conflict and acco- mmodation. Cyclical models do not apply in a short run analysis; the out- comes are multiple and the conditions complex. In all fairness to Park, it should be stated that he was doubt- ful of the inevitability of assimilation and was certainly well aware of alternate schemes and alternate modes of reSponse to ethnic contact. In an article on "The Race Relations Cycle in Hawaii," Park made this statement: . . . it is not possible to predict with any certainty the final outcome . . . race relations will assume one of three configura- tions. They will take the form Of a caste system, as in India; they will terminate in complete assimilation as in China; or the unassimilated race will constitute a permanent racial minority within the limits of a national state, as is the case of the Jews in Europe.107 In spite of his many references to the cycle as inevitable, the scope Of his work allowed for an interdependence of conflict and assimi- lation and for the possibility that each process could be influenced by conditions operating outside of the cycle. Recent evidence on the American ethnic situation supports Park's doubts regarding the race relations cycle. Etzioni,108 Gans,109 Glazer and Moynihan,110 and Gordon,111 seriously question the race relations 107Robert E. Park, Race and Cultpre, Op, cit,, p. 194. 108Amitai Etzioni, ”The Ghetto-~A Re-Evaluation," Social Forces, Vol. 37, March 1959, pp. 225-226. ' 109Herbert J. Cans, The Urban Villagers. (New York: The Free Press Of Glencoe, 1962). 110N. Glazer and D. P. Moynihan, Beypnd the Meltipg Pot. (Cambridge, Massachusetts: The M. I. T. Press, 1963). 111Milton M. Gordon, AssimilatiOp in American Life. (New York: Oxford University Press, 1964). -51- cycle as an adequate theoretical orientation. Eisenstadt's112 analysis of immigrant absorption in Israel could be added to this list. This renewed interest and changing perspective on ethnic relations has itself been conditioned by pressing changes in the society around us. The rapid and seemingly unpredictable develOpments in American race relations113 have directed some social scientists to a conflict as Opposed to a con- sensus perSpective.114 The recent proliferation of urban studies, backed by the programmatic concerns of "wars" on poverty, inadequate municipal government, and tranSportation systems, and so on, have revealed to the surprise of many the presence of functioning national minority groups; the same groups that were doomed to assimilation by the "Chicago School" two generations ago. Assimilation: Igeology or Social Process The term "assimilation" has been employed to describe two quite different phenomena. Louis Wirth, for example, in a 1945 article analy- ses pluralism, assimilation, secession, and militancy as goals or ideo- logical orientations Of ethnic minorities.115 Gordon examines assimilation 11ZS. N. Eisenstadt, The Absopption of Immigrants. (London, England: Routledge and Kegan Paul Ltd., 1954). 113See Everett C. Hughes, "Race Relations and The Sociological Imagination,” American Sociological Review, Vol. 28, December 1963, pp. 879-891. 114L. Killian and c. Grigg, op, cit. 115Louis Wirth, "The Problem of Minority Groups," Talcott Parsons et al., eories of Societ , Vol. I. (New York: The Free Press of Glencoe, 1961), pp. 309-315. -52- anglo-conformity, the melting pot theory, and cultural pluralism as pos- sible ideological orientations of either subordinate or superordinate ethnic groups.116 Others, of course, have used the same concept to refer to a process of cultural and structural fusion which takes place when ethnic groups come into contact. Assimilation, then, may be employed to refer to ideological posi- tions or to changes in cultural and social structures. The distinction is mentioned because in this thesis, assimilation is employed in both senses, referring to the goals Of the parties under study and to the pro- cess of fusion going on between them. Assimilation: Unigimensional or Multidimensional There has been a tendency to treat the process of assimilation in terms of single indices. Assimilation is frequently measured or Opera- tionally defined in terms of language, religious practices, identification, participation in "host" institutions, etc. These indices, singly or in combination, are generalized to the "degree of assimilation" exhibited by a given ethnic or racial group and yet these indices, in various com- binations, suggest possible variables operating with some degree of inde- pendence, rather than the total process itself. For example, it would appear as though the degree to which members of ethnic minorities participate in the institutional structure of the "host" society varies independently from the degree to which they assume 116Milton M. Gordon, op, cit. -53- the value system of the "host" society. This distinction is made by 118 and others.119 Gordon distinguishes between Gordon,117 Eisenstadt, "cultural or behavioral assimilation" and "structural assimilation." The former refers to the extent to which members of one ethnic group take on the beliefs, values, and general cultural patterns of another; the latter refers to the extent to which the members of one group par- ticipate in the "cliques, clubs, and institutions of the host society."120 Eisenstadt makes a similar distinction between "acculturation" which refers to "the extent to which the immigrant learns the various roles, norms, and customs of the absorbing society" and "institutional disper- sion" referring to "the migrant group as such, and its place in the social structure Of the absorbing society."121 These distinctions parallel the traditional distinction between the anthrOpOlogist's concept of acculturation and the sociologist's con- cept of assimilation, although in empirical studies the two frequently 1171b1d. 118 S. N. Eisenstadt, op, cit. 119Peter A. Munch, "Social Adjustment Among Wisconsin Norwegians," American SociolOgical Review, Vol. 14, 1949, pp. 780-787; Erich Rosenthal, "Acculturation without Assimilation," American Journal of Sociology, Vol. 66, November 1960, pp. 275-288; Michel Brunet, La presence anglaise et les Canadiens. (Montreal: Beauchemin, 1964). Frank G. Vallee et a1, "Ethnic Assimilation and Differentiation in Canada," The Canadian Journal of Egonomics and Political Science, Vol. 23, 1957, pp. 540-549. 120 Milton M. Gordon, op, cit., p. 71. 121S. N. Eisenstadt, Op, cit., pp. 12-13. -54- overlap. Many more distinctions may be made. Gordon adds no less than five variables to the two mentioned above.122 The discussion suggests three general categories under which the many variables of assimilation may be grouped. First, there is the cultural dimension or acculturation. Second, there is the social dimension which would include Eisenstadt's "institutional diSpersion" and Gordon's "structural assimilation." Third, there is a psychological dimension involving the process Of identification. The Variables of Assimilation There is little value in making these distinctions unless it can be demonstrated that each varies independently of the others according to certain Specified conditions. The literature suggests this possibility. McKee suggests that ”power may be more significant than assimilation in changing the distribution of life chances in the industrial community."123 The implication here is that though acculturation may have occurred, the exercise of power in the pursuit of position and resources (these being defined in terms of wages and status in a case where ethnicity and class coincide) mitigates structural assimilation, but has little influence on acculturation. A similar conclusion is suggested in van den Berghe's work on South Africa. He points to "the gradual westernization of all non-European 122Milton M. Gordon, Op, cit. 123James B. McKee, "Status and Power in the Industrial Community: A Comment on Drucker's Thesis," The American Journal of Sociology, Vol. 57, January 1953, p. 369. -55- groups [in an] extremely complex process of culture contact [involving] . . . a number of cultures originating on three different continents. "124 Without doubt, the degree of acculturation in South Africa is much less than that exhibited by American ethnic groups. However, the point of van den Berghe's study is that in Spite of a general cultural convergence, deep structural gulfs exist between the four main racial groups. More- over, this structural separation is due to the exercise of power on the part of the dominant EurOpeans. Williams also accepts the prOposition that social assimilation varies independently of acculturation.125 Shibutani and Kwan distinguish between acculturation and social or structural assimilation, although they give more attention to the former. Their basic prOposition is that,j} "any condition that makes possible the establishment of communication channels across the color line facilitates acculturation."126 The above mentioned studies deal with the differentiation of acculturation and structural assimilation and their relative independence. There still remains the psychological dimension of the process. This aSpect includes at least the problem of individual adjustment to the 127 ethnic contact situation, and the problem of identification with one 124pierre L. van den Berghe, South Afripa, op, cit., p. 52. 125Rob1n M. Williams, Jr., Strangers Next Door, op, cit., p. 365. 126T. Shibutani and K. M. Kwan, op. cit., p. 476- 1273. N. Eisenstadt, Op, cit., p. 12- -56- or another of the groups in contact.128 Problems Of individual adjust- ment have little place in the framework adopted for the proposed research. The effects of ethnic conflict, discrimination, and prejudice on the individual psyche are well documented.129 Identification is perhaps more dependent on variations in accul- turation and structural assimilation than on variations outside the total process. At one extreme, a complete lack of acculturation and structural assimilation implies a high level of individual identification with the minority group on the part of its members. At the other extreme, identi- fication with a minority would be an impossibility if such a minority were completely absorbed along the remaining two dimensions. Gordon, in his treatment of the interrelations among the various dimensions of assim- ilation, implies that identification is tied to structural assimilation rather than to acculturation,130 Some level of identification with the in-group can remain only as long as an identifiable group remains intact. Kardiner and Ovesey are quite adamant in distinguishing between acculturation and structural assimilation with respect to the American Negro and locating problems of identification in the social matrix.131 Using Gordon's terminology, 128Milton M. Gordon, op, pit., p. 71. 12EJSee, for example, A. Kardiner and L. Ovesey, The Mark of Qppres- pion. (Cleveland: The WOrld Publishing Company, 1951); and G. Seward, Clipical Studies in Culture Conflict, and Psyghotherapy and Culture Conflict. (New York: The Ronald Press, 1958). 13°M11ton M. Gordon, op, cit., p. 80.) 131A. Kardiner and L. Ovesey, Op. cit., p. 317. -57- "identificational assimilation" is dependent on the level Of structural assimilation, the latter being conditioned by variations Outside the total process of assimilation. To summarize, cultural assimilation or acculturation may occur without any significant changes in structural assimilation. That is, in Spite of a certain level of cultural assimilation, an ethnic or racial collectivity may remain intact vis-a-vis another such collectivity. 0n the other hand, complete structural assimilation implies complete cul- tural assimilation. Cultural assimilation is conditioned by contact and communication between the collectivities in question. Structural assim- ilation iS more dependent on the distribution of position and power among the collectivities in question. Thus, "structural assimilation, . . . rather than acculturation, is . . . the keystone Of the arch Of assimilation."132 In this thesis, the variables Of conflict direct attention toward the structural dimension Of assimilation. Ethnip group gefined. Before proceeding with a discussion of structural assimilation and conflict, a clarification of the key concepts will be helpful. An ethnic group is taken to be a type of quasi-group, (this more inclusive category having been previously defined), objec- tively distinguishable on the basis of common national origins. Ethnic groups may or may not be "minorities." The term "minority" implies the presence of a power differential; numerical differences in membership are 132Milton M. Gordon, op. cit., p. 81. See also 8. N. Eisenstadt. op, cit., p. 15. irrelevant. A minority is a subordinate group in relation to some major- ity or superordinate group. It is in this sense that ethnic groups are conflict groups.133 Thus, if two or more ethnic groups (i.e., parties) are interacting (i.e., interaction between parties) in a subordinate- superordinate relationship (i.e., position and resource scarcity), then according to empirical conditions previously Specified, a state of con- flict exists among them. Aspimilatiog gefined. AssimilatiOp is an interactional process involving two or more ethnic groups in which the social, cultural, and psychological dimensions of each group are in the process of being fused, such that if the process is completed, the groups are no longer identi- fiable as ethnic groups in relation to each other. Structural assimi- lgplpp, the social dimension of the total process, refers to the fusing of primary and secondary social relationships and institutional struc- tures, such that if the process is completed, separate identities cease to exist within the social structure of the groups involved. Cultural ,gggimglppgpp refers to a fusing of values, perSpectives, life styles, etc., such that if the process is completed, such elements cease to be differentiated along ethnic lines. Identificational assimilation, an sapect of the psychological dimension of the total process, refers to the development of a single identity, such that if the process is completed, the original ethnic identification no longer has a saliency in the self- 133R. A. Schermerhorn, These Our People. (BOSt0n3 D. C- Heath and Company, 1949), p. 9. -59- concepts of the individuals involved. Assimilation and Conflict The process of boundary maintenance is a primary source of social conflict. Structural assimilation is the reverse of boundary maintenance. If one ethnic group exhibits a high level of boundary maintenance with respect to another, it follows that it also exhibits a low level of structural assimilation with reSpect to the other group. This is not a prOpositional statement; it is merely the substitution of one term for another, both of which have identical empirical referents. Structural boundaries are structural boundaries whether one is referring to an ethnic group under the rubric of assimilation or to another type of group under the more general concept of boundary maintenance. What are the empirical referents of structural assimilation? The works of Eisenstadt,134 Gordon,135 and Breton,136 suggest three major areas of empirical phenomena into which one should inquire. The degree of structural assimilation (or from the Opposite point of view, structural separation) may be determined first by observing the extent of ethnic closure in major institutions. In the community under study, the fact that the three groups--French Catholic, English Catholic, and English Protestant--operate three distinct school systems would be an instance Of 1348. N. Eisenstadt, Op, cit. 135Milton M. Gordon, op. cit. 136Raymond Breton, op, cit. ethnic closure in major institutions. The two remaining areas of inquiry are the extent to which the associational life of a community divides along ethnic lines and the extent to which primary relationships in cliques and friendship groups are ethnically enclosed. These are the same empirical phenomena that would be investigated should one have wdshed to determine the level of boundary maintenance exhibited by particular ethnic groups. Indeed, if the eupirical conditions for conflict are met in a case involving ethnic groups, a certain level of structural separation will be present. If parties are identifiable, there would be, by definition, some degree of separation. Complete structural assimilation cannot be realized in a situation where a state of conflict exists between ethnic groups. The question posed here is: To what extent do the variables of conflict retard, reverse, or accelerate assimilation? The intensity of a conflict Situation would seem to be related to structural assimilation. As the intensity Of an ethnic conflict increases, there will be, by definition, a tendency for the groups involved to move toward a greater structural separation as the number of conflict agencies, involvement in these agencies, and interest-articulating activ- ity increases. This does not imply that all points of contact where structural assimilation has previously occurred will immediately disinte- grate. It refers merely to an increasing involvement on the part of ethnic group members in ethnic conflict agencies. However, if this ten- dency continues for any length of time, the effect will be an ever- increasing structural separation. Raymond Breton demonstrated that, the presence of formal organizations [i.e., conflict agencies] in the ethnic community sets out forces that have the effect of keeping the social relations Of the immigrants within its boundaries . . . a for- mal structure . . . has the effect of reinforcing the cohesiveness of already existing networks and expanding these networks.137 It is in this sense that intensity in an ethnic conflict situation retards or reverses the process of structural assimilation. Similarly, it can be Shown that the greater the extent to which ethnicity is superimposed on other potential conflict categories, the lower the level of structural assimilation. FOr example, if the member- ship in ethnic conflict agencies coincides with membership in a labor union, in a particular political party, and in a particular religious denomination, the parties would exhibit a very low level of structural assimilation. The relationships between intensity, superimposition, and struc- tural assimilation are true by definition insofar as the indicators of each of the conflict variables overlap wdth the indicators of structural assimilation. In order to determine the level of intensity in a con- flict situation, one notes the extent to which persons are involved in ethnically enclosed conflict agencies. In order to determine the level of superimposition in an ethnic conflict, one would investigate the extent to which ethnicity coincides with class, religious and political affiliations. In both cases, these same indicators would be employed in an attempt to determine the level of structural assimilation. The process of institutionalization has a paradoxical effect on 137mm, p. 202. -62- ' the assimilation of ethnic groups. Institutionalization, by regulating and providing for a continuation Of a conflict relationship, maintains at least a minimum level of structural separation and would, thus, seem to prevent completion of the assimilation process. On the other hand, institutionalization by creating and maintaining channels of communi- cation between the parties facilitates cultural assimilation. Thus, while retarding one dimension of the total assimilation process, institu- tionalization accelerates another. It is doubtful, however, that an increase in the institutionalization of ethnic conflict would result in further separation of the parties. It does not follow that increasing recognition of the parties and their interests, once the initial frame- work has been created, results in an expansion of conflict agents; nor does a decrease in institutionalization necessarily result in an increase in structural assimilation. A complete lack of institutionalization may imply a violent and militant conflict with little routinized procedures. Such a breakdown would lead to further separation rather than assimi- lation. It is in this manner that the variables of conflict are viewed as influencing the course of assimilation and, consequently, it becomes dif- ficult to accept Park's notion of an irreversible, cyclical process involving conflict, accommodation, and assimilation. The level of assim- ilation may be retarded or reversed according to variations in the con- flict variables. No hypotheses are advanced here. Ethnic relations and the process of assimilation are assumed to be special cases of social conflict. Confirmation of the conflict hypotheses will lend support to -63- this position. IV. SUMMARY In this chapter, a model was advanced for the purpose of analyz- ing inter-ethnic relations as a type of social conflict. Social conflict was defined as a structural state. A set of hypotheses were constructed relating the variables of a conflict situation. It iS the interaction of these variables that provides the dynamic of conflict States. The model is based on the related assumptions that (1) social conflict is a result of the interplay of power and position, and of boundary mainten- ance activities; and (2) that all interacting human pOpulations exhibit social conflict. TWO additional sets Of hypotheses were constructed relating aspects of community social structure to the conflict variables. As a convenient means of summarizing the model, the four hypo- theses are listed below. Theoretical Hypothesis 1 The institutionalization of the relations between conflicting parties reduces the level of militancy exhibited by both parties. Theoretical Hypothesis II The greater the extent to which potential conflict categories are superimposed, the higher the degree of intensity exhibited by any single conflict front. Theoretical Hypothesis III The weaker the horizontal pattern with reSpect to decision-making -54- regarding the issues pertaining to a local conflict situation, the more likely will the intensity of a conflict be depressed at the local level. Theoretical Hypothesis IV The nore that members of locally based conflict parties identify with communal interests, the less will be the degree of militancy exhibited by these parties. CHAPTER III THE RESEARCH DESIGN A design is the total process whereby observed phenomena are related to the research problem. It involves three distinct procedures. There is that part which deals with the develOpment of a series of test- able hypotheses, and a set of indicators related to the terms used in these hypotheses. The results of this Operation determine the universe of phenomena to be Observed. There are the various methods of data col- lection employed. Finally, there is the manner in which the resulting materials are analysed. The develOpment of the testable hypotheses and indicators wdll be treated in the first section. Data collection methods will be described in the second section. Methods of analysis are dis- cussed where appropriate in each section. I. THE TESTABLE HYPOTHESES As stated in the previous chapter, the prOposed hypotheses were not developed from the literature alone, but resulted from an interplay between conclusions drawn from field observations and the literature. The research did not begin with a set of fully developed hypotheses. The total process which was comprised of three phases was designed first to arrive at a set of hypotheses and then to submit these to a test. The enphasis in this section of the chapter is on the latter process. The method adopted was to proceed from the general to a set of testable hypotheses by breaking down the terms of the former into their -65- -66- component parts at a lower level of abstraction. For example, regarding the term militancy in Hypothesis 1, the question was: "At a more concrete level just what observable activities comprise the notion of militancy?" The testable hypotheses were stated in terms of these more concrete components.1 Given these components, the next step was to delineate as precisely as possible related empirical phenomena. This required the development of a set Of indicators for each term.2 Thus, the entire pro- cedure consisted of two steps of substitution, each at a lower level of abstraction. Hypothesis I: Institutionalization and Militancy The institutionalization Of the relations between conflicting parties reduces the level of militancy exhibited by both parties. A highly institutionalized conflict system implies (1) that the parties to the conflict reOOgnize each other 322 parties; (2) the pre- sence of established and sanctioned channels of communication through which the claims of each party are articulated; and (3) a mutual recog- nition of each other's claims as negotiable. The relationship between \ these three components correSponds to the basic communications model. 1See H. L. Zetterberg, Op, cit., pp. 79-86. 2See Emile Durkheim, The Division of Labour in Sociegy. Trans- lated by George Simpson. (Free Press Paperback Edition; New York: The Free Press of.Glencoe, 1964), p. 64. Also Paul F. Lazarsfeld, "Prob- lems in Methodology," Sociology Today. R. K. Merton et al., (editors.) (New York: Basic Books, Inc., 1959):_pp. 47-48. ~67- Either party may assume, at a given time, the role of initiator or £2212' 1235.3 Party "A" as an initiator directs a claim or demand, a message, to Party "B" the recipient. The channel through which this message is directed is the vehicle or pgppp of communication. Party "B's" reaction may be viewed as the effect. It is assumed that if the parties recognize each other, it follows that there will be established and sanctioned communication channels. Conversely, non-recognition would imply the absence of communication channels. The presence of conflict agents who are readily identifiable implies both mutual recognition and the presence of communication channels and is, therefore, used as an indicator of the latter. The recipient Of a claim or message may either accept or reject the claim as negotiable. To reject such a claim does not imply non-recognition of the initiator, but it does suggest a temporary reduction in the level of institutional- ization. The presence of implicit or eXplicit agreements will be used as an indicator of claims recognized. Given a point in time, observed can- cellation of such agreements by one party or refusal to negotiate a claim will be used as an indicator of non-recognition of claims. Militancy, the dependent variable, refers to the type of acts directed toward an opponent. It is the type of strategy employed in the pursuit of party interests. At a lower level of abstraction, coercive or intimidating and violent acts imply a militant strategy. An observed use of force to injure or harm an Opponent will be taken as an indicator of 3In a subordinate-superordinate relationship, the minority party usually acts as the initiator. -68- violence. The use of petitions, demonstrations, and non-compliance are taken as indicators of coercive acts. The meaning of petition is taken in a narrow sense as a written request for a specific action directed toward the Opponent and endorsed by a number of party members. Demonstrations refer to diSplays of sup- port and strength through organized or Spontaneous public marches or processions. Men-compliance refers to the refusal on the part of one party to adhere to regulations perceived as being imposed by the other. This does not include regulations brought about by negotiations. Based on this design, the following are proposed as testable hypotheses. Hypothesis I,l: The recognition, on the part of each party, of the legitimacy of each other's demands tends to reduce the use of coer- cive or intimidating measures by one or both parties in the pursuit Of their respective interests. Hypothesis 1,2: The presence of established and sanctioned communi- cation channels through which participating parties may articulate their interests tends to reduce the use of coercive or intimidating measures by one or both parties in the pursuit of their respective interests. Hypothesis 1,3: The recognition, on the part of each party, of the legitimacy of each other's demands tends to reduce the use of vio- lence by one or both parties in the pursuit of their reSpective interests. -69.. Hypoghesis 1,4: The presence of established and sanctioned communi- cation channels through which participating parties may articulate their interests tends to reduce the use of violence by one or both parties in the pursuit Of their respective interests. Hypothesis II: Superimposition and Intensity The greater the extent to which potential conflict categories are superimposed, the higher the degree of intensity exhibited by any single conflict front. Superimposition was defined as the coincidence of social categories significant to a particular conflict system. Religion, ethnicity, and social class have been delineated as categories having a relevance to the conflict system under study. This requires a method for establishing membership in the three categories and assessing the extent of member- ship overlap. Three ethnic-religious groupings, French Catholics, English Catho- lics, and English Protestants are identified as the parties to the con- flict system under study. Membership in these parties at the local level was established in two ways. Sample units ("households") were first classified as either Catholic or Protestant according to the voter's list from which the sample was drawn. A panel of judges consisting of the mayor, the town clerk, and the parish priest then assigned members of the Catholic group to either the French-Speaking or English-speaking sub- community. This process is discussed in detail below. -70- Annual family income and occupational status were used as the indicators of social class. Using the data from question 46 in the questionnaire, reSpondents were divided into three income groups.4 The contingency coefficient (C) was used as an indicator of the extent of overlap or superimposition with respect to the three categories.5 The same Operation was performed using three occupational status groupings, ethnicity, and religious affiliation. The occupations of the respondents to the questionnaire6 were classified according to a scale develOped in Canada in 1951 by Bernard Blishen and revised in 1961 by the Dominion Bureau of Statistics.7 The scores, based on an index which combines educational and income levels, are grouped into seven occupa- tional status classes. Classes six and seven include professional and managerial occupations; classes three to five include "white collar" and skilled occupations; and classes one and two include semi- and unskilled occupations. For analytical purposes, classes six and seven were taken as representing "high" occupational status, classes three to five "medium" occupational status, and classes one and two "low" occupational status. A second but related type of Superimposition was introduced later “Infra, pp. 97-99 for description of questionnaire and see Appendix III, question 46, p. 456. 5See Sidney Siegel, Nonparametric Statistics_for The Behavioral Scieppes. (New York: McGraw-Hill Book Company, Inc., 1956), pp. 196-202. 6See Appendix III, question 16, p. 443. 7Bernard R. Blishen, "The Construction and Use of an Occupational Class Scale." K. D. Naegele et al., (editors) Canadian Society: Sociological Perspectives. (New York: The Free Press Of Glencoe, 1961), pp. 477-485. in the research. Given the three identifiable parties, each exhibiting a certain amount of closure around primary and secondary relations, it was observed that a certain proportion of the members of any one party belong tO associations identified with the other party. This was espe- cially so regarding French participation in English associations. Coleman in his consideration of community conflict refers to this phenomena as the "cross-pressures" placed upon individuals as a result of the multi- tude of ties pulling a person in one or the other direction in a conflict situation.8 The data collected on association memberships from question 34 and 35 in the questionnaire will be used as an indicator of this type of superimposition which, in turn, will be used to assess the extent of over-all superimposition.9 The dependent variable in this hypothesis is intensity--the magni- tude of party involvement. The emergence of conflict agencies, the extent to which agencies pursue the interests of their respective parties, and the viability of the agencies themselves are the components of inten- sity. The Observed appearance of new agencies, and the quantity of agency activity with respect to interest aggregation and articulation are the indicators of the first two components. The size of agency member- ships and the level of participation (i.e., attendance) are the indicators of viability. The following testable hypotheses are advanced within this framework: 8James C. Coleman, Op, cit., pp. 18 and 22. 9See Appendix III, questions 34 and 35, p..449. -72- Hypothesis 11,1: Given an ethnic conflict, the greater the coinci- dence of ethnic, religious, and class affiliations, the higher the level of activity directed toward the pursuit of party interests on the part of ethnic conflict agencies. Hypothesis 11,2: Given an ethnic conflict, the greater the coinci- dence of ethnic, religious, and class affiliations, the more viable will be the participating ethnic conflict agencies. Hypothesis III: Vertical-Horizontal Patterns of Qgganizgtion and Intensigy The weaker the horizontal pattern with reSpect to decision- making regarding the issues pertaining to a local conflict situation, the more likely will the intensity Of a conflict be depressed at the local level. Vertical and horizontal patterns of organization refer to the structural and functional relations of local social units and systems to extra-local units and systems, and to each other. One of the components of this concept is the strength Of one or the other pattern with respect to decision-making on particular issues. Decisions regarding some local issues must be deferred to extra-local systems; others may be brought to a conclusion without involving extra-local systems. For example, during the course of this research, the power to make decisions regarding the distribution of certain welfare funds was removed from the local town council, a unit of the horizontal system, to the county council, a unit -73- of the vertical system.10 Here, the interest is in the strength Of the horizontal pattern. Given a particular conflict system, the decision-making structure with respect to the relevant issues may be located either in the horizontal pattern (i.e., locally) or in the vertical pattern (i.e., extra-locally). In the former case, the horizontal pattern would dominate, in the latter the vertical pattern. However, these are extreme positions. It is assumed that, in a given conflict situation, the strength Of the hori- zontal pattern varies from community to community with more decision- making power deferred to extra-local systems in one than in another. Empirically, there is the problem of measuring the strength of the horizontal pattern. An estimate will be arrived at by delineating the relevant issues at the local level and observing the location of decision- making power with respect to these issues. Based on this method, the following testable hypotheses are advanced using the previously discussed indicators for intensity. Hypothesis III,l: If issues identified as relevant to the local conflict system require decisions beyond the local level, the level of activity directed toward the pursuit of party interests on the part of local conflict agencies will tend to be lower than if such such issues require decisions at the local level. 1Oopservations 009 and 010, December 13, 1965 and January 17, 1966. -74- Hypothesis II;,2: If issues identified as relevant to the local con- flict system require decisions beyond the local level, local con- flict agencies will tend to be less viable than if such issues require decisions at the local level. Hypotgpsis IV: Community Ideptification and Militancy The more that members of locally based conflict parties identify with communal interests, the less will be the degree of militancy exhibited by these parties. According to Warren, community identity involves "the extent of psychological identification with a common locality."11 That is: "To what extent is a locality viewed by its residents as a significant refer- ence group?" Given a community consisting Of ethnic-religious factions, it would appear that the extent to which the residents identify with com- munal as Opposed to factional interests is a coaponent of the broader con- cept of community identification. It is this component that will be dealt with in the proposed testable hypotheses. Two indicators are used. One involves an examination of the vot- ing patterns according to ethnicity on a particular communal issue. In 1959, the citizens of Normanville were required to vote on the question of annexation to Metro Center. The extent to which voting followed ethnic lines will be taken as an indicator of a lack of identity with communal llRoland L. Warren, op, git., p. 13. -75- issues. The second indicator makes use of question 45 in the question- naire.12 Respondents were asked to rank a set of three communal issues and three ethnic-religious issues in order of importance to them. These issues were identified as significant from observation notes. A ratio of the sum of the ranks assigned to the three ethnic- religious issues to the total number of respondents was compared with a ratio of the ranks assigned to the three communal issues to the total number of respondents. The set of issues receiving the lower quotient was taken as more important and as therefore implying a stronger identity than the ratio with the higher quotient. Using the indicators of militancy as discussed above, the follow- ing testable hypotheses are advanced. Hypothesis IV,1: If the members of parties involved in a local ethnic conflict identify more with communal than with ethnic interests, there will be less Of a tendency to use coercive and intimidating measures by one or both parties in the pursuit of their respective interests. Hypothesis IV,2: If the members of parties involved in a local ethnic conflict identify more with communal than with ethnic interests, there will be less of a tendency to use violence by one or both parties in the pursuit Of their respective interests. 12See Appendix III, question 45, p. 455. -75- The Empirical Conditions for Conflict Prior to submitting the above hypotheses to a test, it was neces- sary to determine if a state of conflict existed in the study community. In the previous chapter, the empirical conditions for such a state were defined.13 Here, the methods used to identify the three parties will be described. The methods used in identifying the other conditions are the subject matter of Chapter VII. The three parties, French Catholics, English Catholics, and English Protestants, were quite apparent in the history of French-English rela- 14 Observations pointed to their pos- tions in the Province of Ontario. sible existence at the local level. The first step was to select a sample from the community and classify the units according to membership in one of the three categories. The second step, using a number of indicators, was to determine if these categories meet the criteria required for parties. In order to arrive at the initial classification, the sample units were first classified as Catholic or non-Catholic (Protestant) using information provided on the municipal assessment roll. The problem was to separate French and English Catholics. Objective criteria alone proved unsatisfactory. Whether one Speaks French or English, one's ethnic origin, or one's surname, do not adequately classify people according to member- ship in the assumed French or English sub-communities. Another method 13Sppra, pp. 24-30. “BEE. especially pp. 141-163. -77- had to be found to combine with the above criteria. The method employed was to select a panel of three judges, members of the Catholic sub-community who were active in community affairs and ‘who would be most likely to recognize persons as belonging to one or the other group. The mayor, who had lived in Normanville for some forty years, the town clerk who had been active in town administration since the thirties, and the curé were selected as judges. These three were requested to divide the Catholic portion of the sample into four groups: an English-speaking sub-community, a French-speaking sub-community, a grouping where husband and wife were split between the two, and a group- ing of those whom they did not know well enough to classify. The exact instructions and the criteria for classification appear in Appendix IV. The sample consisted Of 250 units (households), 188 Roman Catho- lics and 62 non-Catholics.15 The judging Operation resulted in seventy- nine French Catholics, three "split households" where the husband was classified as French Catholic, fifty English Catholics, fifteen "split households" where the husband was classified as English Catholic, and forty-one unclassifiable. The classification was based on a majority decision if there was a lack of complete agreement among the three judges. The unclassified group consisted of unknown persons, most Of these being new arrivals in the town, and some where there was complete disagreement among the judges. Of the unclassified group, twenty-one were reclassified using the 15The sample is explained below in detail. ~78- curé's judgement as carrying more weight because of his contacts with newer parishioners, contacts which the other two would not have to as great an extent. The final classification, prior to administering the questionnaire, was as follows: eighty-three French Catholics, six French Catholic "split households," fifty-seven English Catholics, twenty-two English Catholic "split households," sixty-two English Protestants, and twenty unclassifiable. The classification used in the analysis required some adjustments. Of the 250 units, twenty were used in a pre-test and six were eliminated because they had been interviewed on several occasions by this student. This left a total of 224 units. Of these, allowing for refusals, rejects, and persons who had died, moved out of town, or were unavailable, 167 units were used in the analysis. The final party classification consisted of seventy-nine French Catholics, fifty-two English Catholics, and thirty- six English Protestants. The adjustments made to arrive at this final classification were as follows: 1. The "unclassified" group were classified according to their responses to questions 26 and 27 dealing with language Spoken and "mother tongue" respectively.16 2. Twenty-eight units exhibiting considerable discrepancy between the judging results and the reSponses to questions 26 and 27 were also reclassified. 3. The French and English Catholic "split households" were grouped with their reSpective parties according to the original classi- fication of the head of the household. 163cc Appendix III, questions 26 and 27, pp. 445-46, -79- In order to determine if these categories met the requirements set out for parties, the extent of institutional separation, of interpersonal interaction among the members of the categories, and of the subjective awareness of membership were examined. This process is explained in 17 detail in Chapter VII. The method used in analysing the remaining con- ditions for conflict are also dealt with in Chapter VII.18 II. METHODS OF DATA COLLECTION The methods used for this research may be divided into four gen- eral categories: observation, interviewing, analysis of documents, and a structured questionnaire administered to a random sample. Each of these will be described separately. However, an overview of the total process will provide a useful perSpective. As mentioned above, the research proceeded in three stages. Observation and unstructured inter- viewing were the principal methods employed in stages one and two; more structured interviewing and the questionnaire were used in stage three. During the spring and fall of 1964, several trips were made to the research site in order to (l) assess the community's suitability as a research site; (2) explore the structure of the community, locating ethnic sub-systems; and (3) "sample" different aspects of the structure of the ethnic sub-systems to gain an initial insight into the totality. Unstructured interviewing of reSpondents judged as knowledgeable in the 17Infra, pp. 238-253. 181nfra, pp. 253-297. -30- phenomena under study was the method used during this stage. Fifteen persons were interviewed with twenty-five interviews being completed. The second stage began in May 1965 with the investigator taking up residence in the selected community. The period of residence was con- cluded in July 1966. Stage two, lasting until January 1966, was the period during which extensive observation and unstructured interviewing took place. This work resulted in the hypotheses set out above. The third stage involved a series Of structured interviews and the adminis- tration of the questionnaire. Observation The advantages and disadvantages of Observation are well docu- mented.19 The principal danger would appear to lie in the difficulty in maintaining sufficient controls and checks on the data to insure reli- ability and validity, and in the problems involved in describing the actual procedures employed. The latter difficulty can result from neg- lecting to record procedures as the field work is in process. Indeed, the cognizance of procedures adopted by the Observer is one check on the reliability of the data. Throughout the total process, there are four questions which an observer must keep in mind. hhat role or roles are assumed in the field? 19Arron V. Cicourel, Method and MeaSQrement in Sociolggy. (New YOrk: The Free Press of Glencoe, 1964), pp. 39-73; Claire Selltiz et al., Research Methods in Social Relations. (Revised. New York: Henry Holt and Company, Inc., 1959), pp. 199-235; Bernard 8. Phillips, Social Research: Strategy and Tactics. (New York: The Macmillan Company, 1966), pp. 135-143. -31- What should be observed? What procedures should be used to accurately record the Observations? What procedures are used to maintain checks on the data?20 These four questions are used here as a convenient means of describing the total process. The role of the Observer. At the risk of violating a norm of dissertation writing, I will use the first person Singular in this sec- tion in lieu of the recommended third person. The problem of entry was not too great in this case, although it might have been were it not for the peculiar role combination carried into the field. In an investiga- tion of intergroup relations, one's own membership in the particular categories involved can influence reSpondents' reactions. In all first contacts, I introduced myself as a graduate student from Michigan State University studying French-English relations in the local school system. This had certain advantages in that it appeared to elicit a desire-to-help-a-student attitude, an advantage that a non- student investigator would not poSsess. Also, the student role is fre- quently perceived as divorced from vested interests and, therefore, as carrying an aura of objectivity. I never volunteered my own ethnic or religious affiliations. It was, nevertheless, obvious to French-speakers that I was a native English-speaker. In an area where ethnicity is important in day-to-day contacts, I was Often asked such questions as: "Are you an American?" "Are you English?" "Are you Catholic or 20Claire Selltiz et al., op, cit., p. 205. -32- Protestant?" etc.21 The fact that I was a Canadian student, English but nevertheless originating from Montreal, had just returned from a period of residence in the United States and was associated with an American 22 seemed to prevent a university, and had an ability to Speak French, definite classification by respondents. This gave me a fair degree of acceptance in all three ethnic-religious groupings. There were entry problems associated with the English Protestant sub-community. I was, from time to time, reminded that "there are no problems here," the comment being accompanied with an attitude of dis- approval. Two members of the English Catholic community maintained a certain suSpicion about my motives throughout the field work period. On one occasion, at a public meeting, one of these persons showed consider- able anxiety over my noting of a statement he had made. After the meet- ing,he asked me not to "publish" his statement. Occasionally, initial contacts with members of the French Catholic sub-community would begin with a test of my willingness (more than my ability) to Speak French.23 21See for example, Erving Goffman, The Presentation of Self in Everyday Life. (Garden City, New York: Doubleday and Company, Inc., 1959), p. 1. 22An English-speaker speaking French in Canada implies more than simply that the person possesses a skill in another language. It is perceived by French-speakers as implying a whole set of attitudes toward French Canada. 23An excerpt from my field notes illustrates this type of meeting quite well. Observation 027, September 14, 1965. ”. . . introduced me in English, he mentioned Michigan State Uni- versity . . . 's first comment, in a very antagonistic tone, was to ask: 'Are you American?‘ e t o ud- rt -83- The use of French was important from two points of view. It pro- vided a significant means of acceptance, and of equal if not more import- ance, it provided entry into conversations not generally Open to English- speakers. I would not classify myself as bilingual, but my knowledge Of the language was sufficient to enable me to participate in casual con- versation and to conduct interviews. Apart from the problem of entry, there were various roles assigned and assumed during the course of the research. Raymond Gold identifies four types of field work roles: the couplete participant, the participant- as-observer, the observer-as-participant, and the complete observer. 2" I found myself taking all four roles to one extent or another, not sinml- taneously, but at different times and in different situations. The complete participant role implies a high level of interaction with reapondents in some day to day situation (e.g., the work setting, I replied, with equal antagonism, 'NO, I am a Montrealer.’ He thought he had me then. '15:: bien, si vous ates un Montre’alais, vous parlerez frangais, n'est-ce-pas?' 'C'est vrai, je parle franpaisfl He turned to his brother, giving a knowing chuckle; the coolness dis- appeared. I had passed the first test. He then proceeded to talk about the A. P. I. in French. After a few sentences, he asked me (in English) if I would prefer that he speak in English. Test num- ber two. I answered, 'Non, parlez en franpais, je vous couprends.' Again he smiled, glanced at his brother, and away we went." 2[‘Raymntl L. Gold, "Roles in Sociological Field Observations," Social Forces, Vol. 36, March 1958, pp. 219-221, cited by Arron V. Cicourel, 02. 915., pp. 43-44. -34- a play setting, an organizational setting, etc.) where neither the iden- tity nor the purpose of the observer is known to the Observed. This role was never totally adapted. There were many settings, however, in restaurants, taverns, and occasionally at public meetings and events, where I did interact with those present without informing them of my identity or purpose.25 In this sense was the complete participant role assumed, but the level of interaction was low and the situations were terminal. Data gathered in these sessions were used to provide a feel- ing for the total community. A good deal of my initial impressions on cross-ethnic interaction and linguistic behavior resulted from this type of observation. The participant-as-observer role was frequently adapted. Here, the stress is on participation, but the reSpondents are aware of the Observer and his purpose. I found myself playing this role with increas- ing frequency as the field work progressed and I was becoming more widely known. At house parties, dances, and at various informal gatherings with respondents, I was most certainly a participant though my role as student- Observer was known and accepted. The use of this role provided more sys- tematically recorded information on cross-ethnic interaction, linguistic behavior, and attitudes and values with respect to French-English relations. -———- 251a these settings, if I was asked to identify myself and my pur- pose, I did so. Although I was more likely to state that I was a teacher spending some time on getting to know about life in Ontariotowns, than give the usual and more accurate self-description. This was taken by respondents without any additional comments. I had found that the stu- dent explanation in these settings turned the focus on me with a barrage of questions regarding age, source of money, etc. O... G 'I J? R. I“ 0.: n". ‘p’. -35- A similar role, but with the euphasis reversed--the observer-as- participant role--"calls for a relatively more formal observation than either informal observation or participation of any kind."26 This role ‘was assumed in many of the associational meetings attended where I was recognized and accepted as an observer, but participation was neither expected nor desired. Also included here would be the many unstructured and structured interviews arranged by appointment. In both these settings, up orientation was fixed and formal; I was looking for specific information. The complete observer role was utilized only once, but the data gained was of the utmost importance. In this role, respondents are obser- ved, but they are unaware of the Observer; he need not be present. The only point at which this role entered the scheme was in the administra- tion of the questionnaire to the sample. Two observations of some significance should be made in conclu- sion. First, no one role or attitude toward the respondents was adopted throughout the field work period. Different types were assumed for dif- ferent purposes. Indeed, it would appear to be both impossible and hazardous to attempt to assume a Single role in a complete and lengthy field experience. The danger, should such be attempted, lies in the loss of a crucial checking device. In this case, each role served as a reli- ability check on data gathered through each of the other roles. Secondly, apart from these more technical role types, a lengthy 26Raymnd 1.. Gold, pp, cit, p. 221. 12351.: I a. 'AI H OH. lrx -86- exposure in the field involves the Observer as a person with, for better or for worse, all his idiosyncracies. In analysing my experience, I found William Whyte's comments in this regard to be quite accurate.27 People tend to accept or reject the observer on the basis of the personal relationship established regardless of his purpose or his explanation. My exact purpose was unimportant to many pe0ple with whom I came into con- tact. Some provided their own explanation: "I hear you're writing a book about us." "Maybe we'll get on TV along with Peytop Place.” ”You're doing a history of the town, eh? Well, one of my relatives. . . ." My over-all role than was, simply, myself and this was especially the case wdth four reSpondents who became fairly good friends and, in field work terminology, primary informants. OppegyapiOppl categories ppd recorpipg. The question regarding what should be observed and the method of recording are so closely related that they are best treated together. This is not to suggest that the method of recording determines categories of Observation. On the con- trary, categories, either substantive or analytical, determine what is Observed and are a necessity for systematic recording. Obviously, an investigator observes the phenomena related to his research problem. Minus a problw statement, the researcher's observations are Of no more value than the day to day random observations Of the layman. At the same time, it is usually expected that the interaction between observations of 27ct. William myte, Spppet Cpppep Socipty. (Chicago: University of Chicago Press, 1955), p. 300. I Jun 5...! x51: -87- eupirical events and the conceptualization of the research problem will result in changes in and the development of the latter. To fully appreciate this process, it will be necessary to intro- duce the various stages of conceptualization as they occurred throughout the field work period. The stages correspond to the three major research phases outlined above. The initial problem was extremely general and sub- stantive; there were no hypotheses. The interest was in ethnically hetero- geneous coulmnities and the problem of clarifying the concept of assimi- lation. 28 The questions directing observations during the spring and fall of 1964 were designed to explore the extent to which ethnic groupings in such communities form sub-systems or sub-communities and the implica- tions of these sub-systems for the institutional life of the over-all coamnity. At this point, notes were made during interviews and retyped in duplicate afterwards. The amount of recall necessary was at a minimal. Observations regarding the town and its people, few during this first stage, were written up iumediately following a trip. Office space was pro- vided by an organization in Metro Center where appointments were made and notes were put in order before leaving the research site. Using this 28At this point, the problem was influenced by the work of Milton M. Gordon, M" and his notions of the several dimensions or variables of assimilation. A study of Eisenstadt's, op, cit., which had arrived at a similar conceptualization of assimilation was also extremely influential. The initial interest was derived from W. L. Warner's des- cription of ethnic sub-systems in Yankee City. See W. Lloyd Warner, e cia t of rican E ic G on s. (New Haven: Yale University Press, 1945), pp. 220-253. - 88- material, a set of substantive categories were develOped.29 It was from these categories plus additional library work that the focus of the prob- lem shifted from assimilation to social conflict. A theoretical problem began to take shape. Are the processes of assimilation and conflict interdependent? Is assimilation a reversible process?30 These questions dominated my thinking at the point of moving into the community to comp mence field work in earnest. The volume of material accumulating each day presented a record- ing problem. For the remainder of the field work period, a daily journal was kept in which were recorded, at the end of each day or at the begins ning of the next day, the results of interviews and general observations from all other sources. A small notebook or three by five index cards were carried around and used constantly. The taking of notes did not appear to bother respondents during interviews or organizational meetings. Notes were not taken on the spot when playing the complete observer or the participant-as-observer roles. In these cases, recall had to be 29The major categories were as follows: (1) subjective definitions of ethnic groups; (2) voluntary associations; (3) general description of school systems; (4) ideologies, values, and goals re French-English rela- tions; (5) social class; (6) power structures; (7) government; (8) relig- ion and church; (9) economics; (10) history; and (11) local changes affecting assimdlation. 30The first set of observations directed me to the literature on conflict, particularly Lewis Coser, op, cit., Ralf Dahrendorf, 22, cit., and Pierre L. van den Berghe, 22, git. The possible connections between Gordon's variables of assimilation and Dahrendorf's variables of con! flict, as outlined in Chapter II, began to appear at this point. -89- dependedupon.31 In October 1965, following four months of intensive observation, the new material was classified using the original set of categories. As a result of this mid-term assessment, a new set of categories were developed integrating the original assimilation focus with the more recently formulated notions on conflict. The new categories were more analytical, but included no hypotheses.32 For the next four months, until January 1966, both general observations and interviews were explicitly directed by this new set of categories. Included in this period was a good deal of interviewing aimed at verifying past observations. By January 1966, the final concepts and hypotheses began to emerge. For the succeeding two months, observations were reduced to a minimum. All material collected to date was reclassified according to this second set of categories. By this time, the emphasis had switched completely to social conflict. Although assimilation took a secondary place, the inter- est in the relationship between the two was never discarded. The final statement of the research problem and the variables and hypotheses which appear in this report were formulated during this period. The final, 31Thirty-seven nondirective interviews were completed and forty- five organizational meetings were attended during the total field work period. This does not include the innumerable sessions of more formal observation in the complete observer and participant-as-observer roles. 32The major categories in this second set were: (1) issues; (2) vertical and horizontal systems; (3) institutionalization; (4) influ- ence of national dialogue on local system; (5) quasi-groups--religious, ethnic, and class; (6) linguistic behavior; (7) ideologies and values re assimilation, pluralism, and separation; (8) observations on persons. -90- or third major phase of this research, consisted of additional observa- tion, documentary analysis, structured interviews, and the questionnaire all directed by the revised theoretical framework. In summary, it was noted that three stages of conceptualization occurred during the research process, each succeeding the other, each more analytical and less substantive than the former. For illustrative purposes, two examples of the flow or relationship between these sets are presented. One of the descriptive categories emerging from the first phase was "ideologies, values and goals re French-English relations.” This permitted merely a listing of observed values. The second set of categories provided a means by which these values could be classified. The new category was "ideologies and values re assimilation, pluralism, and separation." This category provided the base from which the final category, ”incompatible values as an empirical condition of conflict," was developed. The process leading to the variable "institutionalization" pro- vides another example. Initially, there was a category under which were listed observations regarding the school system and its ethnic-religious divisions. These observations provided an almost perfect prototype of institutionalized conflict. It might be added that it was this set of observations which were primarily responsible for changing the focus from assimilation to conflict. Institutionalization remained as a category in the final analysis, although more recent observations added the sub- category of conflict agents as a vehicle of institutionalization. Checks 23 reliability and validity. Reliability and validity are loosely related. Generally, a highly valid measuring instrument, one which.provides a valid indication of the concepts under study, would also be reliable. However, a reliable instrument, one which yields the same results on repeated.measurements of the same phenomena, is not necessar- ily a valid instrument. The results may not be a true indicator of the phenomena under study. Thus, in dealing with validity one is faced with the extent to which selected indicators reflect the concepts as defined. It calls the whole complex of concepts and propositions into question. Reliability refers only to the adequacy of a given measuring instrument. For this reason, a section on validity per as will appear at the conclusion of this chapter where it will be more convenient to discuss the indicators proposed above in relation to the concepts and the various methods of data collection utilized. Certainly, the checks maintained upon observations throughout the field work period will contribute to an over-all estimation of validity, but the enphasis here will be on reliability. The question is: What checks were maintained to insure a consis- tency in recording procedures, and to prevent errors introduced by observer and respondent bias? Several checks were used. The formulation of a research problem from the very beginning, the continuous revision of these problems, and the development of corresponding observational categories maintained a check on what was observed and on the consistency of these observations. Also, the various field work roles performed placed the observer -92- in differing sets of relationships to his respondents. Each of the four roles may be viewed as a "measuring instrument." In many cases, similar phenomena were observed in different role settings. Thus, data collected in one setting could be checked with data collected in another setting. Observations on cross-ethnic interaction picked up in a setting where the complete observer role was being used were checked out in another setting where the participant-as-observer role came into play. This process pro- vided checks on both consistency and bias. Finally, the use of several different methods of data collection maintained a check on both observer and respondent bias. A constant checking was in process between data derived from general observation and data from unstructured interviews. This material in turn was checked through the use of more structured interviewing which followed the period of intensive observation. Had "pure" observation been the only method used, the problem of reliability would have been much greater. As it stands, given the variety of checks used, reliability is assumed with some confidence. 12mm Two types of interviewing were made use of in this study. The first, an unstructured type, was utilized throughout all three phases. The second, a more structured type, was used during the final phase. The first type is referred to as "unstructured" insofar as identical sets of pre- determined questions were not administered to each reapondent. Some of the interviews were focused in that an appointment was made with a reSpon- dent to discuss specific issues which had arisen during the course of -93- observations made in the community. These interviews were used.primarily for checking out observed data and seeking new leads in building up a universe for observation. Others were aggdiregtive in that there was no particular focus, and the initiative in the interview'was left largely to the respondent. The participant-as-observer and the observer-as- participant roles were related to the focused and nondirective interviews respectively. A total of sixty-two unstructured interviews were conducted over the three research phases. In contrast, the structured interviews made use of a standard set of questions administered to each respondent. Here, six different inter- view schedules were used to obtain data concerning issues, associational structure, conflict agencies, conflict agents, evidence of intensity and 'militaney, and impressions of the respondents regarding particular issues and assimilation as a value orientation.33 Interview schedules 1A and 1B, and 2A and 2334 were designed to secure information on regional bodies to which local associations were affiliated, and local associations themselves. The following were inter- viewed with schedules 1A and 18: 1. President, Regional Board, L'Association canadienne-frangaise d'Education d'Ontario. 2. President, Ontario English Catholic Educational Association. 3. President, Western Ontario Division, La Société Saint-Jean-Baptiste. 4. Public Relations Director, L'Association de la Radio-Television frangaise du and de l'Ontario, (Metro Center Region). 5. Research Director, L'Association de la Radio-Television frangaise du and de l'Ontario, (Metro Center Region). 33See Appendix IV. 3[flies Appendix IV, pp. 459-467. Not all local association officers were interviewed. Data obtained dur- ing phases one and two were used to determine those associations which 'might be acting as conflict agencies. Certain associations were elimi- nated following these interviews. The minutes and other records of the remaining set were used to arrive at a final decision regarding their definition as conflict agencies. The presidents of the following associa- tions were interviewed using schedules 2A and ZB: . Ste. Cécile's Catholic Parent Teachers Association. . The Normanville Catholic Parent Teachers Association. The St. Jean's Catholic women's League. The Knights of Columbus. The Canadian Legion, Nbrmanville. Normanville United Church ibmen's Group. Les dames de Ste. Anne, St. Jean's Parish. . Ushers' Club, St. Jean's Parish. 9. La Société Saint-Jean-Baptiste, Normanville. 10. LiAssociation de Parents et d'Instituteurs, Normanville. ~ oouaxyIJ-‘unr-I Interview schedules 3, 4, and 5 were designed to locate conflict agents and to delineate their relationships and activities in the public school board, the separate school board, and the town council.35 Sched- ule 3 was administered to three of the six public school board members, schedule 4 to all five members of the town council, and schedule 5 to all six members of the separate school board, and one member of the parish high school board. Interview schedule 6 was designed to locate conflict agents and delineate their activities in the school system. The following were interviewed using this schedule:36 353ee Appendix IV, pp. 468-472. 36 See Appendix IV, pp. 473-474. -95- Regional English Inspector. Regional Bilingual Inspector. Principal, St. Jacques' School. Director of French Classes, St. Jacques' School. Teacher responsible to English Inspector, St. Jacques' School. Principal, Ste. Cécile's School. Principal, St. Jean's High School. Principal, St. Jean's Intermediate School. Director of French Classes, St. Jean's Intermediate School. Teacher responsible to English Inspector, St. Jean's Intermediate School. Teacher reaponsible to Bilingual Inspector, St. Jean's Inter- mediate School. Oomuymbunw H H H e A total of forty-one interviews of this type were conducted, bring- ing the total number interviewed over the three phases to 103. The per- sons selected for these interviews should not be construed as members of a probability sample. They were not selected from a universe of individ- uals. In the case of the structured interviewing, respondents were sel- ected because of their positions; the information sought was primarily organizational data. The unstructured interviewing went hand-in-hand with observations. Persons were selected (or they selected the investi- gator) to provide a check on observations and impressions of the commun- ity as a whole. .Qesaesnse Newspapers and associational records were used for both descrip- tive and analytical purposes. In order to arrive at an outline of past events relative to the proposed hypotheses, the content of the following newspapers was analysed: 1. Le Castor, Normanville, Volumes I to V, 1931-1935. 2. The Evegigg Record, Metro Center, 1912-1914. 3. The Evening Star, Metro Center, 1932-1934. 4. The Centuriog, Normanville, Volumes I to VII, 1959-1965. - 4,9.- - I.‘ l‘! -95- The focus was on reported events and on editorials as reflections of opinion. Thus, the question of validity enters in relation to the possible distortion entering the reporting of events through bias and falsification. The only check maintained on such possibilities was the use of other materials dealing with the same periods. Thus, for the period between 1912 and 1927, two additional sources were used.37 The period of the thirties involved, in addition to two newspapers, the minutes of the Normanville Separate School Board, and a published history on that period.38 Material from the Normanville Centurion, used prior to the field work period, was cross-checked with associational minutes, observations, and interviews. Apart from the press and certain records of local associations, secondary sources were used for all historical data. The minutes and correspondence of the following associations were analysed: . The Normanville Catholic Parent Teachers Association. The Ste. Cécile's Catholic Parent Teachers Association. 1 2 3. The Normanville Association de Parents et d'Instituteurs. 4. The Normanville Société Saint-Jean-Beptiste. These were used in conjunction with the data obtained from the structured interviews of associational officers. In addition, the minutes of the 37F. A. Walker, c tholic Edu atio an oliti s 1 tario. (Toronto: Thomas Nelson and Sons, Canada, Limited, 1964); Lucien Brault, Etef egpgsé ge 1's i eme t ili e an e siecle ans l'Ontario at e autres rovigges. (Kingston, Ontario: College Militaire Royal, 1966), and C. B. Sissons, Bi-li a1 8 hools 1 Canada. (Toronto: J. M. Dent audSons, Ltd., 1917) were the main sources used in this connection. 38F.A. walker, 22, 21 . Va!- Norumnville Separate School Board were used for the periods between 1900 and 1911, 1926 and 1936, and 1951 and 1961. School registration records and municipal public records were also made available. The categories employed in analysing all documentary material were identical to those discussed above in relation to observational data. An "open file" using these categories was also maintained on three news- papers during the field work period: Normanville's The Centuriog, Metro Center's The figar, and Le Egoit of Ottawa. Many additional reports, briefs, and news bulletins from several local, regional, and provincial associations were also used. thatTogs A brief mention of the types of notations used in this thesis to distinguish between observational and interview references will be of assistance. Interviews coupleted during the first phases are referred to with a two digit number preceded by an "X” (i.e., Interview X01). Obser- vations made during phases two and three are referred to with a two digit figure preceded by a "O" (i.e., Observation 001). Unstructured interviews completed during phases two and three are referred to with a two digit number preceded by a ”0" (i.e., Interview 001). The structured inter- views carried out during the final phase are referred to with a two digit figure preceded by an "A" (i.e., Interview A01). 1 air A standardized "fixed alternative" questionnaire was administered ..L 133.117. bile-.... I -98- to a sample of Normanville residents.39 The questionnaire was designed for both descriptive and analytical purposes. Data obtained from the following questions were used mainly in Chapters V and VI, describing the community and its people: Quengog Number Igem . 1 Marital status of respondent. 2 to 8 Length of residence and migration. 9 Ownership of residence. 10 and 11 Ethnic origin. 1 12 and 13, 14 and 33 Religious affiliation and attendance at French or English services. _ 23 School system supported. 24 and 25 Number of children and school attendance. 26 Language spoken. 27 Maternal language. 28 to 32 Language behavior, French-speakers. 47 Age of respondent. Data obtained from the remaining questions were related to the proposed hypotheses, as listed below, and used mainly in Chapters VII, VIII, and IX: MW Item Relevagg Vari able 15 to 22 Occupation Social Class-- Superimposition. 34 and 35 Association membership Superimposition. 36 to 40 Cross-linguistic contact Conditions of conflict-- Parties. 41 to 44 Cross-religious contact Conditions of conflict-- Parties. 45 Issue ranking Community identity. 46 Income Social Class-- Superimposition. The questionnaire was pre-coded; responses were punched on holerith cards, and marginals and cross-tabulations were run on a Control Data 3600 computer. 39See Appendix III. A description of the sample appears below. -99- Angsttratiog of The ggestioggair . The questionnaire was admin- istered by both volunteer and paid interviewers. Members of branches of the Catholic women's League from various Metro Center parishes, seven- teen in total, volunteered to act as interviewers under the sponsorship of the General Mission Centre of Essex County. Also secured were seven paid graduate and undergraduate students resident in the area. A train- ing session was held for all interviewers prior to the administration of the questionnaire. A copy of a set of instructions given to each inter- viewer appears in Appendix III, along with a standardized letter of intro- duction and opening comments used by the interviewers. Announcements regarding the interviewing and the project appeared in local church bul- letins in advance. The questionnaire was constructed in both French and English. As far as was possible, French-speaking respondents were assigned to bilin- gual interviewers. Interviewers were instructed to use French if the respondent indicated a desire to do so. Local norms regarding the use of French made it extremely difficult for interviewers to judge which lan- guage to use. Most French-speakers in Nbrmanville respond to strangers in English, and their fluency in English is equal to that of an English- speakert An interviewer would have to judge at the point of meeting or soon after which would be the most apprOpriate language to use. Non- French-speaking interviewers were instructed to turn over respondents who could not be interviewed in English to a bilingual interviewer. The majority of interviews, 83.8 per cent were conducted in English, the balance in French. '. -100- ReTiapTlitg and validity. The problem of validity will be approached in relation to the over-all design below. However, if it can be demon- . strated that the questionnaire possessed an acceptable level of internal consistency, this will contribute to its over-all validity, and may be I; .1. 4 l 1 taken as a measure of internal reliability. That portion of the question- naire dealing with those items listed above as being used for descriptive purposes is assumed to have "face validity." As a rough estimate of reliability, the data appearing in Table II is used. Question 28 asked French-speaking reSpondents: "What proportion of the time do you speak French?"40 The "use of French" frequency scores are a summary of responses by the same subjects to questions 32a to 32s.41 These questions present four situations asking the respondents whether or not they use French in each. A "0" was assigned to those not using French in any of these situations; a "l" to those using the language in one situation only; a "2" to those using French in two situations, and so on to "4". It is assumed that question 28 and the set of questions 32a to 32s measure the same characteristic-~the extent to which French is used by French-speakers. To be sure, each question measures different inten- sities of that characteristic. Question 28 merely asks for an estimate from the respondent on the extent to which he uses French. Questions 32a to 32s ask only for a negative or affirmative reply regarding the use “OSes Appendix III, p. 446. “Ibid. , pp. 447-448. -101- of French in four situations which certainly do not exhaust the situa- tions in which one might use the language. Nevertheless, assuming that both measure the same characteristics and given the differences in inten- sity, if the questionnaire is reliable, the two sets of reSponses should be reasonably highly correlated. TABLE II PROPORTION OF TIME FRENCH SPOKEN BY FREQUENCY OF USE OF FRENCH SCORES W Frequency Scores Fraportion of Per Cent Per Cent Per Cent Per Cent Per Cent Time French Scoring Scoring Scoring Scoring Scoring Number of §poken "4" "3" "2" "1" "0" Cases Most of the time. 21.6 46.0 16.2 16.2 0.0 37 About half the time. 6.7 40.0 33.3 20.0 0.0 15 Occasionally 0.0 23.7 18.4 36.8 21.1 38 Total 10.0 35.6 20.0 25.6 8.8 90 That is, one would not expect those who claim they speak French most of the time to receive a "0" frequency score, nor would those who speak French occasionally be expected to receive a frequency score of ”4". Table II demonstrates this to be the case. Moreover, the majority of those answering "most of the time," 67.6 per cent, received scores of either "3" or ”4," and the majority answering "occasionally" received scores of "0" or "1." The majority between these two extremes received scores of "2" or "3." If the frequency scores are collapsed into three columns (0-1, 2, 'V OI J‘L' -102- and 3-4) to permit a chi-square test,42 the resulting statistic is 22.94, significant at the .05 level with four degrees of freedom. The contin- gency coefficient (C) is .450. Given the upper limit of (C), .816, in a 3 x 3 table, .450 is a reasonably high correlation. These results are therefore taken as demonstrating that the questionnaire possesses an acceptable level of internal consistency. W The questionnaire was administered to a simple random sample selected from a list constructed from the 1965 municipal voter's list. Tb specify the population from which the sample was drawn, it will be necessary to note the types of persons excluded from the voter's list by the Election Act, and those excluded in designing the pOpulation from this list. 2 Only ”British subjects" twenty-one years of age and over are inclu- ded on the voter's list. A British subject would include those born in Canada or in a commonwealth nation, and those born elsewhere who possess Canadian citizenship. There is a second qualification. In order for a person to appear on the voter's list, the property which he occupies must have a minimum assessment of three hundred dollars. If the owner of a prOporty is not qualified elsewhere (i.e., at his own place of residence or business), he takes precedence over a tenant. Therefore, if a property 42It is not advisable to use a chi-square test if there are 20 per cent or more of the cells with an expected frequency of less than five. For this reason, the categories in Table II were collapsed. See Sidney Siegel, 02, cit., p. 178. 'Dl-Iilfili ...IE' i. I'vl .‘vii I -lO3- owner has an assessment of nine hundred dollars, both he and his wife will appear on the voter's list. If he has tenants on this property, only one of these tenants will qualify at the three hundred dollars per voter requirement. Also excluded from the list are roomers, boarders, and others not occupying separate housekeeping quarters. In constructing the pOpulation from this list, the following steps were taken: 1. All those not residing in Normanville were eliminated. 2. This left a total of 2,079 individuals qualified to vote in municipal elections. 3. The sampling unit was a "household" rather than an individual. A household was defined as a husband and wife plus their children, or a single person (unmarried, widowed, separated, or divorced) with or without children. 4. On this basis, the 2,079 individuals were reduced to 1,177 households. The sample was drawn from this universe. Before proceeding with the method of selecting the sample, it would be useful to evaluate the universe in relation to the town's pOpulation with what evidence is available. A bias could enter the universe through the elimination of non-British subjects from the voter's list, tenants disqualified because of lack of assessment, and roomers without separate housekeeping facilities. There is no control over the bias introduced through the elimination of those under twenty-one years of age. Single or married persons under twenty-one were not included with the exception of those spouses under twenty-one whose partners were twenty-one or over. The British subject qualification would not have eliminated too many residents. Using 1961 census data, only 11.2 per cent of the popu- lation were not born in Canada.43 A certain prOportion of these would 43Canada, 1961 Census of Canada, Bulletin 1.2-7. (Ottawa: Dominion Bureau of Statistics, 1963), Table 55, p. 3. *1l .\1Fi\ . all)?! III-"In I u I,» F , ....- . NW -104- be of British origin, and another would have Canadian citizenship regard- less of place of birth and therefore qualify. Very few would have been eliminated on the assessment qualification since, according to the 1961 census, only 16.7 per cent of the dwellings were listed as tenant occupied, leaving a balance of assessment sufficiently large enough to qualify most tenants. On this basis, it is assumed that the population from which the sample was drawn approximates the total adult pOpulation of the town, although technically generalizations can only be made from the pOpulation as defined. The following steps were taken in drawing the sample: 1. Four digit numbers from 0001 to 1177 were assigned to each house- hold.unit, beginning with Polling Division I. Names were listed alphabetically by polling division. 2. Using a table of random numbers44 250 units were drawn. 3. From these 250, six were eliminated because they had been inter- viewed on several previous occasions, and twenty were randomly selected for a pre-test, leaving 224 in the sample. The final sanple used in the analysis consisted of 167 units repre- senting three-quarters of the original 224 units, and 14.2 per cent of the population. The difference between 167 and 224 resulted from twenty- three incompleted interviews due to persons who had moved, were deceased, or were unavailable after three calls; twenty-five respondents who refused to be interviewed, and nine questionnaires rejected due to interviewer error. These incompleted interviews, amounting to 25.4 per cent of the 44H. M. Blalock, SocTal Sgatistics. (New York: McGraw-Hill Book Coupany, Inc., 1960), pp. 430-437. . ICI‘I‘ ”it! n t.“ Kw E llillrllv El \\ I|\‘nh‘ ‘I-k‘ ~‘al‘Qih‘ 1,! has . It‘s. hr...‘ QMNI\ -105- original sample, could well destroy its random quality unless it may be assumed that the incompletes are themselves randomly distributed. This can be partially demonstrated by noting the extent to which a known characteristic was distributed among the fifty-seven incompletes. Party 5. l l t 0 membership was defined prior to administering the questionnaire. Of the fifty-seven incompletes, 35.1 per cent were French Catholics; 26.3 per cent were English Catholics; 28.1 per cent were English Protestants, and the balance of 10.5 per cent were Catholic, ethnicity unknown. This amounts to 71.9 per cent Catholic and 28.1 per cent Protestant which is very close to the distribution of the same characteristic in the original sample which was 74.8 per cent and 25.2 per cent respectively. Another method of evaluating the final sample is to compare the distribution of known characteristics in the sample with the population from which it was drawn. Table III indicates an approximately equal dis- tribution of tenants and owners in the sample as coupared to the population. TABLE III A COMPARISON OF THE DISTRIBUTION OF FOUR CHARACTERISTICS IN THE SAMPLE WITH THE POPULATION W Sample Population Characteristics Number Per Cent Number Per Cent Owners 129 77.2 927 78.7 Tenants 36 21.6 242 20.6 Unknown 2 1.2 8 .7 Total 167 100.0 1177 100.0 Separate school supporters 124 74.3 831 70.6 Public school supporters 43 25.7 346 29.4 Tbtal 167 100.0 1177 100.0 -lO6- A similarly close distribution of public and separate school supporters is also presented. Given this evidence, conclusions from the sample can be approached with confidence. Va 1 a ° A Dis ussio o L mitations Reliability is a necessary, but not a sufficient prerequisite for validity. As stated above, a valid instrument is also reliable, but a reliable instrument is not necessarily valid. Thus, the estimation of the validity of a design lies partly in demonstrating its reliability. Given the purposes of this research, an acceptable level of reliability has been demonstrated for the instruments used. This contributes to a certain degree of confidence in the validity of the design, but not suf- ficiently enough to ignore the problem of validity altogether. Two aspects of validity require attention. Internal validity questions the relationship between theoretical definitions and empirical indicators, and as such, brings into play the whole series of proposit- ions involved in moving from definitions to indicators. External valid- ity questions the ability of the instruments used in determining the values of the various indicators. The question: "Does the instrument measure what it is supposed to measure?" is appropriate here. Also involved is the "truth” of the responses given by interviewees. It is at this point where reliability enters into the picture.45 ASSee a discussion of internal and external validity in Hans L. Zetterberg, op, cit., pp. 114-123. 3.1 .DI-S’ I vi .1 -lO7- Table IV presents the theoretical concepts, their related compon- ents as they appeared in the testable hypotheses, and the indicators of these components. If the design possessed perfect validity, the scope of the content included in the theoretical definitions would be the same as that included in the components of each definition, and the content of the components would be included in their indicators. No claim to perfect validity is made. Indeed, one of the objectives of this research was to arrive at a means for achieving a higher level of internal valid- ity than could be achieved at this stage. Regarding the relationship between the theoretical definitions and their lower order components, the primary problem in each case would appear to be that, although the definitions imply the components, they also imply other phenomena.46 The list of components for each concept is not exhaustive. The danger in this state of affairs is that the unknown components of one concept might well relate to those of another in other than the hypothesized direction. Such a finding would be sufficient to reject a whole set of hypotheses. This problem is partially met by list- ing more than one component for each concept. On this basis, conclusions regarding concepts (1), (5), and (6) wdll possess more internal validity than those regarding concepts (2), (3), and (4). 46Zetterberg notes three types of internal validity problems: (1) those "where the definition implies the indicator and, in addition, something other than the indicator; (2) the indicator implies the defi- nition and, in addition, something other than the definition, and (3) the indicator implies the definition and vice versa." Hans L. Zetterberg, 22. site. pp. 116-117. soe Min I U u. 1: ...-9| n... ..L.....-- -..- lei‘..-ilu I III 0 Ililvien-l ‘ ‘llrliliILI 1" ...— inomi H>whm H< mczz>w& om <>wH>mhmm >ZU HZUHO>Howm Hwooaonwomp nosooonm xopmnmm Hanan ammo no ammnmowo mwoonommom Hsmuomnoem Aav Hsmnpncnuoumppnmnuon Aw.~v > acncmw Hooomnunuou on one: AH.~.HV are oaomosoo om namunoun on onooH-m swunau um ammonunoao. oxmuuown nmamoaosnm monsoon «so assumes. A~.Nv muommsom om mmnuvunurom use A~.~.Hv are anemones om nonmwpon mesonuosma ooaacwuomnnou mmmnnm. arounonu. Amy mcomeuaoomnnwou Am.wv anosn om o ooaosanuou om dunno om new use on mmmnmsoo Hours no nonmu souooumosnm on «snow onrbpoiuo~umwocm nausea awn: nouns ooaasuup unseen. AmV zapunmsow Am.HV Au.mv are sum om noonou