ABSTRACT FACTORS ASSOCIATED WITH PROTESTANT-CATHOLIC MARRIAGES IN THE DETROIT AREA: A PROBLEM IN SOCIAL CONTROL by Paul H. Besanceney, S.J. A review of the patterns of mate selection discovered by social scientists in the United States and elsewhere reveals that the choice of a marriage partner is neither a matter of chance nor purely a personal affair. There are social norms influencing this decision. One of these norms is the rule of religious endogamy: people in this country are expected to marry within their own re- ligious group. When the norm is violated, social control in this matter is said to have been ineffective. It was the purpose of this study to test some hypotheses regarding the factors leading to interfaith marriages, from the vieWpoint of social control ex- erted in support of an endogamous norm. A secondary analysis was undertaken of some data from surveys made by the Detroit Area Study of the University of Michigan in 1955, 1958, 1959, and 1962. These were originally studies of the family or of religion in which some questions suitable for our subject were asked. The first three samples were combined when- ever the same question was asked in each, yielding a total of 1,470 marriages. The additional 185 newlyweds in 1962 were selected by a different sampling method and had to be analyzed separately. Besanceney Only white Protestants and Catholics, plus their marriage choices, were included in this study. No information was gathered con- cerning broken marriages. A study of historical and demographic materials from the Detroit area aided in the anticipation and in- terpretation of the results. A double comparison was made throughout: AProtestants mar- ried to Protestants compared with Protestants married to non-Prot- estants; also Catholics married to Catholics compared with Catholics married to non-Catholics. Since those who had changed religious pre- ference were also identified in the interviews, it was possible to examine several religious types of interfaith marriage, according to whether one partner had changed religious preference and in what direction. Three out of five mixed marriages of both Protestants and Catholics in the Detroit area involved such a change. To study the background factors of interfaith marriages without taking ac- count of this fact would be deceptive. The rate of intermarriage for Protestants in the Detroit area was 36 per cent; for Catholics it was 41 per cent. In terms of individuals rather than of mar- riages, this means that 22 per cent of married Protestants and 26 per cent of the married Catholics in the Detroit area are in an interfaith marriage. A ratio which attempts to control for the re- lative size of groups was also computed. Why do some deviate from the norm of religious endogamy? We sought to explain these deviations through the relationship of the intermarried with the concerned agents of social control, i;£;, the parents, relatives, church, and ethnic group. We also explored the influence of "anomic situations" in which a norm of religious Besanceney endogamy might be weakened. These situations were: wartime mar~ riages, marriage at a late age (emancipated from parents), rural- urban migration, upward educational mobility, and vertical occupa- tional mobility. Our general hypothesis, tested by these specific relationships, was: If the agents of social control do not, or cannot (due to an anomic situation), exercise their influence, the violation of a religiously endogamous rule is likely. In general, this theoretical scheme was more predictive of Catholic, than of Protestant, interfaith marriages. For both, the parents and relatives were the most significant agents of social con- trol so far as religious endogamy was concerned. The church's con- trol in the case of those who attended Catholic schools was more effective for maintaining membership than for forestallingintermar- riage. Occupational mobility, upward or downward, was the only anomic situation which showed a significant (though weak) relation- ship with interfaith marriages. Dedication TO MY PARENTS Copyright by PAUL HAROLD BESANCENEY, S.J. 1963 FACTORS ASSOCIATED WITH PROTESTANT-CATHOLIC MARRIAGES IN THE DETROIT AREA: A PROBLEM IN SOCIAL CONTROL By Paul H. Besanceney, S.J. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology and Anthropology 1963 c; a‘f‘é‘m “OJ/A ACKNOWLEDGEMENTS In expressing thanks to the many who have helped directly in the production of this thesis, it is necessary to leave unnamed many very important people. Who would want to undertake such a project as this without the encouragement and prayerful support of family and friends? Not this writer. My Jesuit Provincial, Very Reverend John A. McGrail generously provided the means for it, while guidance and encouragement came from Rev. Julian L. Maline, S.J., and Rev. Paul V. Siegfried, S.J. The hospitality of friends in Ann Arbor (particularly at St. Joseph Mercy Hospital) and in Lansing (especially at St. Vincent Home) eased the burden and brightened the time. The consideration extended by the adminis- tration of John Carroll University speeded the final stages of the job. May all of these people be rewarded on a more significant occasion. Many have been more directly involved in the research for, and preparation of, the thesis itself. As many as possible are gratefully named here. At the University of Michigan the following professors kindly gave permission for my use of the data for which they were solely or jointly responsible: Drs. Robert 0. Blood, Jr., Robert C. Angell, Gerhard Lenski, Guy E. Swanson, and Ronald Freedman. In the early stages of my research, the stimulation provided by Dr. Harry Sharp was of considerable importance. His successor as Director of the Detroit Area Study, John C. Scott, assisted me in ii. many practical ways with his needed instructions and criticisms. At Michigan State University, Dr. Jay Artis was my cheer- ful guide through the long months spent on analysis of the data and composition of this report. My thanks to him and to my com- mittee members for their patience. In the data processing and use of the MISTIC computer, I would have been much handicapped without the generous and careful work of Mr. Francis Martin's staff and the guidance of Dr. Julian Kateley. Advice from Dr. Frank Sim of the Bureau of Social and Political Research was help- ful, and Miss Hilda Jaffe of the same bureau prepared the map. Between the composition of a thesis and its final appear- ance in a bound volume there are many hours of tedious and concen- trated work. My appreciation is therefore very fresh for the efforts of three ladies who were responsible for the typing of the final c0py at a time when they already had work enough to do: Mrs. Elizabeth M. Catalano, Mrs. Rachel Baskind, and Miss Mary Louise Turk. Similarly, Mrs. Ruth Cheek and her assistants in John Carroll University's reproduction room showed special care and promptness in providing the multilith copies. Mr. Albers of General Bookbinding in East Cleveland also gave his personal atten- tion to the final step. For the c00peration and kindness of each, my sincere thanks. iii. PREFACE TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . viii LIST OF ILLUSTRATIONS . . . . . . . . . . . . . . . . . . . xii Chapter I. II. III. INTRODUCTION: THE PROBLEM . . . . . . . . . . . . 1 MATE SELECTION: REVIEW OF THE LITERATURE . . . . . 9 Marriage Rates: To Be or Not To Be? How Soon? Homogamy: Like Marries Like Residential Propinquity Heterogamy: Complementarity of Needs Endogamy vs. Intermarriage: General Considera- tions Racial Intermarriage Ethnic Intermarriage Religious Intermarriages: Type and Sources of Data Rates and Trends: Jewish-Gentile Marriages Rates and Trends: Catholic Intermarriages Rates and Trends: Protestant Intermarriages Comparison of Rates for the Three Major Groups The "Triple Melting-Pot" Hypothesis Religious Intermarriage Related to the Sex Which Marries Out Religious Intermarriage Related to Nationality The "Old" Factors Related to Religious Inter- marriage The Economic Factor in Religious Intermarriage Class and Caste Hypergamy: Terms Defined Class and Caste Hypergamy: Review of Evidence The Process and Psychological Antecedents of Mate Selection Conclusion SOCIAL CONTROL AND ANOMIA . . . . . . . . . . . . 76 Studies of Social Control and Interfaith Mar- riages Hypotheses to Be Tested in This Study The Assumptions of a Normative Order iv. Chapter Page III. SOCIAL CONTROL AND ANOMIA (cont'd) Institutions The Concept of Social Control Approaches to Deviancy: Anomia Evidence for a Norm of Religious Endogamy in the United States - Conflicting Values'in Selecting the Marriage Partner Parents as Agents of Social Control Family and Relatives as Agents of Social Con- trol The Church as an Agent of Social Control Ethnic Groups as Agents of Social Control Anomic Situations IV. THE RESEARCH: DESIGN AND EXECUTION . . . . . . . . 172 Arriving at a Plan Source of the Data: Detroit Area Study Sampling and Interviewing in the Population Coding and Preparation of Data Cards Restrictions on the Sample Operational Definitions: The Dependent Vari- able Operational Definitions: Independent Variables Stages in the Analysis V. DETROIT AREA SURVEYS IN COMBINATION . . . . . . . . 211 Rates and Ratios of Intermarriage Types of Religious Intermarriage Rates Based on Individuals Rather Than on Marriages Some Characteristics of Those Who Intermarry Social Control of Interfaith Marriages by Parents Social Control by Family or Relatives Social Control by the Church's SChOOlS‘ Social Control of Religious Intermarriage by the Ethnic Group Wartime Marriages Rural-Urban Migration as an Anomic Situation Educational Mbbility and Interfaith Marriages Intermarriage and Vertical Mobility in Occupa- tions VI. 185 NEWLYWED COUPLES IN 1962. . . . . . . . . . . . 248 Special Characteristics of This Sample Operational Definitions V. Chapter Page VI. 185 NEWLYWED COUPLES IN 1962 (cont'd) Rates of Intermarriage by Religion, Sex, and Ceremony Some Unpredicted Characteristics of the Inter- married Parents and Family as Agents of Social Control The Church in Support of Religious Endogamy Close Ethnic Ties and Interfaith Marriage The Influence of Anomic Situation VII. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . 273 How Our Theory of Social Control has Fared in This Study Unpredicted Findings Concerning Protestant . Intermarriages A Composite Sketch of the Intermarrying Prot- estant in Detroit Unpredicted Findings Concerning Catholic Inter- marriages A Composite Sketch of the Intermarrying Catho- lic in Detroit What This Study Tells Future Researchers BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . . 291 Books Articles Unpublished Manuscripts APPENDIX. 0 O O O O O O O O O O I O O O O O I O O O O O O O 324 A. THE HISTORY AND DEMOGRAPHY OF THE DETROIT AREA Personal Impressions of Detroit's History Geography and Early History The Flow of Immigration into the Detroit Area Development of Detroit's Industry The City's Churches Some Earlier Findings of the Detroit Area Study New Expectations for Rates of Protestant-Catho- lic Marriages in Detroit B. DETAILED MARRIAGE CHOICES AMONG PROTESTANT AND CATHOLICS, INCLUDING THE POSSIBLE CHANGES IN RE- iLIGIOUS PREFERENCE C. INTERVIEW QUESTIONS USED IN THIS ANALYSIS OF THE DETROIT AREA STUDIES OF 1955, 1958, 1959, AND 1962 Detroit Area Study #828, in 1955 vi. C. INTERVIEW QUESTIONS USED IN THIS ANALYSIS OF THE DETROIT AREA STUDIES OF 1955, 1958, 1959, AND 1962 ( cont'd) Detroit Area Study #849, in 1958 Detroit Area Study #855, in 1959 Detroit Area Study #882, in 1962 (Zero Parity, i. e., No Children) vii. Table 2- l. 2- 2. 2- 3. 2- 4. 2- 5. 4- 1. 4- 2. 5- l. 5- 2. 5- 3. LIST OF TABLES‘ Journal Articles Published on the Subject of Mate Selection, 1930-1959 . . . . . . . . . . . . Summary of Rates of Ethnic Intermarriages for Selected Nationalities, as Reported from Research, by Percentage. . . . . . . . Summary of Published Rates of Jewish Inter- marriage in the United States, per 100 Jewish Marriages. . . . . . . . . . . . . . . . Summary of Published Rates of Catholic Intermarriage in the United States per 100 Marriages of Catholics. . . . . . . . . . . . Summary of Published Rates of Protestant Intermarriage in the United States, per 100 Marriages of Protestants. . . . . . . . . . Previous Religious Preference of Respondent and Spouse, Reclassified by Mother's Re- jligious Preference in the Formers' Child- hood Days, Detroit Area Study, 1958 and 1959. . . . . . . . . . . . . . . . . . . . . . . Distribution of Misclassified Marriages. According to Religious Type of Marriage, Detroit Area Study, 1955, 1958, and 1959. . . . Actual Intermarriage Rates and Intermarriage Rates "Expected" if Husbands and Wives Were Distributed at Random, for Marriages of Prdtestants and Catholics in the Detroit Area, 1955, 1958, and 1959. . . . . . . . . . . . Types of Interfaith Marriages Among Prot- estants and Catholics in the Detroit Area, 1955, 1958, and 1959. O O O O O O O O O 0 Selected Protestant Groups and the Rate at Which They Intermarry with Catholics and Others in the Detroit Area, by Early Reli- gious Preference of Respondents, 1955, 1958, and 1959. . . . . . . . . . . . . . . . . . . . . viii. Page 10 35 44 47 50 188 196 214 219 220 Table 5-10. 5-12. 5-13. Types of Interfaith Marriages Related to the Sex of Protestants and Catholics in the Detroit Area, 1955, 1958, and 1959, by Per- centage . . . . . . . . . . . . . . . . . . Religious Intermarriage Rate (per 100 Mar- riages) by Early Religious Preference of. Respondents with Selected National Origins on Father's Side, 1955, 1958, and 1959. Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to Total Annual In- come of Family Head, 1955, 1958, and 1959 . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to the Occupation of the Family Head, 1955, 1958, and 1959 . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to Interfaith Mar- riages of Their Parents, 1958 and 1959. Religious Types of Marriage of Protestants and Catholics in the Detroit Area, Related to the Source of Greatest Influence on Respondents' Religious Beliefs, 1958. . Religious Types of Marriage of Protestants and Catholics in the Detroit Area, Related to the Proportion of "Close" Relatives Who Are of Respondent's Faith, 1958 . . . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to Perception that Relatives Would Try to Discourage Respond- ent from Changing His Religion, 1958. . Religious Types of Marriage of Catholics in the Detroit Area, Related to Education in Church Schools, by Sex, 1958. . . . . . Rates of Intermarriage (per 100 Marriages) for Native-Born Protestants and Catholics in the Detroit Area, Related to the Re- spondent's Experience of Migration to Detroit, 1955, 1958, and 1959 . ix. Page 221 224 226 227 229 230 233 236 238 242 Table 5-14. 6- 1. 6- 2. 6- 3. 6" 4o 6- 5. 6- 6. 6- 7. 6- 8. 6- 9. Rates of Intermarriage (per 100 Marriages) for Protestant and Catholic Men in the Detroit Area, Related to Their Vertical Occupational Mbbility Cmmpared to Their Fathers, 1955, 1958, and 1959 . . . . . . . . Month of Marriage for Respondents Related to Present Religious Preference, Detroit Area Study, 1962. O O O O O O O O O O O O O 0 Actual Intermarriage Rates and Intermar- riage Rates "Expected" if Husbands and Wives Were Distributed at Random, for Mar- riages of Protestants and Catholics in the Detroit Area, 1961. . . . . . . . . . .7. ... Religious Types of Marriage Related to the Sex of Protestants and Catholics in the Detroit Area, 1962, by Percentages. . . . . . Religious Types of Marriage Among Protestants and Catholics in the Detroit Area, Related to Officiant at Marriage, 1962. . . ... . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to Total Annual In- come of the Husband, 1962 . . . . . . . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit. Area, Related to Frequency of Mother's Religious Activity When Daughter Was Growing Up, 1962. . . . . . . . . . Rates of Intermarriage (per 100 Marriages) for Protestants and Catholics in the Detroit Area, Related to the Proportion of "Close" Relatives Who Are of Respondent's Faith, 1962 . . . . . . ... . . . . . . . . . Rates of Intermarriage (per 100 Marriages) for Catholics in the Detroit Area, Related to Education in Church Schools, by Sex, 1962. O O O O O O O O C O O O C C O . O O O 0 Religious Type of Marriage Among Catholic.Men in the Detroit Area, 1962, Related to Attend- ance at College by Them and/or Their Fathers . . . . . . . . . . . . . . . . . . . X. Page 245 249 253 255 257 259 262 263 266 271 Table Page 7- l. Protestant Interfaith Marriages in Terms of Social Control: Summary of Findings from the Detroit Area Study, 1955,1958,1959, and 1962. . . . . . . . . . . . . . . . . . 274 7- 2. Catholic Interfaith Marriages in Terms of Social Control: Summary of Findings from the Detroit Area Study, 1955, 1958, 1959, and 1962. . . . . . . . . . . . . . . . . . . . 276 xi. LIST OF ILLUSTRATIONS Figure Page 1. The "Detroit Area" as Defined by the Detroit Area Study. . . . . . . . . . . . . . . . . . . . 178 xii. CHAPTER I INTRODUCTION: THE PROBLEM It was during the academic year, 1958-59, that the writer began to develOp an interest in interfaith marriages as an object of sociological study. He was then a participant in a year-long training program in survey research methods at the University of Michigan, where the Department of Sociology conducts the annual Detroit Area Study for this purpose. As a climax of this experi- ence, each graduate-participant was expected to write up accord- ing to his own interests an analysis of some of the data collect- ed during the survey. Since the writer had long been interested in intergroup relations and since Dr. Guy Swanson's study design had given rise to many questions about religion in the 1959 inter- views, the possibility suggested itself of analyzing interfaith marriages with the data at hand. As it turned out, this analysis in 1959 was a kind of pilot study for the present thesis.1 It taught the writer the need to restrict the sample to characteristics which occurred fre- quently enough in the population of Detroit to permit some cross- tabulations. Therefore the marriages of non-whites, of Jews, and 1It was published, in a revised version, as "Unbroken Prot- estant-Catholic Marriages Among Whites in the Detroit Area," Amer- ican Catholic Sociological Review, 23 (1962), pp. 3-20. 1. 2. of Orthodox Church members were eliminated as presenting special cases. »In considering Protestant-Catholic interfaith marriages, it became evident that a double comparison would generally need to be made, i;g; one between Protestants who were in a mixed marriage and those who were not, and another comparison between Catholics who were in a mixed marriage and those who were not. At the same time, other studies concerned with intermarriage had shown the importance of being clear whether one was counting mar- riages or married peOple when reporting intermarriage rates. For instance, in the data used from the 1959 survey, it was found that of all marriages involving Catholics, 36 per cent were mixed mar- riages, even though only 22 per cent of the married white Catho- lics were partners in these marriages. The reason is simply that it takes two Catholics to make a non-mixed marriage but only one to contribute toward a mixed marriage.2 There are differences, however, between the pilot study and the thesis which grew out of it. The most obvious is that it has been possible, thanks to those responsible for the Detroit Area Study, to make use of data not only from the 1959 survey, but also from those of 1955, 1956, 1958, and 1962. Furthermore, the pilot study was mainly descriptive; the thesis is set in a theoretical framework. The earlier study took both a "before" and an "after" snapshot of the married couples; the thesis is restricted to the "before" view. In other words, we shall here be concerned only 2Ibid., p. 11. More than one well-known sociologist has misread reports of mixed marriage rates. 3. with the background characteristics which may be considered ante- cedent to the marriage; no data will be presented concerning con- sequent religious attitudes or practices. The pilot study did present some hypotheses for testing, which were suggested by the sociological literature but which were unrelated to each other. The prediction of a negative relationship between mixed marriages and a close generational tie with a foreign country was not sup- ported in the pilot study. A similar hypothesis, with a differ- ent rationale, will be tested again in the thesis. The expecta- tions of more frequent mixed marriages among the parents of those who are now in mixed marriages was borne out then and will be re- examined in the thesis. After the writer had transferred to Michigan State Univ- ersity, still intending to pursue this research, he undertook to survey thoroughly the sociological literature on marriage choice, identified as "mate selection" by the biologists who wrote at an early date on the subject. The results of this reading will be discussed in the next chapter. ,At this point it may be sufficient to note that early researchers were impressed with the principles 3 and the "prOpinquity fac— of "like marries like," i.e. homogamy, tor", i.e. that peOple tend to choose marriage partners from among those who live near them.4 Increasingly, sociologists observed 3Cf. Ernest W. Burgess and Paul Wallin, "Homogamy in Person- ality Characteristics," Journal of Abnormal and Social Psychology, 39 (1944), pp. 475-481. 4Cf. Alvin.M, Katz and Reuben Hill, "Residential PrOpinquity and Marital Selections: A Review of Theory, Method, and Fact," Mar— riage and Family Living, 20 (1958), pp. 27-35. 4. that marriage choices between individuals of similar characteris- tics not only de facto occur but that, along certain dimensions, they are expected to. In other words, there appear to be social norms of endogamy, requiring individuals to marry within their racial, ethnic, socio—economic, or religious groups.5 Given these norms of endogamy, social groups must moti- vate their members to conform to them. Most studies on intermar- riage have implied this orientation which considers social control in relation to norms. For example, Joseph Golden took this ap- 6 proach to Negro-white intermarriage. One of the principal con- clusions of David Heer's Canadian study was that an endogamous norm can weaken over time.7 However, it was the Manhattan study by Jerold S. Heiss which made the most thorough use of social con- 8 trol as its orientation. This has had a large influence on the present thesis. Heiss says that his theory of intermarriage assumes that: the ultimate source of barriers to interfaith marriage in American society lie in the family of orientation and for- mal religious organizations....1f, in the case of an inter- married reapondent, it can be shown that these sources did 5See, for example, August B. Hollingshead, "Cultural Fac- tors in the Selection of Marriage Mates," American Sociological Review, 15 (1950), pp. 619-627. 6"Social Control of Negro-White Intermarriage," Social Forces, 36 (1958), pp. 267-269. 7"The Trend of Interfaith Marriages in Canada: 1922-1957," American Sociological Review, 27 (1962), pp. 245-250. 8"Premarital Characteristics of the Religiously Intermarried in an Urban Area," American Sociological Review, 25 (1960), pp. 47-55. 5. not produce effective barriers to integmarriage, we will consider the marriage to be explained. He sets down six general hypotheses to test this theory. In gen- eral, he finds support in his data for five of the hypotheses, in which the intermarried are characterized by: (a) non-religious parents, (b) greater dissatisfaction with parents when young, (c) greater early family strife, (d) less early family integration, and (e)16reater eman- cipation from parents at time of marriage. The Catholic data, he finds, support these five hypotheses. The support is less extensive in the Protestant data and still less in the Jewish. Whether or not a rule of religious endogamy is observed in practice is, therefore, largely a matter of social control. The uses of this concept in sociological literature will be reviewed in Chapter III. Its relevance for the study of interfaith marriages will be found to crystallize in the phenomenon of social institu- tions, esPecially familial and religious.11 When a social institu- tion is in the process of notable change, or when a given situation is ambiguous with respect to a social norm, then we can expect that social control will be weakened. When such a situation could be said to characterize a whole society, it was referred to by Emile 91bid., p. 48. 10Ibid., p. 53. “llMuch reliance will be placed on the treatment of this subject by J. O. Hertzler, American Social Institutions: A Socio- logical Analysis, Boston: Allyn and Bacon, 1961. 6. Durkheim as anomie, normlessness.12 Others have preferred the derivative term, anomia, to refer to the state of normlessness as experienced by the individual.13 It is in the latter sense, prin- cipally, that the concept will be used in this thesis. How, then, do social control and anomia apply to interfaith marriages? In expanding the approach used by Heiss, we ask: Who are the agents of social control in the matter of religious endog- amy? Within the framework of social institutions which we have preposed as a context, we can answer that the parents, the family or relatives, and the church would seem to be the agents of social control in this matter. To the extent that an ethnic or national- ity group is still strong.in our society, it may seek to enforce religious endogamy in the place of a weakening ethnic endogamy.14 When we have given evidence that these agents have probably not ex- ercised control, then we can say that we have given a partial ex- planation of deviancy from religious endogamy. Even if we grant earnest efforts on the part of these agents to exercise social control, there may still be some situations in whieh 2 social norm such as religious endogamy is felt to be weak- ened. Such, for instance, would be wartime, rural-to-urban 12Suicide: A Study in Sociology, translated by John A Spaulding and George Simpson, edited by George Simpson, Glencoe, I11.: Free Press, 1951, p. 258. 1'3Dorothy L. Meier and Wendell Bell, "Anomia and the Achieve- ment of Life Goals," American Sociological Review, 24 (1959), pp. 189- 191. P 1['This is a possible interpretation of Ruby Jo R. Kennedy's findings in "Single or Triple Melting Pot? Intermarriage in New Haven, 1870-1950," American Journal of Sociology, 58 (1962), pp. 56-59. 7. migration, the experience by the son of extreme occupational or educational mobility over the father, or the waiting for marriage until a relatively advanced age. All of these are classified in this thesis as anomic situations and are expected to characterize higher rates of interfaith marriage. Why undertake such a study? The hOpe of an advance in our understanding of interfaith marriages and the opportunity to test a middle-range theory on a mass of data gathered in a metro- politan community by an outstanding survey research agency--these are reason enough. There are additional motivations for the writer, however, Will Herberg and other students of the sociology of religion have given wide circulation to the idea of religious pluralism in our society.15 This is related to the concern of anthropologists and other social scientists with intergroup rela- tions. Is the rule of endogamy an index of intergroup conflict? Conversely, is intermarriage a test of intergroup harmony? There is considerable enthusiasm in the Christian commun- ity for ecumenism and a Protestant-Catholic dialogue. This is quite likely to have its impact on interfaith marriages. In any case, incneased knowledge of interfaith marriages might well pro- vide a context for this "dialogue." It is expected, in fact, that the subject will be discussed by the Roman Catholic bish0ps at the Second Vatican Council. The writer is also concerned about helping the actual , 15Protestant, Catholig,_Jew: An Essay in American Religious Sociology, New York: Doubleday, 1955. 8. participants in interfaith marriages to understand their situation better, so that they can make better adjustments. Those who are considering such a marriage may also be helped toward greater real- ism. If the reader, by now, has begun to share the writer's enthusiam for the importance of this subject, he will find little difficulty in going on to the next two chapters which review the literature relevant to marriage choice as well as to social control and anomia. Chapter IV will describe for fellow researchers the operational definitions of terms and the procedure followed in analyzing the data. The sampling method of the Detroit Area Study makes it reasonable to treat several of the surveys together. How- ever, the small sample of newlyweds interviewed in 1962 were selec- ted from marriage records rather than from city directories. They are therefore discussed in a separate chapter. The reader may find it profitable to turn to Appendix A and Appendix B even before con- tinuing with the next chapter. The first is mostly concerned with providing a demographic and historical setting for this study, and the second will give the reader a synOptic view of the religio- marital composition of the sample. Now let us turn to the time-honored subject of the selec- tion of marriage mates. CHAPTER II MATE SELECTION: REVIEW OF THE LITERATURE The subject: "Who marries whom?--and why?", iafia mate selection, is one which sociologists have, in a special way, claim- ed as their ownJ After a thorough search for journal articles con- cerned in some way with mate selection (and published in English), the writer made a count of their frequency in various professional journals during this century. The two outstanding sociological journals in this country easily rank first and second in this count: American Sociological Review and American Journal of Sociology; Since the purpose of this review is simply to provide a context of relevant research for the analysis of Protestant-Catholic marriages in the Detroit area, the discussion of topics other than intermarriage will be brief. We shall consider only factual re- search with its theoretical interpretations in this chapter, not statements of values or social policies. Although the entire pro- cess which leads to the choice of a mate in marriage will receive some attention, nothing that follows this choice will be discussed. The interest in this subject has been generally increasing over the years since 1930, reaching a peak in 1951. Fluctuations within this trend have occurred, as measured by the number of arti- cles published, but it is the writer's guess that sociologists have simply followed the national trends in this matter: The Depression I 9. 10. years were marked by low marriage rates and low birth rates; the years after World War II were characterized by high marriage rates and a "baby boom." Table 1-1 gives a striking picture of the growth of interest in this subject during the 1930's, 1940's, and 1950's. In each decade the number of articles almost doubled over the pre- ceding decade. Undoubtedly, this is partly due to the appearance of new journals over the years. TABLE l-l.--Journa1 articles published on the subject of mate selection, 1930-1959. Years of Number of Average Publication Articles per Year 1930-39 29 2.9 1940-49 52 5.2 1950-59 fl __9__7 Total: 178 5.9 Such has been the ebb and flow of words on the subject of mate selection. In the sections that follow, we shall look at the literature on marriage rates, on homogamy of mate selection in its various dimensions, on heterogamy as reflected especially in the theory of complementarity of needs, on intermarriage of various types, on hypergamy, and on the antecedents of mate selection. Marriage Rates: To Be or Not To Be? How Soon? Every married person has at least implicitly made two decisions: first, to marry rather than not to marry; then, to marry a particular person with given characteristics. The first decision 11. is logically prior, so we shall give some attention to it before moving on to the second. The P0pu1ation Bulletin of June, 1961, gives an excellent analysis of marriage rates for many countries in the world.1 In this issue it can be observed that since 1920 the United States has had the highest and Ireland the lowest marriage rates of West- ern countries. However, it is difficult to say much more than that about national gross marriage rates because their fluctua- tions are due to so many factors: Periods of war or depression, relative size of the marriage-age p0pu1ation, etc. Few students have said much about those people who do not marry at all. William Ogburn, using United States Census data for 1920 and 1930, stated that "about 10 per cent reach old age without having married."2 Paul POpenoe observed that during the first half of this century in the United States, the trend has been for fewer peOple to remain permanently unmarried, with the exception of the highly educated women.3 .The factor of age at marriage has been widely studied. How- ever, the best data on this point are again provided by the Population Bulletin. In the United States, from 1890 to 1960, the median age at first marriage fell for men from 26.1 to 22.8 years, for women from 22.0 to 20.3 years.4 The greatest dr0p occurred between 1940 and 1950 1Robert Cook, editor, "Spotlight on Marriage," Population Bulletin, 17 (1961), pp. 61-79. 2"Recent Changes in Marriage," 41 (1933), p. 298. American Journal of Sociology, 3"Recent Trends in American Marriages," Eugenical News, 32 (1947), p. 10. 4Robert Cook, editor, op. cit., p. 62. 12. for both men and women, with no change in the following decade. However, this decline was matched or surpassed between 1930 and 1950 by most countries with comparable data.5 Furthermore, in the latest census years, in no recorded country did men marry at an earlier age than in the United States, except in India. However, our women do not marry as young as they do in the recorded coun- tries of Africa, most of Latin America, and half of the Asian countries.6 No doubt, there are many factors at work here, but the writer would guess that our high standard of living (with installment plans), together with the wideSpread practice of birth control, make marriage relatively less difficult to support for American men. In this country it has been observed that, at least at the high school and college level, both men and women who complete the four years marry later than those who drOp out of school before graduation.7 This may directly reflect the influence of the urge to marry and the need to work rather than go to school in order to support a home. However, it is also true that highly educated people are more likely now to marry early, compared to their former practice.8 By contrast, economic depressions make for later 51bid., p. 67. ,f6Ibid., p. 64. 71bid., p. 68. 8Ibid., p. 71. Also see John Hajnal, "Differential Changes in Marriage Patterns," American Sociological Review, 19 (1954), p. 153. 13. marriages.9 Marriage rates have been shown to vary with occupations also. For example, Notestein's analysis of the 1910 United States Census indicated that, on the basis of husband's occupation, the age of women at marriage increases with high social status.10 Women teachers have a low marriage rate, except at forty years or over.11 Farm brides (not grooms) marry about one year earlier, at least in Wisconsin, than non-farm brides.12 Even women who mi- grate from rural to urban areas find that their marriage rate goes down.13 One of the recurring factors affecting marriage rates is an unbalanced sex ratio in a local p0pu1ation. Joseph Greenberg made a study of the marital status in all cities in this country with a population of 50,000 or more. He discovered, not very surprisingly, that the higher the sex ratio (male to female) of an American city, "the greater its percentage of married women. 9Walter C. McKain, Jr., and C. Arnold Anderson, "Assorta- tive Mating in Prosperity and Depression," Sociology and Social Research, 21 (1937), p. 412. 10Frank W. Notestein, "Differential Age at Marriage ac- cording to Social Class," American Journal of Sociology, 37 (1931), p. 22. 11Howard H. Punke, "Marriage Rate among Women Teachers," American Sociological Review, 5 (1940), p. 511. 12G. W. Hill and James D. Tarver, "Marriage and Divorce Trends in Wisconsin, 1915-45," Milbank.Memorial Fund Quarterly, 30 (1952), p. 16. 13Paul H. Landis, "Rural-Urban Migration and the Marriage Rate," American Sociological Review, 11 (1946), p. 158.' 14. and the smaller its percentage of married men."14 'To sum up, countries have their distinctive marriage rates, as do groups or categories within a nation. The trend in recent I decades has been for marriage rates to go up while age at marriage has been going down. Some of the important factors influencing marriage rates and age at marriage are sex ratio, education, and occupation. Homogamy: Like Marries Like In an early unsigned article, attributed primarily to Karl Pearson, two definitions are given which came to be commonly accept- ed and are, therefore, probably worth repeating here: ,Preferential mating is that "in which male or female class- es with certain values of a character find it less easy to mate than other classes with different values." Assortative mating is that "in which, while all classes of males and females find mates, certain classes of males appear to be attracted to certain classes of females." "If a male class of a given character tends to mate with a female class with generally like character, we have a tendency to homogamy."15\ Homogamy is therefore one type of assortative mating. To the writer's knowledge, not much has been discovered to indicate strong homogamy in strictly biological characteristics, 14"Sex Distribution and Marital Status," Sociology and Socggl Research, 33 (1949), p. 371. 15"Assortative Mating in Man," Biometrika, 2 (1903), p. 481. 15. except in age relationships. Burgess and Wallin, using data from 1,000 engaged couples, found some homogamy in height, weight, health, and "physical attractiveness."16 There is somewhat more evidence for homogamy in intellectual abilities. The same authors, reviewing earlier studies, said that, "These studies, using stand- ardized tests, gave an unweighted average husband-wife correlation of approximately .55."17 In their own study of the engaged couples, they tested for homogamy of psychological traits. They found that the positive correlation between the scores of these couples on the Thurstone Neurotic Inventory was .25, and outs self-rating scale of 18 In these aspects twenty-three personality traits it was only .13. "like marries like" is a small truth. Age relationships are a special case of homogamy in biolog- ical characteristics. It is an every day observation in our society that the husband is generally just a little older than the wife, but it is interesting to learn to what extent this is true. The best . study of this factor seems to be that of Click and Landau, based on a sample of 25,000 households by the United States Census Bureau, 1948. They found that the "average" wife was 2.8 years younger than her husband; 78 per cent were younger than their husbands, 12 per , 16Ernest W. Burgess and Paul Wallin, "Homogamy in Personal- ity Characteristics," Journal of Abnormal and Social Psychology, 39 (1944), p. 480. 17"Homogamylin Social Characteristics," American Journal Of Sociology, 49 (1943), p. 110. 18"Homogamy in,Personality Characteristics," op. cit., p. 481. 16. cent were older, 10 per cent were of the same age.19 Socio-economic characteristics have sometimes been treated in studies as just another basis for assortative mating. (In a later section we shall see that some of these are also objects of a rule of endogamy.) .As one example of such a characteristic‘we= can mention the level of education. Hollingshead found in-New Haven, 1948, that for two generations there was a tendency at all levels to marry someone of comparable education. This tendency was strongest for Jews, weakest for Catholics.20 In general, therefore, we can confirm the observation made by Reuben Hill, after reviewing 150 studies, that, in the case of most of the characteristics studied, "people tend to marry people like themselves rather than opposites."21 Residential Pr0pinquity One of the earliest hypotheses to be extensively investi- gated in the matter of mate selection was that marriage partners would tend to have relatively close residences just prior to the marriage. .Perhaps this seems to be an example of sociologists belaboring the obvious. There is, in fact, much to be learned from this approach. However, since we have no data of this sort to analyze in this thesis, we need not spend much time on the subject. 19Paul C. Click and Emanuel Landau, "Age as a Factor in Marriage," American Sociological Review, 15 (1950), p. 523. 20August B.Ikfllingshead, op. cit., p. 626. 21"A Critique of Contemporary Marriage and Family Research," Social Forces, 33 (1955), p. 268. 17. James Bossard gave the initial impetus to this line of research with his study of the distance between residences of couples shown on five thousand marriage license applications in Philadelphia in 1931. He found that one third of all couples lived within five blocks or less of each other, and "the percen- tage of marriages decreased steadily and markedly as the distance betweenjresidents of the contracting parties increased."22 Similar results were reported by other researchers. Fourteen of the studies of residential prOpinquity were reviewed not long ago by Alvin Katz and Reuben Hill. In an attempt to develop a "middle range theory" in this Specific area, they have gone into considerably more detail than is appropriate for this thesis. Let us only note that they call theirs a "norm-interaction" theory, of which the basic assumptions are the following: . 1.: That marriage is normative. 2. That, within normative fields of eligibles, the probability of marriage varies directly with the probability of inter- action. 3. That the probability of interaction is pr0portional to the ratio of Opportunities at a given distance over interven- ing opportunities. Using these assumptions, they have constructed a table by which one can predict whether pre-marriage residential propinquity of couples will be greater or less. They have, in fact, arrived at an out- standing synthesis of many empirical studies. If it were more germane to our thesis topic, we would be obliged to Spend more time, 22"Residential Pr0pinquity as a Factor in Marriage Selection,’ American Sociology, 38 (1932), p. 219. 23"Residential Pr0pinquity and Marital Selection: A Review of Theory, Method, and Fact," Marriagg and Family Livipg, 20 (1958), p. 33. In the original text the quoted material appears in italics. 18. with it. Heterogamy: Complementarity of Needs Residential propinquity was given its place following the section on homogamy in the outline of this chapter because most researchers, following Davie and Reeves, have considered these two factors in mate selection to be closely associated.24 In fact, as the review by Katz and Hill makes clear, it was not simply homogamy, but normative selection, that was at work in residential pr0pinquity. Now we turn to the opposite of homogamy, yig. heterogamy, for not in all respects does it seem to be true that "like marries like." "Opposites attract" may still be a dictum with some truth in it when one is talking about mate selection. Anderson noted, as early as 1938, that "the sole concrete evidence for heterogamy involves per- sonality traits; but the accumulated recent data appear to demonstrate that similarity prevails here as well."25 Let us see. There is a notable similarity between two articles by Anselm Strauss and the later work by Robert Winch, but the present writer does not find that these are acknowledged by Winch among his other sources. Strauss compared the temperamental traits of one's mate and one's father and/or mother. He found that "three temperamental traits had coefficients averaging over .55 for both sexes (mate--both par- ents): (1) gets over anger easily, (2) self-confident, and 24Maurice R. Davie and Ruby Jo Reeves, "Pr0pinquity before Marriage," American Journal of Sociology, 44 (1939), pp. 510-517. 25C. Arnold Anderson, "Our Present Knowledge of Assortative Mating," Rural Sociology, 3 (1938), p. 299. 19. "26 Using the same sample and turning to the (3) sense of duty. question of personality needs, he observes: "Personality needs are a direct outgrowth of early childhood affectional relation- ships with one or both parents"--or, later, with other associates; "the mate fills either the same needs for the individual as were previously filled by a parent, or he fills those needs that a parent left unsatisfied."27 In.1954, Robert Winch, in collaboration with Thomas and Virginia Ktsanes, wrote an article called, "The Theory of Comple- mentary Needs in Mate-Selection: An Analytic and Descriptive Study."28 In the following year they published four more articles on the same subject,29 finally culminating in W'inch's book, 9.1.2.93 Selection: A Study of Complementary Needs.30 The theory repre- sents, therefore, a great expenditure of energy. Let us see what it says, especially as represented in the final version, the book. Winch first takes note of earlier evidence for homogamy in 26"The Influence of Parent-Images Upon Marital Choice," American Sociological Review, 11 (1946), p. 557. 27"Personality Needs and Marital Choice," Social Forces, 25 (1947), p. 332. 28American Sociological Review, 19 (1954), pp. 241-249. 29Thomas Ktsanes, "Mate Selection on the Basis of Personal- ity Type: A Study Utilizing an Empirical Typology of Personality," American Sociological Review, 20 (1955), pp. 547-551. Robert F. Winch, "The Theory of Complementary Needs in.Mate-Selection: .A Test of One Kind of Complementariness," American Sociological Review, 20 (1955), pp. 52-56. Robert F. Winch, "The Theory of Complementary Needs in Mate-Selection: Final Results on the Test of the General Hypothesis," American Sociological Review, 20.(1955), pp. 552-555. .Robert F. Winch, Thomas and Virginia Ktsanes, "Empirical Elaboration of the Theory of Complementary Needs in Mate-Selection," Journal of Abnormal and Social Psychology, 51 (1955), pp. 508-513. 30New York: Harper & Bros., 1958. 20. mate selection, including homogamy of psychological characteristics. But, taking his lead partly from Freud's observation that narcissis- tic (self-loving) persons tend to mate with anaclitic (dependent) persons, he then states his principal thesis as follows: Hence it would seem that mate-selection should follow a princi- ple of homogamy with respect to interests and attitudes. It seems almost as self-evident, on the other hand, that comple- mentariness of motivation (e.g., dominance in one and submis- siveness in the spouse, nurturance in one and receptivity in the other) would maximize gratification at the-motivational level and hence that on the level of such needs mate-selection should follow a principle of complementariness. Without in any way abandoning the Freudians' interpretation of love, Winch makes use of H. A, Murray's definition of "need": a tension which, when reduced, is experienced as pleasure. .In infancy these needs are mainly biological, but in maturity the social aspects become dominant. The author selects some points in the total picture of child development which he thinks give rise to needs which the in- dividual will later seek to gratify in the marital relationship. Starting with Murray's list of needs and general traits, he and his collaborators compiled a list of twelve needs and three "general traits" to cover "the emotional and motivational elements" involved in mate-selection.32 They then "hypothesized for everyone of our fifteen variables that a person high with respect to that variable would select a mate who was low on that variable." Hence the central hypothesis: "In mate-selection the need-pattern of each spouse will be complementary rather than similar to the other spouse."33 31Ibid., p. 10. 32Ibid., p. 92. 33Ibido, pp. 95-960 21. The types of complementariness and other details of the theory need not concern us here. It is disappointing to discover, however, that the authors were able to test their elaborate theory on no more than twenty3five couples, all undergraduates at North- western University, married two years and childless. Winch's con- clusion was that the data generally supported the hypothesis: The degree of association was low, but significant. The initial reactions to Winch's theory were rather nega- tive. However, some have tried to put it to an empirical test. Charles Bowerman and Barbara Day, using a design which was not a close replication of that of Winch, arrived at conclusions contrary to the theory.34 James Schellenberg and Lawrence Bee undertook to reconcile the conflicting results above by analyzing new data from 64 recently married couples and 36 pre-married couples. Contrary to Winch's theory, the results tended slightly in the direction of homogamy; Loop, 73 per cent of the married couples and 61 per cent of the premarried showed positive correlations of need patterns. "The main difference in data-gathering procedure is that.in the pre- sent study scores are derived directly from a standardized personal- ity inventory, while Winch's data are based upon ratings made by investigators after interviews."35 34"A Test of the Theory of Complementary Needs as Applied to Couples during Courtship," American Sociological Review, 21 (1956), PP o 602‘605 o 35"A Re-Examination of the Theory of Complementary Needs in Mate Selection," Marriage and Family Living, 22 (1960), p. 231. 22. Alan Kerokhoff and Keith Davis have just published another test of this theory.36 Their subjects were 94 college couples who were seriously considering marriage. The methodology was objective and reported very carefully. .In October, 1959, all couples filled out questionnaires separately from each other. These included Bernard Farber's index of value consensus and William Schutz's FIRO-B scales, the latter being modified before the analysis so as to conform to the concept of complementarity. The couples were divided into two approximately equal groups: those with "long-term” association (going together eighteen months or more) and those with "short-term” association (less than eighteen months). In May, 1960, each subject was sent another short questionnaire, to establish par- ticularly whether they felt they were nearer to, or farther from, being a "permanent couple". Fifty-six couples thought they were nearer, thirty-eight couples thought they were the same or farther apart. The general hypothesis that the degree of value consensus is positively related to progress toward a permanent union found support in the data, significant at the .05 level or better. The three measured types of complementarity were in the predicted direc- tion, but their tendency toward a positive relationship with progress toward a permanent union was not statistically significant. The effect of the control variable, length of association, had not been predicted; but the analysis showed an interesting pattern. The direc- tion of relationships remains the same, but for the short-term 36"Value Consensus and Need Complementarity ionate Selection," American Sociological Review, 27 (1962), pp. 295-303. 23 0 couples only value consensus is significantly related to progress toward a permanent union, not need-complementarity. On the other hand, for long-term couples the Opposite is true: Two of three measures of need-complementarity are significantly related to progress toward a permanent union, but value consensus is not. The authors' exopost facto interpretation of their data is this: ,A‘series of "filtering factors" Operate in mate selection at different stages of the selection process. Our data general- ly support the idea that social status variables (class, re- aligion, etc.) Operate in the early stages, consensus on values somewhat later, and need complementarity still later. -Our interpretation of the delay in the operation of the complementarity factor is that such personality linkages are“bften precluded by the unrealistic idealization of the -loved one in the early stages of courtship. In summary, the validity of the theory of complementarity of needs is still very much in question. It has been tested only on small college groups selected without systematic randomization, (by researchers who had varying measures of "complementarity", with results which do not consistently support the theory. Although, Kerckhoff and Davis seem to have saved the theory by their concept of "filtering factors", it seems to me that Winch will be disappoint- ed in their findings. If the relationship between need-complementar- ity and a permanent union holds only for long-term couples, this means that "love", as Winch understands it, is a factor in mate- selection only after eighteen months or more of association. If he holds this position, he may have to fight it out with Hollywood. 37Ibid., p. 303. 24. Endogamy vs. Intermarriago; GeneralyConsiderations We come now to an important transition in the progress of this chapter. We have been reviewing the literature concerned with homogamy and heterogamy. These are correlative terms which describe mate selection from the point of view of the individual or, if you wish, from the point of view of the average tendency of individuals in a p0pu1ation to choose mates who are similar to, or different from themselves in some reapect. In this and the following sections we shall discuss mate selection from the point of view of the group, I;_2;2 what impact the presence or absence of a group norm has on patterns of mate selection. The difference between these concepts may be likened to the distinction which William Graham Summer draws between folkways and mores: When this conviction as to the relation to welfare is added to the folkways they are converted into mores, and, by virtue of the philoSOphical and ethical element added to them, they win utility and importance and become the source of the science and the art of living.38 To be more Specific, we are talking now about the group norm of 2292: gopy, defined by Kingsley Davis as: "marriage into a class of per- sons of which I also am a member." He understands intermarriagg al- ways with reference to endogamy as: "the violation of or deviatiOn from an endogamous rule."39 Although most writers have not explicitly 38Folkways: A Studyoof the Sociological Importance of UsagesI Manners, Customs, Mores, and Morals, New York: New American Library, A Mentor Book, 1960, p. 19. 39"Intermarriage in Caste Societies," American Anthr0pologist, 43 (1941), p. 376. Davis defines the correlative term, exogamy, as: "Marriage into a class of which I am not a member." However, this seems not to exist as a group norm in our society, except in the case of the incest taboo. 25. used the terms with this precision even since Davis' clarifications in the 1941 article, the sense of them seems to be implied in, or at least is consistent with, most of the sources discussed below. It is the sense in which the present writer will use the terms. Most empirical studies of intermarriage havebeen almost exclusively concerned with only one of these types of intermarriage: racial, ethnic, or religious. We shall discuss these in turn, re- serving for a separate part of the paper the intermarriages which cross lines of class and caste--because these are-associated with a special type of marriage called hypergamy. Before going on to specifics, let us take a brief look at those sources which give considerable attention to more than one type of intermarriage. Their separate findings will be reviewed, however, in the appropri- ate sections. The pioneering study by Julius Drachsler in New York City is mainly concerned with interethnic marriages.40 However, he also reviews the background of the subject and the data on interracial and Jewish-Gentile marriages which were available to him. flFuther- more, he obviously views intermarriage as we do in this chapter: ,Intermarriage, as such, is perhaps the severest test of group cohesion. Individuals who freely pass in marriage from.one ethnic circle into another are not under the spell of an in- tense cultural or racial consciousness. Consequently, the great- er the number of mixed marriages, the weaker, broadly speaking1 themgroup solidarity. Moreover, such a test is quantitative. 40Democracy and Assimilation, New Yorke ‘Macmillan, 1920. Also Intermarriage in New York City: A Statistical Study of the Amalgamation pofoEuropean PeOples, New York: COpyright by J. Drachsler. 41Democracy and Assimilation, Op. cit., p. 87. 26. Two Andersons,,C. Arnold and Elin L., some six or seven years before Ruby Jo Reeves Kennedy's now-fambus article on the "Single or Triple Melting Pot?" and almost twenty years before Will Herberg's Protestant-Catholic-Jews, pointed out the emer- gence of religious pluralism in our society. .In reviewing ear- ‘lier studies of mate selection, C. Arnold Anderson Observed that "ethnic, economic, and religious boundaries between groups are often coterminous; but when they diverge, the religious are usually strong- est."42 .Elin L. Anderson, in an excellent study of assimilation in Burlington, Vermont, said: It was found that for a considerable number of Burlingtonians social horizons are bounded by the barriers which separate the three main religious groups, Catholic, Protestant, and Jewish. Throughout, this was more true of the women than of the men. Since the Jews were not numerous in Burlington, the data concerned mahfly the other two groups. Of these the author concluded that, "Americanization is being accomplished, but only within two separate camps."44 Lowry Nelson also noted the same in rural Minnesota a year before Kennedy's article: "It is noteworthy that...intermarriage be- tween religious groups is markedly less than among nationalities."45 If the observation of religious pluralism.was not original with Ruby Jo Kennedy, the descriptive term "triple melting pot" apparently was; l'2"Our Present Knowledge of Assortative Mating," Op.cit., p. 298. 43We Americans: A Study of Cleavage in an American City, Cam— bridge: Harvard University Press, 1937, p. 165. 44Ibid., p. 88. 45"Intermarriage among Nationality Groups in a Rural Area of Minnesota," American Journal of Sociology, 48 (1943), p. 590. 27. and she demonstrated the salience of religious ties over ethnic ties by analyzing thousands of cases from the marriage-license bureau of New Haven, Connecticut.46 Bernhard Stern's article-on.91ntermarriage" in the Encyclope- dia of the Social Sciences has‘theAmerit of treating the-subject his- torically and cross-culturally.47 However, the outstanding work on‘ intermarriage in general is Milton L. Barron's PeOple Who Intermarry.48 Although his own data collection in Derby, Connecticut, was not very extensive, his review of earlier studies seems to be-quite-exhaustive. He was concerned with the causes and intermarriage patterns, in the 1 three principal types of intermarriage, racial, ethnic, and religious. The Specific causal factorswhich he discusses here and in his review article will be introduced in their apprOpriate sections. 1 Racial Intermarriago What is racial intermarriage? Our answer to this question- depends, of course, on our definition of poop, and there is not very much agreement on this even among the experts. The writer will be content to follow the usage of the more recent authors, who have stressed the social significance of race, which has been built on its biological base. Race is therefore understood as: "the way 46"Single or Triple Melting Pot? Intermarriage Trends intNew Haven. 1870- 1940, " American Journal of Sociology, 49 (1944), pp. 331- 339. 47Edwin R. A. Seligman, editor, EncyclOpedia of tho Sogial Sciences, 8, pp. 151- 154. Syracuse: Syracuse University Press, 1946. 28. numbers of a society classify each other by physical characteris- tics."49 Negro-white intermarriage in the United States has taken place at a very low rate. This is not surprising, considering the long history of Negro slavery in this country, the long struggle for abolition, and the fact that twenty-eight of the States pro- hibited intermarriage as late as l958--with punishments as high as $1,000 fine and/or ten years in prison.50 The data given by Wirth and Goldhamer are probably the most extensive on this subject. They show from other studies that Negro-white intermarriage in this coun- try reached its peak in Boston for a short period after 1900, when the rate was 13.6 per 100 Negro marriages. Their own data for Boston, 1914-1938, and New York State (excluding New York City), 1916-1937, show rates from 3.1 to 5.2 in Boston, 1.7 to 4.8 in New York. Of course, these rates, if reported as per 100 marriages in- volving whites, would be very much lower because the Negro population in these places was so small.51 Less is known about the intermarriage 49Charles'Wagley and Marvin Harris, Minorities in the New World: Six Case Studies, New York: Columbia University Press, 1958, p. xvi. 50Joseph Golden, op. cit., p. 268. SlLouis Wirth and n. Goldhamer, "The Hybrid and the Problem of Miscegenation," Characteristics of the American Negro, ed. Otto Klineberg, New York: Harper & Bros., 1944, p. 281. .It seems desirable to warn the reader at this point that it is very easy to misread the reported rates of intermarriage«be- cause two pe0ple and at least two groups are always involved. For example, the prOportion of Negroes (individual persons) who have intermarried may be reported instead of the proportion of Negro mar- riages (acts involving two persons), thereby lowering the magnitude of the rate. Similarly, the rate may have as its base each 100 mar- riages involving Negroes 9£_each 100 marriages involving whites‘gg each 100 in the sum total of all_marriages. The same caution should be used in reading the reported rates for ethnic and religious inter- marriage. 29. rates of other non-whites in the United States, but it is clear that a very tiny proportion of the white population is involved in all of these together. ‘ On the other hand, interracial marriage in Hawaii and in Brazil or other countries of Latin America is an accepted part of the culture. Ever since a small number of white men settled in Hawaii about 1790 and married the daughters of the chiefs, it has been unwise to speak publicly against interracial marriages in the Islands.- Although the Japanese in Hawaii were nearly 38 per cent of the p0pu1ation in 1930 and they intermarried less than any other group, still, intermarriage rates have not declined, the general rate in 1945-54 being 28.4 per cent of all marriages.52 Brazil seems to be equally innocent of racial bars in mate selection. What are some of the characteristics associated with inter- racial marriages? Due to an abnormally high sex ratio particularly among Asiatics in this country, intermarriages have been most charac- teristic of males in the minority group. The dimensions of this problem may be illustrated by the fact that, of all the Filipinos in the United States in.l930 who were fifteen years old or more, 40,904 were males and 1,640 were females. In the same year, 76.7 per cent of the females but only 18.1 per cent of the males were married people.53 Negro-white marriages in the United States have some special characteristics. Without the unfavorable sex ratio in the p0pu1ation, 52C. K. Cheng and Douglas S. Yamamura, "Interracial Marriage and Divorce in Hawaii," Social Forces, 36 (1957), p. 82. 53B. T. Catapusan, "Filipino Intermarriage Problems in the United States," Sociology and Social Research, 22 (1938), p. 268. 30. as noted for Asiatic-Americans, Negro males outnumber Negro females in interracial marrages by at least three to one. Negro males who intermarry usually have better-than-average occupations. The 0p- posite is true of white males who intermarry, and these are also likely to be foreign-born.54 In Golden's Philadelphia study, less than half of the interracial marriages were homogamous in the matter of religion.55 What can be said, in summary, that is distinctive about racial intermarriage? The rates are very low in this country with no evidence of a trend up or down. Some credit must be given to the "visibility factor" in allowing social sanctions to be unusually effective in prohibiting this type of intermarriage. Males in the minority group intermarry much more than females, especially where there is an unfavorable sex ratio. Such marriages are also likely to include mixture of religion and to occur somewhat later in life. Why is there no social norm against racial intermarriage in Hawaii, Brasil, and some other societies? The factors are certain- ly complex and not perfectly understood. However, some authors are inclined to stress the long historical view, in which the initial contacts between races were found to be conducive to intermarriage and in which the dominant group(s) have then evolved a favorable 54Louis Wirth and H. Goldhamer, op. cit., p. 281. 55Joseph Golden, "Characteristics of the Negro-White Inter- married in Philadelphia," American Sociological Review, 18 (1953), pp. .177-1830 31. cultural norm.-56 -Ethnic Intermarriage As in the discussion on race, we find in the concept, "ethnic group", a variable of great importance but one which is also difficult to define to everyone's satisfaction. However, the writer will adopt the following description of an ethnic group: ...a collection of people considered both by themselves and by other people to have in common one or more of the following characteristics: (a) religion, (b) racial origin (as indicat- ed by identifiable physical characteristics), (c) national origin, or (d) language and cultural traditions. ~A consequence of this definition is that it is possible to divide the p0pu1a- tion of the United States into ethnic groups in any one of a number of different ways. .A second consequence is that most individuals in the United States are members of several differ- ent ethnic groups....It is a hopeless tgsk to specifythe boundaries of any ethnic group exactly. It follows, therefOre, that racial and religious intermar- riage are special types of ethnic intermarriage, as understood here. Their distinctive aspects and their importance in the literature make it desirable to talk about them apart from other types of ethnic intermarriage. Our focus in this section will therefore be on national origin and language and cultural traditions as constituting self-conscious and recognizable groups. The term is commonly used 56See, for instance, Roger Bastide, "Race Relations in Brazil,“ International Social Science Bulletin, 9 (1957), pp._495-512; and Donald Pierson,lNggroe8'in Brazil, Chicago: University of Chicago Press, 1942. ' 57John Harding, Bernard-Kutner, Harold Proshansky, and Isidor Chein, "Prejudice and Ethnic Relations," Handbook of Social Psycho- logy, ed. Gardner Lindzey, Reading, Mass.: _Addison-Wesley, 1954, p. 1022. ' 32. in this more narrow sense. Avoiding the comparable term, "minority group", also makes it possible for us to include the following dom- inant types in our society as ethnic groups, singly or in combina- tion: whites, Protestants, the English-speaking, Anglo-Saxons. »It will be found, however, that authors do not always accept national origin as an adequate description of a specific ethnic group: e. ., Canadians are French or English Canadians; Russian nationals may be Polish or Jewish ethnics; Swedes, Norwegians, and Danes are sometimes grouped as Scandinavians. Even though the term, ethnic intermarriage, has itself recieved various Operational definitions, we shall not take time to discuss these here. Those who have studied the behavior of immigrants quickly become aware that, besides interethnic marriage, there is such a phenomenon as intergenerational or internativity marriage. ‘In other words, whether or not an immigrant marries someone of differ- ent ethnic origin, he may marry someone who is also an immigrant, or who is the child of immigrants, or who is the child of native- born Americans. Bessie Weasel developed a numerical system for labeling such marriages, which we shall follow here to the extent that it is practical: 1-1 refers to a marriage in which both parties are first generation Americans, or foreign-born. 2-2 refers to a marriage in which both parties are second generation Americans, or native born of foreign parentage. 3-3 refers to a marriage in which both parties are at least third generation Americans, i;_g;, native born of native parentage, whatever the ethnic derivationnalsos8 58An Ethnic Survey of Woonsocket, Rhode Island, Chicago: University of Chicago Press, 1931, pp. 23-24. 33. called "01d Americans." If the marriage partners belong to differ- ent American generations, the first number belongs to the male, the second to the female, as: 2-1, 1-2, 2-3, 3-2, 3-1, 1-3. Bossard is the only researcher who concentrated on this type of marriage rather than interethnic marriages. In upstate New York, .1936, he discovered that 62 per cent of the marriages were between partners of the same generation in America, distributed as follows: 39.0 per cent were 3-3 marriages; 19.9 per cent were 2-2 marriages; 3.2 per cent were 1-1 marriages. The other 37.8 per cent were "internativity" marriages, distributed as follows: 26.5 per cent were 3-2 or 2-3 marriages; 6.6 per cent were 2-1 or-l-2 marriages; 4.7 per cent were 3-1 or 1-3 marriages.59 Because, as we have seen, there are so many ways-to "slice the pie",‘i&_g;, to define categories and analyze data, it is very difficult to find results in one studywhich are comparable with those in another, as far as ethnic intermarriage is concerned. We shall just have to do the best we can. In general, after reviewing many studies, Barron concluded that, "all other factors being equal, it was found that the longer a group's residence and the older its nativity in the United States, the greater was its incidence of intenmarriage."6o. What about the intermarriages of the foreign-born, i. e., the first generation ”James H. Bossard, Marriage and the Child, Philadelphia: University of Pennsylvania Press, 1940, p. 100. 6oPeo 1e Who Intermarr : Intermarria e in a New England Industrial Community, op. cit., p. 319. 34. Americans? Drachsler finds the rate in New York City, 1908-12, to be 10 per cent.61 Wessel established the rate in Woonsocket, 1926, to be 7.8 per cent, and in New London, Conn,, 1938-40, to be 9 per cent.62 .Bossard's figures for upstate New York are so much in ex- cess of these that one hesitates to accept them. .In fact, his rates would be even higher if they were based on marriages instead of on’ grooms and brides. In any case, he reports that the average of ten major nationalities showed in.1929 that 40.7 per cent of males and 43.1 per cent of females crossed ethnic lines, whereas in.1936 that 54.0 per cent of males and 56.3 per cent of females did 80.63 All rates for interethnic marriages in the second generation rise considerably. .Drachsler's rate for New York City is 32 per cent.64 Wessel found it to be 23.6 per cent in‘Woonsocket.65 Again, Bossard's second generation rates for upstate New York are much high- er: 6l.7*per cent of men, 57.7 per cent Of women intermarry.66 Only Wessel's data take us beyond the second generation for interethnic marriages rates. The rate for all types of.3-2 mar- riages is 44.0 per cent. The rate for 3-3 marriages is 31.7 per cent§7 61Intermagriage infiNew York City, Op. cit., p. 42. 62An Ethnic Survey of Woonsocket, Op. cit., Table 35; also "The Ethnic Survey of New London, Connecticut, 1938-44, A Resurvey after Some Twenty Years," American Journal Of Sociology, 50 (1944), p. 90. 63Marriage and the Child, op. cit., p. 115. 64Intermarriage in New York City, Op. cit., p. 42. 65An Ethnic Survey of Woonsocket, Op. cit., Table 35. 66Marriage and the Child, Op. cit., p. 110. 67An Ethnic Survey of Woonsocket, op. cit., Table 35. 35. The higher rate for the 3-2 marriages may reflect the tendency for marriages which cross generational lines to cross ethnic lines also. ,Not all ethnic groups have identical intermarriage rates, of course. Drachsler presents a long table showing intermarriage rates for all nationalities in New York City, 1908-12. Those which can be compared with other studies are presented in Table 2-2. Here is a sampling of the others, from low to high: Syria, 4.6 per cent; Finland, 16.8 per cent; Greece, 22.1 per cent; France, 49.6 per cent; Scotland, 59.8 per cent; Holland, 62.6 per cent; English Canada, 79.9 per cent; Portugal, 88.2 per cent. TABLE 2-2.--Summary of rates of ethnic intermarriages for selected " nationalities, as reported from research, by percentage. Nationality New Yorka Woonsocketb Buffaloc New Havend Buffaloc 1908-12 1926 1930 1870-1940 1960 Irish 21.6 34.3 35.5 German 33.3 52.6 English 62.7 38.1 French Canadian 75.6 12.8 Polish 20.3 3.1 21.0 42.3 67.0 Italian 6.8 13.9 29.0 14.9 73.0 aJulius Drachsler, Democracy and Assimilation, Op. cit., Table V. bBessie Wessel, An Ethnic Survey Of Woonsocket, Op. cit.,} Table 36. cB. R. Bugelski, "Assimilation through Intermarriage," Social Forces, 40 (1961), p. 148. dRuby Jo R. Kennedy, "Single or Triple Melting Pot? Inter- marriage Trends in New Haven, 1870-1940," Op. cit., p. 333. 36. ‘.If we can assume the-studies assembled in Table 2-2 to be roughly comparable, the table suggests several tendencies which have been observed to be at work in ethnic intermarriage rates._ For one thing, Barren and Others have noted that "the larger the group, thelower‘itsrate."68 This seems to operate in keeping the rates of the Irish fairly moderate in Eastern cities. »It also accounts largely for the striking difference between the French Canadians in New York City and in.Woonsocket, since they are easily the largest ethnic group in the latter city. ,Secondly, at least the German, Polish, and Italian cases seem to illustrate the tendency of a group's rate to increase with the length of its residence in the United States, as Drachsler's and Wessel's per- centages showed to be true for general rates also in the preceding sections. Finally, comparing the Irish and Germans as Older immi- grants snd the Polish and Italians as new immigrants, we see some evidence for Anderson's statement that: "Correction for duration of residence in the nation, sex ratios, rural-urban residence, and other extraneous factors does not eliminate the variations in endo- gamoustendency."69 As in racial, so also in ethnic, intermarriage, an unbalanced sex ratio (usually a surplus of men) leads to sex differences in the rate of intermarriage. Drachsler considered this to be the strong- est factor leading tO intermarriage among first generation Americans.7° 68Peo 1e Who Intermarr ,‘ppg_p££., p. 191, 69"Our Present Knowledge of Assortative Mating," Op. cit., p. 2970 ‘ 7ogpgg§gpgpipge in New York City, op, cit., p. 42. 37. It loses its importance in the second generation. Comparing one nationality with another, we can only say that the evidence shows no consistent relationship between sex and ethnic intermarriage. When ethnic lines are actually crossed in marriage, other elements of similarity seem to determine the direction toward which the choices tend. .Sometimes linguistic and other cultural similarities in the ethnic background exert an influence.71 But particularly has religious homogeneity been proposed as a strong influence in this choice, as we shall see. Let us turn, there- fore, to a review of the findings on religious intermarriage. Religious Intermarriage: Types and Sources Of Data Except for the fact that some early studies regarded Jews as an ethnic group among nationalities in this country, for many years little attention was given to religious intermarriages. How- ever, in the last twenty years this type of intermarriage has elicited almost twice as many studies as any other kind. At present, because none of our data comes from the non-Western world, we are concerned with only three major religious groups: Protestant, Cath- olic, and Jewish. There are, of course, many churches and sects among Protestants, several rites among Catholics, and divisions among Jews. However, studies of intermarriage generally ignore these differences, except for a few which are concerned with a specific -Protestant group. Those people with no religious persuasion rarely appear in the results. 71MiltOn Barron, PeOple Who Intermarry, Op. cit., p. 319. Also J. V. DePorte, "Marriages in the State of New York with Special Reference to Nativity," Human Biology, 3 (1931), p. 396. 38. The question of the source of data takes on unusual impor- tance in the matter of religious intermarriage, at least in-the United States. In Canada, in Holland, and seemingly in most Euro- pean countries, the civil registration of all marriages includes information about the religious persuasion of the contractants. In this country a question about religion has generally not been asked when a marriage license was issued. Only in two states so far has this information been sought by the marriage license 72 Only bureau: in Iowa since 1953 and in Indiana since 1959. information obtained in the civil registration is entirely sat- isfactory from the point of view of completeness Of coverage. In some studies, the inference of religious identity has-some- times been made from the ethnic identity secured in the civil 73 registration, as in New Haven. In a small community, this method can be improved by interviewing local people, as in Derby, Connecticut.74 Perhaps the next best source of information about the general p0pu1ation would be a survey based on a probability sample. ‘The only national survey so far reported which includes information about religious intermarriage is the sample inter- 72David M. Heer, "The Trend Of Interfaith Marriages in Canada: 1922-1957," American Sociological Review, 27 (1962), p. 245. 73Ruby JO R. Kennedy, op. cit., and August Hollingshead, Op. cit. 74Milton Barron, Op. cit. 39. viewed by the United States Census Bureau in 1957.75 Samples have also been selected by reliable institutions for surveys conducted in local population centers, such as Manhattan in New York City or the Detroit MetrOpolitan Area.76 The possibility Of Obtaining more information from each respondent may be considered as com- pensating for the sampling error introduced. Another source Of data which is complete up to a point is church records of marriages. Since, to the writer's knowledge, only the Catholic and Orthodox churches regard matrimony as one of the seven Sacraments instituted by Christ, it is only they who claim jurisdiction over these contracts and therefore require that a religious official assist at and register the reception of the sacrament. Consequently, the records of the Roman Catholic Church, which are kept in the local parish and recorded on a diocesan basis in the Official Catholic Directory,77 have been analyzed by, many researchers.78 These studies may also include more intensive surveys of the parish community in the form of a house-to-house 75"Religion Reported by the Civil Population Of the United States: March, 1957," Current Popglation Reports: Population Characteristics, Series P-20, No. 79, Washington: Bureau of the Census. 76Jerold Heiss, op. cit., and Paul Besanceney, Op. cit. 77Published annually by P. J. Kennedy and Sons, New York. 78See, for example, John L. Thomas, The American Catholic Family, Englewood Cliff, N. J.: Prentice-Hall, 1956. Also Edward S. Dunn, "Mixed Marriages in 1947," Ecclesiastical Review, 120 (1949) , pa 34o 40. census.79 The study of Lutheranvinterfaith marriages by Leiffer apparently was based on such a survey.80 The only study which seems to have relied on Protestant church records is the question- naire-survey conducted among 382 pastors in the United Lutheran Church of America by Bossard and Letta.81 The great handicap of all Of these church records and surveys is that they tell us little or nothing about the people who married outside of their church even though they still claim allegiance to it. -It is true that by this act a Catholic, at least, is no longer in communion with his church; but his case is no less interesting to both sociologists and priests. The least desirable source of data from the point Of view of telling us about the general population is a collection of cases that just happen to be available in the circumstances. Baber used the data of those who came to him for counselling at Nevaork University. Slotkin used case records in welfare agencies. Nelson and Landis used questionnaires given to students to gather information about their parents. Prince had his students distri- bute questionnaires only to those who had contracted an interfaith 79Notable examples were: .Joseph H. Fichter, Southern Parish, Vol. 1: Dynamics of a City Church, Chicago: University Of Chicago Press, 1951; Joseph B. Schuyler, Northern Parish: A Sociological and Pastoral Study, Chicago: ,Loyola University Press, 1960; Gerald J. Schnepp, "Three Mixed Marriage Questions Answered," Catholic World, 156 (1942), pp. 203-207. 80Murray H. Leiffer, "Interfaittharriages and Their Effect A on the Religious Training of Children," Lumen Vitae, 4 (1949), pp. 442-454. 81James'H.-Bossard, and Harold C. Letts, "Mixed Marriages Involving Lutherans," Marriage and Family Living, 18 (1956), pp. 308-310. 1'“ It'll 41. or an interdenominational marriage.82 The internal dynamics of such cases may be interesting even though one cannot generalize to a larger pOpulation. Several studies have compared the data Obtained from civil registration of marriages with those Obtained from church records. The best comparison of this type has been carried out in Iowa. The first conclusion was that, in 1955 in Iowa, interreligious marriages were "far more numerous than is generally recognized;" in fact, Catholics were more numerous than church records showed, being 22.5 per cent of the pOpulation by the state's count pp, 13.5 per cent in the estimate of the Catholic Directory for 1950.83 Ecclesiastical and civil records agreed almost perfectly in the count of endogamous Catholic marriages in Iowa, 1953-57; the cor- relation between them was .997. The discrepancies arise when mixed marriages are included in the count. The prOportion of mixed marriages in these years was: according to state records--35 per cent Of all marriages or 30 per cent of first marriages only; according to church records--26 per cent of all marriages. How- ever, the state records also show that 63 per cent Of all remar- 82Ray E. Baber, "A Study of 325 Mixed Marriages," American Sociological Review, 2 (1937), pp. 705-716. J. S. Slotkin, ”Jewish- Gentile Intermarriage in Chicago," American Sociolggical Review, 7 (1942), pp. 34-39. Lowry Nelson, "Intermarriage among Nationality ,Groups in a Rural Area of Minnesota," American Journa1_gf Sociolpgy, 48 (1943), pp. 585-592. Judson T. Landis, "Marriages of Merd and Non-Mixed Religious Faith," American Sociological Review, 14 (1949), a pp. 401-407. ,Alfred J. Prince, "A Study of 194 Cross-Religion Marriages," The Family Life Coordinator, 11 (1962), pp. 3-7. 83Loren E. Chancellor and Thomas P. Monahan, "Religious Preference and Interreligious Mixtures in Marriages and Divorces in Iowa," American Journal of Sociology, 61 (1955), p..233. 4.2 I riages involving Catholics are mixed marriages. The state reports over twice as many mixed marriages as the church, if remarriages are included.84 Such striking differences underline the need to consider the source of data when rates of religious intermarriage are shown. Rates and Trends: Jewish-Gentile'Marriages Some students of ethnic intermarriage have included the Jews as a nationality group like the Swedes or Italians. Apparent- ly they gathered their data about Jewish marriages more by infer- ence based on names or the identity of the Officiant at marriage than from a direct identification in the civil records. Or perhaps school children were led to identity themselves in this way. .In an individual community, Jews may appear as an ethnic group (in the narrow sense) because they have emigrated together from the same country to come to America, and hence they speak the same language and have distinctive customs. Actually, we know that for the United States as a whole there are important differences be- tween Jewish immigrants. The earliest group was mainly Sephardic . and for a long time used Spanish in their ritual. (A much larger Jewish immigration arrived from Germany in the 19th century, being known as Ashkenasim and having their own dialect. Finally, between 1870 and 1914 some two million Jews emigrated from eastern EurOpe and became identified in this country as a Yiddish-speaking ethnic community. As with other immigrants, the national distinctions 84Lee‘G. Burchinal, William F. Kenkel, and Loren E. Chancellor, "Comparison Of State-and Diocese-Reported.Marriage Data for Iowa, 1953-57," American Catholic Sociological Review, 23 (1962), PP. 25 & 28¢ 43. have become lost or blurred. In the United States, being Jewish is much less of a narrowly ethnic identification than it was formerly. It is now almost exclusively a religious identification.85 The persistence of low intermarriage rates, in Spite of Americaniza- tion, may be taken as a justification for considering Jews in this section of the chapter rather than in the earlier one on ethnic intermarriage. The incidence of Jewish intermarriage has been studied for a long time, both in this country and in Europe. Milton Barron summarized these studies in a single table in 1946.86 The rates for EurOpe and Canada will be reviewed only briefly here. The range of intermarriage rates in EurOpe is great: from 0.39 per cent for Lithuania in 1931 to 51.00 per cent for Germany in 1915. The highest recorded rates of intermarriage were found in Germany during the two decades before the rise of Hitler. In Canada, the range is from 2.99 per cent in Quebec for 1926-31, to 6.52 per cent for the rest of Canada in 1920-31. However, David Heer has shown that here also, for the three Provinces of Quebec, Ontario, and Manitoba, the rate of interfaith marriages among Jews has increased by 126 per cent from 1927 to 1957, i:_gzofrom 3.0 per cent to 6.8 87 per cent. To make the record fairly complete for the United States, 85For a fuller discussion of these movements and changes, see Will Herberg, Protestant, Catholic,,Jew, Op. cit., éhap, viii. 86PeOple Who Intermarry, Op. cit., p. 177. 87"The Trend of Interfaith Marriages in Canada; 1922'1957’" 02. Cite, p0 2460 44. Table 2-3 shows all of the rates for Jewish.intermarriage in this country which have been published to this writer's knowledge. -The adata shown here-were-summarized by Barron up to and including his own study. One may wonder about the value of some of these statis- tics, when Barron publishes a rate of 16.67 for Derby, based on TABLE 2-3.--Summary of published rates of Jewish intermarriage in the United States, per 100 Jewish marriages W Stamford, Conn.a 1938 Place Date Rate Place -Date 'Rate New York City8 :1908-12 .2.3 'Derby, Cone.a :1929s30 o.o Cincinnati‘ .1916-19 3.6 Woonsocketb ~1926 1.8 ‘New'Haven‘ 1870 0.0 Derby,.Conn.a 1940 16.7 New Haven8 1900 1.2 .New Havenc 1948 2.9 New Haven‘ 1930 3.0 New Haveng .1950 3.9 New Haven“ 1940 5.7 U. s. A.e -1957 7.2 7.2 Manhattanf 1957 $18.4 aBarron, Op. cit., p. 178. bWeasel, Op. cit., p. 109. cHollingshead, Op, cit., p. 622. dKennedy, op, cit., p. 138. 6United States Bureau of the Census, op. cit., p. 35. fHeiss, Op. cit., p. 49. one case ofintermarriage amogg six Jewish marriages in the community. The data assembled in Table 2-3 give rise to a few observa- tions. .First, the published rates for this country are not so repre- -sentative nor so up-to-date as we would like them to be. ‘Most of .45. our information concerns one state, Connecticut, and in fact one city, New Haven. We have only four statistics which apply to the last twenty years. Secondly, the data seem to indicate a slight, though irregular, increase in the rates. This is true of New Haven, but is particularly noticeable in the national rate of 7.2 per cent in 1957. This is subject to sampling error, however, especially in a relatively small group. It should also be noted that 9 per cent of all marriages were eliminated before this rate was computed, based only on those marriages which were known to include mixed or unmixed pairs Of Protestants, Catholics, and Jews. Finally, the high rate discovered in Manhattan indicates that a Specific community may be quite divergent from the national average. Rates and Trends: Catholic Intermarriagpp, Not many data on Catholic intermarriage in European coun- tries have been available to the writer. However, one excellent study from the Netherlands by Bernard Van Leeuwen shows that 13 per cent of the nation's Catholics are in mixed marriages, varying from a mere 2 per cent in the most Catholic rural area to over 30 per cent in the urbanized coastal area. He notes the tendency for civil records to show more mixed marriages than church records do. .In the large cities of the Holland provinces the increase in inter- faith marriages has been quite striking, in Rotterdam rising from 8.88 20 per cent in the years 1879-93 to 48 per cent in 1936-3 These are the percentages of Catholics in mixed marriages. .If the 88"Le mariage mixte, facteur de deEhristianisation en Hollande," Lumen Vitae, 4 (1949), pp. 428-430. 46. 48 per cent in these terms were converted to the rate of Eiggg marriages involving Catholics, the figure would be in the neighbor- hood Of 70 per cent. Canada also has the advantage of state and church records on Catholic intermarriages. The Dominion Bureau of Statistics shows that the Catholic rate for the country has increased by 60 per cent from 1927 to 1957, i. e. from 7.2 per cent to 11.5 per cent.89 The trend is present even in the province of Quebec, though to a much less extent. A good study by Victor Traynor of one Canadian Catholic diocese also reflected this trend. The rate there rose from 27.0 per cent for the years 1930-33 to 32.5 per cent for 1950-53.90 As we turn to the United States, we have much more data to observe. These are indicated in Table 2-4. As was suggested earlier, the Official Catholic Directory published marriage totals and Catholic pOpulation totals for each diocese year by year. These will not be included in the table, unless, as is likely, it was the source for the first six rates in the table. The rates given by Thomas and by Dunn are from the same source, but they are included because they are compilations of these data. Again, the rates up to and including those of Derby, Connecticut, were assem- bled by Barron. What does this collection of intermarriage rates for Catho- lics tell us? For one thing, there is evidently considerable 89Heer, Op. cit., p. 246. 90"Urban and Rural Mixed Marriages," Social Order, 6 (1956): p. 157. 47. TABLE 2-4.-- Summary of published rates of Catholic intermarriage in the United States, per 100 marriages of Catholics. Place Date Rate Place_ Date Rate Milwaukeea 1929 26.0 United Statesb 1947 26.1 Portland, Ore.8 1929 47.5 *Pittsburghc 1923-48 37.0 Des Moinesa 1929 33.4 Michigan St. U.d 1948 37.6 Fargo,N._D.a 1929 31.5 New Haven8 1948 6.2 Lafayette, La.a 1929 6.4 Connecticutf 1949 . 40.2 0mahaa 1929 19.8 United Staresf ‘ 1930-50 30.0 *Midwest citya 1929 22.3 *Southern city8 1950 15.0 *Parish "B"3 1929 20.5 New Havenh 1950 27.4 *"St. Patrick's"3 1942 50.6 *Midwest cityi 1932-52 5.5 New Haven8 1870 4.7 IowaJ 1953 35.0 New Haven8 1900 14.2 *New York Cityk 1954 21.6 New Haven8 1930 18.0 Iowa1 1953-57 30.4 New Haven8 1940 16.3 United Statesm 1957 21.6 Derby, Conn.a 1929-30 15.6 Manhattann 1957 21.4 Derby, Conn.a 1940 8.8 Detroit° 1959 36.0 *"Lakeport"P 1936-60 18.9 *These studies are confined to one Roman Catholic parish. aBarron, Op. cit., p. 166. bDunn, Op. cit., p. 34. cThomas F. Coakley, "Mixed Marriages, Their Causes, Their Effects, Their Prevention," Lumen Vitae, 4 (1949), p. 455. d Landis, op. cit., p. 402. eHollingshead, Op. cit., p. 622. 48. fJohn L. Thomas, "The Factor of Religion in the Selection Of Marriage Mates," American Sociological Review, 16, (1951), p. 489. gFichter, Op. cit., p. 104. hKennedy, Op. cit., p. 59. 1Francis H. Cizon, "Interethnic and Interreligious Marriage Patterns in Parish X," American Catholic Sociological Review, 15 (1954), p. 255. JChancellor and Mbnahan, Op. cit., p. 237. kSchuyler, op. cit., p. 128. 1Lee G. Burchinal and Loren E. Chancellor, "Catholics, Urbanism, and Mixed- Catholic Marriage Rates," Social Problems, 9 (1962). p. 361. mU. 8. Bureau of the Census, Op. cit., p. 8. nHeiss, Op. cit., p. 49. °Besanceney, Op. cit., p. 11. PAndrew M. Greely, "Some Aspects of Interaction between.Re- ligious Groups in an Upper Middle Class Roman Catholic Parish," Social Compass, 9 (1961), p. 47. 'interest in the subject. Furthermore, the rates are almost uniform- ly higher than we found them to be for the Jews in this country. Is there an upward trend in Catholic rates? The fact of a trend upward in Canada would suggest the same for the United States. Thomas says that the data available from the records of "several dioceses" show "a steady increase in the rates from 1910 to the present."91 Kennedy's data on New Haven may indicate a trend, since the rate of 27.4 per cent for 1950 is quite a jump over the level maintained in the rates she found in 1940 and earlier (also extremely difficult to square with Hollingshead's 6.2 per cent for 91The American Catholic Family, 02. cit., p. 52. 49. 1948). However, the rates displayed in Table 2-4 do not otherwise add evidence to this generalization. In fact, the rate of 21.6 per cent from the national sample for 1957 is quite low relative to other published rates in recent decades. Even if 9 per cent of the marriages had not been eliminated from the calculation, it is hard to believe that this statistic would be altered much. It is un- fortunate that we have no other national samples to compare it with. Rates and Trends: Protestant Intermarrigggp There are relatively few studies showing the rates of Protestant maxed marriages. This is not due to lack of interest, since many Protestant clergymen have published warnings concerning the dangers in marriages to Catholics particularly. It is probably because there is not the Catholic insistence on a churchman Offi- ciating, and perhaps moreso because there are no uniform Protestant church records. However, quite a number of researchers have pub- lished rates for all three major religious groups. Tabel 2-5 shows what these rates are for Protestants, again repeating down to Barron's rates on Derby the data which he first summarized. It should be noted that the rates given by Bossard and Letts apply only to members of the United Lutheran Church of America. In these rates and in the studies of New Haven we see signs of a trend toward more intermarriage on the part of Protestants. Besanceney's data for Detroit also showed the rate of Protestant intermarriage for marriages contracted after 1944 to be significantly higher than the rate before 1944.92 However, it should be evident that these v W 92"Unbroken Protestant-Catholic Marriages Among Whites in the Detroit Area," op. cit., p. 14. 50. trends were discovered in population centers in which the Prot- estant prOportion of the pOpulation has been decreasing relative to Catholics and Jews, hence encouraging more Protestant intermar- riage. We do not know whether such a trend exists in other parts TABLE 2-5.--Summary of published rates of Protestant intermarriage in the United States, per 100 marriages of Protestants. Place Date Rate Place Date Rate New Havena 1870 0.9 New Havenc 1948 25.6 New Havena 1900 8.2 New Havend 1950 29.7 New Havana 1930 21.9 Eastern Statese 1936-40 46.0 New Haven8 1940 20.3 Eastern Statese 1941-45 47.0 Derby, Conn.a 1929-30 50.0 Eastern Statese 1946-50 58.0 Derby, Conn.a 1940 20.3 United Statesf 1957 8.6 Michigan St. U.b 1948 12.2 Manhattan8 1957 33.9 Detroith 1959 31.0 vV—w ~ “...—*— aBarron, Op. cit., p. 174. bLandis, Op. cit., p. 402. cHollingshead, Op. cit., p. 622. dKennedy, Op. cit., p. 59. eBossard and Letts, Op. cit., p. 308. fU. 3. Bureau of the Census, Op. cit., p. 8. gHeiss, Op. cit., p. 49. hBesanceney, op. cit., p. 11. of the country. The low rate of 6.6 per cent for the national sample seems to suggest that there is much less Protestant inter- marriage in the South, Midwest, and Northwest where Protestants 51. are generally a solid majority of the pOpulation. Information about PrOtestant intermarriage in other coun- tries is lacking almost entirely. Barron mentions a single rate of 33.1 per cent given by Fishberg for Hungary in 1903.93 The study by Heer in Canada revealed that Protestant intermarriage rates in that country had increased by 132 per cent from 1927 to 1957, i. e. from a rate of 5.0 per cent to 11.6 per cent.94 —————4 Beyond this, information specific to Protestant rates in other countries has not come to the attention of this writer. Comparison of Rates for the Three Major Groups The most obvious difference between the three groups is that Jewish rates are considerably lower than either the Prot- estant or the Catholic rates in this country, and this in spite of their relatively small numbers in the population. The fact may be explained partly by their geographical concentration and partly by the strength of their group norms. However, the high intermarriage rates noted earlier for Jews in Germany show us that their intermarriage patterns can change. Using the national sample of 1957, Paul Glick of the U. S. Bureau of the Census computed an interesting statistic. He first determined what the rate of intermarriage would be in this country for each religious group if mate selection occurred by chance. He then found out the ratio of each group's actual mixed marriage rate 93People Who Intermarry, Op. cit., p. 174. .94"The Trend of Interfaith Marriages in Canada: 1922-1957," op. cit., p. 246. - 52. to its "expected" rate. The results showed a ratio of .07 for Jews, .26 for Catholics, and .19 for Protestants.95' SO far as this sample is concerned (from.which 9 per cent of the marriages were eliminated as involving someone not of these three groups), it is evident that religious endogamy is still very much the norm in this country. However, Jews stick closest to it, and Catholics depart from it more than the other groups. This may appear strange in view of Barron's conclusion in 1946 that, "Of the majorreligious groups in EurOpe and America, control of intermarriage is most rigid among Orthodox Jews, followed in the order of rigidity by Roman Catholics, Reformed Jews, and Protestants."96 The "Triple Melting;§ot" Hypothesis Is there a trend toward more intermarriage among ethnic groups which are identified with one of the three major religious in this country, with the result that fusion of ethnic elements is going on within each of three religious "melting-pots"? Al- though the data that we have in this country is not conclusive, it seems to this writer that the hypothesis of a "triple melting- pot" needs to be re-examined. Pr0posed by Kennedy and endorsed by Hollingshead, it is contradicted by Kennedy's own increasing rates for Protestants and Catholics in New Haven, 1870-1950, even though Jewish endogamy pretty well held its own. The hypothesis states that ethnic lines are being crossed repeatedly in the United States 95"Intermarriage and Fertility Patterns among Persons in Major Religious Groups," Eugenics Quarterly, 7 (1960), p. 35. 96PeOple Who Intermarry, Op. cit., p. 59. 53. while people continue to marry within their own religious groups. Although it is true that, when ethnic homogeneity is sacrificed in marriage, religious homogeneity will often be preserved, the "triple melting-pot" hypothesis hides two other truths. First, there is more than a little intermarriage going on between these "melting-pots,' especially in the states of the Northeast. Secondly, we have just reviewed some evidence that suggests an increase of intermarriage among Protestants and Catholics at least. Perhaps the trend in Canada is not limited to that north- ern neighbor. There are two almost mechanical differences in the behavior of the data, which should be kept in mind when comparing ethnic and religious intermarriage rates in the "triple melting-pot" hypothesis. For one thing, the very fact that a religious group is larger than any of the ethnic groups which constitute it tends to make ethnic intermarriage rates larger than religious intermarriage rates. This is true because the smaller the group relative to the total pOpulation, the faster its rate goes up with each intermarriage; it can quickly reach 3.223l upper limit of 100 per cent intermarriage rate. A majority group, on the other hand, not only will find its intermarriage rate going up more slowly, but it can never reach a £321 limit of 100 per cent simply because there are not enough mates available outside its group. Therefore, to compare marriage rates without this in mind is misleading. Perhaps the correction for this distortion would be to use Glick's "ratio of actual to 'expected' intermarriage,‘ which seems to standardize the size of the groups in the general pOpulation. However, this assumes that we know both the 54. percentage of each group in the population and its intermarriage rate. .Finally, let us observe the difference there is in the way that ethnic and religious fusions take place in.marriage. .If an 'Italian immigrant marries a Norwegian immigrant in this country, their children are "mixed" rather than "pure" ethnics. This con- dition will not change for their children either. .In time we can .expect that there will be fewer and fewer "pure" ethnics in the marriageable pOpulation of this country. aBut, if this same couple were Catholic and Protestant, respectively, their children would not, therefore, belong to a "mixed" religion. .Each would identify with the religion of one parent or the other or with a third religion -Or with none. There is no necessity in this case, as there is in ethnic intermarriage, which dictates that the membership of any religious group in this cOuntry should diminish in size in the fu- rture. .Doctrinal purity of the members, moral commitment, and a spirit of faith may diminish, but there is no reason why at least nominal membership levels should not be maintained indefinitely. In this light, then, the base for determining a particular ethnic group's rate of intermarriage will grow smaller and smaller (there- fore, its rates larger and larger); but the base for determining a particular religious group's rate Of intermarriage may remain 'about the same, even-though the rate of.intermsrriage year-by-year is increasing. It is conceivable to have 100 per cent intermarriage rates for each religious group and still have no loss of net member- ship by any one of them. If, however, we had 100 per cent intermar- riage rates for all ethnic groups for just one year, there would be 55. no more."pure" ethnic groups left to study. All the contents of the "melting-pot" would be at least partly fused. In the light of these considerations, what sense does it make to talk about " since the the three religious groups as "triple melting-pots, ingredients of each may have been poured back and forth many times? Religious Intermarriage Related to the Sex Which Marries Out Barron observed in his summary of studies in 1946 that "everywhere Jewish men were found to have intermarried more than Jewish women."97 In Stamford, Connecticut, the ratio of male to female was four to three; in New York City, two to one; in New Haven, almost three to one; the same in Wisconsin.98 This is one of the most consistent patterns to be found in all mate-selection studies, yet an explanation has hardly been attempted. The extent to which social stratification plays its part we will explore in a later section. It has also been suggested that Jewish men have greater freedom and mobility. The writer is dubious about this explanation, but he has none better to offer. In the case of Protestants we have two limited studies 1 S i? h conflicting results. In the small town of Derby, Connecticut, 99 '1'! C) r testant women intermarried at lower rates than Protestant men. 97Peoole Who Intermarry, Op. cit., p. 194. o ‘8Milton L. Barron, "The Incidence of Jewish Intermarriage in EurOpe and America," American Sociological Review, 11 (1946), p. 12. Ray E. Baber, op. cit., p. 707. Kennedy, Op. cit., p. 333. Prince, 00. cit., p. 4. 993arron, Peeple Who Intermarry, op. cit., p. 174' 56. On the other hand, Lutheran pastors reported that, of those in their congregations who were in mixed marriages, 64 per cent were women.1001 As for Catholics, Barron had concluded in 1946 that in "most studies" Catholic women intermarried at higher rates than did Catholic men.101 The evidence continues to support this generali- zation. Thomas reports that the ratio was six Catholic women to four Catholic men in mixed marriages in the Bishops' national study, in Whelpton and Kiser's Indianapolis survey, and in Schnepp's study of "St. Patrick's" parish.102 The percentage has gone even higher, to 68 per cent in a "Lakeport" parish, and to 73 per cent in "South- ern Parish."103 However, in Detroit and in the Netherlands, the ratio was reported as only five to four.104 There is one bit of negative evidence: Traynor discovered in a Canadian diocese that, although the Urban ratio was six Catholic women to four Catholic men in mixed marriages, in rural parishes the ratio was reversed at four women to five men.105 We cannot be sure enough of a rural-urban difference to speculate much about it. It is worth noting, however, that most of the studies which show the highest percentage of Catholic women intermarrying depend for their information on parish records. As has been said earlier, 101Barron, People Who Intermarry, Op, Cit}. Po 171- 102The American Catholic Family, Op, cit., p. 154. 1'03Greele)’. 02. Cit-s P- 47° Fichter, 22&—£££&9 P’ 107' 104Besanceney. op. cit;. Po 12° van Leeuwen, 2££—£§£49 p. 431' 1‘05"Urban and Rural Mixed Marriages," Op. cit,, p. 155. 57. these records tell us little about those Catholics who have left the Church to marry. (The same comment could have been made about the Lutheran study above.) We may perhaps assume that most of those who abandon their church at marriage are men, considering the lower religiosity of men in general. Still, since the same relationship is discovered to a lesser degree even in civil re- cords and in a probability sample, it seems that some "explanation" is called for. It has been suggested that Catholic women have less mobility and less initiative in the matter of mate selection.106 However, it would hardly be "cricket” for the present writer to use the same argument to explain female ippgpmarriage in the case of Catholics as was used to eXplain female lpmarriage in the case of Jews. Of course, at this point anything must be considered to be an ad hag explanation. However, this line of thought at least has some merit in logic: Before a mixed marriage at which a Catholic priest officiates, the Catholic and non-Catholic party must each have agreed that the children of the union will be raised as Catho- lics. Since the early training of children is necessarily entrust- ed to the mother and she typically provides the religious influence in the family, such an agreement gives some security to a Catholic woman in a mixed marriage. At the same time, a Catholic husband must rely more on his wife in this matter. If the bride-to-be is not a Catholic, he may become aware of the risk and she may be averse to assuming this alien role, and they may end by calling it 106Van Leeuwen, on._cit., p. 431. 58. Off. Or, if his religious convictions are weak, he may abandon his religion and, therefore, will not appear in parish records Of marriages. However, this is mere speculation until a way of test- ing the "explanation" can be found. Religious Intermarriage Related to Nationality Surprisingly little information has been published on this particular subject. Perhaps it is because quite a large sample is needed to obtain religious intermarriage rates within national groups. Catholic intermarriage rates, as revealed by the parish records of thousands of marriages in midwestern and eastern states, took this distribution in ethnic groups: Polish, 2.6 per cent; Italian, 4.6 per cent; Irish, 9.1 per cent; German, 32.4 per cent.107 In the study of Manhattan, Catholic intermarriage rates were highest among fourth generation Americans, German-Americans, and "Russian" (Ukranian?)-Americans.108 Father Thomas studied the marriage records of fifty-nine parishes in one midwestern urban.area. In parishes which were established on territorial lines rather than on a national basis, the rate of mixed marriages was 15.2 per cent. It was 15.8 per cent for those parishes which he identified simply as "ole ethnic," whereas for the "new ethnic" parishes it was usually less than 4 per cent.109 109"Our-Group Marriage Patterns of Some Selected Ethnic Groups," American Catholic Sociological Review. 15 (1954), p. 17. 59. Two "Old" Factors Related to Religious Intermarriagg The effect that size of group has on intermarriage rates and the importance of residential prOpinquity (or neighborhood segregation) in patterns of marriage choice have already been dis- cussed in other sections¢3f this chapter. At this point it should be sufficient merely to relate these matters specifically to re- ligious intermarriage. John L. Thomas seems to have been the first to specify the size of group as one of the factors influencing Catholic intermar- riage rates. He showed by many illustrations from the Official Catholic Directory that the highest rates occur in dioceses where Catholics have the smallest representation in the pOpulation.110 Recently, Burchinal and Chancellor, using civil records in Iowa, found that mixed marriage rates varied inversely with three levels of Catholic percentages in county populations.111 Using the esti- mates of the Official Catholic Directory, one group of researchers computed a negative rank-order correlations of .76 in 1945 and of .86 in 1955 between Catholic mixed marriages and their percentages in the pOpulations of the forty-eight States.112 This generalization has been tested most extensively in Canada, where religious affiliations are recorded in the national 110"The Factor of Religion in the Selection of Marriage Mates," Op. cit., p. 489. Also The American Catholic Family, op. cit., p. 155. . 111"Catholics, Urbanism, and Mixed-Catholic Marriage Rates," 0p. cit., p. 363. 112Harvey J. Locke, George Sabagh, and Mary M. Thomas, "Inter-faith Marriages," Social Problems, 4 (1957), p. 331. 60. census and in marriage license applications. Locke et al. analyzed official Canadian statistics and found a perfect negative rank— order correlation in 1951 between rates of Catholic intermarriage and their percentages of the local p0pu1ation in the ten Provinces of Canada. The relationship for Anglicans in Canada was "much less striking."113 Traynor's study of a Canadian Catholic diocese con- firmed this relationship (r = - .93) for the EEEEL parishes of the diocese. However, in the urban parishes he found just the opposite, a positive rank-order correlation of .48 between the mixed marriage rates and the percentages of Catholics in the area.114 This remains unexplained. The present writer has tried to make his position clear in earlier pages regarding "size of group" as a factor in intermarriage rates. It seems to be not so much a sociological explanation as a mathematical one. (This is so, even though small groups may be realistically threatened with extinction because of it). This illusory factor apparently can be controlled for comparing different groups by using Glick's method of computing the ratio of the actual to the "expected" (by random selection) marriage rate. Research in this field would, it seems, be much more scientifically meaningful if we were in the habit of comparing these ratios instead of inter- marriage rages. Closely related to this matter is the factor of residential propinquity in religious intermarriage. In fact, when we observe 1131bid., p. 331. I 1']‘-*"Urban and Rural Mixed Marriages," 02. cit., p. 155. 61. residential segregation, we are observing increased density of a particular characteristic of the.p0pulation in a given area. Such segregation prepares the scene for increased interaction among homogeneous types at lower cost. The same applies to the segrega- tion of a school p0pu1ation, or occupational group, or recreation; al center. Mate selection (including religious intermarriage) is a function of frequent social interaction, which (because of time- cost factors) is a function of propinquity. Whether this series of relationships encourages religiously homogeneous marriages or not is a function of the extent to which the members of each re- ligious group live, work, recreate, are educated in close proximity to each other, i;_g;Jare segregated religiously. All of this has already been stated or implied in the earlier discussion of the studies on residential propinquity as a factor in homogeneous mar- riages. The Economic Factor in Religious Intermarriagg As all residents of American cities probably recognize, there is not only a certain concentration of religious groups in neighborhoods, but more noticeably a patterning of economic levels by neighborhoods. We should not be surprised to find some relation- ship between these two types of residentia1\concentration in the matter of mate-selection. For instance, Thomas found that mixed marriage rates for Catholics go up with the increased value of the rental areas.115 It was found also in.Manhattan that Catholic intermarriage is positively related to the high socio-economic 115The American Catholic Family, Op. cit., p. 58. 62. status of the fathers.116 It seems that this relationship for Catholics should be regarded as largely a manifestation of the prOpinquity factor. We would not expect this relationship to exist if Catholics were not concentrated near the middle and bottom of the socio-economic scale in the pOpulation. This means that the more prosperous Cath- olics are likely to find themselves to be decidedly a religious minority in their residential area--if we assume for the moment that economic level exerts a stronger pull in the matter of residen- tial segregation than does religion. However, Barron believed that the economic factor helped Jews in Derby, Connecticut, to overcome the influence of propin- quity and find Jewish mates in other towns, their economic advan- tage providing a facility toward realizing an endogamous goal.117 In contrast to this, Carleton S. Coon found there was a positive correlation between economic prosperity and the rate of Jewish intermarriage.118 Here wealth seems to represent not a facility toward, but a breakdown in social control in the matter of endogamy. Something similar may be said of the finding that Catholic intermar- riage is also associated with being members of the lowest occupation- 116Heiss, Op. cit., p. 51. 117”The Incidence of Jewish Intermarriage in Europe and America," op. cit., p. 13. 118"Have the Jews a Racial Identity?" Jews in the Gentile Eprld, eds. Isacque Graeber and Steuart H. Britt, New York: Macmillan, 1942. Chap. 1. 63. a1 groups.119 This seems to be particularly true of "invalid" marriages, i:_g;, marriages of Catholics outside the Church.120 One can regard this as the lack of an economic facility to keep an endogamous rule, or one may be more inclined to say that social control is weak at the lowest economic levels. However, to pursue this line of argument would take us into the next chapter before we are ready for that subject. The rule of endogamy and its infraction by intermarriage has occupied our attention during many sections of this chapter. We have seen how this aspect differs from the simple problem of homogamy X§° heterogamy in mate selection. We have discussed racial, ethnic, and religious intermarriage as the principal types. Both by design and by necessity, due to the large body or research al- ready completed on this subject, we have concentrated on religious intermarriage. Two subjects must still be considered before we close this chapter on mate selection: the influence of social stratification and something about the psychological process of mate selection. The first of these is a problem more complex and more elusive than we have been talking about up to this point. Class and Caste Hyperggmy: Terms Defined In this part of our chapter on mate selection there is necessarily much overlap with other parts of the chapter. Occupa- 119Thomas P. Monahan and Loren E. Chancellor, "Statistical Aspects of Marriage and Divorce by Religious Denomination in Iowa," Eugenics Quarterly, 2 (1955), p. 164. 120Traynor, op. cit., p. 157. 64. tions, education, race, ethnic group, perhaps even religion--all again become relevant variables in this section. However, the focus is quite different. Whereas before we were talking about like marrying like, about endogamous rules and intermarriage, now we shall be talking about rankings within society and about the marriage patterns of peOple who occupy various ranks. This, in general, is the problem of social stratification as it relates to mate selection. Kingsley Davis, in the most significant article the present writer has found on this subject, states that stratification is "the master-basis of matrimonial choice."121 Whether this is com- pletely true or not, the subject certainly cannot be overlooked in a treatment of mate selection. Stratification is, of course, a major problem-area in sociological theory and research. To go into matters like the number of social classes and how determined, discrete pp, continuous classes, status crystallization, vertical mobility in general, class consciousness--all these would take us far afield from our purpose in this chapter. We shall be content with defining the important terms and reviewing the literature which is most pertinent. What is stratification? Bernard Barber describes it briefly .: "the structure of differential rankings that seems to occur in "122 all societies. K. Davis expresses it in a concrete distinction: Those positions that may be combined in the same legitimate family-~viz., positions based on sex, age, and kinship-~do not f, 1“1"Intermarriage in Caste Societies," Op. cit., P- 376- 2 Social Stratification, New York: Harcourt, Brace, 1957, 65. form part of the system of stratification....Those positions that are socially prohibited from being combined in the same legal family--viz., different caste or f§§ss positions-- constitute what we call stratification. The two principal modes of stratification which are commonly distin- gished are caste and class. The former is simply more rigid than the latter. Caste is inherited from one's parents at birth, and this status is fixed for life. One's class is also inherited at birth, but with adulthood one may achieve a higher (or lower) status. "A cardinal principle of every stratified social order is that the majority of those marrying shall marry equals."124 The functional basis for this principle is this: Since marriage is an institutional mechanism for procreating and rearing children, the requirements of status ascription in a caste order practically require the marriage of equals. A wife reared in a social stratum widely different from her husband's is apt to inculcate ideas and behavior incompatible with the position the children will inherit from their father3 thus creating a hiatus between their status and their role. Coupled with the rule of class endogamy is the practice of hypergamy, i. e., of institutionalized intermarriage whereby a woman may marry a man who has a higher status than her own family's. This may occur even between subcastes in a caste system, e. 3., in India.126 If and when it becomes the rule rather than the exception, it will destroy that part of the system of stratification. However, Davis points out that there is no socially-approved 123Human Society, New York: Macmillan, 1949, p. 364. 124Davis, "Intermarriage in Caste Societies," Op. cit., p. 376. lZSIbid., p. 378. q 1‘6Ibid., p. 381. 66. hypergamy in racial caste systems. This is so because the hybrid child presents an especially acute problem. With his racial features visibly showing his lower caste descent, he could not in- herit the higher status of his father.127 The Opposite practice, hypogamy, is comparatively rare, whereby a woman marries a man who has a lower status than her family's. An example would be the Japanese practice called yoshii, by which a man with no sons would adOpt a lower class son-in-law for one of his daughters. The son-in-law takes the name and title of the father-in-law, inherits his prOperty.128 In American society we call this "marrying the boss' daughter." Having in mind such practices as hypogamy, and especially hypergamy, K.Davis comments: "There are forces that Oppose rank endogamy, but the principle that stratification in itself necessi- tates such endogamy remains firm."129 Class and Caste Hypergamy: Review of Evidence The present writer has not attempted to review all the studies of social class in this country in which marriage patterns may have been incidentally discussed. As samples, however, we may take note of the following. Hollingshead found class endogamy to be the rule in the community he studied: "If we include immediately 127Ibid., p. 387. 12'8Pierre L. Van den Berghe, "Hypergamy, Hypergenation, and Miscegenation," Human Relations, 13 (1960), p. 86. 129"Intermarriage in Caste Societies," op. cit., p. 380. 130"Class and Kinship in a Middle Western Community," American SOCiOlOglgal Review, 14 (1949), p. 475. 67. adjacent classes, then almost all cases are included."130 A recent study in Columbus, in which census tracts were ranked economically and social class was determined by the location of addresses in these tracts, arrived at essentially the same conclusion regarding current marriage patterns.131 Using education as a criterion of class, we are not sur- prised that Landis and Day found more college women than men marry- ing someone of a higher or equal educational status.132 This form Of hypergamy seems to be functional in as much as it corresponds to the male and female roles within marriage and it recognizes the fact that the whole family shares the status of the husband and father. However, at least at the higher levels, this fact is prob- ably modified by later behavior of the couple, as has been observed: ”From our study it appears that the ultimate social position of a couple, originally of different social positions, is dependent upon the behavior adjustments which they make."133 When social class is defined simply by occupation, at least two problems arise. To compare the occupations of husband and wife before marriage, as Thomas Hunt seems to have done, seems to be not very realistic.134 When sex differences are not controlled, the 131Simon Dinitz, Franklin Banks, and Benjamin Pasamanick, ”Mate Selection and Social Class: Changes during the Past Century,‘ Marriage and Family Living, 22 (1960), p. 351. q 1'ZPaul H. Landis and Katherine H. Day, "Education as a Factor in Mate Selection," American Sociological Review, 10 (1945), p. 559. q l"3Allison Davis, Burleigh and Mary Gardner, DeeE.S°uth: Chicago: University of Chicago Press, 1948, p. 199. 134"Occupational Status and Marriage Selection," American Sociological Review, 5 (1940), p. 504. 68. practice of ranking white collar higher than blue collar occupations is not very meaningful. On the other hand, to compare husband's occupation with the occupation of wife's father, is perhaps the best that we can do when we are trying to discover whether occupa- tional hypergamy has taken place.135 However, what we discover in this comparison may be simply a function of the "up-grading" of occupations in this country over the past generation. That is, skilled and unskilled occupations have decreased in number rela- tive to clerical (white collar) and semi-skilled occupations. Given these cautions, the general finding has been that class endogamy is most common, with hypergamy (generally limited to ad- jacent levels) being an alternative.136 Racial intermarriage has already been discussed in an earlier part of this paper, but let us look at it now as a caste phenomenon. Race relations in this country still have the character of caste re- lations. Most States outlaw intermarriage, and the society as a whole forbids it by custam, When it does occur, interracial marriage in this country is more likely to involve a white woman and a Negro man.137 By the criterion of caste this does not fit the pattern of hypergamy. However, it does do so by the criterion of class, for the white woman is generally of the lower class and the Negro hus- band generally has a relatatively high occupation. The woman's 135Paul POpence, "Assortative Mating for Occupational Level," Journal of Social Psychology, 8 (1937), pp. 270-274. 136Richard Centers, "Marital Selection and Occupational Strata," American Journal of Sociology, 54 (1948), pp. 530-535. 137 p. 389. K. Davis, "Intermarriage in Caste Societies," op. cit. 69. caste advantage is traded for the man's class advantage. This is not difficult to understand, since in marriage the man's occupa- tional and economic status belongs to the whole family, whereas the children's caste markings are derived from both parents. A white man marrying a Negro woman would not find this advantage for his children (or this prestige symbol for himself). In fact, we can safely predict that, if the racial caste system ever ceases to exist in our country, we will find the more common pattern to be the marriage of Negro (mulatto) women to white men, as is the case in societies without a racial caste. In the earlier section on racial intermarriage, we noted that there are many societies where this type of marriage is not taboo, where race is not a caste system. For instance, we are told that in Bahia, Brazil, "marriages cross race lines more often "138 than class lines. Expressive of Brazil as a class society, rather than a caste society, is their proverb: "A rich Negro is "139 We should remember a white man, and a poor white is a Negro. also that in these societies "race" is defined by physical appear- ance rather than by genealogy, as in the United States. Dark skin or other Negroid features may then become elements of physical attractiveness, which may make a woman eligible for class hypergamy. Hence, in Puerto Rico, "the available evidence indicated that three white men marry mulatto women for every white woman who marries a 138Pierson, Op. cit., p. 147. 139Bastide, Op. cit., p. 496. 70. mulatto man."140 This is contrary to the pattern in this country. So, in summary, class and caste endogamy are the rule. But, because of the levels involved in stratification, we also find an alternative pattern of hypergamy. Actually, this reflects the advantage of the upper class male, who may select a marriage partner at his own level or reach to a lower level. Rarely does a society approve the Opposite pattern; and, when it does, it is usually to give a status-heir to the bride's father. The Process and Psychologigal Antecedents of Mate Selectigpr Choosing a marriage partner is not, of course, an instant- eous event. The choice is usually preceded by a more or less lengthy deliberation. We call this process dating, courtship, and engagement, as it becomes increasingly serious. Quite a bit has been written about such matters which is not of concern to us here. We shall briefly discuss only what relates to the choice of a mate, not the other aspects of the process. Romantic love, as a strong mutual attraction between male and female, is often regarded as a foil to the forces favoring endogamy. Our earlier discussion of complementarity of needs suggested this possibility. Hugo Biegel approvingly presents an historical review of the influence of romantic love, concluding that "it has mitigated the impact that a too-fast moving and un- organized conversion to new socio-economic constellations has had 140M; W. Gordon, "Race Patterns and Prejudice in Puerto Rico,‘ American Sociological Review, 14 (1949), p. 299. 71. upon our whole culture...."141 Charles Hobart adapted for college couples the Gross Romanticism Scale and reported that he, like Komorosky and Winch, found females to be less independent in their mate selection, not predictable by romantic hypotheses.142 Religious endogamy has received the most attention in studies on dating as a process in mate selection. Burgess and Wallin, in their sample of engaged couples, found strong homogamy (C = .75) of religious affiliation.143 One study of college "rating and dating" behavior found that 34 per cent of the women and 21 per cent of the men "tried to avoid dating someone of a different religion."144 Students at the University of Minnesota reported on the religious affiliations of their first three "affairs". The religious endogamy of those reporting was, in descending order: Jewish females (84.9 per cent), Protestant females (80.8 per cent), Protestant males (79.5 per cent), Jewish males (72.4 per cent), Catholic males (58.5 per cent), and Catholic females (38.0 per 145 cent.) This was a pOpulation comprising almost three Protest- ants to one non-Protestant. 141"Romantic Love," American Sociological Review, 16 (1951), P. 2330 142"The Incidence of Romanticism during Courtship," Social Forces, 36 (1958), pp. 362-367. 143"Homogamy in Social Characteristics.' 02- qLE-. Po 115- 144WilliamM. Smith, Jr., "Rating and Dating: A Re-Study," Marriage and Family Living, 14 (1952), p. 314. 145Clifford Kirkpatrick and Theodore Caplow, "Courtship in a Group of Minnesota Students," American Journal of Sociology, 51 (1945). p..1l6. 72. In general, the small group of studies which we have in- dicate that the same kinds of norms Operate in courtship as we found to be important in mate selection as an end-product in marriage. What about attitudes toward intermarriage, apart form actual dating behavior? In one college sample in Chicago, 49.7 per cent of the men, 65.5 per cent of the women said that they would not consider a mate of a different race. Compare this with those in the same sample who rejected a mate of a different re- ligion (men, 41.6 per cent; women, 42.5 per cent); of different educational status (men 33.5 per cent; women 40.5 per cent); of different social background (men, 27.8 per cent; women, 34.0 per cent).146 In other studies, the norm of religious endogamy re- ceived notable support, although Protestants are less concerned about it than Jews or Catholics.147 At the University of Idaho, Alfred Prince found that 97 per cent Of the students would be willing to date persons of another religious faith, 75 per cent 146Anselm Strauss, "The Ideal and the Chosen.Mate," Ameri- can Journal of Sociology, 52 (1946), p. 205. 147Clifford Kirkpatrick, "Student Attitudes towards Mar- riage and Sex," Journal of Educational Sociology, 9 (1936), pp. 545-555. Reuben Hill, "Campus Values in Mate Selection," Journal of Home Economics, 37 (1945), pp. 554-558. Harold T. Christensen, "Student Views on Mate Selection," Marriage and Family Livigg, 9 (1947), pp. 85-88. Thomas C. McCormick and Boyd E. Macrory, "Group Values in Mate Selection in a Sample of College Girls." Social Forces, 22 (1944), pp. 315-317. Victor A. Christopherson and James Walters, "Response of Protestants, Catholics, and Jews concerning Marriage and Family Life," Sociology and Social Research, 43 (1958), pp. 16-22. 73. to "go steady", 50 per cent to marry such a person.148 Some information is available about the correlates of attitudes toward interfaith marriages. Three studies found women to be more opposed to these than men.149 One found a slight tendency in the Opposite direction.150 In two samples, those who attended church less frequently were more favorable to intermar- riage.151 Prince found Catholics to be more willing to intermarry than Protestants, but they also were twice as likely to specify the condition that the children be raised in their religion.152 Willingness to intermarry was also related to the factor of being relatively young.153 There have been a few indications of changes in mate- selection standards over time. A survey of marriage offers in Yiddish newspapers showed the influence of the economic climate: Economic factors predominated in these offers in 1935, but person- 148"Attitudes of College Students toward Inter-Faith-Mar- riages," Marriage and Family Living, 19 (1957), p. 120, abstract from The Coordinator, 5 (1956), pp. 11-23. 149Ibid.; also Christensen, Op. cit.,; Leo G. Burchinal, "Membership Groups and Attitudes toward Cross-Religious Dating and Marriage," Marriage and Family Livigg, 22 (1960), pp. 248-253. 150Prince, "Factors in Mate Selection," pp;_pi£, 15l‘Prince, "Attitudes of College Students toward Inter- Faith Marriages," op. cit.,; Lee Burchinal, Op. cit. 152Ibid., p. 120. 153Burchinal, Op. cit., p. 253. 74. ality traits did so in 1950.154 Robert McGinnis repeated in~l956 a study which had been conducted in 1939 on the same campus. The differences were: "Similar religious background and interests" went up three rank positions. "No previous experience in sexual intercourse (chastity)" went down three rank positions. The rank of health and sociability also went down.155 It is difficult to find much in these attitude studies which we have not already seen in studies of marriage choices. We do note that personality traits are more prominent than social categories in the conscious motivation while selecting a mate. These were not stated, however, in terms which are readily trans- lated into "complementarity of needs." It seems to this writer that the qualities sought in a marriage partner are those which are most in doubt in the given socio-economic setting. For in- stance, the younger adolescents place emotional development on top of their requirements. White college students look for good; character, intelligence, insight and understanding as the discrimin- ating marks. Negroes rated health and economic security higher than whites did. Being predominantly Baptists and Methodists, they were less concerned about religious intermarriage than whites generally were. In other words, when certain characteristics can be taken for granted in the circles in which one moves, these are inot likely to appear on a list of traits most desirable in a marriage 154Emil Bend, "Marriage Offers in a Yiddish Newspaper--l935 and 1950," American Journal.of Sociolggy, 58 (1952), pp. 60-66. 155"Campus Values in Mate Selection: A Repeat Study," Social Forcesy 36 (1958), p. 369. 75. partner. This analysis is simply a recognition of the necessity of controlling the important variables in the situation before making comparisons. Conclusion This has been a long review of some of the literature on mate selecion as studied and analyzed by social scientists, with special reference to religious intermarriage. It should be clear to the patient reader who has persevered this far that no one author, no single orientation can provide "all the answers." Many have proposed or tested hypotheses, but only a few have con- structed even limited theories. However, when we put all of these together we have reason to say that we are at least approaching a scientific understanding of one complex social phenomenon. The approach taken in this study hinges on the concept of social control, which is the concern of the next chapter. CHAPTER III SOCIAL CONTROL AND ANOMIA The plan of this chapter is as follows: So that the read- er may be clear now about the scope and direction of this study, 3 we shall begin by taking notice of some studies, especially that of Heiss, which have used social control as an approach toward- understanding Protestant-Catholic marriages. This will give us an opportunity to state without further delay the hypotheses of this study. Then we shall turn to a general discussion of the princi- pal concepts used. After this, it will be necessary to see what ‘ evidence there is for a norm of religious endogamy in the United States and to consider what other norms may be in conflict with it. Finally, we shall review the previous research findings whiCh seem. rrelevant, at least indirectly, to the hypotheses of this study. Stgdies Of Social Control and Interfaith Marriages ' In the chapter just concluded, we.defined endogamy as h social norm and intermarriage as a deviation from that nonm.~ Rates ' of any type of intermarriage represent degrees of deviation from that particular endogamous norm, or, when they reach a certain undefined point, they may represent the group's abandonment of that norm. For instance, Chancellor and Mbnahan felt justified in saying, "The role of religious affiliation, in the final analysis, may not be any more-important than the other elements in marital 76. 77. selections."1 So far we have been accepting low intermarriage rates as evidence of a social norm. Later in this chapter other indications of a norm of religious endogamy will be reviewed. In passing, it may also be worth recalling that Katz and Hill, in their "norm-interaction" theory which grew out of their review of studies of residential propinquity, gave as the first assumption of their theory that "marriage is normative." The research which has best organized data on religious intermarriage from the point of view of social control is found in an article recently published under the unpretentious title, "Pre- marital CharacteriStics of the Religiously Intermarried in an Urban Area." We have referred to it earlier as the Manhattan study.2 Actually, it is a notable middle-range theory of intermarriage. To repeat.this part from our introductory chapter, the theory as- ‘sumes that: the ultimate sources of barriers to interfaith marriage in American Society lie in the family of orientation and formal religious organizations....lf, in the case of an intermarried respondent, it can be shown that these sources did not produce effective barriers to intermar- riage, we will consider the marriage to be explained.3 . Heiss sets down six general hypotheses to test this theory, each being particularized by several specific hypotheses. He concludes from his data that the intermarried have: 1Loren E. Chancellor and Thomas P. Mbnahan, "Religious Preference and Interreligious Mixtures in Marriages and Divorces in Iowa,” American Journal of Sociology, 61 (1955), p. 259. 2Jerold S. Heiss, Amerigan Sociological Review, 25 (1960), pp. 47-55. 31b1d., p. 48. 78. (a) non-religious parents, (b) greater dissatisfaction with parents when young, (c) greater early family strife, (d) less early family integration, and (e) greater eman- cipation from parents at time of marriage. Taking our lead from this study, but including findings from other studies of interfaith marriages, let us see what can now be said about the following as agents of social control for the norm of endogamy:‘ parents, family, ethnic group. It should also be helpful to consider briefly the social conditions of secu- larism and anomie as they relate to interfaith marriages, before we state the hypotheses for the present study. First, what are the characteristics of the parents of the religiously intermarried? In Heiss' Catholic sample it was re- ported that "religion was not at all important to both parents."5 Catholic respondents also reported dissatisfactiOn with their early relationship with their parents, that they argued with par- ents over religion, that they often disagreed with their parents, that their parents quarreled often. The Protestant-intermarried reported that their parents never attended church. In the Jewish sample, two statements were made concerning parents which are hard to square with each other: The intermarried were significantly different from the intramarried in that fewer of them reported living with both their real parents up until age sixteen, and yet a higher proportion of them said that their parents were not di- vorced, separated, etc. If their parents were living together, 79. yet the respondents were not living with them both, we must assume that the children lived away from home before the age of sixteen in significant numbers. Mysterious? However, for the sample as a whole Heiss found that the parents of the intermarried were more likely themselves to be intermarried. This agrees with the earlier finding of Gerald J. Schnepp and of the present writer.6 Emancipa- tion from parents by being at least thirty years Old at the time of marriage was a characteristic of the intermarried among Catho- lics and Jews, but among Protestants there was a tendency in the opposite direction.7 The failure of the family to exercise social control in the rule of religious endogamy was also indicated in Heiss’study by these results: For the sample as a whole, intermarriage was positively related to having a low Family Integration score, and negatively related to having grandparent's living with them. Sur- prisingly, intermarriage was positively related to being the young- est or only child and negatively to being the oldest child. So far as the present writer is aware, too little is known to make a case for differential social control related to a child's birth- order in the family. Heiss found that the Catholic intermarried also had a high Intra-Familial Conflict score, whereas fewer of the intermarried in the Jewish group reported seeing their rela- 6"Three Mixed Marriage Questions Answered," Cgtholic World, 156 (1942), p. 204. Also: Paul H. Besanceney, "Unbroken Protest- ant-Catholic Marriages among Whites in the Detroit Area," American Catholic Sociolggical Review, 23 (1962), p. 15. 7Heiss, op. cit., p. 50. 80. tives fairly often. In an earlier day a person's "people" were thought of as his larger family. Especially when a group is surrounded by a con- trasting majority group does their own shared sub-culture hold them; together, their national originaor language or customs. What sob cial control do ethnic groups exercise in the matter of religious intermarriage? In the Jewish case, religious and ethnic lines are practically coterminous. However, in the area around Derby, Con- necticut, the same could be said about the equally small groups of Greek Orthodox and Greek Catholics (distinguished from Roman Catho- lics not by church but by rite). Yet, the intermarriage rate of the Jews was much lower than that of the other two groups. Barron was probably correct in identifying one important differential factor as being the Jews' long history of migration and status as a minority group, which conditioned them to develop attitudes and techniques resistant to intermarriage.8 In the Catholic case, Heiss found that in Manhattan intermarriage was negatively related to being first generation Americans. Some ethnic groups have re- mained cohesive much beyond the first generation in America. Thomas attributes to this factor the low rate of Catholic mixed marriages in some large cities.9 Heer feels that the trend toward increased religious intermarriage in Canada may be due to a lessen- ing of in-group feeling among the French Catholic and British 8Milton L. Barron, "The Incidence of Jewish Intermarriage in Europe and America," American Sociological Review, 11 (1946), p. 13. 9John L. Thomas, The American Catholic Family, Englewood Cliffs, N. J.: Prentice-Hall, 1956, p. 50. 81. Protestant elements in the population.10 He also asks whether, in the larger society itself there has not been a change of atti- tude toward intermarriage. It is certainly difficult to assess the impact of ethnic cOhesiveness on religious intermarriage; but, to the extent that this exists, it belongs in the category of so- cial control. We now turn from the agents of social control--parents, family, ethnic group--to the conditions in which it is ineffective to restrain religious intermarriage, yig. secularism and anomie. The first refers to the abandoning of distinctively religious values in some degree. It puts a this-world goal in place of an other-world goal. "Rationality'supplants faith as a guide to ac- tion. ’Worldly "success" is not to be hindered by religious or moral considerations. The radical secularism is religious indif- ference, which we saw already manifested in the experience of Manhattan's religiously intermarried by the Protestant parents who never attended church and the Catholic parents for whom religion 11 Secularist values may be seen in the was not at all important. positive correlation, noted earlier, between economic prosperity and religious intermarriage for Jews. (Above, p. 62 ). In the data whictheiss presents, secularism is probably also displayed in both the choice and the experience of those Catholics in mixed marriages who had not attended parochial schools and for whom their 10David M. Heer, "The Trend of Interfaith Marriages in Canada: 1922-1957," American Sociological Review, 27 (1962), p. 250. 11Heiss, Op. cit., p. 50. 82. interfaith marriage is not their first marriage. The qualifica- tion "probably" is made because inaccessibility may explain the first, and widowhood the second type of behavior. Finally, let us consider briefly the influence of anomia on the rates of religious intermarriage. We are talking about the normlessness an individual may experience when the agents which exercised social control over his behavior are no longer present and his own internalization of these norms has been quite super- ficial. Examples might be found among those who have left the farm to go to the city, or have left home to go to college. We are talking about the normlessness that is likely to be felt in stressful circumstances, such as wartime or oppressive poverty or unemployment or unusual vertical mobility. In the literature on religious intermarriage we find some indications of such a rela- tionship, all of them based on Catholic data. One student in Holland cited urbanization as one of three major factors involved in increased religious intermarriages.12 In Iowa there was a slight positive association of mixed marriages with urbanism, when controlling for the percentage of Catholics in the population.13 Data from Iowa also showed a negative association between inter» faith marriage and having the occupation of farmer.14 In Manhattan 12Bernard Van Leeuwen, "Le mariagg mixte,,facteur de dechristianisation en Hollande," Lumen Vitae, 4 (1949), p. 430. 13Lee G. Burchinal and Loren E. Chancellor, "Catholics, Urbanism, and Mixed-Catholic Marriage Rates," Social Problems, 9 (1962), p. 363. 14Thomas P. MOnanan and Loren E. Chancellor, "Statistical Aspects of Marriage and Divorce by Religious Denominations in Iowa," Eugenics Quarterly; 2,(1955), p. 164. 83. intermarriage was negatively related to being born on a farm (prob- ably in a foreign land).15 As illustrations of the normlessness which characterized a people at war, we have data on intermarriage in one parish over twenty-six years, showing the rates to be the highest during World War 11.16 Thomas estimated from records in the Official Catholic Directory that diocesan mixed marriage rates increased by five or ten per cent during both world wars.17 That anomia assOciated with poverty also has its influence is suggested by the fact that the males who married outside the Church in one Canadian city were al- most all in unskilled or semi-skilled occupations.18 At the other end of the economic scale, we may see anomia in the positive assoc- iation of Catholic mixed marriages with those who are upwardly mo— bile in education.19 With this much for background, the reader is now in a better position to perceive the rationale for the hypotheses of this study. It will be noticed that we follow Heiss in specifying the parents and family as agents of social control. However, Heiss did not include the church or the ethnic group as direct agents of control 15Heiss, op. cit., p. 51. 16Thomas F. Coakley, "Mixed Marriages, Their Causes, Their Effects, Their Prevention," Lumen Vitae, 4 (1949), p. 456. 17John L. Thomas, "The Factor of Religion in the Selection of Marriage Mates," American Sociological Review, 16 (1951), p. 489. M 18Victor J. Traynor, "Urban and Rural Mixed Marriages," Social Order, 6 (1956), p. 157. 19Heiss, Op. cit., p. 51. 84- in his hypotheses. Neither did he make explicit use of the con- cept of anomia, although some of his data would lend itself to this kind of conceptualization. Hypotheses To Be Testedyln This Study The foregoing studies provided the germ for the follow- ing system of hypotheses. However, both Opportunities and limita- tions for testing this theory of social control in relation to Protestant-Catholic marriages were dependent on the data at hand. As has been stated in Chapter I, the Detroit Area Study surveys for 1955, 1956, 1958, 1959, and 1962 appeared to be suitable for the study of interfaith marriages because they included questions concerning the religious preference of husband and wife. By examr ining the interview schedules used in these surveys, the writer believed that the following hypotheses could be tested. The phras- ing of the sub-hypotheses, in fact, reflects the wording of these interview schedules. A. General hypothesis: If the agents of social control do not, or cannot (due to an anomic situation), exercise their influence, the viola- tion of a religiously endogamous rule is more likely to occur. B. Specific hypotheses: 1. If parents do not exercise social control, the viola- tion of a religiously endogamous rule is more likely to occur. ‘ a. Protestant-Catholic marriages will be positively associated with having parents of mixed religions. b. Protestant-Catholic marriages will be negatively associated with mother's frequent talking about or participating in religious activities with the family when self was growing up. c. Protestant-Catholic marriages will be negatively associated with naming parents as having had "the greatest influence on your religious beliefs." 85. 2. Ifgfamily or relatives do not exercise social control, the viOlation of a religiously endogamous rule is more likely to occur. a. Protestant- Catholic marriages will be negatively associated with having a high proportion of rela- tives who are of one's own faith. b. Protestant-Catholic marriages will be negatively related to the statement that friends or relatives would feel "unhappy or disturbed" or "would try to discourage" one from changing to Protestant or to Catholic. c. Protestant-Catholic marriages will be negatively related to more visits by parents and their chil- dren with relatives than friends when self was growing up. 3. If the church does not exercise social control, the violation of a religiously endogamous rule is more likely to occur. a. Protestant-Catholic marriages will be negatively related to one's education in church schools. b. Protestant-Catholic marriages will be negatively related to frequent attendance at religious serv- ices the year before marriage. 4. If the ethnic group does not exercise social control, the violation of a religiously endogamous rule is more likely to occur. a. Protestant-Catholic marriages will be positively related to being an American of the second and subsequent generations. 5. If an anomic situation exists, the violation of a re- ligiously endogamous rule is more likely to occur. a. Protestant-Catholic marriages will be positively related to being married during war years--Wor1d War I, World War II, Korean War, determined from the question on length of marriage. b. Protestant-Catholic marriages will be positively related to getting married at a late age, with consequent emancipation from home. c. Protestant-Catholic marriages will be positively related to migration from a farm to Detroit before age 18. d. Protestant- Catholic marriages will be positively related to the husband's extreme upward mobility in education, compared to his father. e. Protestant-Catholic marriages will be positively related to the husband's extreme upward mobility in occupation, compared to his father. f. Protestant-Catholic marriages will be positively related to a brief courtship. These are the hypotheses to be tested in our study. How- ever, it turned out that two of the interview questions were not 86. coded by the directors of the original studies, yig. hypotheses 2-c and S-f. Consequently, these two hypotheses will not receive any further attention, since they cannot be tested with the data we have. As a point of clarification, it may also be worth stat- ing here that the anomic situation in the fifth specific hypo- thesis does not necessarily infer a state of normlessness in the entire social system. It refers to a situation which is thought to induce anomia in the individual personality system, with spe- cial reference to the norm of religious endogamy. The Assumption of a Normative Order It has been stated that this research project is concerned with a concrete action, yig. the choice of a marriage partner, insofar as this conforms with the norm of religious endogamy. The writer's initial interest in this was practical, not theoreti- cal. Nevertheless, the social scientist who reads the hypotheses which are to be tested here will immediately be aware that the theoretical ramifications of this research problem are very exten- sive. In general, the writer's orientation in the study of this problem has been that of structural-functional analysis. Many so- cial theorists have something to contribute to our understanding of the subject of this study. However, it has been the writer's experience that the works of Talcott Parsons---because of his broad Theory of Action framework and its articulation on the psycho- logical, social, and cultural levels of analysis---these works seem to provide more points of contact with our far less abstract theory of interfaith marriages than do the works of any other author. On specific concepts or hypotheses, of course, other so- 87. cial theorists will be better able to further our understanding. An overview of the hypotheses proposed in the preceding section surely indicates the need for us to rely on the most fund- amental of all sociological assumptions, iy_gy, the assumption of a normative order. If there were no order or regularity in human activities, then behavioral science would be an impossibility. If social norms, iy_g., shared standards of behavior, were non-exist- ent, then sociologists would have to go out of business.‘ Let us consider briefly how this basic concept has arisen and been de- scribed. In all Social phenomena we perceive the working of the physio- logical laws of the individual; and moreover something which modifies their effects, and which belongs to the influence of individuals over each other---singu1arly complicated in the case of the human race by the influence of generations on their successors.20 In this pregnant sentence we find suggested the interrela- tionship of psychological, social, and cultural systems. Yet, they . are the words not of a contemporary social theorist but of the "grandfather" of them all---Auguste Comte. He has not made expli-p cit, however, the concept of "norms." In fact, this writer is not prepared to say who among so- cial scientists first made explicit use of the concept of social or cultural norms. Emile Durkheim certainly was using this notion when he described his "social facts" as "ways of acting, thinking, and feeling, external to the individual, and endowed with a power 20Auguste Comte, The Positive Philosophy of Auguste Comte, trans. Harriet Martineau, New York: Calvin Blanchard, 1858, p. 46. 88.” of coercion, by reason of which they control him."21 We find the same content with different labels in the folkways and mores of William G. Sumner: In fact, the real process in great bodies of men is not one of deduction from any great principle of philosophy or ethics. It is one of minute efforts to live well under existing con- ditions, which efforts are repeated indefinitely by great num- bers, getting strength from habit and from the fellowship of united action. The resultant folkways become coercive. All are forced to conform, and the folkways dominate the societal life. Then they seem true and right, and arise into mores as the norm of welfare. Among anthropologists, Ralph Linton used almost an equivalent con- cept in his "ideal patterns:" "The sum total of the ideal patterns which control the reciprocal behavior between individuals and be- tween the individual and society constitute the social system under which the particular society lives."23 Whoever introduced the pre- cise concept of cultural or social norms, we can see that similar ideas have been around for a long time. George C. Homans has a clarity of style which will be particu- larly helpful at this point. A norm, he says, "is an idea in the minds of the members of a group, an idea that can be put in the form of a statement specifying what the members or other men should do, ought to do, are expected to do, under given circumstances."24 21The Rules of Sociological Method, 8th ed., trans. Sarah A. Solovay and John H. Mueller, ed. George E. Catlin, Glencoe, 111.: Free Press, 1958, p. 3. 22Folkways: A Study of the Sociological Importance of Usageg, Manners, CustomsL,Moresy_gnd Merals, New York: Mentor Book, 1960, p. 49. 23The Study of Man: An Introduction, New York: Appleton- Century-Crofts, 1936, p. 105. 24The Human Group, New York: Harcourt, Brace, 1950, p. 123. ~89. His stress on norms as ideas enables him to make a basic distinc- tion: "By our definition, norms are a part, but only a part, of what social anthropologists call the culture of a group."25 They are statements of what ought to be done, not expressions of actual behavior in a group. The interplay of these two aspects is gener- ally of great interest to a sociologist, as it will be for us in the matter of interfaith marriages. Underlying all norms are the values of a group. The dis- tinction between these two concepts is rather subtle. We may say, in general, that norms express a group's consensus on what should be £922, whereas values express what should be esteemed or desired. Since individuals belong to various groups, there can easily be conflict between the norms or values of these groups, as we shall find in a later section to be the case with respect to marriage choices. In Parsons' terms, we are concerned not simply with the cognitive or the appreciative modes of value-orientation, but with the synthesis of these two in the moral standards, "that aspect of value-orientation which is of greatest direct importance to the sociologist."26 One might think that values express the goals of a group, and norms the means to those goals. However, this is not viewed as an appropriate distinction. As Robin Williams expresses it, cultural norms include both cultural goals and the approved means for reach- 25Ibid., p. 125. 26The Social System, Glencoe, 111.: Free Press, 1951, pp..13-14. 90. ing those goals. To be cultural, the norms have only to be acquired by learning and to be shared by individuals, but2 norms shade‘into one another in many complicated nuances. 7 He goes on, in the same place, to specify--- some major dimensions of variation in cultural norms. 1. Prevalence within the given collectivity. 2. Enforcement and imputed sources of authority. 3. "Formal characteristics" of the norm itself, e. 8°: specificity, rigidity. 4. Relation to other norms. These dimensions will each be examined in our study of Protestant- Catholic marriages. These remarks begin to give a theoretical context to our research problem. Before moving on, it may be worth noting that similar ideas are generally found in courses on sociological prin- ciples and in the sociological approach to social psychology.28 The concept of norms also has been productive of small group re- search29 and of cross-cultural studies.30 27émgrican Society: .A Sociological Interpretation, New York: Alfred A. Knopf, 1959, p. 25. 28For example, Ronald Freedman, et a1., Principles of Sociology, rev., New York: Henry Holt, 1956, pp. 122-124; also Tamotsu Shibutani, Society and Personality, Englewood Cliffs, N.J.: Prentice-Hall, 1961, pp. 40-46. 29Dotwin Cartwright and Alvin Zander review the work of Sherif, Ascn, and others in their editorial comments in Group Dynamics: Research and Theogy, 2nd ed., Evanston, 111.: Row, Peterson, 1960, pp. 165-188. Two good examples of recent research using this concept are found in the same volume: Leon Festinger, Stanley Schacter, and Kurt Back, "The Operation of Group Standards," pp. 241-259; and Stanley Schacter, "Deviation, Rejection, and Communications," pp. 260-285. 30See, for example, Society and Self: A Reader ipySocial Psychology, ed., Bartlett H. Stoodley, New York: Free Press of Glencoe, 1962, pp. 203-215, "Normative Attitudes of Filipino Youth Compared with German and American Youth." Also: Ephraim H. Mizruchi and Robert Perucci, "Norm Qualities and Differential Effects of Deviant Behavior: An Exploratory Analysis," American Sociologi- cal Review, 27 (1962), pp. 391-399. 91. Institutions There has been considerable disagreement among sociolo- gists about the use of the concept, "institution," so much so that Homans chooses to avoid it altogether31 and Florian Zmaniecki asks whether sociology will drop the term and leave it to philosophy.32 This is unfortunate, since the concept can be a powerful one which lies at the heart of a theory of social organization. The diffi- culty seems to arise from the uncritical acceptance of usages found in the ordinary English dictionary, where the term receives at least three distinct meanings: 1) an established practice or system.of practices; 2) an organization having a social, educational, or re- ligious purpose; 3) the building which houses such an organization. It seems to this writer that MacIver and Page have done well to restrict the meaning of institution to "established forms or conditions of procedure characteristic Of group activity."33 They thereby are able to distinguish associations, such as the family or state, from institutions, such as marriage or government. The notions are related, Of course: "Every association has, with respect to its particular interest, its characteristic institutions," e. g..a church has its sacraments, etc.34 In this sense, we can 31The Human Group, op. cit., p. 269. .32"Social Organization and Institutions," Twentieth Centugy Sociology, eds., George Gurvitch and Wilbert E. Moore, New York: Philosophical Library, 1945, p. 199. 33Robert M; MacIver and Charles H. Page, Society: An Intro- ductory Analysis, New York: Rinehart, 1949, p. 15. 34Ib1d. 92. belong to an association but not to an institution. We shall try to use the concept consistently in this way. Although Robin Williams restricts the meaning of "institution" in the same way,35 others who have written extensively on the subject do not exclude the notion of association from the concept.36 Talcott Parsons, although using the term "collectivity" in place of "association," agrees closely with MacIver and Page, and incidentally specifies our particular interest in this study: An institution in this sense should be clearly distinguished from a collectivity. A collectivity is a system of concretely interactive specific roles. An institution on the other hand is a complex of patterned elements in role-expectations which may apply to an indefinite number of collectivities. Converse- 1y, a collectivity may be the focus of a whole series of insti- tutions. Thus the institutions of marriage and of parenthood are both constitutive of a particular family as a collectiv- ity.37 Parsons distinguishes three classes of institutions: rela- tional, regulative, and cultural. The first of these, which defines statuses and roles of the parties in interaction, is what will con- cern us in the problem of mate selection according to a role of re- ligious endogamy. In discussing the "pattern variables" of role- definition at this point, he notes that "the relative primacies given to choices between them can be treated as constitutive of the 35American Sociepy, op. git., p. 29. 36Cf. William G. Sumner, op. cit., p. 61; Robert c. Angell, The Integration of American Society: A Study of Grgpps and Insti- tutions, New York: McGraw-Hill, 1941, p. 25; Joyce 0. Hertzler, American Social Institutions: A Sociologipal Analysis, Boston: Allyn and Bacon, 1961 p. v; Constantine Parunzio, Major Social In- stitutions: An Introduction, New York: Macmillan, 1939, p. 546; Everett C. Hughes, "Institutions," Principles of Sociology, ed. Alfred M. Lee, 2nd ed. rev., 1959, p. 230. 37The Social System, op. cit., pp. 39-40. 93. patterning of relational institutions."38 The second (later enumer- ated as the fifth) of the "pattern variables," which he focuses on the system of relational institutions, is the "Private yp. Collec- tive Interest Dilemma" or "Self-Orientation yp, Collectivity-Orien- tation."39 This expresses aptly the dilemma of a Protestant or Catholic faced with the alternative of selecting a mate outside his religious group. ~One might think Parsons had this situation in mind when he wrote: There is a moral issue only when the alternatives involve a presumption of relevance to the "integrity" or the "solidar- ity" of an interaction system when the preservation of that integrity or solidarity is itself a value....The case of self-orientation is the case where, in the choice in ques- tion, which alternative is chosen is felt to be or defined as indifferent as far as the integrity of a valued social system of action is concerned. That of collectivity-orienta- tion on the other hand is that where such integrity is defined as being involved, so that the actor who chooses one side is violating his responsibilities, to the system as a unit and its participant members. It is onfly when an action system in- volves solidarity in this sense that its members define cer— tain actions as required in the interest of the integrity of the system itself, and others as incompatible with that in- tegrity----with the result that sanctions are organized about this definition. Such a system will be called a'hollectivity." Collectivity-orientation, as it were, involves posing the "Question of confidence;" "are you one of us or not? your attitude on this question decides."40 It is well to keep this question in the general context of collectivities, since we are interested not only in society-wide norms with respect to mate-selection but also in the viewpoints of the smaller groups which are named in our "agents" of social con- trol, namely the parents, family, church, and ethnic group, insofar 381bid., p. 59. 391b1d., p. 67. 401b1d., p. 97. 94. as they support a norm of religious endogamy. Therefore, when Hertzler says that, "The control of mating in the interests of society constitutes the essence of the institution of marriage," we can read "religious collectivity" for "society" and "religious endogamy" for "marriage."41 Finally, Robin Williams describes the "three main prob- lems" in the study of social institutions in such a way that more than enough work is cut out for us. The third of these will be our principal concern: First, one must describe and analyze the normative structure itself: the existing patterns, their causes and interrela- tions, the sources and mechanisms of institutional integra- tion, and consequences of the norms. Second, one must dis- cover the processes of change in institutional patterns: their causes, mechanisms, and results. Third, one must study the relation of individual personalities to the normative structure; this is the area of "social psychology" dealing with culture-and-personality problems and facing the complex- ities of "social control" and of motivations for conforming, innovating, or dissenting. The Concept of Social Control Whether we think of the rule of religious endogamy simply as a cultural norm or as an integral part (for the members Of a re- ligious group) of the institution of marriage, it clearly is of some concern to the religious group whether the rule is observed or not. To discuss the ways in which its observance is ensured is to enter into what Williams has just told us are the "complex- ities of 'social control.'" Since it is this concept which gives unity to the hypotheses of our study, we shall hardly be amiss in 41American Social Institutions, op. cit., p. 225. 42American Society, Op. cit., p. 32. 95. giving it considerable attention. First, a few words about the changing popularity of this concept and then an effort to define it satisfactorily. It seems that the first use of the expression "social control" by a sociol- ogist occurs in an early text by Albion W. Small and George E. Vincent.43 However, it became popular a few years later when it was made the subject of a book by Edward A. Ross.44 It even pro- vided the theme for the Twelfth Annual Meeting of the American So- ciological Society at Philadelphia in 1917, although the papers - given there were only loosely related to this concept.45 Since Ross' work, several authors have contributed books on the sub- ject."6 However, it seems to this writer that Only the texts by Landis and Roucek could be used for teaching a course in this area now; the quality of the latter is also very uneven because of its many contributors. Of the several review articles on this subject, 43An Introduction to the Studyyof Society, New York: American Book, 1894, p. 328. 44Social Control: A Survey of the Foundations of Order, New York: Macmillan, 1901. 4SPepers and Proceedipgs of the American Sociological Society, 12 (1917). 46Luther L. Bernard, Social Control in Its Sociological Aspects, New York: Macmillan, 1939; Jerome Dowd, Control in Human Socigtiep, New York: D. Appleton-Century, 1936; Paul H. Landis, Social Control: Social Organization and Disorganization in Process, rev., Philadelphia: Lippincott, 1956; Richard T. LaPiere, A Theory of Control, New York: McGraw-Hill, 1954; rFrederick E. Lumley, The Means of Social Control, New York: Century, 1925; Social Control, ed. Joseph S. Roucek, New York: D. Van Nostrand, 1947; somewhat revised in 1956. 96. the best seems to be that by Georges Gurvitch.47 In spite of the centrality that is often ascribed to this concept, it is not fre- quently used in research or theoretical articles. Sociological Abptracts sometimes has an entry under this heading appended to the Sociology of Law. The American Sociological Review indexes fewer articles on this subject than one finds under such headings as Social Class, Social Participation, or Marital Adjustment. Many wellsknown theoretical works in the areas of Social Organiza- tion and Social Psychology include one or more chapters with "So- cial Control" in the title.48 However, only occasionally does a study appear in which this concept is the unifying theme.49 What is the meaning of "social control"? Almost a score of definitions have come to the writer's attention. Meat of these stress the relationship of social control to system maintenance or social prganization. Social control is defined as one of the "Master Processes" in the theory of Charles Loomis: "the process by which deviancy is either eliminated or somehow made compatible with the functioning of the social groups."50 Hellingshead gives 47"Social Control," Twentieth Centugy Sociology, eds. Georges Gurvitch and Wilbert E. Moore, New York: PhilosOphical Library, 1945, pp. 267-2965 See also: August B. Hollingshead, "The Concept of Social Control," American Sociological Review, 6 (1941), pp. 217-224; Joseph S. Roucek, "The Development and Status of Social Control in American Sociology," American Sociological Review, 20 (1959), pp. 107-123. 48For instance, Hemans, The Human Group, op. cit.; Parsons, .The Social Systgm, op. cit.; Shibutani, Sociepy and Personality, op. cit. 49For example, Roscoe Pound, Social Cpptrol Through Law, New Haven: Yale University Press, 1942; Gordon C. Zahn, German Catholics and_§it1er's Wars: A Stpdy inySocial Control, New York Sheed and Ward, 1962. 50Social Systems, Princeton: Van Nostrand, 1960, p. 35. 97 us a clue instead of a definition: "The essence of social control is to be sought in the organization of people."51 Hemans is one of those who call attention to what social control means to the individual person. His definition is almost operational: "the process by which, if a man departs from his existing degree of obedience to a norm, his behavior is brought back toward that de- gree, or would be brought back if he did depart."52 Rarely, the term "social control" has been used to designate an outgoing pro- cess, such as "to control one's environment."53 For the purposes of our study, the definition given by Loomis is satisfactory. Associated with the rule of religious_endogamy, as we shall see, there are regulations by which deviance is "somehow made compat- ible with the functioning of the social groups." The means of social control may be conscious, such as re- wards or punishment, advertising or propaganda, or even laughter. Or the means may be unconscious, such as the folkways, mores, and societal roles. Law is a conscious as well as a formal means of control by the state. Landis is probably right in his judgment that, "Unconscious forces, imbedded in the culture, are at least as strong as the conscious, formal control devices."54 Since the elements of social structure, such as roles and 51"The Concept of Social Control," op. cit., p. 221. SZThe HMman Group, op. cit., p. 301. 53WilliamT. Liu and Frank Fahey, "Delinquency, Self- Esteem, and Social Controls: A Retreductive Analysis," American CatholpggSociological Review, 24 (1963), pp. 3-12. 54Social Control. 09- C1t-. P- 8- ii.- 98. institutions, work.for social control by their very existence, it follows that the socialization of the members of society into these roles and institutions is a key process contributing to social con- trol. This has been a special interest of social psychologists and anthropologists.55 Parsons also stresses its relationship to social control: There are such close relations between the processes of socialization and of social control that we may take cer- tain features of the processes of socialization as a point of reference for developing a framework for the analysis of the processes of control.5 He sees both processes as involving adjustment to strains through mechanisms of defense and adjustment. This leads him to describe the process of psychotherapy as "a prototype of the mechanisms of social control.57 The salient mechanisms used by the therapist, in Parsons' view, are those of permissiveness, support, denial of reciprocity, and manipulation of rewards. Later Parsons retains the same terms for the "learning-social control phases" in the socialization of the child.58 The consistency and scope of Parsons' theorizing are-capable of inducing amazement in the unsophisticated, such as the present writer. In our study of interfaith marriages, we shall not be undertaking a thorough and systematic study of socialization as 55For a good review of such studies, see: Irvin L. Child, "Socialization," Handbook of Social Psycholpgy, ed. Gardner Lindzey, Reading, Mass.: 1954, pp. 655-692. 56The Social System, op. cit., p. 298. 57Ibid., p. 301. 58Talcott Parsons and Robert F. Bales, Family, Socializa- tion and Interaction Process, Glencoe, 111.: Free Press, 1955,p.39. 99. relatedtto social control. However, it is not difficult to see that socialization is involved in our hypotheses relating to the influence of parents and of education in church schools. It would take us tOo far afield to go into a detailed dis- cussion of the qualities and functions of social control. Let us note for future reference, however, that the rigidity and effec-' tiveness of a system of social control seem to depend largely on two factors: 1) the homogeny of the group;59 and 2) the social and physical opportunities for escaping from.the group.60 In an excellent discussion related to this, William R. Gatton hypothe- sizes that conformity of a group's members to its norms will in- crease with the in-group's ethnocentrism. What consequences this may have for the group as a whole is suggested by this well-phrased observation of his: "If ethnocentrism.promotes social control with- in the group, it may make the group more effective at adaptation pf environment even when it interferes with the group's adapta- tion £9 environment."61 Approaches to Deviancy: Anomia Deviance and the mechanisms of social control may be de- fined with reference to the social system involved, or with the focus on the individual actor. We are more concerned about the latter in our research problem. Hence the definitions given by 59Landis, Social Control, op. cit., p. 148. 6OHomans, The Human Gropp, op. cit., p. 290. 61"The Functions and Dysfunctions of Ethnocentrism: A Theory," Social Problems, 8 (1960-61), p. 204. 100. Parsons are appropriate: ...deviance is a motivated tendency for an actor to behave in contravention of one or more institutionalized normative pat- terns, while the mechanisms of social control are the motiva- ted processes in the.behavior of this actor, and of the others with whom he is in interaction, by which these tendencies to deviance tend in turn to be counteracted.62 The very complementarity of these definitions suggests the inter- play between deviance and social control. It reminds us of' Durkheim's well-known passage on the function that crime performs in stimulating a reaffirmation of collective sentiment with respect to the violated norm.63 As Homans has expressed it more recently: "Crime---not too much of it---is needed; it keeps the controls in good working order. [A control is not effective unless it is test- ed...."64 In relating these ideas to Protestant-Catholic marriages, it seems to this writer that two points should be made at this time: 1) "Crime" and "criminal" have acquired connotations which 'make them inappropriate to apply to the violation Of a norm like religious endogamy. Even the concept of "nonconformist," which Merton carefully delineates from "criminal," is not always the proper word.65 We shall generally use the non-restrictive terms, "deviant" or "deviance." 2) The point about crime aiding the so- cial system by stimulating sanctions should not be accepted with- 62The SocialySystem, op. cit., p. 250. 63Ihe Division of Labor, trans; George Simpson, Glencoe, 111.: Free Press, 1947, pp. 102-103. 64The HumanyGropp, op. cit., p. 310. For a fuller discus- sion of the same point, see Kai T. Erikson, "Notes on the Sociology of Deviance," Social Problems, 9 (1962), pp. 307-314. 65Robert K. Merton, Social Theory and Social Strpgture, rev., Glencoe, 111.: Free Press, 1957, pp. 360-363. 101. out a distinction. If violation of a norm is a way for a non- conformist to bring about its revision or removal, as Merton ob- serves, it may indeed help the social system toward defining its own boundaries and making new adjustments to the environment. How- ever, the idea that crime, properly speaking, is functional for the system (always "in limited quantities") because it tests the sanc- tioning machinery---is, simply from the point of view of the so- cial system, a point to be studied rather than assumed. Is the social structure stronger or weaker piggy the violated norm.has been reaffirmed? If weaker, than how is the criminal act "func- tional" for the system? If stronger, then why should not each punished crime have the same effect? Why do authors avoid this cumulative notion by adding a phrase such as "not too much of it"? EnOugh for the functions of deviancy. we are more-concern- ed in this study with predicting its occurrence. This brings us to the question of its relationship to alienation and an ie, a sub- ject to which very many have contributed in recent years. It is important for us to be as clear as we can about this, since sev- eral of our hypotheses concern "anomic situations." Historically, the problem stems from.Emile Durkheimis use of the term anomie both in his study of suicide and in an earlier work.66 To get to the root of the meaning, this exposition seems. to help: I Thus, social disorganization; discarded and discredited norms; 66The Ndvision of Labor in Society, op. cip., p. 368; Suicide: A Stpdy in Sociology, trans. John A. Spaulding and George Simpson, ed. George Simpson, Glencoe, 111.: Free Press, 1951, pp. 241-276. 102- a flat unwillingness to accept in any form a checkrein on pleasures, appetites, production, Or prosperity: this con- stellation of signs Durkheim translated into Greek. To the negative prefix 3, he added the plural of laws, npgpus, and turned the "no-laws" of anomous into French as l'Anomie. That is why Durkheim' s first mention of the term occurs as "the state’of unbri edness" (to put it awkwardly), or, in French "L'etat de d feglement ou 'anomie. ... "67 Some have translated Durkheim's term as "normlessness," but it is‘ important to note that for him it originally designated a state of a society or a group. Loomis, following Parsons here, defines anomie as "under-institutionalization...in which persons are pro- vided with no effective norms to guide them, no meaningful status- roles, no sanctions, and other elements to standardize behavior."68 One distinction made in the concept as used at this level may be applied to our study. It is attributed by Merton to Sebastian De Grazia, viz. the distinction between simple and acute anomie: Simple anomie refers to the state of confusion in a group or society which is subject to conflict between value-systems, reSulting in some degree of uneasiness and a sense of separa- tion from the group; acute anomie, to the deterioration and, at the extreme, the disintegration of value-systems, which reSults in marked anxieties.69 As we shall see, the principal value-systems which are in conflict in interfaith marriage are the secular and the religious. Anomie at the societal level has its counterpart in the experience of the individual personality. To avoid confusion in using the term at these two levels, we are told that MacIver used the spelling "anomy" to designate the state of mind of the individ- 67Isabel Cary-Lundberg, "On Durkheim, Suicide, and Anomie," American Sociological Review, 24 (1959), p. 250. 68Social Systems, op. cit., p. 36. 69Social Theory and Social Structure, op. cit., p. 163. ,103. ual.7O A few years later, Leo Srole-suggested the spelling "anomia" to distinguish the psychological concept from the sociological con- cept.71 His suggestion was taken up by others and is followed in this study also. Apart from Spellings, what does this psychologi- cal concept actually mean? It seems to this writer that clarity of meaning was lack- ing in the use of this term and the related concept of alienation. A significant contribution toward eliminating such confusion was made by Melvin Seeman in an article published three years ago.72 He discussed "five basic ways in which the concept of alienation has been used"---power1essness, meaninglessness, normlessness, iso- lation, and self-estrangement. If we assume that these exhaust the meanings which the term has received and if we restrict anomia to its Durkheimian usage of "normlessness," then we can see that anomia as a psychological concept (not as a property of a social system) constitutes one ppgp of the literature on alienation. One can hope (without any real assurance) that these clear distinctions will be followed by future writers on the subject, but it is neces- sary to be cautious in interpreting earlier publications. One modification of Seaman's article, proposed by several men at Whittier College, seems to have merit. They reformulate the five meanings somewhat and consider them to represent stages of a pro- 70Ibid., p. 162. 71"Social Integration and Certain Corollaries: An Explor- atory Study," American Sociological Review, 21 (1956), pp. 710-712. 72"0n the Meaning of Alienation," American Sociological Review, 24 {1959), pp. 783-791. 104. pgpg of alienation. Seeman, in his reply to them, judges that this view of alienation would be an "over-commitment which leaves too little room.for historical circumstances, situational pressure, or personality type in shaping the specific form of sequence that alienation will take."73 This criticism is prob- ably correct with regard to the concept of alienation, but the process idea as the Whittier group proposed it could be regarded as an hypothesis to be tested. Related to the meaning of anomia and alienation is, of course, the problem of measuring these concepts operationally. In 1959, Leo Srole proposed a set of attitudes which he thought would provide a scale of alienation and would "place individuals on a eunomia-anomia continuum representing variations in interpersonal integration with their particular social fields as 'global enti- ties."74 Gwynn Nettler followed with a measure of alienation, which correlated with Srole's anomia scale + .309.75 When one examines the five attitude statements in Srole's scale it is dif- ficult not to agree with Nettler's judgment that what they measure is largely despair. (In view of what was said above about the meanings of alienation in Seeman's article, it is regrettable that this scale has been used by a number of researchers who considered 73Charles J. Browning, Malcolm F. Farmer, H. David Kirk, and G. Duncan Mitchell, "On the Meaning of Alienation," American Sociological Review, 26 (1961), pp. 780-781. 74"Socia1 Integration and Certain Corollaries: An Explor- atory Study," op. cit., p. 711. 75"A Measure of Alienation," American Sociological Review, 22 (1957). Pp. 670-677. 105. themselves to be measuring anomia; the Durkheimian sense of "norm- lessness" is conceptually not the same as "despair.") In this writer's eatimation, the most thoroughgoing approach to the measure- ment of alienation so far published has been reported by Dwight G. Dean. In selecting items for his three subscales of Powerlessness, Normlessness, and Social Isolation, he enlisted the aid of seven sociologists as judges of 139 attitude-items, finally retaining 24 items.76 Conclusions resulting from the use of these scales will be cited toward the end of this chapter if they relate to the "anomic situations" of our study. Here we may mention two characteristics which have been thought to be associated with anomia, yig. socio- economic class and religion. Interest in the question as to whether anomia is related to religion stems from the finding by Durkheim that anomic suicide was more characteristic of Protestant than of Catholic populations (divorce rates being his index of anomy).77 Using Srole's anomia scale in San Francisco, Meier and Bell report that, when socio-economic status and age are controlled, "the Protestants generally show lower anomic scores than the other religious groups," though "the relationship is more obscure."73 Dean and Reeves used the "normlessness" subscale developed by the former and found that women at a midwest Protestant college had 76"Alienation: Its Meaning and Measurement," American Sociological Reviewy 26 (1961), pp. 753-758. 77Suicide, Op. cit., p. 260. 78Dorothy L. Meier and Wendell Bell, "Anomia and Differ- ential Access to the Achievement of Life Goals," American Socio- lpgicpl ReviewI 24 (1959), p. 200. 106. significantly higher normlessness scores than women at a midwest Catholic college.79 Since divorce rates and normlessness seem more relevant to interfaith marriages than does despair (these being the criteria for anomia used in these three studies), we may have some slight reason here to expect our hypotheses concerning "anomic situations" to be better predictors for Protestant, than for Catholic, mdxed marriages. Interest in the relationship of socio-economic status to anomia was stimulated by Robert Merton's study of the subject. In a society such as ours, in which the achievement of "success" is held out as the goal for all and in which those in some socio- economic strata find the means to this goal relatively unavailable to them, these strata will experience a "strain toward anomie," "the breakdown of the regulatory structure."80 (This not quite the same as the hypothesis that vertical mobility is related to anomia.) \Durkheim.had found anomic suicide to be most character- istic of those with independent means, among those in professional, industrial, and commercial occupations.81 Although Merton takes account of "white collar crime" in one deviant adaptation (Inno- vation) to this conflict between the "success" goal and limited accessibility of culturally-approved means toward the goal, his analysis suggests, in general, that "the greatest pressures toward 79Dwight Dean and Jon A. Reeves, "Anomie: A Comparison of a Catholic and a Protestant Sample," Sociometry. 25 (1962), p. 211. 80Social Theory and Social Structure, op. cit., p. 157. “Mel Me. 913- 257-258. 107. deviation are exerted upon the lower strata."82 Nevertheless, the association between social class and anomia is still in doubt, as Ephraim.H. Mizruchi observes: Is the association an inverse one, as Srole and Bell report, or is it "negligible," as Roberts and Rokeach suggest? Dif- ferent indices Of class were used in these three studies, which probably accounts for some of the discrepancy between the findings. No standard index of social class has as yet been applied systematically in the various attempts to eval- uate its role in relation to anomia. His own data from a small city in upstate New York show a signifi- cant inverse relationship between scores on Srole's anomia scale and Hollingshead's two-factor Index of Social Position as well as the respondents' subjective class identifications. In applying this generalization to the interpretation of interfaith marriage rates, we must remember the large element of "despair" in the Srole scale which was used for each of these studies. This may not be related to the "normlessness" which we look for in our hypotheses concerning anomic situations. In Merton's study of social structure and anomie, he goes to considerable trouble to develop and describe a typology of modes of individual adaptation to cultural values, based on their accept- ance or rejection.of the goals and institutionalized means.84 Parsons modified the paradigm by adding the dimension of activity ‘yp. passivity.8S Robert Dubin greatly expanded the typology and 82Socingheory and Social Structure, op. cit., p. 144. 83"Social Structure and Anomda in a Small City," American Sociological Review, 25 (1960), p. 647. 8€§ggipl_1heory gpd_§ocipl Structure, op. cit., pp. 139 ff. 85The Social S stem, My pp. 256 ff. 108. ended with fourteen categories of deviant behavior.86 At this point in our study it does not seem worthwhile to go into detail concerning these typologies. Our hypotheses are not intended to validate them, and our data are only partly suitable for this. We shall call upon these typologies later if they seem to contri- bute to our interpretation of the results of our study. Finally, having considered at length sources of deviant behavior and its typologies, let us bring this section to a close by mentioning the study of some reactions to deviancy. we have already called attention to the point that crime and nonconform- ity are thought to stimulate the reinforcing of the social struc- ture. Small group studies have shown that deviates do not fare well in sociometric choices, that they also experience pressures to change in the form of increased communication received.87 On a more theoretical level, HOward Becker's Presidential address to the American Sociological Association in 1960 dealt with reactions not only to deviates but to the condition of normlessness itself.88 It will be more appropriate to consider the application of his ideas to our study, however, when we come later to talk about secu- larism as a value system. 86"Deviant Behavior and Social Structure," American Socio- logical Review, 24 (1959), pp. 147-164. Merton gives his reactions in the same issue of the Review: "Social Confonmity, Deviation and Opportunity-Structures: A Coment on the Contributions of Dubin and Cloward," pp. 177-189. 87Leon Festinger, Stanley Schacter, and Kurt Back, "The Operation of a Group Standards," and Stanley Schacter, "Deviation, Rejection, and Communication," op. cit. 88"Normative Reactions to Normlessness," American Socio- logical Review, 25 (1960), pp. 803-810. 109. Eyidence for_p Norm of Religious Endogamy in the Upgped States Through most of this chapter we have been trying to lay the theoretical groundwork for this study of Protestant-Catholic marriages, talking about norms, institutions, social control, deviantbehavior, anomia. The reader may be ready to exclaim at this point: "This is all well and good. But how do you know that these concepts have any application in logic to your study? YOu have not even shown that there exists a norm of religious endogamy." Well, that is precisely what we are going to turn to now. We have already interpreted low rates of interfaith marriage as some evi- dence of an endogamous norm, of course, but it is time now to ex- amine the explicit expressions of the norm. We shall first take an historical view, then review current statements of the official position of the churches, followed by warnings stated by religious guides of young people. Finally, we shall show that the norm is shared or at least expressed by others in the general population, even some who have themselves intermarried. As a starting-place in the history of religious endogamy, the command given by God through Moses to the Isrealites as they were about to enter the promised land is clear enough. Regarding the other peoples they would find living there, God said: "Nei- ther shall you take their daughters as wives for your sons; other- wise, when their daughters render their wanton worship to their 8°d32 they Will make your sons do the same."89 89Exodus: 34/16. Translation by the Confraternity of Christian Doctrine, Washington, D. C.: Confraternity of Christian Doctrine, 1952. 110. Jewish opposition to marriage with Gentiles through later centuries is well known. However, our concern in this study is with norms of Catholics and Protestants regarding intermarriage. In the Christian era, the first known prohibition was stated by the Council of Elvira, Spain (306 A.D.), which "for- bade Christian girls to marry pagans, Jews, heretics, or pagan priests." Girls are singled out since "under the civil law of the time (ig_g., in the Greek-Roman empire) the man had the right to dictate the religious profession of his wife and children." In 339 A.D., the Emperor Constantius "forbade the marriage between Christians and Jews under the penalty of death."90 One council at Constantinople in 692 declared that, if marriage is attempted by a Catholic with a heretic, it is null and void.‘ However, the Scholastic teaching in the 13th century was that such marriages were forbidden but not invalid. No provisions were made during the Middle Ages for ecclesiastical dispensations for such mar- riages.91 This practice arose (for the newly discovered lands, especially in the Far East) in the late 16th century. For Europe the dispensation had to be obtained from the Holy See, although "a custom developed in Germany, England, Poland, and the Nether- lands to marry freely mixed couples without a papal dispensation and without the abjuration of heresy."92- The tenor of these times is also represented by the Presbyterian Confession of Faith, 90Stanislaus Woywood, "Marriages between Catholics and Non- Catholics," Homiletic and Pastoral Review, 40 (1940), p. 411. 91Ibid., pp. 412-414. 921b1d., p. 416. 111. Chapter XXIV, Section III, which reads: It is lawful for all sorts of people to marry who are able with judgment to give their consent; yet it is the duty of Christ- ' ians to marry in the Lord, and, therefore, such as profess the true reformed religion should not marry with infidels, Papists, or other idolaters....9 How did the much discussed promises required of partners to a mixed marriage in the Roman Catholic Church arise? When Pope Pius VI (1775-99) was reigning, Belgian civil law required that Catholic pastors witness mdxed marriages. The Belgian bishops asked the Holy See what they were to do if the Catholic party could not be dissuaded and the Protestant party did not want to become Catholic. The answer: ...the pastor could be present as a mere material witness to the declaration of the marriage consent, and only under the following conditions (called the-cautelae): (1) that the marriage be not celebrated in a sacred place, that the pastor wear no sacred vestments, recite no ecc1esiastica1 prayers over the parties, and shall not bless them; (2) that the pastor demand and receive from the heretical party a written declaration under oath and in the presence of two witnesses, who together with the party shall sign the declaration bY, which the Protestant binds himself to permit the Catholic party the full and free exercise of his religion and that all the children without distinction of sex will be raised in the Catholic Faith; (3) that the Catholic party give a written declaration under oath in the presence of two witnesses and subscribed by him.and the witnesses that he will never aposta- tize from the Catholic Faith, that he will educate all the children to be born in the Catholic religion, and endeavor to induce the non-Catholic partner to join the Catholic Church; (4) as to the proclamation of the banns of mixed marriage which the imperial decree has ordered, the Holy Father does not ap- prove of such proclamation and directs the bishops to petition the emperor that he exempt them from making such announcements, at least that they may not be in church. (July 13, 1782)94 93Algemon D. Black, If I Marry Outside My Religion, New YOrk: Public Affairs Pamphlets, 1954, p. 14. 94Woywood, "Marriages between Catholics and Non-Catholics,‘ op. cit., pp. 417-418. 112. ‘ The Code of Canon Law of the Roman Catholic Church, pro- mulgated on May 19, 1918, and still in effect, reads as follows in 8 the pertinent sections: The Church most strictly and everywhere forbids mar- riages between a Catholic and a person enrolled in an hereti- cal or schismatical sect. If there is a danger of perversion (i. e., "danger of loss of faith or of Christian life," p. 634) for the Catholic party and the offspring, such marriage is forbidden also by the divine law. (Canon 1060) The Church does not dispense from.the impediment of mixed religion except under the following conditions: (1) there must be good and weighty reasons; (2) the non-Catholic party must promise to avert all danger of perversion from the Catholic party, and both parties must promise to have all the children baptized and raised in the Catholic Faith; (3) there must be moral certainty that the promises will be kept. The promises are, as a rule, to be made in writing. (Canon 1061) The Catholic party has the obligation to work prudent- ly for the conversion of the non-Catholic. (Canon 1062) Though the Church has granted the dispensation from the impediment of mixed religion, the parties are forbidden either before or after the Catholic wedding to approach either in person or by proxy a non-Catholic minister as a minister of religion to give or renew the matrimonial consent. If the pastor knows for certain that the parties will violate or have already violated this law, he shall not assist at their mar- riage except for very serious reasons until he shall have con- sulted the Ordinary and all danger of scandal has been removed. If the civil law demands it, the Church does not censure parties for appearing before a non-Catholic minister who is acting merely as an official of the Government, provided that their purpose is solely to comply with the civil law and to get civil recognition of their marriage. (Canon 1063)95 These canons do not exhausc the regulations of the Catholic Church regarding interfaith marriages, but they express the main princi- ples. Cardinal Cushing of Boston was recently quoted in an inter- view as hoping that Vatical Council II will make-some changes in these rules. He said: As it is now, the requirement that a non-Catholic partner make the famous promises before marriage is an irritant to many, QSStanislaus Woywood, "Marriage Impediment of Mixed Reli- gion," Homiletic and Pastoral Review, 40 (1940), p. 634. 113. and some, it is clear from what happens subsequently, make the promdses.in bad faith. If we no longer required the promises, we would not be revoking any divine law; we would not be chang- ing any dogma of the Church. There are good reasons for con- sidering such a move. Remember when mixed marriages could not be performed in church, but had to take place in the rectory? , we changed that and permitted the use of the church. Instead of separating that ceremony from a holy environment, we now start these marriages off in a context of church, which opens up the possibility of many actual graces being given instead of tpg generating of feelings of frustration, hostility, etc. The reactions of America's readers to these statements were immediate. The editors thought it necessary to interpret the Card- inal's mind for him: They thought the Cardinal meant the Church should a1- 1ow mixed marriage couples to bring up their children in dif- ferent faiths. He intended no such thing.... Would dropping the law mean denying the obligation of bringing children up in the true faith? Not at all. The ob- ligation would remain, squarely confronting the individual Catholic conscience. If Cardinal Cushing discerned that Catho- lics did not want to be left without the force of a Church law backing up a basic moral requirement, he would no doubt with- draw the question he raised. However, that would mean limit- ing a development he has encouraged for some years: increas- ing the range of responsible action by the laity. In Cardinal Cushing's opinion, one result of the Church's intensive educational efforts in recent generations should be some understanding of the distinction between divine and man-made law, and an increased measure of trust in the laity. His question about the premarital promises was intend- ed to suggest a practical way for the laity to exercise their co-responsibility in the Church.97 Needless to say, churchmen of other faiths are disturbed by the requirement of these promises. In fact, a new awareness of these regulations seems to have prompted the most recent declara- tions of Protestant bodies against marriages with Catholics. One 96Walter‘M. Abbott, "Two Council Fathers: Cardinal Leger and Cardinal Cushing," America,'108 (1963), p. 865.: .. 97"Mixed‘Marriage Question," Americpplo9 (1963), p. 8. 114. of the first on record was the 1948 General Convention of the Protestant Episcopal Church, which unanimously adopted a resolu- tion patterned after that adopted by the last Lambeth Conference: Resolved, that this convention earnestly warns members of our Church against contracting marriages with Roman Catho- lics under conditions imposed by modern Roman Catholic canon law, especially as these conditions involve a promise to have . their children brought up under a religious system.which they cannot themselves accept; and further, because the religious education and spiritual training of their children by word and example is a paramount duty of parents and should never be neglected nor left entirely to others, we assert that in no circumstances shall a member of this Church give any under- standing as a condition of marriage, that the children should be brought up in the practice of another communion. It should be clear already that the rule of religious en- dogamy gets much of its strength from concern about the religious education of the children of a mixed marriage and the fulfilling of that obligation by parents who are not agreed on the direction this education should take. Except for this problem, the churches might very well leave the mixed couple to their own devices. As it is, the following church assemblies echoed (each with its own specific motivations) the resolution just cited: 1950 - General Assembly, Presbyterian Church of the United States. - American Baptist Convention (Northern). - International Convention of Disciples of Christ. 1951 - Southern Baptist Convention. 1953 - Convention of the Lutheran Church--Missouri Synod.99 The shnction imposed by the Catholic Church in these mat- ters is excommunication reserved to the Ordinary (i. e., the bishop of the diocese), applied to Catholics who contract marriage before 98Black, op. cit., p. 15. 99James A. Pike, If You Margy Outside Your Faith; New York: Harper, 1954, pp. 93-98. 115. a non-Catholic minister contrary to Canon 1063, or who marry with the explicit or implicit agreement to educate some or all of their offspring outside the Catholic Church. (Canon 2319) In addition, for members of the Catholic Church, only those marriages are valid which are contracted before the parish priest or the local Ordir nary or a priest delegated by either Of them.and before two wit- nesses. (Canon 1094) Consequently, the presence of a priest is a good index that the marriage satisfies the other requirements of the Catholic Church and that the Catholic who is marrying is in good standing with the Church. The larger Protestant denominations have not specified any consequences which follow for members who do not observe the reso- lutions cited above. However, "several small Protestant sects excommunicate members who marry outside of their particular churches."100 The reader may be convinced by now that a norm of religious endogamy exists in the official statements of the churches of the United States. How is this norm kept before the eyes of their members? There are occasional high-level conferences, such as that held for two days by the Herzl Institute of New York on “Inter- marriage and Jewish Life," February, 1960. There are teaching- aids such as the 27-minute film, "One Love--Conflicting Faith,". produced by the Dept. of the Christian Family, the General Board of Education, and the Television, Radio and Film.Commission of the 100James H. Bossard and Eleanor S. Boll, One Merriage, Two Faiths: Guidance on Interfaith Marriagg, New York: Ronald Press, 1957, p. 84. 116. Methodist Church. We can assume, no doubt, occasional reference' to it in sermons, Sunday school classes, and family counselling sessions of individual congregations. However, there are two means of dissemination which are rather easy to identify: textbooks for high school and college-course, and "guide" books for young people and families. Sister M; Evodine McGrath made a study of twenty-eight Catholic colleges for women and five for men in the Midwest, to determine their offerings relative to marriage and the family. One of her findings was that mixed marriages were discussed in a variety of courses (usually a religion or family course)---a total of 47 courses for the 55 colleges.101 Father Edgar Schmiedeler has written one text for high school students and another for col- lege use, each warning against mixed marriages.102 Even non-sec- tarian.marriage and family textbooks discourage such unions.103 Protestant authors are more prominent in the production of "guide" books for young people and families. Besides the three cited earlier by Black, Bossard and Boll, and Pike, we should men- 101The Role of the Catholic College in Preparing for Mar- riage and Family Life, Washington: Catholic University Press, 1952, Table 2, p. 22. 102Lookipg Toward Marriage, Washington: Family Life Bureau, National Catholic Welfare Conference, 1948; and Marriage and the Family, New York: McGraw-Hill, 1946. See also: Alphonse H. Clemens, Marriage and the Fgmily: An Integggted Approach for Catholipp, Englewood Cliffs, N. J.: Prentice-Hall, 1957; Clement S. MHhanovich, Gerald J. Schrepp: and John L. Thomas, A Guide to Catholic Marriage, Milwaukee: Bruce, 1955. 103For example, Henry A. Bowman, Marripge for Mpderns, 4th ed., New York: 'McGraw-Hill, 1960, pp. 171-205. 117. tion that John C. Wynn has written a good one.104 In a slightly different class is the scholarly comparison of Roman Catholic and Protestant views on marriage, with special reference to mixed mar- riage, written by Mario Colacci.105 Apparently an Evangelical Protestant, the author does not fully understand some Catholic teachings, but he uses official Catholic sources (often with the most "conservative" commentators) and tries to be accurate and fair. Among Catholic authors of marriage "guide" books, perhaps the best known is Father George A. Kelly of New York.106 All of these warn against interfaith marriages. Granted that the churches try by various means to maintain the norm of religious endogamy, is the general population aware of it? It would be possible to cite a dozen references from.non-re- ligious sources which show an awareness of this norm in our coun- try---which is the only point we wish to make just now. To show the variety of contexts in which this may occur, we mention Alfred M, Lee's discussion of socialization, J. O. Hertzler's book on so- cial institutions, and George Simpson and Milton Yinger's master IOQHQw Christian Parents Face Familngroblems, Philadelphia: westminster Press, 1955. 105Christian MarriageToday, Minneapolis: Augsburg Press, 1958. 106The Catholic Marriage Manual, New York: Random House, 1958; and The Catholic Family Handbook, New York: Dell Publishing, 1962. See also: John A. O'Brien, Happy Marriage, New York: Popular Library, 1957. 118. work on minority groups.107 One curious indication of awareness of the norm of religious endogamy is found in the complaint by Glenn M. Vernon that social scientists show their own bias against interfaith marriages by consistently reporting the increased per- centage of divorces which follow such marriages instead of compar- ing the percentages of unbroken marriages, in which the contrast is not so great.108 Although one cannot base a statistical generalization on a small set of case reports, it is of interest that a number of people who have entered interfaith marriages and stuck it out have written about them afterwards to dissuade others from arriving at the same choice.109 So far no similar report has come to this writer's attention in which a couple in a mixed marriage urged others to do the same. Another indication of the norm of religious endogamy can be found in studies of attitudes toward interfaith marriages. A number of these studies have already been reviewed in Chapter II, 107 ‘ Alfred M. Lee, "Socialization of the Individual," Principles of Sociology, ed. Alfred M. Lee, New York: Barnes and Noble, 1959; p. 309; J. O. Hertzler, firican Social Institutions, op. cit., p. 228, George Simpson and Milton Yinger, Racia1:pnd Cultural Minor- ities, rev., New York: Harper, 1958, p. 568. - 108"Interfaith Marriages---Bias in Professional Publica- tions," Religious Educating, 55 (1960), pp. 261-264. 109For example, see: Anonymous, "What it Means to Marry A Catholic," The Forum.I 81 (1929), pp. 339-345; Thomas W. Beesley, "What It Means to Marry a Protestant," The Forum, 82 (1929), pp. 226-230; A Catholic Husband and a Protestant Wife, "I Love You, But....," Sign, 31 (1952) pp. 12-15, 74-76; Maxwell Hamilton, "We WOuldn't Marry Each Other Again," Catholic Digest. 19 (1955), pp. 8-11; Carl Bakal, "The Risks You Take in Interfaith Marriage," Good Housekeeping, 149 (1959), p. 62. 119. pp. 71-73. The reader is now better able to appreciate the find- ing by Alfred Prince (whose study was cited there) that Catholic college students were more willing to intermarry than Protestants, but also twice as likely to specify the condition that the chil- dren be raised in their religion. To the studies reviewed in the last chapter, two more pieces of-evidence for endogamous atti- tudes should be added. Father Harry E. Hoover made a study of the attitudes of Catholic high school students toward mixed mar- riages, drawing on nine Master's theses written earlier at Catho- lic University on the same subject. His analysis was based on 4,000 questionaires returned from eleven Catholic high schools in eight dioceses. His conclusions: Girls are more opposed to mixed marriages than the boys are; senior boys are more opposed than the 110 When he matched the seniors on seven other three classes. variables and correlated their attitude scores (regarding mixed marriages) with an eighth factor, "the following significant results" were obtained: Those were most opposed to mixed mar- riages--a)whose homes were most Catholic (an index based on ten behavioral items), b) who had above average I. Q.'s, c) who were high achievers in school.111 Another indication of the existence of the norm of re- ligious endogamy comes from one of the Detroit area surveys being used for the present study. Lenski asked his sample of Detroiters whether they thought it wiser for members of their group to marry 110Attitudes of High-School Students toward Mixed Marriages, Washington: Catholic University Press, 1950, p. 51. 111Ibid., p. 58. 120. within the group. "The overwhelming majority of white respondents said that it 33; wiser. This was true of 92 per cent of the Jews, 81 per cent of the Catholics, and 75 per cent of the Protese tants."112 This brings to a close our evidence for the existence of a norm of religious endogamy in the United States. The historical development of the norm is continuous with its current status, al- though there has been a mollifying of the sanctions involved. Recent statements of the norm by the churches and religious ad- visers have been clear and frequent. The endogamous norm is al- so reflected and shared, at least to some extent, by the general population. It has not been our intention here to try to show just how strong or universal the norm is, but rather that it is a viable element of the current American scene. Conflicting Values in Selectingythe Marriage Partner Now let us consider very briefly what conflicts may be involved in choosing a marriage partner. Here we find the stuff of which great dramas and novels are made: the impulsive individ- ual struggling with his own ideal self-image and with the many- voiced social structures around him.and Lg him. Any treatment of this subject will seem too brief because it will always be outdone by the complexity of human motivation and especially of modern society. MacIver and Page, whose treatment of this subject is excellent, state that there are two main types of conflict between 112Gerhard Lenski, The Religious Factor: A Sociological Study of Religion's Impact on Politics, Econpmgcs, and Family Life, Garden City, N. Y.: Doubleday, 1961, pp. 48-49. 121. the individual and the code: (1) that in which personal interest or personal valuation is opposed to a prevailing code, and (2) that in which the indi- vidual is pulled opposite ways by the prescriptions of dif- ferent codes when two or more are applicable within the same situation.113 Most of what is said in this section of the chapter will be relat- ed to the second conflict, always understanding that the choice made in terms of this conflict may be a kind of rationalization to resolve the first conflict. At the end we shall consider ex- plicitly some aspects of the-self-collectivity dilemma. Two concepts have been developed in social psychology which are relevant to the momentous choice we are considering: reference group and marginal man. George Mead's "generalized other" and Charles Cooley's "social self" are usually given credit as forerunners to the reference group concept. However, Georg Simmel's "web of group-affiliations" was at least as perceptive in this matter.114 The actual term "reference group" was coined by Herbert Hyman in 1942, stimulated by Roper's opinion polls. After this, interest in the concept received new impetus from Newcomb's well-known Bennington study, which contrasted the rel- ative influence of campus ypg home environment in shaping pOliti- cal ideology. What is meant by "reference group"? One dictionary de- fines it as: "any group with which a person identifies and/or “compares himself to such an extent that he tends to adopt its 113Society: An Introdpctory Analysis. 02- Cit°9 Po 199- lléghe Web of Gropijffiliations, trans. Reinhard Bendix, Glencoe, 111.: Free Press, 1955. 122. standards, attitudes, and behaviors as his own."115 The term has been used loosely, and will be so used here, to include member- ship and non-membership groups, collectivities, social categories, and individuals. Robert Merton, relating this concept to social stratifica- tion, proposes the general hypothesis that some similarity in status attributes between the individual and the reference group must be perceived or imagined, in order for the comparison to occur at all. Once this minimal similarity obtains, other similarities and differences per- tinent to the situation, will provide the context for shapin evaluations....The underlying similarities of status among members of in-groups, singled out by Mead as the social con- text, thus appear as only one special, though obviously impor- tant, basis for the selection of reference groups. Out-groups may also involve some simdlarity of status. 15 The status-striving which he implies here certainly operates in much of social interaction, and mate selection is no exception, as we observed in Chapter 11. This striving, together with the need to have an identity, to belong, may, as he says, be explained at least partly in terms of the concept of reference groups. Hence we must not overlook the concept's potential for interpreting inter- faith marriages. The "marginal man" is, in Merton's paradigm, just one cat- egory of persons in reference group theory, viz. those who aspire to membership in a group but are regarded as ineligible by the group itself.117 The concept of marginal man, however, antedates 115R. B. and Ava C. English, A Comprehensive Dictionary of Peychological and Psychoanalytic Terms, New York: Longmans, Green, 1958, p. 232. 116Social Theory and Social Spgppture, op. cit., pp. 242-243. 117Ibid., p. 290. 123. that of reference group. Robert E. Park invented the term, and Everett V. Stonequist ensured its fame by writing a book about it. The latter's definition of it is: The individual who through migration, education, marriage, or some other influence leaves one social group or culture with- out making a satisfactory adjustment to another finds himself on the margin of each but a member of neither. He is a "mar- ginal man."118 The concept has often been applied to the offspring of intermarriages of every type, but it is mentioned here because the writer believes that it sometimes characterizes the partici- pants in religious intermarriage. For instance, the interfaith marriage choice of an upwardly mobile individual or of a second generation American may be influenced by this socio-psychOlogical condition. In any case, its relationship to conflicting values and reference groups, and its similarity to anomia should be evi- dent. The social changes which have taken place in recent cen- turies easily breed "marginal men" and give rise to value conflicts which would not have to be faced in a simpler, more homogeneous culture. Urbanization, which began about five thousand years ago and has become dominant in our country only in the last couple of generations, has brought with it much heterogeneity in a dense population and a large degree of anomymity at least in the central city. Industrialization adds its impetus to value confliCts by the spatial and occupational mobility which it fosters. These may be considered the chief historical forces which have been estab- 118The Magginal Man: A Study in Personality and Culture Conflict, New York: Charles Scribner's Sons, 1937, p. 3. 124. lished the conditions for more frequent value conflicts, whether in the choice of a marriage partner or in any other important decision. If urbanization and industrialization sum up the chief conditions fostering value conflicts in modern societies, then the chief yplpg orientations which sum up the opposition to religious endogamy would seem to be secularism and individualism. Let us consider each of them in turn. Lenski shows some concern with secularism, but his mean- ing is somewhat different from that which we will give to the term. He says: In short, norms of tolerance and secularism inevitably arise in urban centers. Norms peculiar to one religious group or another are de-emphasized, leaving a common core of moral norms which are shared by all the various faiths represented in the community. Then he adds in a footnote: '"Secularism is used here in the sense of religious neutralism."119 The "common core of moral norms" seems to assume the "common values" of a society, which are the most basic aspect of the moral order in Robert Angell's discussion of moral integration.120 This use of the concept "secularism" seems to make it a characteristic of society, a kind of moral "common denominator" in which religious differences in the popula- tion are submerged. 1n the present study the term will be used rather as a value orientation for the individual member of a so- ciety, in which this-world considerations take precedence over the 119The Religious Factor, OP- Cit-s P- 9° 120Free Society and Moral Crisis, Ann Arbor: University of Michigan Press, 1958, pp. 16 ff. 125. requirements of a faith in an other-world goal.121 Ours is there- fore closer to the meaning which Paul Landis gives to the term, although we shall not follow him in the rest of his analysis: In the long view of history, two prevailing concep- tions of authority are seen: (1) the sacred, and (2) the secular. The trend away from the mystical interpretation of life toward the rational interpretation of it is known as secularization. Will Herberg has discussed at length and with remarkable insight the problem of secularism in American religion itself. "Under the influence of the American environment," he says, "the historic Jewish and Christian faiths have tended to become secular- ized in the sense of becoming integrated as parts within a larger whole defined by the American Way of Live." Citing Robin Williams' observation that religion in the United States tends not to be re- garded as an ultimate value but as a good because it is useful in furthering other major values, reversing the former endsemeans re- lationship, Herberg goes on: Insofar as any reference is made to the God in whom.a11 Ameri- cans "believe" and of whom the "Official" religions Speak it is primarily as sanction and underpinning for the supreme 121Even if we were to focus on secularism as a characteris- tic of a society, it would not be in terms of Lenski's "religious neutralism" but in the sense employed by J. Anthony Samenfink: "At the present writing the word 'secular' is used to denote a so- " ciety whose morality is 'based solely upon regard to the well-being of mankind in the present life, to the exclusion of all considera- tions drawn from belief in God or in a future state.'" In "A StUdY of Some Aspects of Marital Behavior as Related to Religious Con- trol," Marriage and Family Living, 20 (1958), p. 164. 122Social Control, Op.cit., p. 14. For a criticism of this "commitment to a belief in 'the tOtal historical process,'" see Christopher B. Becker's introduction to his father's posthumous Presidential address to the American Sociological Association, Howard Becker, "Normative Reactions to Normlessness." op. cit., p. 804. 126. values of the faith embodied in the American way of Life. Secularization of religion could hardly go further.123 Later he contrasts European with American secularism. The former is militant, generates anti-religious sentiments; the latter is covert, makes anti-religion almost impossible because it would have to be expressed by religious people. Herberg concludes: The witness to authentic Jewish-Christian faith may well prove much more difficult under these conditions than when faith has to contend with overt and avowed unbelief. The spirit of secularism has always been pervasive and powerful and has always had its effect on religious institu- tions. The unique feature of the present religious situation in America is that this secularism is being generated out of the very same conditions that are, in part at least, making for the contemporary religious revival. The sociological fac- tors that underlie the new urge to religious identification‘ and affiliation are also factors that enhance the seculariza- tion of the religiousness they engender. It is not secularism as such that is characteristic of the present religious situation in this country but secularism within a religious - framewbrk, the secularism of a religious people.124 A number of writers have noted also the secularization of marriage and family life, a tendency which makes the norm of reli- gious endogamy less meaningful. Elliott and Merrill approach the matter from the viewpoint of social disorganization. After stat- ing that the "secularism of the modern world has fostered a corre- sponding secular attitude toward marriage," they observe that the Catholic Church "mobilizes every element in its powerful institu- tional machinery toward the continuance of the religious concep- tion of marriage." However, the sacramental nature of marriage was not tenable to the Protestant reformers, even though they "generally maintained the essential sanctity of the monogamic 123Protestant - Catholic - Jew: An Essay in American Reli- ‘gious,Sociology, New York: Doubleday, 1955, p. 96. 1241b1d., p. 288. 127. relationship." The Protestant Episcopal Church long held a firm line against divorce, "but the general increase in divorce among non-Catholic groups eventuated in the relaxation of the Episcopa- lian Divorce Canon in 1946, by allowing greater latitude to the Bishop in deciding whether remarriage is possible."125 Bossard and Boll studied data concerning rituals in more than four hundred families, data from various sources and covering a span of eighty years. The first trend which they say "may be clearly identified" for these years is the trend "from predomib nance of the secular."126 The same authors, in their book on interfaith marriages, state the following as the third type (out of four) of lay reaction to church opposition to these marriages: that religion and the church are no longer important--at least, not important enough to be considered in so personal and seri- ous a matter as choice of a lifelong mate. This is a secular age, an age of change, dominated by science, we are told.127 Is secularism, as we are using the term, now more charac- teristic of one religious group than another? Only two studies attempting to answer this question have come to the attention of this writer. Samenfink's study was based on a sample at a state college in Louisiana. After matching the subjects on five vari- ables, he concluded: When contrasted with the equivalent findings for a matched group of Protestants, however, the Roman Catholic respondents were less accepting of the secular position, with 125Mabel A. Elliot and Francis E. Merrill, Social:Dis- porganization, 3rd ed., New York: Harper, 1950, pp. 357-358. 126James H. Bossard and Eleanor Boll, Ritual in Family Living, Philadelphia: University of Pennsylvania Press, 1950, p. 22. 127One Marriage, Two Faiths, op. cit., p. 91. 128. one exception. In the matter of equalitarianism in family living, the Roman Catholic position was more secular than was that of the Protestant respondents.128 Another study compared the attitudes of college students belonging to the three major religious groups (N - 156). In answer to the question, "To whom do you owe your greatest loyalty?" 74 per cent of Catholics said "God" and 24 per cent said "Family;" for Prot- estants, the percentages were 50 and 39; for Jews, 11 and 74, respectively--these two responses being the most frequent for each group. When asked "What is most important in life?" the first and second choices of Catholics were these: "Live in accord with re- ligious beliefs," and "Make world a better place." For both Prot- estants and Jews, the first and second choices were, respectively: "Make world a better place" and "Own happiness."129 Behavior may be far different from expressed beliefs, of course. If these differences hold up in other studies, they will be consistent with the finding that Catholics in this country are somewhat less successful in the competitive struggle for this world's goods than are the other two religious groups, since these differences imply that they may not be so highly motivated to suc- ceed in this matter. Lenski, for instance, cites Neil J. Weller's doctoral dissertation to this effect and from his own data con- cerning attitudes toward work, after controlling for class posi- tion states this conclusion (in italics): 128"A Study of Some Aspects of Marital Behavior as Related to Religious Control," op. cit., p. 169. 129Joseph Maier and William.Spinrad, "Comparison of Reli- gious Beliefs and Practices of Jewish, Catholic, and Protestant Students," Phylon, 18 (1958), pp. 356-358 129. To sum up, it appears that Protestantism is conducive to more positive attitudes toward those positions in society which are more demanding (and also more rewarding), while Catholicism is conducive to more positive attitudes toward the less demand- ing (and hence less rewarding) position.130. In a similar vein, he observes that Catholic-Protestant differences in fertility rates (Detroit, 1952-59) have been greater in the middle class than.in the working class, which he suggests "could well be a factor contributing to our earlier finding that differs ences in rates of mobility are greater in the middle class than in the working class."131 Large families do not make for upward mo- bility.in an urban society. 4 Z‘Albert J. Mayer and Harry Sharp, using data from the Detroit Area Study from.l954 through 1959, apply a "handicapping" system (based on rural, foreign, and specific Detroit background) to the members of religious denominations and conclude that Catholics rank lowest in achievement of worldly success of all denominations in both white and Negro populations.”2 In Iowa the tendency seems to be in the same direction, though not so pronounced except in professional occupations, when comparing "church-Protestants" with Catholics. Other Protestants (ig_gg, those who do not specify a denomination) and those of no religion had lower occupational status levels than Catholics.133 130The Religious Factor, op. cit., p. 86. 131Ibid., p. 219. 132"Religious Preference and WOrldly Success," American Sociological Review, 27 (1962), p. 226. 133Lee G. Burchinal and William.F. Kenkel,'"Religious Identification and Occupational Status of Grooms," American Socio- logical Review, 27 (1962), p. 531. 130. It seems to this writer that the surprising results report- ed after a thematic apperceptive measure of achievement motivation was administered to.a national sample of 1,620 adults also have relevance for our discussion of secularism as a value orientation. The authors of this study, contrary to expectation, found that Catholic g Achievement scores were higher than those of Protes- tants in middle age, even with several controls.134 They reject the contrary findings of B. C. Rosen and of D. C. McClelland gg. ‘gl. because they were based on quota samples (tending to overselect higher status groups) and were samples of the population in the Northeast, which Veroff et. al. found to be the only region in the national study in which Protestants (in the high income group) are clearly stronger than Catholics inqg Achievement scores. They then looked more closely at the middle-age group and found that Catholics with larger families have higher n Achievement scores than those with low families, true at each income level. Income was independently and negatively related to these scores for Cath- olics. For Protestants there was a slight positive relationship between income and g Achievement scores, but no association with the number of children. The authors of this article seem to arrive at more precise statements than are warranted by the size of their Catholic sample, but our purpose now is not to review their long analysis. Rather, let us note that these findings fit our secularism theory. Catho- lics do not seem to be particularly motivated toward success in .134Joseph Veroff, Sheila Feld, and Gerald Gurin, Achieve- ment Motivation and Religious Background," American Sociolo ical Review, 27 (1962), pp. 205-217. 131. this world as such. Their other-world orientation generally re- quires them to look for a reason for more-than-average striving for this.world's goods, such as family duties or special economic stress. If these differences represent variant degreds of secular- ism as a value-orientation on the part of Protestants and Catho- lics, we need not see this simply as a choice between this world and the future world. It may be, instead, a perceived difference in relationship between success in the two worlds, as Mayer and Sharp remind us: The powerfully reinforced and traditional Roman Catholic Church tends to orient its members toward the hereafter; successful performance in the market place and the acquisition of the symbols of economic achievement age of relatively little impor- tance as an indication of the Catholic's status after death. On the other hand, adherents of Protestantism are assumed to be highly concerned with worldly success and the attainment of material possession, status, and the prestige that is assoc- iated with upward social mobility. These things often are viewed as indications that salvation is assured, or at least is more probable.135 , What does all of this have to do with religious endogamy? *1 Our discussion has been long enough to obscure the point. To the extent that secularism represents an orientation toward this world in opposition to orientation toward a future-life, it is a value orientation which will often conflict with adhering to the norm of religious endogamy. If Protestants, in this country at least, are more influenced by the values of secularism than are Catholics, as some studies just cited seem to suggest, then we should expect that 135"Religious Preference and Wbrldly Success," op. cit., p. 218. See Chapter II, pp. 51-58, above, for the ways in which striving for status shows itself in mate selection in general. 132- Protestants are less likely than Catholics to adhere to the norm of religious endogamy--other things being equal. Now let us turn to the other major value orientation which is in conflict with religious endogamy, viz..individualism. Paul Landis has stated the problem in its broadest terms: To what extent the personality of the individual shall be submerged into the social unit is a key problem of all so- cial orders, and of every social institution. The question involves not only the degree of uniformity and variation in behavior, but also that of the extent to which the-individual will seek out his own interests or that of the group.135 The rapid rate of social change, as expressed in the forces of urbanization and industrialization, has certainly contributed to the importance-of the individual 1;. the group in recent genera- tions. These forces were accentuated by the fusion of cultures that accompanied mass immigration into the United States. After describing the stable peasant culture of Poland, Thomas and ' ‘Znaniecki turn to various forms of adaptation to social changes which have entered peasant life. One adaptation, they say: gives birth to a particular kind of individualization, found among the bulk of young immigrants of both sexes in America and among many season-immigrants in Germany....In both America and Germany this is due, in the first place, to the higher wages, but in democratic America the Polish social life gives the immigrant a feeling of importance which in Polish communal life is the privilege of a few influential farmersv.v.Former1y the individual counted mainly as a member of a family; now hev counts by himself, and still more than formerly. The family ceases to be necessary at all.137 That this growing individualism will manifest itself in 135$ocial Control, op. cit., p. 898. 137William I. Thomas and Florian Znaniecki, MiG—h, Peasant in Europe and America: Mpnograph of an Immigrant Group, Vol. I: Primary-Group Organization, Chicago: University of Chicago Press, 1918, pp. 101-102. 133. the choice of a marriage partner is to be expected. Hertzler is one of many who have taken note of this change: Marriage is more and more a matter of individual choice of mate. It is usually regarded as an arrangement between the two.individuals immediately concerned rather than between the two families or even kinship groups as was to a considerable extent the case a century ago. Whom a person marries is con- sidered to be a maqur of personal concern....Marriage, instead of serving group ends, is aimed at immediate personal satisfac- tions, sexual gratification, affection, and companionship.138 As every American movie-goer knows, romantic love-is the only reason for marrying. This emphasis on personal attraction was probably an inevitable development to fill the vacuum left by the decline of the earlier system of arranged marriages, which itself could not persist in the absence-of a land-based stable community with a clear-cut status system and kinship ties. How young people are still emotionally dependent on their parental family when they are urged to undertake the responsibilities of an independent house- hold, yet have no clearly defined mate-selection process to make the transition smoother for them, Hence-we should not be surpris- ed that, as Robin Williams observed, "An almost compulsive empha- sis upon romantic 1ove emerges in part from this situation."139 When it comes to interfaith marriages, the story is the same, as Bish0p Pike states it: "But no amount of reasoning, no number of case histories necessarily downs the aphorism so typical of American culture: 'Love will find a way.'"140 138American Social Institutions, op. cit., p. 241. For a similar statement, see James H. Tufts, America's Social_Morality, New York: Henry Holt, 1953, p. 94. 139American Society, op. cit., p. 74. 140If on Marr Outside Your Faith, op. cit., p. 63. 134. To the extent that romantic love is stressed, there is probably little chance for a young couple to discuss the practical aspects of marriage. However, many sociologists and others have taken the position that American courtships are not quite as quick and thoughtless as Hollywood would have us think.141 Companion- ship in marriage seems to be the norm.most often expressed. This may take any number of forms in actual married life, as Robert Blood's study of married couples in the Detroit area showed. One of the bases for companionship is, of course, religious homogamy. That study found that the greater the difference in religious back- grounds, the less was the wife's marital satisfaction.142 In one sense, therefore, the strain toward companionship in marriage should contribute toward observance of the norm of religious en- dogamy, rather than be a force in conflict.with it as is romantic love. However, in another sense, it may manifest the secularism of religion that Herberg was writing about. This is suggested, perhaps, by Lenski's finding that "one fifth of the homogeneous marriages had been contracted by persons raised in different faiths." Moreover, conversion of one partner or the other was more frequently found among third-generation than among first- and second-generation respondents.143 Americanization seems to pro- mote religious homogenization. Herberg's point is validated in 141One of the most consistent expressions of this theme is found in Ernest Burgess and Harvey Locke, The Family: From Insti- tution to Companionship, 2nd ed., New York: American Book Co., 1960. 142Robert 0. Blood, Jr., and Donald M. Wolfe, Husbands and Wifes: The Dynamics of Married Living, Glencoe, 111.; Free Press, 1960, p. 257. 143The Religious Factor, op. cit., p. 49. 135. Lenski's finding only to the extent that these conversions were motivated by the desire to have religion serve as a means toward marital companionship rather than as an end in itself. This brings to an end our discussion of conflicting values in selecting the marriage partner. We have seen that the two con- cepts, reference group and marginal man, have some potential for explaining the operation of these conflicting values. Urbaniza- tion and industrialization have been given credit for providing the conditions which foster value conflicts in American society. The value orientations which are in conflict with the norm of re- ligious endogamy were summed up in the forces of secularism and individualism, the latter being manifested first in romantic love and then in the norm of companionship in marriage. Now let us move on to the agents of social control which find a place in the hypotheses of this study. Parents as Agents of Social Control We cannot simply assume, of course, that all parents will want to, or are in a position to, support a norm of religious en- dogamy. Many parents lack any absolute standards in morals and re- ligion. Consequently, they often present several alternatives, "expecting the child to choose his own course, a thing which no child is capable of doing," as Paul Landis sees it.144 It may be, as Melvin L. Kohn believes, that middle-class parents value inter- nalized standards of conduct, "acting according to the dictates of 144Social Control, Op- Cit-: P- 200' 136. one's own principles"--whatever they may be.145 For Catholic par- ents also, it is problematic how much responsibility they accept themselves for the religious training of their children. Father John L. Thomas found that, in a large sample of beginners in the first grade of Catholic schools, any one of the five most common prayers and the five most common doctrines had been taught at home only to about three out of ten of the children in the Mid- west.146 However, there is evidence that children accept their par- ents as reference persons in religious matters. Bernard C. Rosen found that this was true, and also that in religious matters there was generally close agreement between the parents' attitudes toward beliefs, practices, and behavior--and their children's attitudes toward these beliefs, practices, and behavior.147 Even when the peer-group was presented as an alternative reference for decisions to be made, teenage girls were found to prefer the views of their parents on such questions as, "Which one of two boys to date.148 For a thousand college students in Atlantic and Southern states, Putney and Middleton arrived at the following conclusions regard- ing parent-child agreement on religion: 145"Social Class and Parental Authority," American Socio- logical Review, 24 (1959), p. 352. 146"Religious Training in the Roman Catholic Family," gaggi- can Journal of Sociology, 57 (1951), p. 182. 147"The Reference Group Approach to the Parental Factor in ttitude and Behavior Formation," Social Forces, 34 (1955), pp. 137-144. 148Clay V. Brittain, "Adolescent Choices and Parent-Peer Cross Pressures," American Sociological Review, 28 (1963), p. 388. 137. Three distinct types of relationships were found between the convictions of the students and those they attributed to their parents: the students tended to conform to a religious ideol- ogy held jointly by their parents; when the parents disagreed with each other the students inclined toward the parent who was closest to the modal position; and the students also in- clined toward the position of the mother when the parents dis— agreed.149 If parents are to be effective agents of social control regarding the norm of religious endogamy, they must have some re- ligious sentiments as well as maintaining a parent-child relation- ship which is conducive to their exerting influence. It has some- times been suggested that high religiosity in the parents is likely to lead to rebellion in the children. Judson T. Landis questioned 904 male and 1,750 female unmarried students in family sociology courses at eleven colleges across the nation and arrived at several conclusions relevant to these ideas: 1) There was a positive assoc- iation between family religiousness and "success in family living." 2)‘Re1igiousness of parents is positively associated with the child's reported feeling of closeness to mother and father." 3) A devout family is characterized by "the 50-50 pattern of domi- nance in the home as reported by these students, while indifference to religion is associated with the extremes of a fatherwipg mother- dominated home."150 John E. Mayer has recently published a book called, 149Snell Putney and Russell Middleton, "Rebellion, Conform- ity, and Parental Religious Ideologies," Sociomentry, 24 (1961), p. 135. 150"Religiousness, Family Relationships, and Family Values in Portestant, Catholic, and Jewish Families," Marriage and Family Living, 22 (1960), pp. 341-347. '- 138. Jewish-Gentile Courtships: An Exploratory Study of a Social Pro- gppp, which is very pertinent to this part of our subject.151 He sought out 45 Jewish-Gentile married (or engaged) couples and inter- viewed 86 of these individuals, recently married and almost all for the first time. On the basis of these interviews, he classified 29 individuals (in 25 couples) as having been "Reluctants" toward intermarriage before meeting their mates. Mayer found that a large majority of the parents were opposed to the marriage, many bitterly, and that this bothered the children. They gave numerous reasons for their opposition, mostly on religious and/or cultural grounds. However, at first many "merely protested" and did not intervene, thinking that "nothing serious would develop."152 Parents were thought to have been restrained by the cultural norm to leave mate selection to the individual. One-third of the intermarried felt that their par- ents had expressed less opposition than they actually felt. How- ever, after verbal objections failed, many of the parents tried in a variety of other ways to separate the couple. Yet, in only about half of the cases did the parents‘who objected tell the ppigpg of their opposition. Since these 45 are married couples, obviously the opposi- tion of the parents was ineffective for them. Why? These are some of the reasons, as Mayer interpreted the interviews: The image the parents had of what was happening often lagged behind 131New York: The Free Press of Glencoe, 1961. See especially Chapters 7, 8, and 9. 152Ibid., pp. 135~136. 139. the actual course of events, in many cases because the child did not inform them. The child sometimes avoided the parents or hastened the marriage to forestall opposition. Often the off- spring felt they were morally right in insisting on a free choice and that religious prejudice was wrong. Often they condemned their parents' motivation ("selfishness," etc.), and some felt that parents would eventually withdraw opposition. Even in this small sample we can see how complex are the motivations~and the relationships between parents.and children in the matter of mate selection. As we have seen earlier, Jews are likely to be more opposed to intermarriage than others in this country. Yet, even those parents who were strongly opposed and whose children professed affection for them did not in these cases succeed in preventing the intermarriage. We are not going into the motivation of those who intermarried and the stages in the process, but, insofar as the arguments of the parents were met openly, it was in terms of the moral right to a free choice and the principle that religious prejudice is wrong—-principles which may have seemed irrelevant here in the view of the parents. The reader may wish at this point to refer again to the first pages of this chapter, in which the findings of Jerold Heiss and others regarding parents, family, and ethnic group as agents of social control in the matter of interfaith marriages are pre- sented. In general, Heiss' interpretation of the Manhattan data was that the parents of the intermarried either were not interested in supporting a norm of religious endogamy or their early relation— ship with their children was such as to forestall their effective- 140. ness as agents of social control. From several other sources we find that the attitudes of parents are more opposed to interfaith marriages than those of their children. Ray Baber's sample of parents gave most importance to these characteristics in their child's spouse: health, same religious faith, moral standards, disposition and personality. Their children rated these third, fourth, fifth, and first, re- spectively.153 In Virginia Staudt's college sample, 49.2 per cent of males, 54.2 per cent of females reported that their family would oppose interfaith marriage.154 Alfred J. Prince discovered that at the University of Wisconsin parents wanted their sons to rate "same religious faith" figpp and their daughters to rate it second, whereas these actually rated it eleventh and twelfth, re- spectively.155 Does this generational difference in attitudes indicate the value of experience, the conservatism of advancing age, or a liberalizing trend? At this stage in our knowledge, it is not possible to say. Family and Relatives as Agents of Social Control The parents were singled out as being of special importance in our subject matter. In this section we shall consider the family in general, including siblings and less closely related 153"Some Mate Selection Standards of College Students and Their Parents," Journal of Social Hygiene, 22 (1936), p.-124. 154"Attitudes of College Students toward Marriage and Related Problems. II. Age, Educational, Familial, and Economic Factors in Marriage," Journal of Psychology, 34 (1952), p. 100. 155"Factors in Mate Selection," Family Life Coordinator, 10 (1961), p. 57. 141. kin. Of course, what is said here cannot be applied without modification to all of the many types of family in the United States. Families at different socio-economic levels, or classes, have different patterns of family interaction.156 There are re- gional differences, rural-urban differences; and national patterns tend to persist in the ethnic groups of this country.157 When the word "family" is used, Americans almost always think of the immedi- ate, the nuclear famdly.. Although thought to be increasingly important, this unit is also regarded as a weakened agent of so- cial control. Robert C. Angell describes its plight in the-com- munity: Although the various.reasons given for the failure of the family as an agency of moralization are fairly independent, they often relate to the point previously discussed: the instability of the surrounding community. Because modern urban life is not congenial to the large kinship unit so com- mon in primitive and peasant societies, the conjugal family has become the prevailing form. Because conjugal families start anew in each generation, and because they are small so- cial units, they tend to take the impression of the environ- ing conditions more easily than would a large family composed of several generations.158 Robin Williams-relates this situation to American individ- ualism in the choice of a marriage partner: ’"The-individualistic 156Strong endogamous tendencies in the "established fam- ilies" are especially noted by August B. Hollingshead, "Class Differences in Family Stability," Class, Status and Power: A Reader in Social Stratification, eds. Reinhard Bendix and Seymour M. Lipset, Glencoe, I11.: Free Press, 1953, pp. 284-292. Class differences in religious and other family rituals are described by James H. Bossard and Eleanor Boll, Ritual in Family Living, op. cit. 157The American Journal of Sociology devoted an entire issue in Vol. 53 (May, 1948) to these variations. See also Norman D. Hhmphreys, "The Changing Structure of the Detroit Mexican Fame ily: An Index of Acculturation," gaggican Sociological Review, 9 (1944), pp. 622-626. 158Free Society and Moral Crisis, op. cit., p. 72. 142. mate-choice in America is thus partly accounted for by the iso- lated conjugal unit, the discontinuity of generations, the devalu- ation of kinship, and the extensive social mobility found in our society."159 Is the nuclear family an isolated unit in urbanized Ameri- ca? There is still a lively discussion in sociological circles regarding the extent of this social change. Certainly no one believes that we still generally have a strong extended kinship group such as is described in The Polish ngppnt in Europe and America. However, Eugene Litwak presents evidence to show that a "modified" extended family still exists in urban communities, even without the former geographical propinquity, and he defines it as: "a series of nuclear families joined together on an equal- itarian basis for mutual aid."160 For instance, he says, "There are three independent studies which indicate that in large urban centers (Buffalo, Detroit, Los Angeles)...a1most 50 per cent of the muddle class individuals saw relatives at least once a week. or more."l61 Some others who agree with Litwak are the following: Harry Sharp and Morris Axelrod found mutual aid among relatives to be conmon in Detroit, especially between parents and children 159American Society, op. cit., p. 49. 160"The Use of Extended Family Groups in the Achievement of Social Goals: Some Policy Implications," Social Problems, 7 (1959-60), p. 178. See also his "Geographic Mobility and Extended gamily Cohesions," Apprican Sociological Rgview, 25 (1960), pp. 85-394. ' 1611bid., p. 179. 143. and sibilings.152 Studies in Chicago and Cleveland revealed that a high percentage of people had relatives living in the same metro- politan area, many within a few blocks.163 Prominent in these two samples were migrants from the lowlands of Tennessee and Kentucky, who maintained the-rural and smalldtown visiting patterns in their new urban setting. What do these studies have to do with inter- faith marriages? They indicate, at least, that close kinship ties have still (or have regained) some importance in the interpersonal relations or urbanities. If visiting and mutual aid take place along these lines, then we are safe in assuming that these rela- tives urge conformity to group norms in matters which are important to these types of interaction. Religious endogamy is probably one aof these norms. The Church as an Agent of Social Control As with each of the sections in this chapter, one could write a book around this subject alone--and some people have. However, we shall try to limit our treatment of it by using only sociological sources and focusing on aspects somehow related to 162"Mutual Aid Among Relatives.in an Urban Population," in Ronald Freedman, et al., Principles of Sociology: A Text with Baggingp, rev., New York: Henry Holt,.1956, pp. 433-439. 163Leonard Blumberg and Robert R. Bell, "Urban Migration and Kinship Ties," Socigl Problgmp, 6 (1959), pp. 328d333. Also IMarvin B. Sussman, "The Isolated Nuclear Family: Fact or Fiction," Social Problems, 6 (1959), pp. 333-340. 144. religious endogamy.164 Religion certainly does not occupy the central position in the United States now that it did in the Middle Ages of Europe or in Colonial America. Before world War II, many would have agreed with Robert 0. Angell's statement that "the church in con- temporary America is not so much.rejected asignored."165 Never- theless, since that time and even over a longer span of years, church membership in this countryihas been growing almost twice as fast as the general population: During these fifty years the estimated population~in- creased almost 100 per cent, while the officially reported church membership of all bodies in l956-—100,162,529-;repre- sented a gain of about 190 per cent over the 1906 figure of 35,068,058, which is the corrected figure published in the 191678ensus. In 1906, 186.denominations reported; in 1956, 258 reported.166 As.we saw in our discussion of secularism, the bare figures for growth in church membership may be deceptive. The seculariza- tion of religion itself in this country was one of Herberg's prin- cipal observations, and Paul Landis discussed the same point in 164For a good historical sketch of the-churches in America, see Will Herberg, Protestant - Catholic - Jew, op. cit., Chapters 6-8. For the current scene and trends, see "Religion in American Society," Annals of the American Acadgpy of Political and §ocial Science, 332 (November, 1960), the entire issue. On religious intergroup relations, see Charles F. Marden, Minorities in Ameri- can Society,-New York: American Book Co., 1952,-Chapters 14‘15. Also: Leo Pfeffer, "Issues That Divide," Journal of Socigl Igsues, 12~(l956), pp. 21-39; Robin M. Williams, Jr., "Religion, Value- Orientations, and Intergroup Conflict," Journal of Social Issues, 12 (1956), pp. 12-20. ‘165The Integration of Aggrican Socigty: A Study of Groups and Ipstitutiopg, New York: McGraw-Hill, 1941, p. 164. 166Benson Y. Landis, "Trends in Church Membership in the United States," Appals of the épgrican Academy of Political 5nd So- cial Science, 332 (1960), p. 5. 145. terms of a continuum from eternity-centered to moralitywror society- centered churches. In addition, Landis observes that "secular institutiéns have encroached upon what was once church territory."167 This is true at least in the spheres of social service, of educa- tion, and of family living and recreation. Under such circumstances, one may question whether reli- gion has any influence on other aspects of life. Fortunately, Gerhard Lenski's book, The Religious Factor, was directed at ant swering precisely this question. Most of the associations he dis- cusses will not concern us; but when these-correlates represent moral norme, then we have reason to interpret the relationship as indicating the church as an agent of social control. For instance, when he concludes that religion is as good a predictor of attitudes as is social class on 35 items, the only ones which clearly relate to social control are those which state a moral judgment, such as: "Gambling wrong," "Divorce wrong," ”Birth control wrong," "Moderate drinking wrong," "Sunday business wrong." On these matters, he found religion to be a much better predictor of attitudes.than was social class.168 A key distinction in Lenski's analysis is that between re- ligious group as community and as association. This is an appli- cation of T3nnies' Gemeinschaft und Gesellachaf . He combines these two aspects in the term, "socio-religious groups," and is able to show a different profile for each major religious group on these two dimensions. As might be expected, one index of this 167Social Control, op. cit., p. 221. 158The Religious Factor, op. cit., Table 50, pp. 293-294. 146. "community" dimension is the group's degree of adherence to reli- gious endogamy. One point he wishes to stress by making this distinction is that "these subcultures ought never be interpreted as merely by-products of, or derivations from, theology."169 Lenski's predictions for the future, based on his findings in Detroit, are that associational vitality in these groups will increase, and that differences in the characteristics now associ- ated with belonging to a specific socio-religious group are likely to increase.170 This seems to say that the church is pp; declin- ing as an agency of social control. Let us now consider some aspects of the Protestant and Catholic churches separately. First we should note that Protest tantism is not a homogeneous entity. For instance, in a variety of small communities and controlling for social class, "higher participation patterns existed among members of Continental Europ- ean origin churches than among Anglo-American origin churches."171 Using data from Indianapolis, Lenski found religious interest to be more characteristic of denominations which stress doctrine and the religious education of the young.172 In a small town in up- state New York, in which there were four Protestant churches, only the Baptists tended to be endogamous. Episcopalians, Congregation- 1691bid., p. 19. 1701bid., pp. 288-291. 171Victor Obenhaus, W. Widick Schroeder, and Charles D. England, "Church Participation Related to Social Class and Type of Center," Rural Sociology, 23 (1958), p. 298. .172Gerhard E. Lenski, "Social Correlates of Religious Interest," American Sociological Review, 18 (1953), p. 435-43. 147. alists, and Methodists.intermarried freely, with the wife affili- ating with the church of her husband.173 Urbanization is of special concern to Protestant churches, since their patterns of social control in this country have been -largely those of rural communities.174 Even the old urban churches must adjust.to-the movement to the'suburbs. Consequently, studies of the urban church and its prospects for the future are quite common.175 Another point that affects Protestant churches as agents of social control is presented in Gerhard Lenski's Detroit study. White Protestant churches seem to weaken kinship ties, at least with the-extended family. Marginal Protestants were more likely to be involved in the kin group, but less likely to participate in avoluntary associations than were the churchgoing Protestants. In fact, Lenski sees the Protestant church, in its characteristic American form, as a training ground for participation in voluntary associations, with consequent strains.on family ties.175 If this 173Arthur-J. Vidich and Joseph Bensmnn, Small Town in Mass Society, Princeton, N. J.: Princeton university Press, 1958, p. 230., 174See Rockwell C. Smith, The Church in Our Town: A Study of the Relationship between thg Church and the Rural Community, rev., New York: Abingdon Press, 1955, especially p. 22. 175For.instance,'MZurray H. Leiffer, City and Church in Transition: AA:Study of the MediumrSized City and Its Organized figligious Life, Chicago: Willet, Clark, 1938. Also: F. Stuart Chapin, "The Protestant Church in an Urban Environment," Cities and Society: The Revised Reader in Urban Sociology, eds. Paul K. Hatt and Albert J. Reiss, Jr., Glencoe, I11.: Free Press, 1957, ,pp. 505-515; Truman B. Douglass, "Ecological Changes and the Church," Annals of the American Academy of Political and Social Science, 332 (1960), pp. 80-88. 176The Religious Factor, op. cit., pp. 219-223. 148. is so, its application to religious endogamy should be evident. On the other hand, Lenski found the opposite relationship to hold in the case of Detroit's Catholics. "A_high degree of involvement in the Catholic church was positiVely correlated with a high degree of involvement in the kin group.177 This was shown by several indices. The involvement of active Catholics with voluntary associations was mild (compared to Protestants), but this involvement was thought to be due at least partly to the Ameri- canization process. This church-family relationship seems to be illustrated in the Catholic Church's opposition to the use of contraceptives. Its considerable effectiveness in this matter shows that it can func— tion as an agent of social control. One national study found that, in spite of the majority practice to the contrary in this country, 69 per cent of Cathblics conformed to Catholic doctrine regarding~ contraception, among all couples (787) with Catholic wives--the most deviant categories being those of blue collar husbands, the wife not finishing high school, and the husband's income in the $5,000-6,999, range.178 In another study of the seven largest metropolitan areas in this country, one conclusion was that "Re- ligion is clearly a better predictor than class. For all comparé isons, the hypothesis of a rank ordering of Catholics, Protestants, 1771h1d,, p. 223. In this respect, Lenski says, the Catho- lic pattern is like a "church-type" religious organization, treat- ing the kin group as an ally rather than as a competitor in the gradual transformation of the world. 178Ronald Freedman, Pascal K. Whelpton, and Arthur A. Campbell, Family Planning, Sterility and Population Growth, New York: McGraw-Hill, 1959, p. 201. 149. and Jews in that order with preference for large families, less success in planning, and shorter intervals is confirmed."179 The differences were stronger for the white-collar than for the blue- collar class. The existence of the huge Catholic school system in this country is concrete evidence of the intent of the Catholic Church to be an agent of social control for its members. In Detroit, Lenski found that 35 per cent of Catholics received more than half of their education in Catholic schools, a drop in enrollment being notable in the Depression of the 1930's. Among those so educated were 46 per cent of Catholics with background in north-western Europe, and 28 per cent of those with background in southern and eastern Europe.180 As for the "consequences" of this experience, it was most closely associated with attitudes unfavorable toward birth control and divorce. Regarding religious endogamy, the dif- ference was small; 84 per cent of Catholics with a Catholic educa- tion being in favor of endogamy, compared with 79 per cent of other Catholics.181 Finally, a little bit can be said about the effectiveness 179Charles F. Westoff, Robert G. Potter, Jr., Philip C. Sagi, and Elliot G. Mishler, Family Growth in Mptropolitan Ameri- ca, Princeton, N. J.; Princeton University.Press, 1961, pp. 191-' 192. 180The Religious Factor, op. cit., pp. 240-241. For a general discussion of religious education, see Rolfe L. Hunt, "Religion and Education," Annals of theggggrican Academy of Polit- ical and Special Science, 332 (1960), pp. 89-100. Regarding Catholic Catholic schools, see Peter H. and Alice S. Rossi, "Some Effects of Parochial School Education in America," Daedalus, 90 (1961), pp. 300-328. 1811bid., pp. 245-246. 150. of the Catholic Church's regulations on interfaith marriages. In the study by westoff.pp_pl., "Of the total 140 Catholic individ- als married to a non-Catholic, 86 reported having been married by a priest----Marriage by a priest was almost twice-as frequent among couples where the wife was the Catholic."182 0f Catholic mixed marriages in one Canadian city, 62 per cent were pefformed by a priest.183 As for the prenuptial agreement required for a Catho- lic mixed marriage, Murray H. Leiffer reports that, "In the major- ity of the cases interviewed where such an agreement had been signed, both partners felt bound by it." However, "In many in- stances, in spite of the signed agreement, even the Catholic part- ner displays little interest in rearing the-children in the Catho- lic £aith.184 To sum up this part of the chapter, we have seen that over a span of centuries the churches have lost ground to secularism, granting that there are important differences, especially among Protestant churches, in‘this and other respects. ‘Lenski's data seemed to show Protestant churches as pulling their members away from the family toward voluntary associations, with the opposite pattern holding in Catholic families. We have looked at Catholic schools as they relate to social control; and we have found some evidence, in general, of the moderate success of the Catholic 182Family Growth in Mgtropolitan AmericaI op. cit., p. 181. 183Victor J. Traynor, "Urban and Rural Mixed Marriages," 0p. cit., p. 157. ‘ 184"Mixed Marriages and the Children," Christian Centur , 66 (1949), p. 106. 151. Church.in the-matter of birth.control practices and the require- ments for Catholic mixed marriages. Ethnic Groups as Agents of Social Contrpl As we have observed earlier, when a group is surrounded by a contrasting majority group, their own shared subculture is likely to hold them together, their national origin or language or customs. However, we may expeCt that the longer an immigrant group.resides in a host country, the more its in-group feeling will be weakened--other things being equal. Therefore we have hypothesized in this study that the rule of religious endogamy is more likely to be violated with each-succeeding generation of residence in America--until a kind of national pattern (or urban pattern) is reached for this country. When we ask what social control ethnic groups exercise in the matter of religious inter- marriage, we are assuming both that an ethnic group can function as an agent of social control and that it would be-concerned about enforcing a rule of religious endogamy.185 Let us call each of these assumptions into question now, and examine them in the first three generations. Is the ethnic group an effective agent of social control, in general? This seems to depend on the speed with which it Americanizes and merges its differences with the general population. The greater the cultural 185Related aspects treated elsewhere-in this thesis are as follows: ‘Chapter II--"Ethnic Intermarriage," "The 'Triple Melting-Pot' Hypothesis," "Religious Intermarriage Related to Nationality;" Appendix A, on immigrant groups in the Detroit Area. .152. differences--such as in race, language, religion, peasantry among turban dwellers--the more-slowly these differences will dissolve and cease to be a rallying point for in-group sentiments. The largeness and residential segregation of a group will also slow this~rate of change. To the extent that the group is distinctive and self-conscious,it will have the potential of exercising its own social control. What about the first generation as an agent of social con- trol? When the waves of immigrants arrived in this country in the 19th and early 20th century, they hoped and strove to set up life the way they knew it in the Old Country--usually in peasant vil- lages.186 The fortunate few were able to obtain land where they could grow things with the work of their hands. Most lived in crowded city dwellings and were lucky if they found work in fac- tories or mines. They learned to find their identity in the new land through common language rather than by village or district of origin. Soon there were associations to take proper care of the dead; there were schools and newspapers in their own language, and the stage. For the immigrants as a body there was quite an effective agency of social control in operation. When the immigrants tried to act as socializers or moral- izers for their children, the second generation, it was often a different story. There were varying degrees and sources of con- flict. With the Polish the conflict was sometimes "almost incred- 186For a general treatment of these personal struggles, see Oscar Handlin, The Uprooted, New York: Grosset & Dunlap, 1951. 153. ible" in its bitterness, and it centered often enough on economic disputes, with the parents insisting on maintaining the Old World practice of receiving all of the child's earnings.187 In the-case of the Italians, there was often sharp conflict between these two generations, which became heightened when the parents wanted to chaperon the courtship and do the match-making as their parents. had done.188 On the other hand, there seems to have been much less friction between first and second generation Belgians, partly due to the good favor of native Americans that resulted from their role in WOrld War I.189 However, in most national groups there was a large measure of such conflict, even though it might be postponed and diffused over another generation or two because of the group's insulation from the larger community. The father's authority was easily challenged (especially if he were out of work) because his children knew far more about American culture than he, and this left him dependent on his child in hwmiliating ways. The first generation, as an agent of social control for the second genera- tion, was at a severe disadvantage and seems often to have been ineffective. What about the internal strength of the second generation? As we saw earlier, they are considered the typical "marginal men," 187William.I. Thomas and Florian Znaniecki, The Polish Peasant in Europe and America, op. cit., p. 104. 188One America: The History4_Contributions, and Present WWW. eds. Francis J- Brown and Joseph S. Roucek, rev., New York: Prentice-Hall, 1945, p. 264. 189lbid., p. 92. 154. neither fully foreign nor fully American in their cultural habits nor in their view of themselves. But again, the extent of this malaise wbuld depend on the remoteness of their cultural patterns from.American ways and the sharpness of their conflict with the first generation. One-clear indication of their marginal status is that they "display a lower percentage of married persons than either the foreign-born or the native-born of native parentage."190 In the first generation, this might be explained by an unfavor- able male-to-female sex ratio,.but not so in the second genera- tion. These children of immigrants were less capable of establish- ing norms and social control among themselves since they took dif- ferent directions,.tending either to identify with, or to reject, their ethnic origins. We know little about their relationship to their offspring in the third generation, whether they could func- tion as an agent of social control. This was probably a function of whether they themselves had attended foreign-language or English- language schools and the extent to which they could serve as models of a type of American cultural patterns for the third generation. Are the third generation themselves able to act as an agent of social control? Being the native-born children of native-born parents, the likelihood is small that they would identify them- selves as part of a distinct ethnic group, unless they lived in a large segregated sub-community. Barring this, we should expect 190William C. Smith, Aggricans in the Making: The Natural .fiiptory,of the Assimilation of Immigrants, New York: D. Appleton- Century, 1939, p. 358. For comparison of different ethnic groups in this regard, see David M. Heer, "The Marital Status of Second-Generation Americans," American Sociological Review, 26 (1961), pp. 233-241. 155. them to become merged with the general population, with some modi- fications which we shall come to. Now our second major question: Would an ethnic group be -concerned about enforcing a ruleaof religious endogamy? There is little doubt about the feelings of the first generation in this, matter. Many students of the subject have-called attention to this in one way or another: The more thorough the separation from the other aspects of the old life,.the greater was the hold of the religion that alone survived the transfer....the immigrants directed into their faith the whole weight of their longing to be connected with the past.191 The church structure to an ethnic group threatened with the loss of identity serves more than any other struc- ture to organize the group as a community system.192 Intermarriage is a serious matter in the eyes.of many parents, particularly when religious lines are crossed.193 The relative importance-of religious and ethnic intermarriage may be illustrated by the following percentages from a first and second generation community of relatively isolated Norwegians: More than 90 per cent of all four groups place membership in a Lutheran church as a first essential. Over three-fourths of the two rural and older urban generations and slightly less than two-thirds of the second urban category regard stemming from a Norwegian family as a primary requisite for a marriage partner.19 Is the second generation interested in maintaining a rule of 191Oscar Handlin, The Uprooted, op. cit., p. 117. 192WilliamL. Warner and Leo Srole, The Social Systems of American Ethnic Groups, New Haven: Yale University Press, 1945, p. 218. 193William C. Smith, Americans in the Making, 02- ci£-: p. 362. 19"John and Ruth H. Useem, "Minority-Group Pattern in Prairie Society," American Journal of Sociology, 50 (1945), p.384. 156. religious endogamy for their children? Except in segregated com- munities, the picture here is not clear. Will Herberg tells us that those who rejected their ethnic identification rejected their re- ligion also, but that in most cases "the ties with the old reli- gion were never completely broken."195 So far as religious act- ivity is concerned, Lenski's Detroit data show that this falls off in the second generation only in the-case of Jews: other- wise there is a steady rise from generation to generation.196 Perhaps the clue to the inconsistency has been suggested by William Smith when he observed: "Much of the religious conflict [between first and second generationg7 centers about language."197 It may be that large numbers of the disaffected second generation simply shifted to English-speaking congregations of their own denomina- tion, whereas in the Jewish case language and religion were more intimately bound up--and it was more difficult for the marginal second generation to reject their ethnic ties without at the same time rejecting their religion. How about the third generation? Will they accept a norm of religious endogamy for themselves, and will they pass it on to their children? Herberg believes that, since they are no longer anxious about their Americanness, they seek to confirm the tie 195Protestant - Catholic - Jew, op. cit., pp. 31, 42. 196The Religious Factor, op. cit., pp. 41-43. Another apparent exception to this were the Southern-born white Protestant migrants to Detroit, who showed no more religious activity by his measures than first-generation foreign immigrants, even though the former were at least sixth—generation Americans. '197Americans in the Making, op. cit., p. 328. 157. that binds them to their ancestors in the only important way that remains to them. He quotes Marcus Hansen to this effect: "What the son wishes to forget, the grandson wishes to remember."198 He also believes that the kind of religion that they return to is more secularized, one which functions to identify its members as one of three religious sub-communities--all equally parts of the "American Way of Life." His analysis may fit the Jewish sub- community best, but we have no reason to doubt that it applies in some degree also to other third-generation Americans. How can we sum up this web of relationships for ethnic groups in general? If we remember that large and relatively seg- regated groups such as the Polish in the Detroit area will be able to slow down the transition, the following may be our ordinary expectations with respect to religious endogamy in ethnic groups: If the first generation were not themselves married in the Old Country, their attitudes toward intermarriage were probably domin- ated by memories of their early years and the status of religious intermarriage in their country of origin. For themselves religion was most often the hub of their self-identity in the New World. Since homogeneity of religion was the most common experience in their communities "back home," they would assume, and insist on, religious endogamy for their children. However, try as they might, they would often fail to be effective agents of social control because of their children's acculturation and the consequent strained relations between generations. The'second generation is marginal; 4 198Protestant - Cgthblic - Jew, op. cit., p. 43. 158” many do not marry at all or marry late. Protestants and Catholics among them do not change their religious identification in large numbers if they can find a compatible English-speaking congrega- tion to shift to. Nevertheless, their hybrid socialization tends -to make them anomic as well as anxious strivers-for-status, with the consequence that they are more susceptible to interfaith mar- riage than the first generation. Identification with a religious group comes more naturally to the third generation. Still, their increased Americanization makes them.more mobile, horizontally. and vertically, with more frequent interfaith contacts in the general population through their school and work life particularly. This mobility and interaction will by itself increase the occur- ence of interfaith marriages, even though they value-religious .identification and activity more. Furthermore, the more secular their own view of religion has become, the-more easily they will convert to the partner's religion at marriage, in the interests of family harmony. Such is our summary, based on reports of previous research of the relationship of ethnicity to religious endogamy. Anomic Situations Now we can turn to the final section of this chapter, a look at the research relative to the last set of hypotheses in our study--those gathered under the heading of "anomic situations." The reader may wish, at this point, to refer to our general dis- cussion of anmmia earlier in this chapter. It seems to this writer that the five situations included in our hypotheses--wartime, late 159. marriage, rural-urban migration, educational mobility, occupational mobility--are authentic derivations from.Durkheim's notion of anomie as described above by Isabel Sary Lundberg. The situations apply to the individual as he finds himself located in the social struc- ture, as he experiences his environment. Each situation is viewed as personal, hence our preference here for the psychological term; anomia, derived from Durkheim's sociological concept, anomie. Only the first situation, wartime, could profitably be thought of as a society-wide situation also. The fact that the last three situa- tions also have something in common with reference group theory does not make them less eligible for inclusion in a broader context. Let us consider each of these situations briefly, in turn:. Wartime That wartime brings about disorganization, and possible normlessness, in the lives of a nation's citizens hardly needs demonstration. It should be sufficient to summarize the situa- tion in the words of Ernest R. Mowrer: The war as a crisis period is characterized in part by: enhanced social consciousness and identification with the goal of the group (the nation), instability and mobility of both soldiers and civilians, economic well-being particularly of those engaged in war production, high tempo of life, and grief and depression for those involved directly and indirectly in the casualties of war.199 Our hypothesis that Protestant-Catholic marriages will increase in wartime is based on the assumption that norms which are not directly related to the common war effort will be weakened (if not discarded) 199"Social Crises and Social Disorganization," American Sociological Review, 15 (1950), p. 65. 160. at this time. Hasty marriages between departing servicemen and their sweethearts, with the fearful thought that he may not come back anyway, have been a common observation--even though they do not seem to have been documented in the sociological literature. In World War II, at least, it was not uncommon for these marriages also to eventuate in divorce before even a battlefront death could dissolve them. As for the actual incidence of Catholic mixed mar- riages during wartime, we have already summarized above the only two such studies that have come to this writer's attention. Late Marriage It was only in the Manhattan study of Jerold Heiss that this writer found any mention of late marriage as being related to deviance from the norm of religious endogamy. This represented for Heiss one of two classes of people who were relatively emanci- pated from parental influence, the other being those who had been married previously. He found that both classes of people seemed to enter into mixed marriages with relative frequency in the Catho- lic and Jewish populations. However, "the data on Protestants do not show the expected relationship relating to age at marriage."200 We have classified late marriages with anomic situations in this study because it would seem, not only that the offspring's advancing age would give the parents less influence, but also that those who are not married and are still eligible for marriage have thinned out by the late twenties, especially for women. Under such 200"Premarital Characteristics of the Intermarried," 22; cit., p. 52. 161. circumstances, it seems that the strain to overlook the rule of religious endogamy would increase. Nevertheless, since we know little about the correlates of late marriages, we must be very tentative about this. A case might also be made for an anomic situation in the very early marriages, if this represents a rejec- tion of parental influence. Rural-Urban Migration MMch more has been written on this situation than on the two preceding. In fact, we cannot pretend to be thorough in our discussion of this subject. Its history and variations are too great. Just thirty years ago Niles Carpenter was able to write: This shock:effect of migration into the city is discussed at length elsewhere. Here it may merely be pointed out that,' in the author's opinion, many of the disorganizing effects of the city upon the individual....are to be interpreted not so much as effects of city life as such, but rather as the effects of the sudden igpact of the characteristically urban set of conditioning influences upon a personality that has been accom- pdated to a gharacteristically non-urban set of influences. In short, certain individuals break down as a consequence of their failure to become adequately reconditioned to the city.201 This is precisely the kind of situation we have in mind in predicting that those who move from a rural environment to a city before they marry are likely to depart from.the norm of religious endogamy. The same might be said for those who originate in a small town, and for the same reasons. However, the last three decades have seen an increasing urbanization of our entire society, even the rural areas. This can be credited to the electrification of 201The Sociology of City Life, New York: Longmans, Green. 1931, pp. 217-218. 162. rural areas, the prosperity of many of those who remain on the farm, innovations and expansion in mass communication, and growth of the means and routes of transportation. The percentage is be- coming smaller of those who have not undergone a large measure of urbanization while living right on the farm. During these critical thirty years, the experience of anomia on the part of farm-to-city migrants has probably become less and less pronounced. The movement off the farm has largely taken place already, as shown by the fact that one out of three adults now living in a non-farm place has grown up on a farm. But this migration cannot be very significant in the future, since "only 13 per cent of the adult population remains on our farms."202 It is well to remember also that "the farm-reared are over- whelmingly Protestants." According to a national study, these farm- reared Protestants attended church "with slightly more regularity than other Protestants. This higher rate of church attendance is concentrated in low-income groups."203 The reader may recall that Lenski found Southern-born white Protestant migrants to Detroit to have a low rate of church attendance relative to other Protestant migrants. Can these apparently inconsistent findings be reconciled? Lenski had just forty-five cases; so his sampling error may be large for the Southern-born. Furthermore, we are quite sure that almost all of his Southern-born were rural, but we do not know what 202Ronald and Deborah Freedman, "Farm Boy in the City," in Ronald Freedman et al., Pripciples of Sociology, op. cit., pp. 462- 463 0 203Ibid., pp. 469-470. 163. percentage of his other Protestant migrants were rural or urban in origin. Still, if the national study is correct, this would not make much difference. It may be that the Southern-born are simply a special case in this regard. Perhaps, like foreign immigrants, they find the northern and urban churches uncongenial and cannot as yet finance their own institutions. As we saw earlier, the Southern white migrants to Chicago and Cleveland have maintained their kinship and community ties as well as their rural patterns of visiting. This would operate a- gainst an experience of anomia for this group. In spite of their lower rate of church attendance (temporary, perhaps), those who have migrated should not be anomic when they have this kind of res- idential pattern in the urban setting. (Their children, however, may be nearly as anomic as second-generation Americans.) Other urban migrants apparently have high religious participation and therefore could be expected to maintain religious endogamy on that account. These are current conditions which indicate that our hy- pothesis will probably not hold up. If we can examine the older marriages among the farm-to-city migrants in our data, we may find the kind of disorganization and anomia that Carpenter was writing about. If so, in that group our hypothesis should be confirmed. Educational Mobility By educational mobility we mean the situation in which a son (or daughter) acquires more years of schooling than his father (or mother). When the child of an immigrant does this, it is likely to accentuate the strains between parent and offspring already pres- ent and make for a particularly anomic situation for the offspring. 164. When parents with a college education send their children to col; lege also, this is not viewed as an anomic situation, since the offspring in this case are not so likely to consider the norms of their parents irrelevant to themselves in college. Educational mmbility means that a young person who has been socialized in an environment in which norms were stated in simplistic (and prob- ably negative) terms acquires the capacity to make new distinctions and thereby judge the old norms to be irrelevant to himself. As it stands, this is mostly speculation. The writer has not found any published material which clearly points to.our hypothesis.that those who experience educational mobility will be more likely to violate a norm of religious endogamy. Jan Hajda has written an interesting article somewhat re- lated to this hypothesis. It is based on a national sample of graduate students, identified by themselves as intellectuals or non-intellectuals and cross-classified also as alienated or not (from the dominant patterns of the society as a whole). In the context of the entire society, the criteria he used for alienation are mostly consistent with our concept of normlessness. He found that: Infrequent or no attendance at religious services, loss of religious faith, self-exclusion from membership in religious bodies, and profession of secular beliefs, are associated with both the feeling of alienation from non-academic society and the intellectual self-image....the high representation of‘pgpg- inally Protestant students in the category of the alienated-- is difficult to account for; it appears to be partly due to the fact that a high proportion of Protestant students lose the original faith.204 204"Alienation and Integration of Student Intellectuals," American Sociolo ical Review, 26 (1961), pp. 765-766. 165. This finding would argue for a high rate of Protestant mixed marriages, since by our operational definition they would classify themselves as presently "other'" their tendency to select a Protestant marriage partner would constitute a Protestant mixed marriage by our definition. With respect to Catholics, since they are underrepresented in the college population, an application of“ the generalization of residential and socio-economic propinquity would indicate the occurrence of higher rates of interfaith mar» riages for them. It seems to this writer that their educational mobility would weaken for them their earlier values and norms, thereby giving rise to increased violations of religious endogamy. Occupational Mobility The condition of society which for Durkheim was almost the ideal type of anomy we here apply to the situation of an individual. In Durkheim's view, man's appetites are unlimited unless he is re- strained by the social conscience. In times of social crisis or abrupt transition, when the public conscience is reclassifying men and things, society is unable to supply this restraint; it is in a state of de-regulation or anomy. "The less limited one feels," he says, "the more intolerable all limitation appears."205 It is the contention of this thesis that extreme upward mobility in occupa~ tion places the individual in such a situation so far as his per- sonal life is concerned. He must reclassify everything that is associated with his unaccustomed economic status; hence religious endogamy is likely to be discarded with other rules of restraint. 205Suicide, op. cit., pp. 246-254. 166. The condition of downward mobility, on the other hand, would seem to give rise to a type of normlessness more closely associated with the "despair" of Srole's anomia scale. we have included the downwardly with the upwardly mobile in our hypothesis partly because of Lenski's finding that in the Detroit area: "Those who were highly involved in their socio-religious sub-communities were less likely either to have risen pp to have fallen in the class system than were those who were more marginal to the group."206 To define our terms before we go on, by occupational mobil- ity we are referring to vertical, not horizontal, mobility, ip_gp, the movement of an individual from one level to another in occupa- tional ranks. We make only the gross distinction between blue- collar and white-collar occupations, the latter being-ranked higher for males than the former, so that we can use the terms "upward" and "downward" with reference to these ranks.207 Our hypothesis will be tested only for intergenerational mobility, ;p_§p, compar- ing a man's occupation with that of his father, rather than with his own earlier occupation, which would be intragenerational mobil- ity. Lipset and Bendix tell us that in the United States from 31 to 35 per cent of the-sons achieve upward mobility in this sense. A smaller minority is generally found to be downwardly mobile. However, when comparisons are restricted to urban occupations for 206The Religious Factor, 09- cit., P- 110- 207For reasons supporting this assumption of ranks, see Seymour Lipset and Reinhard Bendix, Social Mobility in Industrial Society, Berkeley: University of California Press, 1959, pp. l4- l6. 167. both fathers and sons, those reported upwardly mobile have-ranged from 17 to 21 per cent of the total and the downwardly mobile from 8 to 13 per cent.208 Is occupational mobility associated with religion? At- least in the United States, Jews "have markedly higher rates of upward mobility than non-Jews, according to a number of studies."209 However, in the study at hand we are more concerned with Protestant- Catholic differences in mobility. Although the Catholic sample is too small for detailed comparisons to be trustworthy, data for the United States from the Survey Research Center of the University of Michigan show almost no difference in manual-nomaanual mobility between Protestants and Catholics of the second generation, but somewhat more upward mobility for Protestants than for Catholics in the third and later generations. Investigations in four countries indicated "no significant differences between mobility rates of Catholics and Protestants...."210 However, as we saw above when discussing secularism, there is evidence of greater upward mobil- ity for Protestants than for Cathdlics, at least in the Detroit area. Furthermore, Lenski finds fault with the analysis in the study of mobility in the United States by Lipset and Bendix, just cited, as well as with a study by Mack, Murphy, and Yellin, which also reported no differences in upward mobility for Catholics and 208Ibid., pp. 17, 88. 2091bid., p. 56. 21OIbid., pp. 51-54. 168. Protestants.211 The existence and qualifications of a relation- ship between occupational mobility and Protestant-Catholic reli- gions are still in doubt and need further investigation. Are there consequences of occupational mobility which would seem to be relevant to religious endogamy? The trouble with dis- cussin "consequences" is that this assumes that we can establish the order of causality, whereas this is generally impossible. For instance, Merton writes of "anticipatory socialization," adopting the values of a group to which.one aspires but does not belong. This would make mobility a consequence of having acquired the char- acteristics of a higher occupational level.212 Richard F. Curtis, on the-other hand, finds in the mobile male a "gradual assumption of the culture of his stratum of destination," which seems to make these characteristics the consequents or concomitants of his mobil- ity.213 Whether consequences or not, some variables have been found to be associated with occupational mobility for males. Curtis and Morris Janowitz suggested a few years ago that social mobility "is likely to have disruptive consequences on primary group structures, 21lThe Religious Factor, op. cit., pp. 76-77. 212Social Theory and Social Structure, op. cit., p. 265. However, the fact that Merton also finds this anticipation in a group's "isolates," who are not mobile, means that mobility does not necessarily follow from anticipatory socialization. Ibid., p. 305. 213"Occupational Mobility and Church Participation," Social Forces, 38 (1960), p. 316. 169. such as family, clique, and friendships."214 However, the investi- gations of Eugene Litwak indicate that extended family cohesion is not weakened by occupational mobility itself but by geographical mobility, the criterion for family cohesion being family visits.215 To the extent that his conclusions are valid, they would argue some- what against our hypothesis that violations of religious endogamy are associated with occupational mobility, due to anomia. Lipset and Bendix report that: "People who are upward mobile, but pp; those who are downward mobile or geographically mobile, have higher rates of mental disorder than those who are stationary."216 They believe that the downwardly mobile would be at least equally vulnerable to mental disorder if this were due to an anomic situation. We have little more to rely on regarding this association. However, in analyzing intrageneratignalimobility, Richard L. Simpson and H. Max Miller found greater anomia (Srole's scale) among the upwardly mobile than among the stable--although the "prevailing attitudes are more anomic" in the lower-status groups, apartfrom.mobility.217 These two studies show anomia and mental disorder both to be associated with upward mobility. This writer suspects that downward mobility, because it is more rare, 214Seymour Lipset and Reinhard Bendix, Social Mpbility in Industrial Society, op. cit., p. 65. 215"Occupational Mobility and Extended Family Cohesion," American Sociological Review, 25 (1960), pp. 9-21; "Geographic Mobility and Extended Family Cohesion," American Sociological Review, 25 (1960), pp. 385-394. 216Social Mobility in Industrial Society. 02- CiEo: P- 252- 217"Social Status and Anomia," Sogial Problemp, 10 (1953): pp. 262-263. 170. is more difficult to analyze properly and that it too will be found in time to be associated both with anomia and mental dis- order. Apart from.mobility and anomia, how about the relation- ship of mobility to religious interest? Lenski, in analyzing data from Indianapolis, discovered that for Protestants (no Catholics in the sample) upward mobility (in income) was associated with little interest in religion.218 0n the other hand, Richard F. Curtis, using the Detroit Area Study surveys of 1952, '54, and '55, concluded that upward and downward mobility were pp; related to church participation of females or of Catholics.219 These two studies would suggest, then, that religious endogamy might be impor- tant to Catholics but not to Protestants who are mobile. However, these studies both deal with associational criteria, in Lenski's terms, which are not necessarily related to a communal‘type vari- able such as interfaith marriage. How can we tie these various findings together and relate them to our hypothesis that occupational mobility is an anomic situation which will increase Protestant-Catholic marriages? That anomia is associated with upward mobility is fairly well established; perhaps it is also associated with downward mobility. Upward mobil- ity itself is probably more characteristic of Protestants than of Catholics. The anomia experienced by these men seems likely to 213"Social Correlates of Religious Interest," 2£;_££E'D p. 546. 219"Occupational Mobility and Church Participation," M p- 319- 171. weaken for them the social control applied to the norm of religious endogamy, even though Litwak's data show that this is not an all- or-nothing proposition. The reader who has read all of this chapter deserves con- gratulations at this point because the chapter is finally at an end. The subject we are dealing with in this thesis has so many ramifications that it has been impossible for this writer to arrive breezily at a quick conclusion. Social control, norms, institutions, deviancy and anomia, evidence for a norm of religious endogamy in this country, the conflicting values that operate in.mate selec- tion, the agents of social control which are relevent to this choice, .the anomic situations which may weaken the endogamous normr-all these are-complicated subjects in the sociological literature. Taken together, they encompass a great many of the problema which are central to the-study of sociology and social psychology in general. Now we are ready to go on to see what the data of the _ present thesis may contribute to their solution. CHAPTER IV THE RESEARCH: DESIGN AND EXECUTION In Chapter I, the reader found a description of the behav- ioral and motivational development of this research plan; in the chapter we are now beginning he will find something more in the order of an intellectual development of this particular research design and the methods used to test it. In Chapters II and III we have immersed ourselves in past research on marriage choices and in the concepts which provide a theoretical context for our research. Now, as we move into the concrete measures used to apply our hypo- theses to the data available, we can better appreciate the complex- ities of this task. In this chapter we shall consider briefly how the plan of this study came to take shape. Then the procedures of the Detroit Area Study will be described and evaluated. Next, we shall see what is distinctive about the data and sample boundaries of this study. We shall give considerable attention to the operational definitions of the dependent and independent avariables. And, finally, the stages followed in the analysis will be outlined. Arriving at a Plan When one undertakes to design a study which involves a secondary analysis of data gathered for other purposes, it becomes something like planning a vacation auto-trip. You have your own 172. 173. ideas about what you would like to see, based partly on what you have heard from.other travelers; you have limited resources for such traveling; and you can visit and examine only what is there on the map, not what your imagination may suggest to you. So it was in this research design. Actually, the only way in which the writer has felt limited as to resources is in the matter of time. Even though he has been involved with this project off-and-on for over four years, there is always more that could be done to improve the product. As for other resources: Many kind friends have-re- moved the concern about living expenses; the directors of the Detroit Area Study at the University of Michigan-have graciously made the data available; and various members of the faculty and staff at Michigan State University have donated their time, talents, and the use of the University's facilities toward the analysis of these data. However, the process of matching ideas with the con- tours of the data available was more difficult. Among all the suggestions provided in the literature on mate selection for the study of interfaith marriages, the approach which seemed to this writer to be most fruitful was that taken through the concept of social control by Jerold Heiss.in his Man- hattan study. In our study it was hoped to test all of the.r21evant agents of social control in the best way possible, using the data of the Detroit Area Study. It was also intended to examine some situations which would seem to hinder the operation of social con- trols supporting endogamous marriage. These were the ideas with which the writer approached the map of the Detroit Area Study inter- view schedules to see what could be discovered there. 174. The first requisite for our use of one of these surveys was that the religious preferences of ppph husband and wife (Protr estant, Catholic, Jew, Orthodox, other) must have been determined. Otherwise, we could not identify an interfaith marriage. This in- formation was available in the surveys of 1955, '56, '58, '59, and '62.1 Once the usable surveys had been identified, it became a question of selecting interview items which would fit into our theoretical framework. It was only at this point that the hypo- theses of this study began to be specific. The decision had al- ready been made to limit the study to factors, rather than conse- quences, of interfaith marriages. While examining the questions in these interview schedules, the possibility was kept in mind of testing each of the generalizations which emerge from the litera— ture on mate selection. In fact, before social control impressed itself on the writer as the key concept to use, he had attempted to organize the data in terms of a paper by James Schellenberg 1Permission to use these data was obtained first from Dr. Ronald Freedman, then Chairman of the Executive Committee of the Detroit Area Study, and from John C. Scott, Director of the Detroit Area Study. Permission to use the data specific to each survey was obtained from at least one of the faculty participants respon- sible for each of the following annual studies: l955--Robert 0. Blood, Jr., and Morris Axelrod; 1956--Robert C. Angell; 1958-- Gerhard Lenski; 1959--Guy E. Swanson and Harvey Brazer; 1962-- Ronald Freedman and David Goldberg. Religious preference of husband and wife were known also from the survey of 1957 on political participation (Daniel Katz and Samuel Eldersveld), but the other questions asked did not pro- vide suitable material for this study. 175. called "Models of Mate Selection."2 The cultural and ecological models described in that paper gave a broad context to the problem but did not suggest any new hypotheses. By contrast, the concept of social control and its correlative anomde did uncover some new hypotheses in the interview schedules as well as organize a number of the generalizations found in the literature. Once this insight was achieved, it was relatively easy to line up the hypotheses as they appear in Chapter III.3 However, even then the matching of theory with data was not complete because, as has already been noted, two of the questions in the interview schedules which appeared suit- able for this study had not been coded by the directors of the orig- inal studies. §pprce of the Data: Detroit Area Study We have already referred to the Detroit Area Study often, including a brief description in Chapter I. Some of the pertinent results of these surveys will also be found in Appendix A. At this point we will take a better look at this facility of the University 2Unpublished paper read at the annual meetings of the Ohio Valley Sociological Society; East Lansing, Michigan, May 4, 1962. 3After these hypotheses had been specified, an unsuccess- ful attempt was made to apply Dr. Louis Guttman's "facet analysis" to this study plan. He was then Distinguished Visiting Professor at Michigan State University. It seems to this writer that the attempt would have been more appropriate at an earlier stage of the planning, before the plan had crystallized. However, for a fruit- ful application of facet analysis to a completed study, cf. Louis Guttman, "A Theory for Intergroup Beliefs and Action," American Sociological Review, 24 (1959), 318-328. 176° of Michigan and how it operates.4 This was established in 1951 to aid the research efforts of faculty members in the social sciences and to provide training for graduate students in a cooperative sur- vey research project. It was begun through the support of the Ford Foundation but is now integrated in the budget of the Department of Sociology at the University of Michigan as a continuing graduate training facility. Each year one or more research proposals, sub- mitted by interested faculty members, are selected by an interdepart- mental executive comm ttee for use in the DAS. The faculty member so involved takes responsibility for seeing the research project through to its final analysis. Student training is supervised by the staff of the DAB, supplemented by the personnel of the Institute for Social Research. Graduate students participate in designing the interview schedule, pre~testing it, and conducting the inter- views in Detroit an d suburbs (ordinarily doing the interviewing during the vacation between the fall and spring semesters). They then work on coding the responses and con nclude by analyzing a seg- ment of the data around a research problem which they have them- selves proposed. Numerous books and monographs, doctoral dissertations, articles, and reports have been produced over the years through the activities of the Detroit Area Study.5 Of those surveys being 4For a fuller description, see Ronald Freedman, ”The Detroit Area Study: A Training and Research Laboratory in the Community," American Journal of Sociology, 59 (1953), pp. 30- 33; and Harry Sharp, "Graduate Training Through the Detroit Area Study," Amprippn §pgiplpgigal Reyipw, 26 (1961), pp. 110-114. 5A complete bibliography of materials based on Detroit Area Study reSeazoh,;1951-l962, can be obtained from the Department of Seciology, Survey Research Center, Un diversity of Michigan. 177. reanalyzed for the present study, the results of those of 1955, and 1958 have appeared in book form.6 The data gathered in the 1956 survey were reported in A Sociaerrofile of Detroitll956.7 The results of the studies conducted in 1959 and 1962 have not yet been published. In general, the surveys of 1955 and 1962 were family studies; that of 1956 centered on moral issues; those of 1958 and 1959 concerned religion. Sampling and Interviewing in the Population The selection of the annual sample for the Detroit Area Study is carried out in consultation with the Sampling Section of the Survey Research Center at the University of Michigan.8 With the exception of the 1962 sample, which will be described in Chap- ter VI, the other surveys used in our study were based on the follow- ing sampling procedures: The population to be sampled in each case consisted of all adults living in private dwelling units within the areas of Macomb, Oakland, and Wayne counties, tracted in 1950. 6Robert 0. Blood, Jr. , and Donald M..Wo1fe, Husbands and Wives: The Qypamics of Married Living, Glencoe, 111.: Free Press, 1960; Gerhard Lenski, A Sociolo ist' 8 Inquiry, Garden City, N. Y., Doubleday, 1961. 7A Social Profile of Detroit‘l956: A Report of the Detroit Areg Stpdypof_the University of Michigan, Ann Arbor; University of Michigan, 1957. 8For detailed descriptions of the sampling procedure, see: John Takeshita, "Selection of a Sample of Dwelling Units for the Detroit Area Study, 1954-55," #1070, mimeographed; A Social Profile of Detroitll956,.op. cit., pp. 4-7; "The Sample Design for the 1957- 58 Detroit Area Study;9 #1443, mimeographed; "Selection of Sample: Detroit Area Study, 1958-59," #1458, mdmeographed; all are avail- able from the Detroit Area Study, Survey Research Center, university of Michigan. 178. Macomb Co. {3 e Oakland Co . WAZFZENTWF? a} $3.: BER * SourHFIELD L Y ____ I" _17, _ "V" 3’1. -r\" it?“ OAK ,, [ l PARK '. FERN- P Oakland Count _ _ _______ i- _ 1‘ 1.- _LMacemh Ce. Wayne County Wayne County Hunrt~cro~* *PLEASANT‘ rupee WOODS \_—‘\ LIVONIA “'Gfip‘g" REDFORO \AX“ TWP. A TRAMCK L H V DETROIT om: r2 51v; M f? ARDEN CITY 1: ‘ ‘DEARBORN NANKIN ‘r P. INKSTE TAYLOR TWP. Figure 1. The "Detroit Area" as defined by the Detroit Area Study. 179. The area, shown in Figure l, on February 1, 1958, included approx- imately 87 per cent of the total population of the Detroit Standard Metropolitan Area, with boundaries determined according to the 1950 Census of the United States. After primary sampling units were determined and stratified according to size of population, geo- graphic location, and economic values, selection from each stratum was based on a table of random numbers. Wherever possible, city directories were used (after checking for completeness) for system- atic sampling of pages and then --- of addresses, in clusters of three. A 'block supplement" method was used to augment the city directory sample. For those areas not actually covered by any city directory, the chunk and segment method (a form of area-probability sampling) was used. Each dwelling unit was given the same prob- ability of being selected for the sample, a sampling fraction from l/900 to 1/1100, depending on the year of the survey. In 1955 and 1962, only the wives were interviewed. In the other years the selection of the adult to be interviewed within a given dwelling unit was randomized, not left to the discretion of the interviewer. Except in 1962, the size of the sample selected was about 900 in each survey, with the expectation of obtaining about 750 interviews in each year. How this combined sample of three thousand was re- stricted for this study will be detailed in a later section of this chapter. Because the entire population was not interviewed we must, of course, take account of "sampling errors." The size of the sam- pling error will depend on how large a sample was selected and how large the variation is found to be for any given characteristic. 180. The sampling error for reported percentages can be found.in pub- lished reports for the 1955 and 1958 surveys:9 lIn'theusurvey of 1959 it was equal to that of 1958,.since the sampling method for these two surveys was practically identical.10 The sampling error for 1955-is sometimes one per cent larger than these later surveys for a given sample size and sample percentages. Similar statements could be made regarding the statistical significance (with 95 per cent probability) of differences in percentages in each sample. The problem of combining theselsamples into one study has been care- fully investigated by C. Michael Lanphier, former research assist- ant for the Detroit Area Study, but his results are not yet avail- able. Non-sampling errors are of two types: "reporting errors" and'hon-response errors." The size of such errors cannot be deter- .mined mathematically. The directors of the Detroit Area Study have regularly tried to keep reporting errors at a minimamtby careful training of the interviewers with the aim of motivating the respond- ent to answer questions to the best of his ability. Interview re- sponses are also checked for inconsistencies. The response rate in these surveys has been about 87 per cent. .It is over 90 per cent for subgroups, such as those in the age group, 21-34, and those who are married and have-children present. An average of 2.9 9Robert 0. Blood, Jr., and Donald M. Wolfe, Husbands and Wives, op. cit., pp. 276-280; Gerhard Lenski, The Religious Factor, op. cit., pp. 331-340. 10"Measures of Smpling Error," Project 855, Detroit Area Study, Survey Research Center, University of Michigan, April, 1959, mimeographed. As will be explained a little later, the survey of 1956 will not be used in the final analysis. Hence its sampling error is not of concern to us now. 181. calls are made to obtain a single-interview. Some demographic information was obtained concerning almost all possible respondents, and a comparison-between respondents and non-respondents on such items showed relatively small differences.11 One way to check the reliability of results in such surveys is to compare the samples for successive years on various.charac- teristics. Another way is to compare a Detroit Area Study sample with the-reports of the 1950 united States Census for the Detroit Standard Metropolitan Area. When this has been done, the samples show only slight variation, and each is a close approximation to the census reports on demographic variables.12 Coding and Preparation of Data gprds On the basis of long discussions by the participants, as well as at least one pre-testing of the interview schedules, many of the questions used in the Detroit Area Study interviews are precoded. In other words, the interviewer needs to write only a number or check one of a set of possible responses to the question. In the present study almost all of the questions used were precoded, .the major exception being the husband's occupation. All of the questions which enter into Our analysis can be found in Appendix C. In each survey a check for coding reliability was made by the staff of the Detroit Area Study. This was done by having two persons 11For a detailed evaluation of the surveys of 1956, 1957, and 1958, see Harry Sharp and Allan Feldt, "Some Factors in a Probability Sample Survey of a Metropolitan Community," Americpn W. 24 (1959). pp. 650-661. 12A Social Profile of Detroit 1956, op. cit., pp. 75-76. 182. -independently code the same set of interviews. Their judgments were compared and an index of agreement computed. Where there was comparatively low agreement concerning the coding of responses, the code index was reconstructed before becoming part of the anal- ysis.13 The writer used the code books of previous Detroit Area Study surveys at the Ann Arbor office to write-out directions for the construction of two decks of IBM cards which would include all the information from the DAB surveys of 1955, '56, '58, '59, and '62 needed for this study. The Data Processing staff of the Survey Research Center at the University of Michigan prepared these new decks for the writer's use at Michigan State university. On the latter campus the writer intended to use two computer programs pre- pared for the MISTIC computer. To do so, a recoding was required to get rid of 11- and lZ-punch codes in all columns before using MISTIC's Library Routine Kl6-M and to reduce all variables to single- column fields before using Frank Sim's CHI-Square Program for MISTIC. This was done according to the writer's directions by the Data Processing staff of Michigan State University. Later checking un- covered only one small error in this process. Before testing the rhypotheses of this study, it was also necessary for the writer to .combine the responses to interview questions into various.indices. How this was done will be explained when welcome to operational definitions. 13For a general discussion of this subject relative to the;l955 Detroit Area Study survey, see Ruth Searles, "Coding as an Element in Research Design," #1068, mimeographed, Detroit Area Study, Survey Research center, Unitersity of Michigan. »l83. Restrictions on the Sample The very title ofour study indicates that we do not intend to generalize to the entire population of the Detroit area. When the writer arranged for the data cards to be made up in Ann Arbor, he specified that only married people were to be included in the sample, since this is a study of marriage choices. The broken marriages of the divorced or separated or widowed could not be analyzed be- cause questions were generally not asked about the former spouse. 'in such cases. This is unfortunate for opr purposes. It would be better if our respondents could have been interviewed at the time -of their marriage, when couples later to be separated would not be removed from our sample. However, we have-no choice but to catch .our respondents at some point in their life experience after the honeymoon is over. I Other categories were eliminated from our sample for spe- cial reasons. Negroes were not included because of the strong norm of racial endogamy even in Northern cities which prevents free interracial marriages. It would not be worthdwhile to analyze Protestant-Catholic marriages among Negroes alone, since Catholics constitute only 5.0 per cent of all Negroes in the Detroit area. Similarly, the Jews, Eastern Orthodox, and those with no religios preferences are not numerous enough for separate analysis--consti- tuting 3.0, 2.2, and 3.1 per cent, respectively, of the white pop- ulation of the Detroit area.14 These will appear in our sample 14Derived from.Albert J. Mayer and Harry Sharp, "Religious Preference and WOrldly Success," American Sociology Review, 27 (1962), p. 221. 184. only if they have married a Protestant or a Catholic. In this case, they will be designated here simply as "other." Originally,.it had been intended to analyze only the fippp marriages-among our‘unbroken marriages. However, it was decided to retain all unbroken marriages in the first analysis. Then the .gigpp marriages could be analyzed later as a sub-class. Actually, this sub-class was not analyzed separately. At best, this would not have been wholly satisfactory, since only in 1958 was the re- spondent asked whether this was the first marriage for his (her) spouse also. In the surveys of 1955 and 1959 we would be unable to.remove‘pll later marriages. How'much.difference will this make _in our results? In the three surveys combined for the analysis in Chapter V, this was pp; the first marriage for 17 per cent of the respondents (whatever it was for their,spouses). Among these-re- ‘married respondents, the first marriage ended by death for 31 per cent and by civil divorce for 69 per cent. Although most studies of interfaith marriages do not try to separate the first from later marriages, there would be an -advantage in doing so. Itseems that remarriages.involving Catho- lics are much more likely to be mixed marriages than are first mar- riages.15 Therefore we can expect our sample to have a somewhat 4 15"State-records in Iowa show that 63 per cent of all re- 'marriages:involving Catholics are-mixed marriages, compared to 30' per cent of the first marriages; Lee G. Burchinal, William F. Kenkel, and Loren E. Chancellor, "Comparisons of State- and Dio- cese-Reported Marriage Data for Iowa, 1953-57," American Catholic §ociological Reviewa 23 (1962), p. 25. -185. higher rate of mixed marriages then it would if it were restricted to first marriages. Opgrational Definitions: The Depgndent Variable Our dependent variable, the characteristic whose occurrence in the population we are trying to predict, is the religious type -of marriage entered into by those who are either Protestant or Catholic. As measured by the norm of religious endogamy, a mar- riage may be either mixed or unmixed religiously. Since every-re- spondent in each of the surveys included in this study was asked, "What is your religious preference?" and was asked the same ques- tion about his (her) spouse, it is a simple matter to discover which Protestants are now married to non-Protestants and which Catholics are now married to non-Catholics. However, as is gener- allyrecognized, many couples share the same religion after mar- riage because one of them has changed his (her) preference. To classify married couples simply by their present religious pref-- erence would put us in the position of having many in the unmixed category who were-socialized in different religious traditions. Furthermore, our group-of mixed marriages would be greatly reduced. The writer did, in fact, begin-his.analysis of the-data by using this classification based on present religious preference, since the survey of 1956 gave him no other choice in this matter. However, in making-cross-tabulations against other variables, he soon discovered that including the data of 1956 for this reason was contributing far less to this study than would the use of "previous religious preference" as a means of identifying interfaith marriages 186. in the surveys of 1955,.'S8, and '59. Protestant or Catholic inter- faith marriages,-based on present religious preferences, constitué ted less than’lO per cent of all marriages in these four surveys. When classified by previous religious preference, mixed marriages of Protestants and Catholics constituted 29 per.cent of all our 4 selected marriages in the surveys of 1955, '58, and '59. There- fore the writer decided to leave the sample of 1956 out of the analysis altogether, both because he would thereby gain in the to- *tal number of mixed marriages to examine, and more particularly because previous religious preference would give us.a better pic- ture of the situation at the time of marriage. Let us see why ‘this is so.- Suppose we base our classification of interfaith marriages on responses to a question about previous preference. If a respond- ent says that he has changed his religious preference, does this mean that the change took place near the time-Of marriage? Not always, of course. However, we are-going to assume that the change generally does take place about that time if it means crossing the line either way between Protestant and Catholic affiliations. In- formal observation would give us some confidence in making this assumption. More concretely, Gerhard Lenski reports that in the survey of 1958: One of our important findings was that it is most un- wise to assume that spouses who are currently of the same faith were always of the same-faith. Although 85 per cent of the white Protestants and Catholics in our sample reported that “ they and their spouse were of the same major faith (LLJEo: Protestantism or Catholicism), a check of their religious back- ground revealed that only 68 per cent had been reared in the same faith. In other words,.one fifth of the now homogeneous 187. marriages had been contracted by persons raised in different faiths.16 Although Lenski.did inquire about the reasons respondents had for changing religious preference and about the year of the change, he does not.report its relationship to the year of marriage. The present writer did not obtain the data concerning the length of marriage and is therefore unable to make this computation for this sample. However, in the small 1962 sample of couples who were all married in 1961, there were thirty individuals among the 185 couples who changed either to Protestant or to Catholic. One changed after the wedding; twenty-two changed in 1961 or 1960; only seven changed before 1960. So it appears that, for more than three-fourths of these cases, changing religious preference-was associated with meeting one's marriage choice. vFrom this point on, we work on the assumption that these married persons who changed their religious preference did so about the time of their marriage--that they rejected, in this sense, the -social controls of the religious group in which they were socialt ized. Therefore the survey of 1956, which had no information about previous religious preference, will not appear after this in our analysis. For the surveys of 1958 and 1959 we have direct evidence to identify the original religious group, since respondents were asked (about themselves and their spouses) whether their religious preference had always been what it is now--and, if not, what it was previously. Changes between-Protestant denmminations will not enter into our analysis, only those between Protestant.and Catholic. ‘ 16The Religious Factor, op. cit., p. 49. 188. TABLE 4-1.--Previous religious preference of respondent and spouse, reclassified by mother's religious preference in the formers' child- hood days, Detroit Area Study, 1958 and 1959. L ”— fl PIEV1OUS R31181008 Mother's Religious Preference Preference of Respon- dent ofi Spouse Catholic Other Protestant All Catholic 386 4 9 399 Other 20 5 16 41 Protestant 1 6 _1_1_ 418 445 Total 422 20 443 885 In the survey of 1955, we have only indirect evidence about previous religious preference. The respondents were.asked (about themselves and their husbands), not what their previous religious preference had been, but what their mother's religious preference was when they were growing up. To test whether we might assume that these two questions are equivalent, the writer analyzed the date of 1958 and 1959 to see how many cases would have to be reclassified as to marriage type if he were to classify these on the basis of mother's religious preference instead of on the respondent's Sand sppuse's) previous religious preference. As is clear from Table 4- 1, it would make little difference to reclassify by mother's reli- gious preference those who say that their former religious prefer- ence was either Protestant or Catholic. The classification of 3.3 per cent of the Catholics and 6.7 per cent of the Protestants would be affected by this. However, to reclassify those who say their former religious preference was "other" would make a difference in 189. 88 per cent of the forty-one cases. Why this is so will be explained before we come to the end of this section. Meanwhile, let us simply note that, for the two large religious groups in the surveys of 1958 and 1959, we will make few mistakes in assuming that the respondent's (or spouse's) previous religious preference and their mother's re- ‘ligious preference are equivalent. Now, if we compare the present with the previous religious preferences of all respondents and their spouses, we shall discover several religious types of marriage. Not only are there the two types which are currently mixed or unmixed religiously, but there are also the two types in which married persons who were pgeviously either Protestant or Catholic have changed their religious prefer- ence. From the point of view of the social control exercised by either religious group, these four types may take on contrasting appearances. Some months before proceeding to the analysis of the data, the writer recorded his ideas on this matter. It may be worth- while to insert these paragraphs here: Interfaigh marriages: Is there a difference (in the endogamous social norm) between firofiégtanf and Catholic viewpoints? Preface: It seems to be agreed from~both Protestant and Catholic view- points, that it is preferable to marry someone who has shared one's faith. Lacking this, a match is preferred with one who is a recent convert to one's faith. From both viewpoints, to marry one who remains a convinced adherent of another faith is less desirable than the first two alternatives. The question remains whether one would prefer, in general, to see another adopt the faith of his (her) marriage partner rather than allow the family to remain divided in this matter. Dr. Robert Blood suggested that this would be the social norm expressed by many Protestants, whereas the writer had (apparently fram the Catho- lic point of view) rated this alternative as less desirable than a marriage that was to remain mixed. Is there this dif- ference in norms? If so, what might explain it? ngp rglgvan; liberal Proteptang Attitudes: Although Protestants do not generally count matrimony as one 190. of the sacraments established by Christ, many of them do think of it as a sacrament (i. e., an instrument of grace) through the love that the partners bring to each other. This can hardly be achieved fully unless the married pair worship together. The choice of the place or manner of worship is one which they should make according to their own likes, not being obliged to one or the other church as such. Normally they may be expected to find the church of one or the other's childhood to be more congenial, but they may just as well choose a third denomina- tion. This position seems to be based on the primacy of the individual's judgment in matters of faith and ethics. Sppe releyapt fundamentalist Protestapgfiattitpdes: The fundamentalist Protestant is not so likely to think of mat- rimony as a sacrament in either of the senses above. However, he is prepared to believe whatever the Bible says. He joins a specific denomination or congregation because he is in general agreement with the way its ministers and members interpret the Bible. Similarly, he feels that married people should also agree and worship together. However, in the event that one of his own denomination wants to marry someone of another faith, the norm would probably be not to change to the other's denomi- nation or to a third choice because the individual's faith takes precedence over marital unity. Some relevpnt Roman Catholic aptitudes: In the Catholic view Matrimony is one of the seven sacraments instituted by Christ and entrusted to His Church. Hence the church may be expected to regulate the reception of this sac- rament. Although marriage is a sacred contract when entered into by any human.creature of God, it is a Sacrament in this sense only when received by a validly baptized person entering the contract according to the requirements of the Catholic Church. Since it is Catholic tradition that Christ established a church which He would protect from error in preaching His gospel, and that the Roman Catholic Church is this church, the social norm would disapprove of a member of this church "aban- doning his faith" even for the sake of establishing religious harmony in his marriage. Catholics acknowledge that one may be following his conscience in doing this and therefore not sinning, but as a general moral norm they would not approve of it. (With slight modifications, the Orthodox viewpoint would probably be equivalent to this.) The four basic religious types of marriage, from the point of view of social control by a religious group, are therefore the following: 1) Marrying someone from one's own religious group; 2) marry- ing someone who has become affiliated with one's own religious group; 3) marrying someone of a different religious group, without either partner changing; 4) marrying someone of a dif- ferent religious group and becoming affiliated with the partner's religious group. 191. These four types are arranged in rank-order, as a Catholic would view their decreasing desirability. The writer has distinguished these four types in his analysis of the data, but he has not at- tempted here to treat this variable as an ordinal scale--partly because of the disagreement about the endogamous norm of religious groups, as just stated. With these types of marriage in mind, the writer sorted out for inclusion in the study all the marriages in the surveys of 1955, '58, and '59 which involved at least one partner whose previous religious preference was either Protestant or Catholic. The detailed distribution of marriages is shown in Appendix B. Because the MISTIC computer program for CH12 and other values re- quired single-column fields for analysis, these detailed categories were collapsed into the following index: Religious Type of Marriage Both Catholic at marriage Catholic husband; wife changes Catholic wife; husband changes Catholic husband Protestant wife - no change Both formerly Protestant or Catholic; changed together Protestant husband, Catholic wife - no change Protestant wife; husband changes Protestant husband; wife changes Both Protestant at marriage All marriages involving one (not two) other than Protestant or Catholic at marriage, except those converting to Prot- estant or Catholic (in #‘2, 2, 3, 7, 8, above).17 Oxoooxnoxmwav—I 17The correspondence between these ten categories and the numbered categories in Appendix B is as follows, if the reader should wish to check any of them: #1 - ll, 14, 51, 54, #2 - 12, 17, 52; #3 - 21, 61, 71; #4 - l3: #5 — 22, 44, 55, 66; #6 - 31; 192. The index just presented was found to be unsuitable in that form for carrying out the cross-tabulations required in this study. The reader may recall that we are trying to make a double comparison: Marriages involving Protestants which are not mixed compared with those that are mixed; and marriages involving Catho- lics which are not mixed compared with those that are mixed. Therefore the writer constructed an index of marriages for each religious group, assigning all marriages in the sample to one category in each index, as follows: Marriages of Protestants Marriages of Catholics 9. Both Protestant at 1. Both Catholic at marriage marriage 8. Protestant with 2. Catholic with Prot- Catholic or other estant or other (mixed) (mixed) 5. Neither partner Egg 5. Neither partner Egg ' Protestant Catholic This simplified index for the marriages of each religious group made it possible to employ a dichotomy between mixed and unmixed marriages in each religious group by leaving #5 out of the analysis. (The code numbers used on the data cards have been retained here.) #7 - 43, 53, 83; #8 - 34, 35, 38, 68; #9 - 33, 32, 63; #0 - 18-19, 27-29, 37, 39, 47-49, 57-59, 69, 73-74, 81, 84, 9l-93, 95-96. . Categories in which no cases occur are not shown here. In collapsing the categories, a change in religious preference in the direction of one's spouse was assumed to have occurred at the time of marriage; the few cases of changing away from the preference of one's spouse were assumed to have taken place some- time after marriage. The categories shown are intended to re- construct the situation at the time of marriage. 193. The earlier index called "Religious Type of Marriage" served, in general, as a guide for the classification of marriages in these new indices--but not exclusively. In some categories of the old index, cases were examined individually to be sure that they were classified correctly in the new index. Always the previous reli- gious preference (if there was a change) was the principal basis for classification, since we are concerned with predicting the marriage choices of those who were socialized (reared) as either Protestant or Catholic. Perhaps it should also be noted here that a marriage involving a Protestant and a Catholic partner would occur twice, i;_g;, once in each of these indices. To allow for finer distinctions, the mixed marriages were subdivided in each index as follows: Marriages of Protestgnts 'Marrigges of Cgtholics 1. Both Protestant at marriage . Both Catholic at marriage 2. Protestant and changing . Catholic and changing partner partner 4. Mixed marriage still . Mixed marriage still 5. Protestant partner changes . Catholic partner changes 9. Neither partner was Protestant . Neither partner was Catholic NH \OUi-P As should be clear, #'9 2, 4 and 5 of each expanded index are simply three divisions of the mixed marriage category in the corresponding simplified index. Again, many of these were checked individually to be sure that they were classified as indicated. In spite of this care, the definition of our dependent vari- able is not yet complete. Although cross-tabulations were obtained while using the indices we have just described, an unforeseen prob- lem was uncovered. It was not enough to know, for instance, that 1we were analyzing in a given instance only marriages involving Prot- estants, whether mixed or unmixed. It was also necessary to know 194. whether the person answering a given question was a Protestant; in a mixed marriage we might be talking to the non-Protestant partner. This would be undesirable when we are trying to predict from back- ground characteristics of individuals to the religious type of mg;- giggg. Therefore it was necessary to construct an index for the religion (former) of the respondent, and a second index for the religion (former) of the spouse, so that these could be used as controls before attempting a cross-tabulation of the religious type of marriage with any other variables. Each of these indices is simple and looks like this: I ReligiOn (former) bf the - . * Religion (former) of the Respondent R's Spguse 1. Catholic I 1. Catholic 2. Other 2. Other 3. Protestant 3. Protestant Where the religion of the respogdent was given as "other," this ‘marriage would not get into the analysis at all--except for those few variables in which a background characteristic of the spouse is involved. Incidentally, after using this control, no mixed mar- riage will be counted twice (as would have been the case when using only the two indices for marriages of Protestants and Catholics), since each respondent will appear only once. This naturally gives us about half as many mixed marriages in any relationship as we would have had before. All cross-tabulations were rerun after using this control. The operational definition of our dependent variable is now «:omplete: we are trying to predict the marriage type of all mar- riages involving at least one partner who was (if not also now) 195. either Protestant or Catholicx-The independent variables are back- ground characteristics which have been selected because-of their relevance to the theoretical context of social control. Regretfully, the writer must report at this point an error in the classification of some of the marriages, which was not re- alized until the analysis was all but complete. It came about in this way: Early in the processing of the data cards, those mar- riages were eliminated from.the sample for whom the code was "Not ascertained" in response to the question, "What is your religious preference?" in the case of either respondent or spouse.' When it was discovered a little later that some of those whose present re- ligious preference was known were coded as "Not ascertained" in answer to the question about previous religious preference, the procedure to be followed did not seem so obvious. The decision was made to classify these in the general category of "Other," together with those few who had responded "No preference." This seemed reasonable if we assume that this lack of information was due to the indifference of the respondent rather than to the inefficiency of the interviewer, this being one assumption that does not hold up under examination. The writer had his misgivings much later and was not convinced-of the error until the comparison was made between the respondent's previous religious preference and his Cher)’ mother's religious preference (See Table 4-1). From these results we see that if the respondents of 1958 and 1959 had been classified by their mother's religious preference, only five out of the forty- one now classified as "Other" would still be "Other;" the rest ‘would be Catholic or Protestant. This means that these marriages 196. TABLE 4-2.--Distribution of misclassified marriages according to religious type of marriage, Detroit Area Study, 1955, 1958, and 1959. (Original) Code Religious Type Year of Survey Category) Number of Marriage .1955 1958 1959 Total (Total) 2. Catholic husband; wife ' changes 6 10 16 (75) 3. Catholic wife; husband changes 8 12 20 (65) 7. Protestant wife; husband changes 6 2 9 17 (65) 8. Protestant husband; wife changes 3 2 l 6 (49) 10. All marriages involving one other than Protest- 8 l 7 16 (57) ant or Catholic; except those in #‘s 2, 3, 7, 8, of code ___ ___ ___ Total 31 5 39 75 (311) were classified as "mixed" whereas they should have been classified as "unmixed.” ' If we Could undertake.a‘similar comparison between the former religious preference of the spouse and his (her) mother's religious preference, the results would probably be about the same. An examination of other characteristics of these cases has suggested the same conclusion, though not so conclusively: Almost all of I these have been misclasSified. Their distribution, by religious type of marriage, is shown in Table 4-2. How does the writer propose to correct for this error? In ‘two ways: When reporting simple rates (or ratios) of mixed mar- Iriages, these seventy-five cases will first be reclassified as sixty- Ifour unmixed and eleven mixed. (This is the distribution discovered ‘Vohen we assume that those whose previous religious preference was 197. "Not ascertained" have always held their present religious prefer- ence.) Secondly, when the results of cross-tabulations are reported; the writer will specify whether, and in what direction, the re8ults would be altered if these cases were reclassified. This will be possible because the writer separated out these seventy-five cases and obtained their distribution on every variable which will come into the analysis. Actually, the effect on the analysis should be slight, since the writer's practice of removing those classified as "other" by former religion means that forty-four of the misclass- ified cases will not appear in a cross-tabulation when we are ana- lyzing the respondents' characteristics, and the remaining thirty- one of them.will not appear when we are analyzing a characteristic of the respondent's spouse. Qperational Definitions: Independent Variables In this section we shall consider each sub-hypothesis in turn, defining the independent variable operationally if the hypo- thesis is to be tested by the data of the 1955, 1958, or 1959 sur- veys. If the hypothesis is to be tested only by the data of the 1962 survey, g;_gg, Hypothesis lb, we shall save all discussion of it until Chapter VII This latest survey is sufficiently unique to make it desirable to discuss it as a whole: sampling, operational definitions, and findings. There is one aspect of the concept of social control, as operationalized here, which should be mentioned before passing on to the first hypothesis of the study. We are considering social control as it is exercised upon members, not nonemembers, of a group. Groups do, of course, also influence non-members (e. g., 198. in Merton's "anticipatory socialization"), but we do not define such influence as social control. Social control is akin to the concept of systememaintenance, an internal mechanism; it is not a mechanism which is directly adjustive to the environment. Neither is the social control of a sub-society directly-adjustive to its larger society; gg_g., the social control exercised by a religious group on its members does not directly express its relationship to the larger'society.18 Our Hypgghesis la states: Protestant-Catholic merriages will be positively associated with having parents of mixed religions. Respondents in 1958 and 1959 were asked about the religious prefer- ence of their father and mother. When the same preference was re- ported for both (Protestant, Catholic, Jewish, Eastern Orthodox, etc.), their marriage was classified as unmixed. All others were classified as mixed. Different types of interfaith marriages were identified in these samples, but these cases were not numerous enough to be distinguished in our analysis. H thesis 1 , which can be tested only with the data of 1958, reads: Protestant-Catholic marriages will be negatively associated with naming parents as having had "the greatest influ- ence on your religious beliefs." The other precoded responses to this question were: friends; teachers; husband, wife, or chil- dren; ministers, priests, or rabbis; books, TV or radio; none of 18These comments were prompted by a suggestion from.John C. Scott (in private correspondence) that a type of marriage-in which one partner changes to the other's religidn "may result from effective control by the church of destination." There mey be such influence on a nonémember, of course, but in this thesis we do not define it as social control. 199. these. (The "not ascertained" were, of course, not included.) All of these-were combined into a Single category of "others." The writer's reason for this dichotomy was that the parents would most safely be assumed to represent agents of control for the re- ligious group in which the respondent was socialized. The "others" may also have been such, but they might just as easily have exerted their influence toward a different religious group. Hypothesis 2a lacks the element of prior time which would ‘make-it'satisfactory as a predictor of the religious type of mar- riage. It states: Protestant-Catholic marriages will be negatively associated with having a high proportion of relatives who are-of one's own faith. Since the respondents were asked about the reli- gion of the relatives they felt "really close to" pfpgp, rather than before, their marriage, the findings are more likely to repre- sent a consequence of, rather than a factor in, interfaith mar- riages. Nevertheless, this hypothesis was tested by fixing a di- chotomy between respondents who answered "all of them" and those who gave any other answer. In Hypothesis 2b we predict that: Protestant-Catholic mar- riages will be negatively related to the statement that friends or relatives would feel "unhappy or disturbed" or ”would try to dis- courage" one from changing to Protestant or to Catholic. In the interview, those who said that their friends or relatives would not try to discourage them were then asked whether these would feel at all unhappy or disturbed about it. The responses to the second question were not used in this study, since they were too few for analysis. A large majority of both religious groups said 200. that their friends or relatives would try to discourage them (65 per cent of Protestants, 86 per cent of Catholics); these were compared with the "unsure" and "would not try" responses. Again we have the disadvantage of interviewing our respondents gfpgpythe marriage about a matter closely related to the dependent variable which we are trying to predict. The fact that "friends and relatives" are combined in this question is a weakness also. However, in a study of Jewish-Gentile courtships we find reason to believe that friends do not generally support an endogamous norm: In conclusion, the interest we have taken in disapprov- ing friends should not obscure the fact that the large majority of our subjects were the nuclei of approving circles. Typically, the Reluctant's friends were sympathetic, and they did not. hesitate to display their feelings. Their support, in turn, was of some consequence in bringing about the marriage. Therefore we might expect positive responses to the question used in this hypothesis to refer to relatives rather than to friends. Regarding the church as an agent of social control, we stated in Hypgthesig 3g: Protestant-Catholic marriages will be i negatively related to one's education in church schools. This is one hypothesis, at least, in which there is no problem with the time element. However, as the writer was reminded in a letter from Dr. David 0.-Mbberg, "There is a wide variation of church schools."‘ But it seemm that our respondents were not confused by this varia- tion. In the survey of 1958, respondents were asked simply: "Did you get any of your education in parochial schools or other schools 19John E. Mayer, Jewish-Gentile Courtships: An Exploratory Study of a Social Process, New York: The Free Press of Glencoe, 1961, p. 198. 201. run by your church? (IF YES) How many years did you attend church schools?" It seems almost certain that those interviewed under- stood these questions to.refer to schools of general education operated by churches, rather than to Sunday schools or other types, since only 7 per cent of the 174 respondents or spouses who attended such schools for as long as four years were;ppp Catholic. There- fore only the responses of Catholics were analyzed for this hypo- thesis. Those who had attended church schools for less than four years were placed in the category of those who had not attended at all because this was judged too short a time for any appreciable degree of social control to be experienced by comparison to an equal or greater number of years in public schools. Hence we have two categories for Catholics attending church schools: four years or more yp. less than four years. 'The same index was constructed for Catholic men as for Catholic women, whether they were respond- ents in the interviews or their spouses. Therefore a control for sex is built into the test of the hypothesis. Our Hypothegig 5a predicts that: Protestant-Catholic mar- riages will be positively related to being an American of the second and subsequent generations. As we have noticed earlier, writers about immigrants and their children have not all agreed in their definition of second- or third-generation Americans. How- ever, there seems to be a growing consensus to use the terms in the following way, as we do in this thesis: First-generation American-- any immigrant born in another country; second-generation American-- anyone born in this country, whose father and/or mother were born elsewhere; third-generation American--anyone born in this country, 202. both of whose parents were also, with one or more grandparent being an immigrant. Therefore we are using the minimum bond to a foreign country as the standard for our definition. These definitions made it impossible to use the data of the 1955 survey for this hypothesis, since the interview schedule asked for information about birthplaces only on the paternal, not the maternal, side of the family. Now we come to the anomic situations. Hypothesis 5a reads: Protestant-Catholic marriages will be positively related to getting married during war years. The years selected as war years were those in which the united States was at war (in the Korean War as the major participant in the United Nations effort), not the total: length of each war--since the effect on the behavior of young people in this country is what we have in mind. Therefore the war years, as defined in this study, are: World War 1, 1917-18; World War 11, 1941-45; Korean War, 1950-53. For this hypothesis it was essential first to remove all but figpp marriages from consideration, since the religious type of the marriage which the respondents were in when interviewed is what we want to predict in our data. The date of marriage had to be determined by subtracting the number of years each respondent was married from the year in which the interview took place. This hypothesis could be tested only with the data of 1955 and 1959 because the writer had failed to get the information about the length of marriage which was obtained in the 1958 inter- views. In the hypothesis concerning age at marriage also, only first marriages could be included--for the same reason as in the preceding one. We are predicting in Hypothesis 5b that: Prot- 203. estant-Catholic marriages will be positively related to getting "married at a late age, with consequent emancipation from home. In the coding process, ages were grouped into five-year intervals. Hence, to determine the age at marriage by subtracting the years of this marriage from the age-of the respondent when interviewed, we had to make use of the*mid-point in each five-year interval as a starting point. The intervals coded and the mid-points used were as follows: Interval Mid-point Interval Mid-point Interval Mid-ppint 21-25 yrs. 23 40-44 yrs. 42 60-64 yrs. 62 26-29 yrs. 27.5 45-49 yrs. 47 65 and 30-34 yrs. 32 50-54 yrs- 52 over 67 35-39 yrs. 37 55-59 yrs. 57 To randomize the effect of the decimal in the mid-point 27.5, the writer first subtracted the years of marriage from this number and then "rounded up" for half of the cases in this category and "rounded . down" for the other half--to get a whole number each time. Another factor which made it particularly tedious to construct this index of age at marriage was the fact that the surveys of 1955 and 1959 gave slightly different values to the code numbers they used to express the length of the marriage. For example, in 1955 the code "01" meant "I to 1.9 years," whereas in 1959 "01" meant "1 year or less." To make these codes more nearly equivalent, the writer added "1" to the Egg; number for each category of the 1955 distribution. Furthermore, the interval "65 years and over" caused trouble. A mid-point of 67 was selected to be consistent with the other cate- gories and because it was thought that the rising death rate at these ages would warrant this mid-point. Using the mid-point of 67 years for those who had been married over fifty years made for 204. some very early marriages. A check of the 1950 United States Census revealed that 71.3 years would be a more reasonable median age for those "65 years and over." However, it did not seem worth- while to change the data cards for these relatively few cases. Through these steps a code giving the "exact age" at first mar- riage had been constructed, but this had to be collapsed into a "single-column field" in order to use Sim's program on the MISTIC computer. After this was done, the frequency distribution was still quite abnormal for age at first marriage, so it was decided not to try to test the hypothesis with these data. We must blame the five-year intervals in our data for making our index of age at first marriage a decidedly unreliable base for testing our hypo- thesis in this case. However, we can still test the hypothesis by using the data of the 1962 survey, when the respondents were interviewed within a year after their marriage. We shall consider these data and the results of the test in Chapter VI. Our Hypothesis 5c states: Protestant-Catholic marriages will be positively related to migration from a farm before age 18. Our data are well suited for testing this hypothesis. Not only do we have migration data from the three large surveys, but we have appropriate categories, leaving out the foreign born: Detroit area native Urban migrant (from place of 50,000 or more) Small town migrant (from place of 49,000 or less) . Rural migrant before age 18 Rural migrant after age 18. UIJ-‘(JJNl-l Only for the rural migrants do we have the important determination of age, which would lead us to expect more intermarriage in cate- gory #4 than in #5, since most of the latter are likely to have 205. married before experiencing the anomic situation of rural-urban migration. However, even without knowing the time-of migration for categories #2 and #3, the first three categories will give us inter- esting points of comparison. The surveys of 1959 and 1962 give us data for testing Hyppthesig 5d, but they cannot be combined in analysis. In each case we shall be faced with a sample size that is less than satis- factory. The hypothesis states: Protestant-Catholic marriages will be positively related to the husband's extreme upward mobility in education, compared to his father. The categories used are fairly self-explanatory: Father and son both grade school (or none) Father and son both high school Father and son both collegej Father grade school, son high school Father high school, son college Father grade school, son college Father college, son high school Father high school, son grade school mNGUIJ-‘UJNH One year or more of education at a higher level of education was considered enough to place the individual at that higher level. One possible category--father college, son grade school--did not occur in either sample; only four cases of category #8_occurred in 1959, none-in 1962. Categories #5 and particularly #6 are those A in which we look for more interfaith marriages. Finally, Hypothesis 5e predicts that: Protestant-Catholic marriages will be positively related to the husband's extreme up- ward mobility in occupation, compared to his father. As we saw above, Lenski's study prompts us to make the same prediction for the downwardly mobile. Again we have the pleasant possibility of testing this in all three of the large samples. Our criterion for 206. vertical mobility is based on a higher rank for white collar than for blue collar occupations. In the case of either fathers or sons, white collar occupations include these categories: profes- sional, technical, and kindred workers; managers, officials and proprietors, except farm; clerical and kindred workers; sales workers. The blue collar occupations are: farmers and farm.man- agers; craftsmen, foremen and kindred workers; operatives and kindred workers; private household workers, service workers; farm laborers and foremen; laborers, except farm and mine. Putting these together, we arrive at the following index of occupational mobility: 1. Father white collar, son white collar 2. Father blue collar, son blue collar 3. Father white collar, son blue collar 4. Father blue collar, son white collar If these gross distinctions support our hypothesis, it may be worthwhile to examine finer distinctions among occupations in the future. Stages ip_the Analysis One reason for including this last Section in the chapter is to show what difficulties may occur when one tries to combine the data of several different surveys, even assuming that the ‘sampling methods make it legitimate to do so. It should become :clear through this narrative that the great advantage of having access to data which have already been gathered for other purposes is considerably diminished by the problems of adjusting these data to a new research design. Another reason for this section is to present in simple succession the steps of the analysis. 207. we have already noted the necessity of restricting the sample to first marriages when testing hypotheses 5a and 5b. If we had been able to gather our information at the time of marriage, this restriction would not have been necessary for hypothesis 5a. We have also noted that some questions asked (or the answers coded) are not perfectly suited to our purposes, such as the mix- ing of "friends and relatives" as agents of social control. There are numerousother shortcomings in the data. Of course, even so- cial researchers who gather their own data must often adjust to something less than the ideal. Leaving aside the survey of 1962, which must be treated separately, we began with four surveys which included information about the religious preferences of husbands and wives and which, therefore, might be useful for this study of interfaith marriages. The figgt stage in the analysis was to list each of the variables for which there were measurements in the data, whether these variables were expressed in the hypotheses or might be used as controls. The writer compiled a separate list of the variables which appeared in all four surveys, other lists for those that appeared in each combination of three years, each combination of two years, and each single year. This was a kind of bookkeeping operation to display what it was possible to do with the data. Next to each variable was indicated the column number of the data cards where it could be found. Furthermore, the level of measure- ment which each variable represented was noted, 1. e., whether it 208. was a nominal, ordinal, or interval scale.20 This was felt to be necessary in order to allow for consideration-of the various sta- tistical tests and measurements which might be appropriate for each hypothesis. Most of our variables are only nominal scales, so it seemed best to proceed with the CHI2 computer program. If the distribution in the cells of any of these tables seemed to warrant it, we could later apply a more sophisticated test or measurement. The second stage in our analysis was an effort to see what would be-gained or lost by including the data of 1956 in our study, since previous religious preference had not been determined in that year and consequently no evidence about those who had changed re- ligious preference. This survey could be-combined with the other three in testing the hypotheses concerning occupational mobility and rural-urban mobility, and it could be combined with the sur- veys of 1958 and 1959 in testing the hypothesis that interfaith marriages will increase among Americans of the second and sub- sequent generations. Each of these hypotheses was, in fact, tested by a CHI2 table for all of these surveys individually. The data from 1956 did not show a consistent pattern of extreme distribu- tions, which might have made it important to retain this informa- tion. It was decided to leave this survey out of the study and to use only those in which interfaith marriages could be identified by "previous religious preference." The ppipg step of the analysis involved combining variables into the indices which were needed to test the various hypotheses. 20For a description of these terms, see Hubert M. Blalock, Jr., Social Statistics, New York: McGraw-Hill, 1960, pp. 11-15. 209. Then each index or important variable was cross-tabulated with our index for Religious Type of Marriage. This was not a test of the hypotheses in any strict sense, but it was to give us an idea of the approximate distribution of the variables when all cases were considered together. In just a few instances it was evident at this point that the number of cases would not be adequate for testing one or another hypothesis. The next step was to construct a separate index for mar- rriages'of Protestants and for marriages of Catholics, both a simple and an expanded index. After CHI2 tables had been obtained with each index of this type, there came the realization which led to the fifth and final stage of the analysis. The realization was that we must distinguish whether the information we have in any variable applies to the respondent or his (her) spouse. Since a ‘ Protestant-Catholic mixed marriage involves one partner.of each faith, it is essential that we know'which one is responding to a given interview question. Otherwise we would find ourselves try- ing to predict the marriage choices of Protestants from the char- acteristics which belong to Catholics. Therefore Protestant re- ‘spondents (or spouses) were separated from Catholic respondents (or spouses) when CHI2 tables were again obtained through the -computer program. Those respondents (or spouses) who were in the "other" religious category but were married to a Protestant or a Catholic did not enter into the analysis at this stage. By this time we were quite sure of the interpretation of our data. Some other cross-tabulations, not called for in the hypotheses, were obtained also out of curiosity. 210. The description of procedure which we have concentrated in this chapter do not make for easy reading. ‘Onlyma reader who is himself interested in doing some‘similar research is likely to persevere this far. However, it is important to go into such details to allow for scientific evaluation of this study and per- haps for its replication. For this purpose we have reviewed here 'the intellectual journey which arrived at the plan of the study; we have described the Detroit Area Study and the characteristics of its survey data; we have specified how the sample was restricted for our study; we have presented operational definitions for the dependent and the independent variables; finally, we have outlined the principal stages in the analysis of the data. Now nothing re- ‘mains for us to do but to examine the findings of this study. CHAPTER V DETROIT AREA SURVEYS IN COMBINATION In this chapter we shall present the findings of this study relative to the interfaith marriages of Protestants and Catholics, based on the data of the Detroit Area Study surveys of 1955, '58, and '59. The last two were principally studies of religion, for which the sampling methods were almost identical. There can be little doubt about the validity of combining these data. The survey of 1955 was a family study in which only the wives were interviewed. There is no reason known to the writer why this fact, that husbands were not also interviewed, should argue against its being combined with the other two surveys in testing the four specific hypotheses for which its data are used in this chapter. As was stated in the preceding chapter, the sampling method used for the 1955 survey gives little reason to fear distortion in combining its results with those of 1958 and 1959. Let us recall that we are really studying two sub-samples in the p0pu1ation of the Detroit area: white Protestants who are now married and living with their Spouse, and white Catholics who are now married and living with their Spouse. We are concerned to describe and in some measure explain the extent of interfaith mar- riages among those who have grown up in these religious groups. Since the norm of religious endogamy does exist in these two 211. 212. Christian sub-societies of our country, we shall consider its violation partially explained if we find that the deviants were less under the influence of the agents of social control than were others or if they lived in situations which would be likely to weaken norms such as this religious one. Let us begin with the rates of intermarriage against the background of the literature we have reviewed in Chapter II. Rates and Ratios of Intermarriagg Even in the introductory chapter of this study, the read- er was alerted to the danger of misreading intermarriage rates and applying_them to individuals rather than to marriages. Again in Chapter II we called attention to the factor of size of group as being more of a mathematical than a sociological explanation of intermarriage rates. Now, as we come to present the patterns of interfaith marriages found in our data, the reader is again reminded that there are many ways of expressing these facts and that caution is required before comparing (without adjusting) the rates reported-in different studies. This applies as well to the rates of religious intermarriage displayed in Tables 2—3, 2-4, and 2-5. We believe that Paul C. Glick used an effective way of pre- senting intermarriage rates--by computing also the "expected" per cent of intermarriage if the members of each group had married randomly, from which he arrived at theurggig of actual to expected mixed marriages for each religious group.1 Click apparently made w v V v—v— 1"Intermarriage and Fertility Patterns among Persons in Major Religious Groups," Eugenics Quarterly, 7 (1960), pp. 31-38. 213. his calculations from a CHI2 table of eXpected frequencies. The source of his observed frequencies was a report by the United States Bureau of the Census from its national survey of religion in March, 1957.2 Even though we admire Glick's method of computing_a ratio of mixed marriages, the differences in our data will not allow us to present a directly comparable statistic. His respondents re- ported only their present religious preference; we are more in- terested in the mixed marriages of people raised in different religious groups, including those whose present religious prefer- ences are not mixed. His sample includes only couples in which both partners report one of the three major religions; ours includes all married Protestants and Catholics, but only those of other re- ligious categories who are their spouses. His sample does not exclude non-whites; ours does. The Manhattan study by Jerold S. Heiss, which has frequently been cited in earlier chapters as influencing our study design, can also be contrasted with ours.3 He reported intermarriage rates for Jewish reapondents (who are relatively numerous in that community), whereas Detroit's Jewish p0pu1ation was too small for analysis in this study. Heiss combined the Eastern Orthodox and others into a fourth, residual category, and computed intermarriage rates for this category also, which we have not done. However, his sample 2"Religion Reported by the Civilian POpulation of the United States: March, 1957," Current POpulation Reports: Population Charac- teristics, Series P-20, No. 79, Table 6. 3“Premarital Characteristics of the Religiously Intermarried in an Urban Area," American Sociological Review, 25 (1960), pp. 48-49. 214. includes only whites, and the rates are computed according to "early religion"--so our data for Protestants and Catholics should be comparable to his. Table 5-1 presents the findings of this study after the manner Glick used in his national report for 1957. There are a number of things for us to notice about this table. The overall mixed marriage rate for Protestants in this combined sample of the TABLE 5-1.--Actual intermarriage rates and intermarriage rates "expected" if husbands and wives were distributed at random, for marriages of Protestants and Catholics in the Detroit area, 1955, 1958, and 1959. Married Couples "Expected" Per Ratio of Actual Early Religion of Actual Actual Cent if Random to "expected" Husband and Wife Number Per Cent Intermarriage Mixed Marriages One or both Protestant 935 100 100 -—- Protestant-- Protestant S99 64 32 Catholic or other 336 36 68 .53 One or both Catholic 838 100 Catholic-- Catholic 495 59 27 Protestant or other 343 41 73 .56 Zetrcit area is 36 per cent; for Catholics it is 41 per cent.4 These rates are both five per cent higher than those reported 4The chances are 95 out of 100 that each of these percent- ages is accurate for the p0pu1ation within four points. 215. earlier for the 1959 data alone by this writer.5 They are also higher than were reported by Heiss for his Manhattan sample, which were 33.9 per cent for Protestants and 21.4 per cent for Catholics. These differences may be due largely to the relative size of the religious groups in these metrOpolitan areas. Let us now consider the-expected per-cent of intermarriage if Protestants and Catholics had chosen their marriage partners randomly among whites in the Detroit area. Table 5-1 shows that the intermarriage rates then would be 68 per cent for Protestants and 73 per cent for Catholics. The writer believes that he has replicated Glick's statistic reported in the article already cited, although the latter did not give the details of his procedure. Since in our study anyone who had not married a white Protestant or Catholic was eliminated from the sample, we did not have a true basis in our data for computing the "eXpected" intermarriage rate for each religious group. Fortunately, another study using the data of the Detroit Area Study, 1954-59, has categorized the re- igious preference of all the adult respondents by race and denom— ination. These data are the basis for the expected intermarriage rates for white Protestants and Catholics in the Detroit area as 5"Unbroken Protestant-Catholic Marriages Among Whites in the Detroit Area," American Catholic Sociological Review, 23 (1962), p. ll. The corrected rates for mixed marriages found in this re— analysis of the 1959 data alone (27 per cent for Protestants, 29 per cent for Catholics) are lower than were reported in that article (31 per cent for Protestants, 36 per cent for Catholics). These differences can be accounted for by two facts: The data of that article do not include forty-nine interviews which were completed after the others. The correction for those whose former religious preference was "not ascertained" was not made in that article, but was made in the remanalysis as explained in Chapter IV. The correc- tion lowers the rate of mixed marriages for both Protestants and Catholics. 216. found in Table 5-1.6 The Catholic p0pu1ation in the Detroit area-is more nearly equal to the Protestant p0pu1ation than it is in the country as a whole. Nevertheless, it is smaller. Computing the ratio of actual to eXpected mixed marriages for the two groups has the effect of narrowing the difference between the measures of their intermar- riage behavior. The five-point difference in the actual percent- age is reduced to a three-point difference in the ratios. This is the adjustment for the "size of group" factor which the writer believes is needed before comparing intermarriage rates for differ- ent groups and different localities. The ratio of actual to eXpected mixed marriages reported by Click for Protestants was .19; for Catholics it was .26. Our ratios are .53 and .56, reapectively. Why so much larger? If we use the same method to compute these ratios for our data when con- sidering only the present religious preference of the respondents 6See Albert J. Mayer and Harry Sharp, "Religious Preference and Worldly Success," American Sociological Review, 27 (1962), p. 221, Table 1. They report 3,789 Protestants, 3,307 Catholics, and 642 others, besides the Negro demoninations. As an example for the reader's guidance, the following are the calculations used to determine the expected intermarriage rate for white Protestants in the Detroit area: Religion of Wives Religion of Husbands pl,895 x 1,895 = 928 Protestant Other All 3,870 Protestant 928 967 1,895 Marriages of Protestants: 928 = .32 928 Other 967 1,908 1,975 2,862 967 967 All 1,895 1,975 3,870 1.00 - .32= 68 2,862 217. and their spouses, as Click did for the national data, then the ratio becomes the same for both white Protestants and white Catho- lics in Detroit--.22. This is certainly comparable to the ratios which Glick computed from the national data. .If he had not includ- ed non-whites in his computations (heavily Protestant), there is no doubt that our Detroit ratios and his national ratios for Protestant and Catholic mixed marriages would be even closer. It would be possible to look for a trend in the rates of interfaith marriages in the Detroit area.> In the article already cited, this writer did report a moderate trend toward more inter- marriage, statistically significant at the .05 level for the Prot- estant sample, almost so for the Catholics. This trend appeared only when the marriages of 1944-58 were compared with all earlier marriages. As was noted at that time: If we could better approximate a truly longitudinal study, we might find that the trend in the gross statistics is an arti- fact of the high rates of divorce and separations or deser- tions that occur in mixed marriages. In other words, the ranks of unbroken mixed marriages (which are the only ones we have been considering) may have become "thinned out" by the time they reach their 16th anniversary. As noted above, three studies have shown a marked increase in the divorce rate of mixed marriages.7 The point we wish to make now is: Although there may be a trend toward more interfaith marriages in Detroit and in the United States, this should be established from information (such as marriage records) dating from the time of marriage rather than from a sample of couples now living together. Furthermore, changes in the relative size of religious groups in a given locality must be taken into 7"Unbroken Protestant-Catholic Marriages Among Whites in the Detroit Area," op. cit., p. 14. 218. account in interpreting what appears to bars trend in the rate of intermarriage. In other words, if Protestants have become a relatively smaller group in a given metrOpolitan area such as Detroit, we should not be surprised that their unadjusted rate of intermarriage has increased. Types of Religious Intermarriggg As was explained in the preceding chapter, we did distin- guish between types of religious intermarriage within the general category of Protestant or Catholic mixed marriages. In terms of social control, the three descending steps in Table 5-2 represent increasing weakness of social control on the part of the religious group in question--according to the Catholic and the fundamenta- list Protestant view of the endogamous norm which we described in Chapter IV. We can see that there seems tobe a slight tendency for Protestants to change their religion when they intermarry, more than Catholics do; but this difference is not statistically signi- ficant. On the other hand, about three out of five marriages in each religious group which were mixed originally have become re- ligiously homogeneous, through one partner's change of religious preference. According to the liberal Protestant view of the endog- mous norm which we also described in Chapter IV, this pattern is preferable to having married partners remain religiously mixed. While we are on the subject of types of interfaith marriage, we have separated out the Protestant groups which are more numerous and have computed the rate at which they intermarry with non-Prot- estants. Table 5-3 shows what these rates are. The rates of Pres- byterians, Methodists, and Lutherans are intermediate and almost 219. TABLE 5-2.-~Types of interfaith marriages among Protestants and Catholics in the Detroit area, 1955, 1958, and 1959. Types of Interfaith Marriage . Number Per Cent Protestant Respondents: Protestant unchanging; spouse changesa - 87 26 Protestant with Catholic or other; no change - 138 41 Protestant changes; Spouse unchangingb - .111 33 Total of Protestant mixed marriages - 336 100 Catholic ReSpondents: Catholic unchanging; Spouse changesc - 105 31 Catholic with Protestant or other; no change — 133 39 Catholic changes; Spouse unchangingd - lg; 31 Total of Catholic mixed marriages - 343 101 8‘Does not include the nine cases in which both were Prot- estant, but one changed. These were classified as unmixed Protestant. blncludes five cases in which both were Protestant, but both changed together. CDoes not include the seven cases in which both were Catho— lic, but one changed. These were classified as unmixed Catholic. dIncludes fourteen cases in which both were Catholic, but both changed together. identical with each other. Protestants of no denomination and Episc0pa1ians are most likely to marry non-Protestants; the Bap- tists, least likely. We shall not go beyond this in analyzing the intermarriage patterns of Specific Protestant denominations because this would constitute another Study in itself. 220. TABLE 5-3.--Se1ected Protestant groups and the rate at which they intermarry with Catholics and others in the Detroit area, by early religious preference of respondents, 1955, 1958, and 1959. Early Religion of Spouse Early Religion of ReSpondent Protestant Non-Protestant Total Protestant Per Cent - 56 44 100 (no denomination) Number - (23) (18) (41) Episc0palian ‘Per Cent - 65 35 100 Number - (44) (24) (68) Presbyterian Per Cent - 72 28 100 Number - (67) (26) (93) Methodist Per Cent - 72 28 100 Number (120) (47) (167) Lutheran Per Cent - 74 26 100 Number - (107) (37) (144) Baptist Per Cent - 80 20 100 Number - (100) (25) (125) Rates Based on Individuals Rather Than on.Marriagg Rates of interfaith marriages are almost always expressed as a percentage of marriages involving adherents of a particular religious group. However, they are sometimes interpreted by , others as the percentage of individuals in the religious group who intermarry, as when Bossard and Boll misinterpreted the rates of John L. Thomas' study by saying, "It is conservative to say that today, each year, at least one-half of all Catholics marrying find their mates outside the Roman Catholic Church."8 Therefore 8James H. Bossard and Eleanor S. Boll, One Marriage, Two Faiths: Guidance on Interfaith Marriage, New York: Ronald Press, 1957, p. 55. 221. TABLE 5-4.--Types of interfaith marriages related to the sex of Protestants and Catholics in the Detroit area, 1955, 1958, and 1959, by percentage.a Types of Interfaith Marriage Husbands Wizgg Protestant Respondents: Protestant unchanging; Spouse changes - 27 25 Protestant with Catholic or other; no change - 46 A 37 Protestant changes; Spouse unchanging - _21_ '_31 100% 99% Number of cases - 149 187 Catholic Respondents: Husbands Wiggg Catholic unchanging; spouse changes - 35 26 Catholic with Protestant or other; no change - '37 41 Catholic changes; Spouse unchanging - _28_ _23 100% 100% Number of cases - 175 168 aThe notes for Table 5-2 also apply to this table. we had better express our data in terms of individuals also, in order to forestall such a misunderstanding. It should be evident that the 599 unmixed Protestant mar- riages in our combined sample involve twice that number of Protes- tants, whereas the 336 mixgd marriages of Protestants involve just 336 Protestants. Therefore these 336 make up just 22 per cent of the total number of married Protestants--even though, as we have stated earlier, the rate of mixed marriages for white Protestants in the Detroit area is 36 per cent. In the former percentage we 222. are talking about individuals; in the latter, about marriages. Similarly, the 343 Catholics in mixgg marriages (compared to the Catholics in the 495 unmixed marriages) comprise just 26 per cent of all married Catholics--considerab1y lower than the 41 per cent mixed-marriage rate for white Catholics in the Detroit area. Keeping these rates of individuals in mind makes it easier to evaluate the sex differences in rates of intermarriage. Among Protestants, women seem slightly more inclined to intermarry than the men, having an intermarriage rate of 24 per cent‘gg. 20 per cent for men. Among Catholics, the intermarriage rates are almost identical-~26 per cent for men and 25 per cent for women.9 Neither difference in percentages is statistically Significant. There appear to be some interesting sex differences within the category of those who have intermarried. AS we see in Table 5-4, the largest percentage differences for men are these: 27 per cent of Protestant husbands remain Protestant while their wives change religious preference, compared to the 35 per cent of Catho- lic husbands who remain Catholic while their wives change. There is practically no difference between ProteStants and Catholics in the percentage of husbands who change their religious preference. The balancing difference appears when we see that Protestant husbands are more satisfied to keep the marriage a religiously mixed one (46 per cent) than are the Catholic husbands (37 per cent). For 9The revised analysis of the data for the 1959 survey by itself Shows no sex difference whatever. Therefore the earlier re- port by this writer that Catholic women outnumbered Catholic men in intermarriage by a ratio of five to four must be revised to Show no difference. Reasons for the new adjustment are probably the same as those stated in footnote #6, above. 223. the wives there is almost no difference between Protestants (25 per cent) and Catholics (26 per cent) in the rate at which they maintain their religious preference while their husbands change. However, the Protestant wives seem to Show a slight tendency to change their own religious preference (37 per cent 13. 33 per cent), whereas Catholic wives are a little more inclined to con- tinue in a mixed marriage. (41 per cent‘xg. 37 per cent). It must be noted, nevertheless, that none of these sex differences is statistically Significant, since the size of the Sub-samples is relatively small. Some Characteristics of Those Who Intermarry When Ruby Jo Reeves Kennedy wrote her famous article on the "triple melting-pot", she was saying that people who marry outside their ethnic group tend to choose marriage partners with- in their religious group, thus giving rise to a melting-pot for each of the three major religions.10 In her data She classified all Irish, Italians, and Polish as Catholic; but all British-Amer- icans, Scandinavians, and Germans as Protestant. Her data did not distinguish between Protestant and Catholic members of the same nationality. This we can do in the Detroit data, as is shown in Table 5-5. Perhaps She could safely assume in New Haven that all Irish were Catholic and all Germans were Protestants, but we cer- tainly could not do so in Detroit, where these nationalities are well represented in both religious groups. 10"Single or Triple Melting Pot? Intermarriage Trends in New Haven, 1870-1940," American Journal of Sociology, 49 (1944), pp 0 331-339 0 224. TABLE 5-5.--Religious intermarriage rate (per 100 marriages) by early religious preference of respondents with selected national torigins on father's side, 1955, 1958, and 1959. National Origina Catholic Respondents Protestant ReSpondentS of ReSpondent , on Father ' S S ide R3123 M _R_a_t_:_e_ ME English 25.2 154 Scotch, Scotch-Irish 22.0 59 Irish 37.1 70 22.6 7 93 German 33.3 63 17.5 154 Polish 19.2 182 Italian 32.9 76 8This table includes only those nationalities which were claimed by more than fifty descendants in one of the major re- ligious groups. Statistics which are comparable to those in Table 5-5 are rare in the literature on intermarriage. As we have seen earlier, John L. Thomas found, in an extensive Study of Catholic parish records, about the same rate of intermarriage as we have for Germans; but, for the Italians, Irish, and Polish, it was below ten per cent. The rates for religious intermarriage in Manhattan presented in the doctoral thesis by Jerold S. Heiss are directly comparable to ours for Detroit. Of those intermarriage rates which he reports, the following are based on sub-samples of more than fifty: Irish Catholics-~12.5, German-Austrian Catholics--29.5; German-Austrian Protestants--42.7, Italian Catholics--15.3.11 11"Interfaith'Marriage in an Urban Area," unpublished Ph. D. dissertation, Indiana University, 1958, p. 107. 225. Perhaps the principal factor in detenmining these rates is the re- lative size of the ethnic group in the population. Unfortunately, we cannot compute a ratio of actual to expected rates of intermar- riage as we did for interfaith marriages as a whole, since we do not have information about the numerical distribution of national- ity groups in the Detroit area. As we expected (See Appendix A), the Polish Catholics are a large enough group in Detroit to exert an influence toward minimizing the general Catholic rate of inter- faith marriages. It is surprising to this writer, however, to discover that the German Protestants in Detroit exert the same influence on the intermarriage rates of Protestants. In an earlier part of this thesis we discussed evidence showing that, for Catholics, interfaith marriage is positively associated with high rental values and with high socio-economic status, as well as being associated with belonging to the lowest occupational groups. Our data allow us to make some comparisons now for the variables of income and occupation. Table 5-6 shows the income levels of the heads of families in our combined sample. The intermarriage rate for those at each level is diaplayed accord- ing to the religion of the respondent. (It would not seem to in- fluence the table whether husband or wife gives us the situation, since we are here concerned with the relationship of the type of marriage to the level of income for the family.) Perhaps the most striking thing about Table 5-6 is that there is so little variation among the Protestant rates of inter- marriage (ranging from 21.4 to 27.0), by contrast to the maverick Catholic rates (ranging from 13.6 to 42.1). However, the Catholic 226. TABLE 5-6.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to total annual income of family head, 1955, 1958, and 1959. Rates of Intermarriagg Income of Family Head Respondent is: Protestanta Catholicb $8,000. and over. 24.9 (177)c 39.2 (120)c $7,000 - $7,999. 27.0 ( 74) 42.1 ( 57) $6,000 - $6,999. 26.0 ( 96) 31.8 (107) $5,000 - $5,999. 23.5 (145) 29.5 (139) $4,000 - $4,999. 24.2 (120) 22.0 ( 91) $3,000 - $3,999. 23.0 ( 61) 15.3 ( 59) $2,000 - $2,999. 25.7 ( 35) 29.4 ( 34) $1,000 - $1,999. 25.0 ( 28) 13.6 ( 22) Under $1,000. 21.4 ( 14) 33.3 ( 15) Totals - 24.7 (750) 30.0 (644) cm2 (8 d. f.) .65 20.76 p <:.99 <:.01 v2 .001d .032d aTwelve cases in this column, normally distributed as to income, were misclassified as mixed marriages. bFifteen cases in this column, normally distributed as to income, were misclassified as mixed marriages. CNumbers in parentheses indicate the number of cases in each category. dV2, like the more conventional PHIZ, gives us a measure of Strength of relationship which standardizes for the number of cases contributing to each CHIZ. Its upper limit is 1.0. Unlike PHIZ, its upper limit will still be unity if we come upon a table with more than two columns. .Cf. Hubert M. Blalock, Jr., Social Statis- tics, New York: McGraw-Hill, 1960, p. 230. 227. TABLE 5-7.-—Rates of intermarriage (per 100 marriages) for Protest- ants and Catholics in the Detroit area, related to the occupation of the family head, 1955, 1958, and 1959. Rates of Intermarried Husband's Occupation Respondent is: Protestanta Catholicb Professional, technical workers 24.2 (124)c 53.2 ( 79)c Managers, prOprietors, except farm 25.9 (108) 30.9 ( 81), Clerical workers 23.7 ( 59) 29.8 ( 57) Sales workers 22.5 ( 49) 25.6 ( 39) Craftsmen, foremen 23.6 (220) 25.5 (208) Operatives 26.1 (161) 25.2 (139) Private household and service workers 34.6 ( 26) 33.3 ( 33) Laborers, except farm and mine 38.5 ( 13) 17.7 ( 11) Totals - 25.1 (750) 30.0 (644) CHIZ (7 d. f.) 3.15 25.53 p (.90 < .001 v2 .001 .040 aTwelve cases in this column, quite normally distributed as to occupation, were misclassified as mixed marriages. bFifteen cases in this column, quite normally distributed as to occupation, were misclassified as mixed marriages. CNumbers in parentheses indicate the number of cases in each category. rates are not simply erratic. After varying irregularly at the lower income levels, where the number of cases is uniformly small, they show a tendency to climb with the level of income. This re- lationship is statistically significant at the .01 level of proba- bility. It is true that the number of cases shows that Catholics 228. are underrepresented at the highest levels of income, and this helps to account for their higher rates of interfaith marriage at those levels. However, the regular climb in rates seems to say that there is something involved here more than simply the relative size of groups. Almost the same comments can be made about Table 5-7 as were made about Table 5-6, except that here both religious groups show somewhat more variation in their rates of intermarriage at different occupational levels. Of course, the concentration of cases in the categories of craftsmen, foremen, and Operatives is a reflection of Detroit's industrial complex. The intermarriage rates at the lowest occupational levels are unusual for both re- ligious groups, but the small number of cases gives us a basis only for a suspicion that Protestants tend to intermarry more in these occupa- tional groups. The climbing rates of interfaith marriage appear again for Catholics in the white collar occupations, especially in the professional and technical category. (This distribution for Catho- lics is statistically Significant at the .001 level of probability. Their relatively small numbers in the two most prestigeful occupa- tional categories help here also to explain the increased rates of intermarriage, without giving a full explanation. Social Control of Interfaith Marriages by Parents One agent of social control which, as we have found reason to believe from other Studies, can be expected to support the norm of religious endogamy is the parents. As one test of the more general hypothesis that the norm is more likely to be violated when parents do not exercise social control, we hypothesized: 229. TABLE 5-8.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to interfaith marriages of their parents, 1958 and 1959. Religious Type of Rates of Intermarriage Parents' Marriage Amonngheir Children A ReSpondent is: Protestanta Catholica Parents not mixed - - 24.2 (385)b 28.8 (365)b Parents mixed 25.0 ( 44) 57.1 ( 28) Totals 24.2 (429) 30.8 (393) cm2 (1 d. f.) .02 9.83 p - <.99c <.002c v2 .002 .025 f 8The ten Protestant and fifteen Catholic marriages of the children which were misclassified as mixed almost all occur in the 1959 sample, in which our hypothesis is not supported for either religious group. They may obscure the relationship here somewhat. The 1958 sample, in which the hypothesis is supported for Catholics at the .001 level of probability, would be unaffected by reclassfi- cation. b . . Numbers in parentheses indicate the number of cases in each category. cThis level of probability is based on a one-tailed test of significance, since the direction of the relationship had been pre- dicted. Cf. Hubert M. Blablock, Jr., Social Statistics, New York: MCGraw-Hill, 1960 , P a 2 18 o la. Protestant-Catholic marriages will be positively associat- ed with having parents of mixed religion. The reasoning behind this statement was that parents who were them- selves in a religiously mixed marriage would not be in a good posi- tion to support a norm of endogamy, even if they wanted to. We find confirmation of our hypothesis for the Catholic sub- sample in parents who were in mixed marriages was really quite small. 230. TABLE 5-9.--Religious types of marriage of Protestants and Catholics in the Detroit area, related to the source of greatest influence on reSpondentS' religious beliefs, 1958. Source of Greatest Influence Religious Types of Marriggg on Religious Beliefs Parents All Others Protestant Respondents :3 _1.___ L _Z.___ __N— Both Protestant at marriage - 72.6 (69) 69.4 (68) Protestant unchanging; Spouse changes- 9.5 ( 9) 9.2 ( 9) Protestant with Catholic or other - 17.9 (17) 13.3 (13) Protestant changes; Spouse unchanging- 0.0 ( 0) 8.2 ( 8) Totals - 100.0 (95) 100.1 (98) CH12 (3 d. f.)= 8.50 p.<.05 v2 =- .044 Catholic Respondents:a Both Catholic at marriage - 68.9 (51) 65.4 (53) Catholic unchanging; Spouse changes- 13.5 (10) 11.1 ( 9) Catholic with Protestant or other - 9.5 ( 7) 9.9 ( 8) Catholic changes; spouse unchanging - 8.1 ( 6) 13.6 (11) Totals - 100.0 (74) 100.0 (81) 01112 (3 d. £.)= 1.31 p. (.80 v2 = .008 aThe two Protestant and two Catholic marriages in this table which were misclassified as mixed have little impact on the results. We must remember that these cases probably represent only about 40 per cent of the marriages of parents which were originally mixed, since we found this to be the fact concerning the marriages of our respondents. Three out of five marriages involving couples from religiously different backgrounds actually become religiously 231. homageneous marriages. Hence these would appear in our table as "Parents not mixed," even though such parents would be no more likely to give effective support to religious endogamy than our "Parents mixed." This is not a very satisfactory way to test our hypothesis. It is therefore all the more notable that we still find a Statistically Significant relationship in the Catholic sample. Another hypothesis concerning parents as agents of social control, which we can test with data from one of the surveys being used in this chapter, is the following: 1c. Protestant-Catholic marriages will be negatively associ- ated with naming parents as having had "the greatest influence on your religious beliefs." Our reasoning for this hypothesis was that parents would use their influence in religious matters to urge the observance of religious endogamy within the group with which they identified themselves and their children, whereas other religious influences in their children's lives might come from other religious groups and be at least indifferent to this particular moral norm. Table 5-9 shows that, among both Protestants and Catholics, those who named parents as having the greatest religious influence tend only slightly more than others to marry within their own re- ligious group. Therefore, as stated, our hypothesis does not re- ceive statistically significant support in our 1958 data for either 2 (1 d. f.) - .2, Protestants or Catholics. (In each case, CHI p < .70, when direction has been predicted.) However, the detailed distribution of rates for the types of marriage which we have in this table is statistically signifi- cant for Protestants at the .05 level of probability. This is 232. accounted for mainly by the distribution of cases in the category of marriages in which the Protestant changes religious preference while the spouse is unchanging. .None of these eight cases named their parents as having had the greatest influence on their re- ligious beliefs. The same tendency appears in the distribution of Catholics, but not so strongly. Therefore, the role of par- ents as agents of social control for their religious group is shown not so much by whether their children violate the norm of religious endogamy as by whether they continue in their religious preference when they do so. This finding is certainly in agree- ment with our theoretical approach through the concept of social control. Social Control by Family or Relatives We have data in the 1958 survey for testing two hypotheses which we have stated concerningzthe family and relatives as agents of social control. The first of these reads: 2a. Protestant-Catholic marriages will be negatively associa- ted with having a high proportion of relatives who are of one's own faith. As we observed in Chapter IV, this hypothesis cangg£_be prOperly tested by asking people about their relatives 3:555 their marriage. What the writer had in mind was the situation before marriage, in which having close relatives who belonged to one's own religious group could be expected to influence one's observance of the norm of religious endogamy. Nevertheless, there are some aspects of the data we have for this hypothesis which may be worth noting. Let us notice first in Table 5-10 that there is no fourth category of religious types of marriage for each religious group, 233. TABLE 5-10.--Religious types of marriage of Protestants and Catholics in.the Detroit area, related to the proportion of "close" relatives who are-of respondent's faith, 1958.- Proportion of Relatives Who Religious Type of Marriagg Share Respondent's Religigg ’ All Less Than All iRespondents Now Protestant:a __Z__ N __%__ N Both Protestant at marriage - 86.0 (86) 61.5 (51) Protestant unchanging; spouse changes - 7.0 ( 7) 13.3 (11) Protestant with Catholic or other . 7.0 _(_Zl"2§;3_ (21) Totals - 100.0 (100) 100.0 ‘ (83) CHIZ (2d.f.) . 15.38 p .001 V2-.084 Respondents Now Catholic:a Both Catholic at marriage - 88.1 (74) 56.4 (31) Catholic unchanging; spouse changes - 8.3 ( 7) 20.0 (11) Catholic with Protestant or other - 3.64 _(_§). 2§;§_ - (132 "Totals - 100.0 (84) 100.0 (55) 01112 (2d. f.) - 19.6 p 4:.001 v2-.141 8The two Protestant and two Catholic_marriages in this table which were misclassified as mixed have little impact on the results. the type in which the respondent has changed religious preference but the spouse has not. The reason is that only those who were Protestant at the time of the interview were asked the question: "Of those relatives you feel really close to, what proportion are Protestants?" Similarly for CathOlics. Consequently, one who had changed religious preference would be questioned according to his present, not his previous, religious preference. Those respondents 234. who had changed religious preference would be eliminated from this table altogether due to the writer's practice of sorting respond- ents according to their early religious preference before proceed- ing with the crossetabulation. For example, the Protestant who changed to Catholic would be classified in Table 5-10 as a Prot- estant respondent. But he would not be asked about his Protestant relatives, as would be the Protestants who did not change. There- fore no respondent who changgd religious preference appears in Table 5-10. The next thing to observe in this table is that this distri- bution of types of marriage is statistically Significant at the .001 level. This would still be true for both religious groups if we combined the two types of intermarriage shown here and just con- sidered the gross rate of intermarriage. (Protestants: CHI2 - 14.53, p. (.001, V2 - .079. Catholics: C1112 - 18.11, p.< .001, V2 - .130. 1 d. f. for each.) It is also not surprising to find that for Protestants at least, there are more who feel close to their Protestant relatives among those in the second type of marriage than in the third type. In other words, when the other partner also maintains his (her) re- ligion, we find more cases of £1235 relatives across religious lines (21) than when the spouse has changed to the respondent's religion (11). Again, whether this is factor or function we cannot tell from interviews taken some time after the marriage. A somewhat better index of social control on the part of relatives is the expression of pressure which is felt by the reapond- ent to maintain religious ties. Hence we hypothesized: 235. 2b. Protestant-Catholic marriages will be negatively related to the statement that friends or relatives would feel "unhappy or disturbed" or "would try to discourage" one from changing to Protestant or to Catholic. As we stated in Chapter IV, we have some reason to believe that friends as a category do not exercise social control for religious endogamy; hence relatives were probably the ones intended by those who answered affirmatively to this question. AS in Table 5-10, those who have already changed religious preference are not in- cluded among the respondents found in Table 5-11. The hypothesis, as stated, is supported by the data of Table 5-11 for both Prot- estants and Catholics. The respondents who answered that their relatives would try to discourage them from changing religions also were likely 22£,t° have contracted an interfaith marriage. We cannot be sure how much the experience of marriage has influenced the respondents' perceptions of their relatives' atti- tudes toward maintaining religious ties. Nevertheless, the detailed categories of marriage type (not shown in this table) indicate that those who are in marriages which are still mixed are more likely (than are those who have married a convert) to respond that their relatives "would Egg try to discourage" them from changing religions. In other words, in tenms of social control, those in a marriage that is still mixed seem more likely to think of themselves as being close to leaving their religious group. To the writer this seems quite significant. Even though none of these respondents has changed his (her) religious preference, those who are in currently mixed marriages said that relatives would not try to discourage them from changing religions. This suggests that interfaith marriage can be thought of as a step toward leaving one‘s 236. TABLE 5-1l.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to perception that relatives would try to discourage reapondent from changing his religion, 1958. Whether Relatives or Friends Rates of Intermarriggg Would Try to Discourage Res- “ _pondent from Chagge in Religigg Protestanta Catholica Would try to discourage change - 18.0 (117)b 21.2 (118)b Unsure - 22.2 ( 9) 33.3 ( 3) Would not try to discourage - 32.3 ( 56) 53.3 S 15) Totals - 24.7 (182) 34.0 (136) CHI2 (2 d. f.) 9.3 ,7.5 p (.005c {.025c v2 .051 .055 8The two Protestant and two Catholic marriages in the table which were misclassified as mixed have little impact on the results. bNumbers in parentheses indicate the number of cases in each category. cNote b of Table 5-8 applies here. religious group. Hence social control is an appropriate concept to refer to endogamy as well as to maintenance of membership. Social Control by the Church's Schools It is, perhaps, the religious community as such rather than any segment of it which is most likely to seek to uphold a norm of religious endogamy; and the organization representing this community is the church. (However, it was difficult to find in our surveys any variables which would test the influence of the church in this matter. In 1958, those interviewed were asked whether they had received any of their education in parochial Schools or 237. other schools run by their church. Almost all who answered af- firmatively were Catholic. So we have hypothesized for Catholics: 3a. Protestant-Catholic marriages will be negatively related to one's education in church schools. As explained in Chapter IV, we have used four years as the minimum experience in which we would expect church schools to make their influence felt in such matters. If we were to collapse the three types of intermarriage in Table 5-12 into one class of mdxed marriages, we would find that the rate of intermarriage is not significantly lower for those who attended church schools than for those who did not-—although it tends in that direction. (Catholic men: CI-II2 '- 3.00,6p.,<_.10, V2 = .017. Catholic women: CHI2 - .56, p<.50, V2 - .004. l d. f. for each.) However, if we examine the part of the table concerned with Catholic women,-we see a significant tendency for those who have attended church schools to maintain their religious ties if they intermarry. Those who have not attended such schools are much more likely to change their religious preference when they.inter- marry. The same tendency is present, but not so strongly, for Cath- olic men. It seems, therefore, that in this sample the Catholic schools did notsucceed noticeably in selling religious endogamy as such to their students, by comparison to Catholics not in these schools. However, Catholic schools were effective agents of social control in the matter of their alumni and alumnae maintaining re- ligious ties if they did intermarry. This was evidently the message they communicated. Again, as noted in the preceding section, the independent variable shows a stronger relationship with leaving one's religious group than with intermarriage itself, suggesting that the 238. TABLE 5-12.--Re1igious types of marriage of Catholics in the Detroit area, related to education in church schools, by sex, 1958. Years of Education Religious Type of Marriage .in Church Schools 4 or more 'Under‘é Catholic ‘Men:a _J:_ __L L L Both Catholic at marriage - ~ 61.6 (53) '57.5 (50) Catholic unchanging; spouse changes - 20.9 (18) 13.8 (12) Catholic with Protestant or other - 9.3 ( 8) '11.5 (10) Catholic changes; spouse unchanging - ‘_§;l ‘(_11 (11;; ‘(léz Totals - 7 I -99.9 (86) 100.0 1(87) c1112 (3 d. f.) = 4.41 p4=.30 v2 - .025. I Catholic Women:b Both Catholic at marriage - 67.1 (51) 60.4 (32) Catholic unchanging; spouse changes — 14.5 (11) 5.7 (,3) Catholic with Protestant or other - ' 17.1 (13) ‘11.3 ( 6) Catholic changes; spouse unchanging - ‘_1;§_ $.11. _Zg;§ ‘(121 Totals - ‘ - 100.0 (76) 100.0 (53) 01112 (3 d. f.) - 17.26 p<.001 v2 - .138 aNine marriages of Catholic men were misclassified in the second type of marriage rather-than in the first. They were concen- trated among those who attended church schools, and therefore show more intermarriage for this group than was actually the case. bSix marriages of Catholic women were misclassified in the second type of marriage rather than in the first. They were evenly divided between those who were or were not educated in church schools. Hence the relationship here-should be unaffected. latter is a kind of mid-point on a continuum representing the~effec- tiveness of social control with respect to maintaining membership. 239. Social Control of Religious Intermarriage by the Ethnic Group In Chapter III we discussed at length what we would expect to be the relationship between the norm of religious endogamy and each generation of Americanization. We explained why we included among the hypotheses of this study the statement that: ha. Protestant-Catholic marriages will be positively related to being an American of the second and subsequent gener- ations. Our data for 1958 and 1959 were suitable for testing this hypothesis. However, we found no statistically significant relationship of this sort among either Protestant or Catholic respondents. Neither do the detailed categories for mixed marriages reveal any special tendency. It does not seem.worthwhi1e to reproduce the table by which we sought to test this hypothesis. A few summary statements should give the reader an adequate picture of the shape of these data. First, the Protestant sample is concentrated in the fourth (or later) and second generations, with few in the first. Almost two-thirds of the Catholic sample are second-generation Americans, with only a few in the fourth-or-later generations. Rates of intermarriage for Protestants range from 19.2 to 31.5, for Catholics from 27.4 to 38.8. Protestant rates rise in the second and third, then fall again in the fourth generation, possibly reflecting the presence of Southern white immigrants. Catholics rates of intermarriage rise somewhat in the third, then fall again in the fourth (or later) gen- eration. In general, the Catholic data are quite far from support- ing our hypothesis the Protestant tendency is in the predicted. direction, if we are right in attributing the low intermarriage rate in the fourth generation to the ethnic influence of Southern 240. whites. Wartime Marriagg We have considered wartime to be a normless period for young people especially, so far as anything which is not related to the war effort is concerned. Hence our hypothesis: 5a. Protestant-Catholic marriages will be positively related to getting married during war years. Our data come from the 1955 and 1959 surveys. Only those_respond- ents who are in their first marriage are included. Most of our wartime marriages took place during World War II, with about two- thirds as many in the Korean War, and just nineteen in World War I. There is simply no relationship in our data between inter- faith marriage and being married in wartime. We have previously cited two earlier studies which showed such an association to be present in the parish records of Catholic marriages. It is possible that many of the wartime marriages among the Detroit area A p0pu1ation were unstable and did not last until the time of our interviews. At any rate, interfaith marriages are not overrepre- sented in the wartime marriages that survived. Rural-Urbaaniggation as an Anomic Situation Earlier students of urban sociology,have led us to expect manifestations of anomia on the part of those who migrate from.the farm to the city. As we suggested in Chapter III,this may no longer be true as the society as a whole becomes increasingly urbanized. : Nevertheless, we have hypothesized: 5c. Protestant-Catholic marriages will be positively related to migration from a farm before age 18. 241. .This specific age was chosen partly because the coders of the Detroit Area Study data had already made use of it, and partly because it seemed an appropriate age at which to try to separate those who migrated before their marriage from those who migrated afterwards. Our rationale was this: Those who had married while still living on a farm would probably have the low inter- marriage rates which are associated with homogeneous, stable communities; those who left the farm and then found their marriage mate in the city might be expected to feel emancipated from the norms of their place of origin and to tend to contract interfaith marriages. Table 5-13 shows us some interesting aspects of migration and intermarriage rates among the American—born current residents of the Detroit area. First we must note, however, that the hypothesis as we stated it is not supported by our data. There is no difference in intermarriage rates between the two types of rural migrants. But let us look at the distribution of cases across the types of migration. Small town and rural migrants con- stitute the majority of all Protestants in the Detroit area, where- as native Detroiters (probably children of the foreign-born) make up a majority of all Catholics in Detroit. In neither group are city-to-city migrants a very large category, nor do they show much intermarriage. This latter finding is rather puzzling in view of the fact that the Detroit area natives, the only other urban peOple in our sample, show high rates of intermarriage--most striking in the Protestant sub-sample. In fact, it is the remarkably high rate of intermarriage among the Protestant Detroit area natives, the 242. TABLE 5-13.--Rates of intermarriage (per 100 marriages) for native- born Protestants and Catholics in the Detroit area, related to the respondent's experience of migration to Detroit, 1955, 1958, and 1959. Rates of Intermarriage Types of Migration to Detroit Respondent is: Protestanta Catholica Detroit area native — 36.3 (212)b 30.5 (295)b Urban migrant - 19.8 ( 86) 22.6 ( 62) Small town migrant - 23.6 (157) 33.3 ( 81) Rural migrant, before age 18 - 16.0 (156) 37.1 ( 62) Rural migrant, after age 18 - ‘11;;‘(_§g2, 4;;1_(_1g; Totals - 24.9 (663) 31.1 (512) 01112 (4 d. f.) 24.3 4.0 p 4:.001 <150 v2 .038 .008 aTen Protestant and ten Catholic marriages in this table were misclassified as mixed marriages. They are distributed quite normally among the types of migration and should not influence the relationship in the table. bNumbers in parentheses indicate the number of cases in each category. largest single category of Protestants, which mainly accounts for the statistically significant distribution of marriages in the Protestant sample. Here again the relative size of the group seems to be an important factor. Among native Detroiters, the peOple in our sample who grew up to marriageable age in the same locality, Protestants are a smaller group than the Catholics. Hence we should expect the former 243. to have a somewhat higher rate of intermarriage than the latter.12 In general, the Catholic-sub-sample comes closer to fulfill- ing our prediction than does the Protestant sample. .For them, mi- gration to a more urbanized center seems to be associated with high- er rates of intermarriage, whatever the age of migration. The inter- marriage rates for these categories among Protestants, however, are much lower. If these categories of Protestants and Catholics (so different in actual size) could be prOperly standardized to remove some of the arithmetic influence of the~size of group, we might find that Protestant and Catholic migrants are not so different from each other. In the 1962 sample (not included in this chapter), it appears that most Protestant migrants to Detroit come from the South and from "outstate" Michigan. If they maintain their small town and rural visiting patterns, as we found earlier to be the case in Chicago and Cleveland, this would help to account for their low rates of inter- marriage. Educational Mobility and Interfaith Marriages It is a commonplace observation that parents are likely to lose some of their influence over their children if the latter ac- quire a college education which the former did not have. .With this 12If, for the moment, we assume that white Protestants and Catholics constitute the total population of the Detroit area, we can compute a statistic such as we used for the overall intermarriage rates before (Table 5-1 and footnote #6, above). Then the "expected" intermarriage rate for Protestants born in Detroit would be .732 and for Catholics .590. The ratio of the actual to the expected rate of intermarriage for Protestants then becomes .497 compared to .517 for Catholics. According to this rough method of controlling for the size of the group, Protestant native Detroiters intermarry a little less often than do the Catholic native Detroiters. 244. in mind, we called this an anomic situation and predicted: 5d. Protestant-Catholic marriages will be positively related to the husband's extreme upward mobility in education, com- pared.to his father. Perhaps there would have been as much reason to predict the same consequence in the mother-daughter relationship. As explained in Chapter IV, we do not have enough cases in the 1959 data to test this hypothesis. Even before we separated Protestant from Catho- 'lic respondents, we had only forty-one cases of this kind of up- ward mobility-ewith.just eight of them being in interfaith mar- riages. Consequently, we can say nothing decisive about the valid- ity of this hypothesis. Intermarriage and Vertical Mobility in Occupations In Chapter III we haVe discussed at length the literature which attempts to relate occupational mobility to anomia and to religious affiliation and behavior. In view of these studies and for reasons similar to that given for the preceding hypothesis, we have predicted the f01lowing: 5e. Protestant-Catholic marriages will be positively related to the husband's extreme upward mobility in occupation, compared to his father. As we have noted earlier, Lenski reported that those who were highly involved in their religious sub-community were less likely to be either upwardly g£_downwardly mobile. This has prompt- ed us to include the downwardly mobile in our hypothesis. Since we are re-analyzing Lenski's data, a word of explanation about this is called for. Religiously homogeneous marriage was one of the criteria he used for involvement in the religious community (together with the criterion of having all or almost all of one's relatives 245. TABLE 5-l4.--Rates of intermarriage (per 100 marriages) for Prot- estant and Catholic men in the Detroit area, related to their vertical occupational mobility compared to their fathers, 1955, 1958, and 1959. Rates of Intermarriagg Relative Occupational Positions of Father and Son Son is: Protestanta Catholica White collar father and son - 22.7 ( 88)b ~l7.0 ( 47)b Blue collar father and son - 17.1 (222) 27.0 (244) White collar father and blue collar son (downward) - 26.7 ( 45) 42.8 ( 32) Blue collar father and white collar son (upward) - 23.3 (133), 3Q;l.(1141 Totals - '20.7 (488) 28.1 (437) 0312 (3 d. f.) 3.9' 7.2 p <.lSC <.05C v2 .008 .016. 8The eight Protestant and eight Catholic marriages in this table which were misclassified as mixed marriages are quite norm- ally distributed in the occupational categories of the table. bNumbers in parentheses indicate the number of cases in each category. cNote b of Table 5-8 applies here; and friends in the same religious community). One might conclude that we are predicting what he has already demonstrated with the same data. We are, of course, reversing the order of prediction from his communal involvement leading to the lack of occupational mobility (for daughters as well as sons) to our predictive order of occupational mobility leading to interfaith marriage (comparing sons with their fathers). As Lenski would readily grant, there is 246. no way at present to choose between these two orders of causality, except to see how they fit into a broader theoretical scheme. A careful longitudinal study may someday clarify the situation. We must also call attention now to the fact that we are using not only Lenski's 1958 data in Table 5-14, but also Blood's 1955 data and Swanson's 1959 data. Actually, the number of cases from the 1955 survey is larger in this table than the other two surveys combined. So we are not simply restating old evidence in a new form. Special care had to be taken in combining the data of 1955 with those of the other two years in Table 5-14. In 1955, informa- tion was obtained from the wives abggg_the occupations of their husband and husband's father. In the other two years this informa- tion was obtained only when the husband was the respondent. As it turned out, the relationship we predicted was in the two later surveys stronger for Catholics and weaker for Protestants than it was found to be in the 1955 survey. Combining the three has the effect of making Protestant and Catholic intermarriage rates more similar to each other. As the table presents it, the relationship is statistically significant for Catholics at the conventional level of probability, not quite so for Protestants. Table 5-14 allows us to examine upward and downward mobility separately, as related to interfaith marriages, and we note that the downwardly mobile are even more likely than the upwardly mobile to intermarry. However, our hypothesis (revised) does not require us to distinguish between types of mobility, nor between white collar and blue collar non-mobile pairs. If we, therefore, combine the two non-mobile categories and contrast their intermarriage pattern 247. with that of the two categories of occupational mobility, we are actually testing the hypothesis as revised. When we do this, it turns out that the Protestant distribution has a CHI2 of 2.1 (1 d. L), p. <.04, and v7- - .004; whereas the Catholic 01112 becomes 3.1 with p = .002, and V2 - .007. Therefore the predicted associa- tion between intermarriage and occupational mobility is statistic- ally significant for both religious groups, although not strong. This is the last of the hypotheses of our study which can be tested in these three large surveys. In Chapter VII, we shall attempt to pull together and interpret the findings which have been presented in this chapter. But first let us review the infor- mation given to us by the smaller 1962 survey. CHAPTER VI 185 NEWLYWED COUPLES IN 1962 In 1962 the Detroit Area Study, using a design constructed by Drs. Ronald Freedman and David Goldberg, was concerned with "family growth" and its motivation. For their purposes they wanted a cross section of four particular groups: people who got married, or had a lst, 2nd, or 4th baby in July, 1961. Consequently, theirs is not a probability sample of the ggneralgpopulation, in the Detroit area. However, this writer considered the interview ques- tions asked of those just recently married to be of particular relevance to the study of interfaith marriages. So he was happy to be granted access to the data of this first group in the four used for the 1962 study, especially since the authors of the study were still engaged in their own analysis of the data. As we have indicated earlier, there are special features about this survey which make it impossible to combine its analysis with the earlier Detroit Area Study surveys, which we have just discussed in the preceding chapter. Special Characteristics of This Sample In the sample used for this chapter, all of the women were interviewed from six to eight months after their marriage. At first just 101 interviews were sought and obtained with those who had been married in July, 1961. Later it was decided to double 248. 249. TABLE 6-1.--Month of narriage for respondents related to present religious preference, Detroit Area Study, 1962. Month of Marriage' . Present Religious Preference of Wife Protestant Catholic £33 Orthodox ‘51; July, 1961 ' 38 56 4 3 101 October, 1961 - __28_ _52_ _g_ _l_ ‘__QQ Totals - 66 ' 115 6 4 191 the size of this sub-sample in the larger study. Hence we have 90 interviews with wives who had been married in October, 1961. All of these were selected from the marriage records of the State of Michigan from among those married at ceremonies within the bound- aries of the Detroit Area Study. The age limits, 15-34 years, were established by studying the age distribution of the newly-mar- ried in the Detroit area recently. However, no marriages of non- whites or second and later marriages were included. Within these restrictions, the sample was drawn on a random basis.1 Interviews with these wives averages about 80 minutes in length. The over-all response rate in the study as a whole was 92 per cent, which is very high. Table 6-1 shows the distribution of the wives in this sample by religious preference and month of marriage. (The distribution 1Ronald Freedman and David Goldberg, "Family Growth in the Detroit Area: The prospectus for the research program of the Detroit Area Study for 1961-62," unpublished paper for Project 882, Detroit Area Study, Survey Research Center, University of Michigan, September, 1961. At the time of present writing, the authors had not yet written up a paper describing in detail the sampling pro- cedure used. 250. of husbands by religious preference is nearly the same, except that it includes six who are of "no preference.") As the reader will recognize at once by recalling our statistics in Chapter V, Catho- lics are disproportiénately well represented in this small sample.. Why should this be? There are two penitential seasons in the year, the four weeks of Advent just before Christmas and the six weeks of Lent before Easter, in which Roman Catholics may not have a wedding in church. This means that the weddings of Catholics which might have taken place during this one-fifth of the year will be distributed over the other weeks. Furthermore, the months of May and October, being officially dedicated to Mary, the mother of Jesus, and to her rosary, are popular choices among Catholic women as a wedding time (for devotional reasons). These factors may not entirely explain the large number of Catholics in this sample, but they do contribute something to it. Operational Definitions As stated in Chapter IV, operational definitions and other procedures were not different for this sample unless they are said to be so in this chapter. The first change that we had to make was in our definition of the dependent variable, religious types of marriage. The wives interviewed were asked regarding themselves (and similarly for their husbands): "And now, I want to ask you about your religious preference--that is, are you Protestant, Catho- lic, Jewish, or something else?" Then the question: "Have you always been (Protestant, Catholic, Jewish etc.)?" Finally: 251. "When did you make this change?"2 Since they were not asked to specify what their previous religious preference was, but only whether they had changed, we cannot know for certain in what re- ligious group the changers were socialized. However, in view of the religious composition of the Detroit area population, we should make but few errors if we make the following assumptions: A wife now Protestant, who says that she has changed religious preference, is assumed to have been Catholic before the change. One may say the equivalent to this for the husbands and for changes to Catholic. If we keep this difference in mind, we shall find that the indices developed for this sample are directly comparable to those described in Chapter IV and applied to the three larger samples. In the process of classifying marriages by religious types according to these criteria, just six marriages were found to include neither a Protestant nor a Catholic and therefore to be ineligible for our study. There were 185 marriages still meet- ing our requirements for analysis. As the reader may recall from Chapter IV, we need to know what the early religion of the respondent is if we are going to relate any background characteristic of an individual to the type of marriage. The early religion of those who changed was identi- fied in the way just stated. There were four respondents and ten spouses who were found to have been neither Protestant nor Catho- lic, although they were married to someone of these faiths. The 2Of the thirty who changed to Protestant or to Catholic (eight husbands and twenty-two wives), twenty-three did so within a year of the calendar year of their marriage. 252. distribution of our sample by early religion was: Religion of Respondent Religion of Spouse Catholic --101 Catholic - 110 Other - 4 Other - 10 Protestant - _§Q Protestant - _og 'Total - 185 Total - 185 When a characteristic of the Respondent is the independent variable, those whose religion was "Other" are excluded from the analysis; similarly, if a characteristic of the Spouse is the independent vari- able. All of the interview responses used for this chapter were precoded. Two or three variables or indices are slightly different for this sample from.the corresponding ones found in thepreceding chapter. These small points will be noted when we come to the hypo- theses in question. Rates of Intermarriage by Religion, Sex, and Ceremony In Chapter V, we explained the method used by Paul Click to compute the ratio of actual to "expected" rates of mixed marriage. In this chapter our procedure will be the same. Table 6-2 corres sponds in its form to Table 5-1. First we notice that, although the Catholic rate of mixed marriages is practically the same as it was for our combined sample, the Protestant rate is 53 inter- faith marriages per one hundred involving Protestants, compared to the rate of 36 found in the data of Chapter V. We can hardly attribute this sharp increase to the relative size of the Prot- estant group in the general population of the Detroit area in 1962. 253. TABLE 6-2.--Actua1 intermarriage rates and intermarriage rates "expected" if husbands and wives were distributed at random, for marriages of Protestants and Catholics in the Detroit area, 1961. ‘ g . "Expected" Ratio of Early Religion Married Couples. Per Cent if Actual to of Husband , Actual Actual Random.Inter- "Expected" and Wive Number Per Cent“ marriage ~ Mixed Marriages One or both Protestant 97 100 100 , -- Protestant-- Protestant 46 47 32 .. Catholic or other 51 53 68 .78 One or both , Catholic 133 100 100 -- Catholic-- Catholic 80 60 27 Protestant or other 53 40 73 .55 The mixed marriage rate would go up with the numerical decline of the group; but, in the three to eight years' time since the other Detroit Area Study surveys acquired information about religious preferences, Protestants have certainly not ceased to be the largest major religious category among whites in the Detroit area. It seems to this writer that we must go back to what has already been said about norms and practices among Catholics regarding appropriate times for a wedding. In other words, for the two months of July and October, in which our respondents were married, Catholics may well constitute the majority of the marrying white population. Hence the Protestant rate of intermarriage would rise on this ac- count. In Table 6-2, we have used the same "expected" per cent for random intermarriage which appeared in Table 5-1. We have no better 254. basis than was used there for determining this statistic. However, if what we have just said about the marrying population of July and October is true, then the ratio of .78 for Protestants shown in Table 6-2 (compared to .53 in Table 5-1) is not an accurate measure of this group's intermarriage behavior. The ratio for the Catholic group is almost identical in these two tables. As the reader may recall from Chapter V, the ratio of -actual to expected mixed marriages which Click computed from.U. S. Census data was only .19 for Protestants and-.26 for Catholics, much lower than those in Table 6-2. However, if we base our computations on the present religious preference of the respondents, our ratio. for Protestants becomes .38 and for Catholics .23. The Protestant sample is still not very close to what has been found at other times in the Detroit area not to the national survey. If we compare Table 6-3 with Table 5-4, both showing inter- marriage by religion and sex, we see first of all that the unmixed marriages are included only in Table 6-3. This was done because, (without them, the number of cases in the 1962 survey is too small to justify the reporting of percentages. However, it is rather surprising to find almost the same patterns of intermarriage in this small sample as we had in the larger, combined sample. Prot- estant women and Catholic men intermarry a little more than their counterparts. “Protestant.men and Catholic women are likely to re- tain their religious affiliation. Protestant women are more likely to change their religious affiliation in favor of the one which their husband has. It appears, therefore, that the attitudes which we attributed to liberal Protestants and to Catholics are reflected 255. TABLE 6-3.-- Religious types of marriage related to the sex of Protestants and Catholics in the Detroit Area, 1962, by percent- ages. m Religious Type of Marriage Protestant'Respondents: Both Protestant at marriage Protestant alone unchanging Mixed still Catholic-alone unchanginga Totals. 4 Number of cases Catholic Respondents: Both Catholic at marriage Catholic alone unchanging Mixed still Protestant alone unchanginga Totals Number of cases Husbands 73 100% 63 73 15 9 3 100% 110 Wives 59 15 22b 100% 78 79 5 11 5 100% 101 aThis fourth category is stated accurately, even though it appears out of place with respondents of a different Christian group. If we included the assumption about the changers, as stated in the text, this fourth category for Protestant respondents would read: "Protestant alone changes," and for Catholic respondents: "Catholic alone changes." However, we are certain only of what they changed go, not what they changed from. bThere were two cases, not included in this table, in which both husband and wife changed to Catholic. In other parts of this chapter these two are-classified with the fourth category of Protestant marriages. in our data. If it is true that a unified family is particularly important to the wife and mother, it seems that Protestant women 256. do not object to achieving this through their own religious con- version, whereas Catholic women are less likely to find this to be an agreeable solution to the problem. Neither Protestant nor Catho- lic husbands are so inclined to change religious preference. There is one type of information available from the 1962 survey which was not present in the other surveys, namely the type of ceremony which was used in the marriage. Table 6-4 shows us not only the percentage for each type of Officiant at marriage {bottom row), but also the percentage for each religous type of marriage (last column) among all marriages of Protestants and of Catholics.3 We notice, first, in the last column that there is not much difference between the Protestant and Catholic groups in the percentage of marriages which remained religiously mixed (20 go. 16). The differences appear in the other two categories of mixed marriage, with Catholics showing more tendency to hold to their religious preference (18 yo. 8 per cent).4 Looking now at the bottom row of Table 6-4, we see a sur- prising statistic: One third of the marriages involving Protestants in our small sample were performed by a Catholic priest. Even if 3The number of mixed marriages, 51 among Protestants and 53 among Catholics, seemed too small to justify the computing of percentages for the three types of interfaith marriage, apart from the unmixed marriages, as we did for Table 5-2. 413 the fourth religious type of marriage, as stated in our operational definitions, we are not certain what those who changed to Catholic or to Protestant were before their change-- only that they did change to one of the major religious groups. We have assumed that such people changed from the other major religious group. There is no such problem with the second partner-- whatever that of the changer may have been. 257. TABLE 6-4.--Religious types of marriage among Protestants and Catho- lics in the Detroit area, related to Officiant at marriage, 1962. 4 Religious Type Officiant at the Marriage8 of Marrioga Protestant Catholic Per Marriages of ‘Clergygan Priest inil Other All Cent Protestonts: Both Protestant - 45 l 46 47% Protestant alone unchanging - 8 8 8 Mixed still ’ - 4 12 2 1 19 20 Catholic alone unchangingb - l 21 2 24 5 Totals - 58 33 5 1 97 (L Per cent - 60 34 5 1 .g 100% Marriages of Catholics: Both Catholic - 76 4 80 60% Catholic alone unchanging - l 21 2 24 18 Mixed still - 2 15 3 l 21 16 Protestant alone unchangingb- 8 8 6 Totals - 11 112' 9 1 133 1 Per cent - 8 84 7 1 ._9 100% a"Rabbi" was one alternative Officiant which was precoded for this question. However, this response was not given for any of the marriages in this table. bSee footnote "a" for Table 6-3. we distrust the assumption underlying the fourth religious type of marriage, it is notable that about two-thirds of the nineteen mar- riages of Protestants which are still mixed were witnessed by a Catholic priest. Turning to the marriages involving Catholics, we 258. find that two-thirds of all the mixed marriages of Catholics in this sample had a Catholic priest as Officiant. Perhaps to explain these facts we should advert again to the months in which these mar- riages all took place, July and October. There is some selection here of practicing Catholics (as reflected by their not marrying in a penitential season), and there may also be a selection of the ‘more devout ones (as suggested by a marriage in October). §ooo Unpredioted Characteristics of theilntegoarrioo One of the characteristics of the intermarried couples which we looked at in Chapter V was the income of the family head. Table 6-5 is presented here to allow for comparison with Table 5-6. How- ever, we do not state a CHI2 value nor a level of probability for the small sample in Table 6-5. This would be misleading, since the number of cases in many cells of the full table is very small. For the same reason, the intermarriage rates for the individual cate- gories of this table should be viewed cautiously. Our Table 5-6 showed intermarriage to lack any relationship to income for Prot- estants but to increase somewhat with income for Catholics. In Table 6-5 the tendency (more apparent if you compare the top half with the bottom half of the categories) is for intermarriage to be positively related to income for Protestants, but negatively for Catholics. In the latter case, this is the reverse of the rela- tionship in our combined sample of Chapter V. Information was gathered during the 1962 survey concerning ‘the husband's occupation. There is no striking relationship found in our small sub-sample, and we shall not present a table of per- centages. There is the same problem cf small numbers when we use 259. TABLE 6-5.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to total annual income of the husband, 1962. Rates of Intermarriage Inoome of Husband- Respondent is: Protestant Catholio1 $9,000 and over. 50.0 ( 2)a 50.0 (6)a $8,000 - $8,999. 50.0 ( 4) 25.0 (.4) $7,000 - $7,999. 100.0 ( 1) 00.0 ( 7) $6,000 - $6,999. 40.0 ( 5) 15.4 (13) $5,000 - $5,999. 30.0 (10) 20.0 (25) $4,000 - $4,999. 42.9 ( 7) 50.0 (20) $3,000 - $3,999. 30.8 (13) 23.1 (13) $2,000 - $2,999. 9.1 (11) 33.3 ( 9) $1,000 - $1,999. 20.0 ( 5) 12.5 (.8) under $1,000. . 14.3 (‘7) 40.0 ( 5) Totals - 19.0 (65) 27.3 (110) aNumbers in parentheses indicate the number of cases in each category. many categories of occupations as there was with income. Besides, these are newly married men who probably have not yet reached the peak of their occupational career. The third variable which, in addition to income and occu- pation, is usually regarded as an indicator of socio-economic status is education. We shall later come to the hypothesis of this study which concerns the educational mobility of the husbands. Here we can summarize rather quickly the data which relate to the education of the wives. Fifty of the eighty Protestant wives (by 260. early religious preference) were high school graduates, andexactly half of these were in interfaith marriages. The intermarriage rate for all other categories combined was.thirty per cent. The full distribution 0f the table was.statistically significant at the ten per'centrlevel of probability. Those who changed their religious preference were_also highly concentrated within this category of high school graduates. The pattern was similar but not so pro- nounced for the Catholic women. Nearly two-thirds of these were high school graduates, and they had a somewhat higher rate of intermarriage than the other educational categories. The writer thought it might be worthwhile to examine the education of the fathers of these respondents, on the hunch that - low rates of intermarriage may be associated with increased educa- tion for the fathers.) Perhaps the better educated fathers would be more-able to communicate the norm of religious endogamy to their children. Such a relationship is evident (at the .05 level of probability) only among the fathers of the Protestant'wives. Here the largest category of fathers, the grade school graduates, also represent the daughters with the highest rate of intermarriage. However, this tendency does not appear for the Catholic wives. There is a suggestion of the-same relationship for the fathers of Protestant husbands; but, for the Catholic sample, the fathers who have attended or graduated from.high school find that their sons have slightly higher rates of intermarriage than the others. Finally, without knowing what to expect, the writer sought to discover whether any relationship exists between rates of inter- ~marriage and the number of siblings reported by husbands and wives. 261. Among the Protestant husbands and wives no pattern whatever appeared. However, among both Catholic husbands and Catholic wives, there 9 appeared to be a tendency (not statistically significant for the whole distribution) for a little more intermarriage to occur among those who claim to have just one sibling or none. (Heiss also found that the only child had a higher rate of interfaith marriages in Manhattan.) There seems to be no clear interpretation which would link such a pattern with the concept of social control. Parents and Family as Agento of Social Control As we turn now to the hypotheses of our study which can be tested with this small sample, we find just one which refers to parents as agents of social control, yio.: lb. Protestant-Catholic marriages will be negatively associ- ated with the mother's frequent talking about or participating in religious activities with the family when self was growing up. Since our respondents in 1962 were all female and this question was not asked concerning the spouse, we are here touching on a mother- daughter relationship. Are daughters less likely to contract interfaith marriages when their mothers have manifested more re- ligious interest within the family of orientation? Table 6-6 answers this question as well as Our data per- mit. If we look only at the rates of intermarriage of those whose mothers were religiously active, compared with those whose mothers were not, we seem to find a strong relationship between interfaith ‘marriages and this variable. However, the number of respondents who answered "never" or "almost never" is too small to make this tendency statistically significant. In fact, except for our desire 262. TABLE 6-6.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to frequency of mother's religious activity when daughter was growing up, 1962. Rates of Intermarriage Frequency of Mother's Religious Activity Respondent is: Protestant Catholic Never, almost never - 62.5 (16)8 37.7 ( l4)a Sometimes, quite often - 37.5 (64) 18.4 (787) Totals - 42.5 (80) 20.8 (101) CH12 (1 d. f.) 2.33b .38b p .10C .60c v2 .029 .004 aNumbers in parentheses indicate the number of cases in each category. bThese CH12 values have been corrected for continuity be- cause of the small number of cases in the top cells. CThis level of probability is based on a one-tailed test of significance, since the direction of the relationship had been predicted. Cf. Hubert M. Blalock, Jr., Social Statistics, New York: McGraw-Hill, 1960, p. 218. to show the direction of the tendency, these "rates per hundred" would not be reported for such a small number of cases. We simply have a situation here in which the distribution of responses is skewed in the direction of frequent religious activity. Further- more, among Protestants there is almost no difference in the rate of intermarriage for those who answered "sometimes" and those who answered "quite often." However, this is not true for Catholics. If we compare the number of intermarriages of the Catholic women who answered "quite often" with those who gave any other answer 263. TABLE 6-7.--Rates of intermarriage (per 100 marriages) for Prot- estants and Catholics in the Detroit area, related to the Propor- tion of "Close" relatives who are of respondent's faith, 1962. ' Rateszof-Entepoogriage'w Proportion of Close Relatives Who Are-of Respondent's Faith Respondent is: Protestant Catholic All of Respondent's Faith - 13.3 (30)8 7.0 (43)8 Less Than All of Respondent's Faith - 35.5 (31) 24.5 (53) Totals - 24.6 (61) 16.7 (96) 9 01112 (1 d.f.) 3.0b 4.2b p a,03c -,.o4c v2 ‘.049 .044 ,aNumbers in parentheses indicate the number of cases in each category. bThe CHI2 values have been corrected for continuity because of the small number of cases in two of the four cells for each table. cNote-c of Table 6-6 applies here. (about half), we find that the CHI2 value is 4.66 in the predicted direction, for which the level of probability is .03 and v2‘- .046. In other words, so far as interfaith marriages are concerned, the significant dividing line in the responses for Protestant women is between "almost never" an "sometimes;" for Catholic women it is between "sometimes" and "quite often." One of the hypotheses intended to explore the influence ‘of the family or relatives as an agent of social control in inter- faith marriages was: 264. 28. Protestant-Catholic marriages will be negatively associ- ated with having a high proportion of relatives who are of one's own faith. This hypothesis was tested with the data of 1958 (Table 5r10), but we stated our dissatisfaction then with the fact that this variable was measured 飣££ the marriage and therefore could hardly be used to predict the occurrence of interfaith marriages. In the data of 1962 we have a somewhat better test for the hypothesis. These wives were interviewed 6-8 months after their marriage. Therefore, when a Protestant (for example) was asked, "Of the relatives you feel close to, about how many are-Protestants? would you say all, nearly all, more than half, or less than half?" the influence of the association with her husband's relatives would be of short duration and probably not very.strong. Since "feeling close to" someone is generally thought to be associatedwith protracted inter- ~action, we are probably getting in these responses a pretty good representation of the preamarriage situation. It is likely that those who did not answer "All" to this question were themselves the children of interfaith marriages. Hence this hypothesis is quite’similar to Hypothesis la, having parents of mixed religions. Such a mixture of religions among the relatives would tend, we have said, to make‘thelfamily or relatives less effective-as an agent of social control regarding interfaith;marriages. we mmst note again, as we did for Table 5710, that no .respondent who changed religious preference appears in Table 6f7. (See the explanation given for Table 5-10 as to why this happened.) This is unfortunate, since it particularly eliminates many of the Protestant mixed marriages (because these wives were more inclined 265. to change religious preference). It is all the more remarkable that, in spite of these missing interfaith.marriages, we still find a statistically significant relationship for this table. The Church in Support of Religious Endogamy As in the survey of 1958, we again have the opportunity in the 1962 data to test the following hypothesis: 3a. Protestant- -CatholiC‘marriages will be negatively related to one '8 education in church schools. By contrast with 1958, the question regarding education in church schools in the 1962 interviews was asked only of those wives who gave their religious preference as Catholic. Similarly for their (husbands. 0n the face of it, this would seem to be an economical procedure, since there would not be enough Protestants who had attended their own church-related schools to warrant separate anal- ysis.’ However, the fact that only those who identified themselves as currently Catholic were asked this question gives rise to.the same kind of problem that we have just noted for Table 6-7. Those who had changed religious preference‘oo Catholic would not have been Catholic in their youth and would therefore not be expected ,to have-attended Catholic schools. The writer eliminated such cases from Table 6-8 by including only those men and women who, by our operational definition, would be identified as Catholic in their youth. However, this procedure still leaves out of Table 6‘8 those wives and husbands who were Catholic in their youth but are identified as currently Protestant. These individuals would simply not have been asked the question in the interview concern- ing-attendance at Catholic schools. 266. TABLE 6-8.--Rates of intermarriage (per 100 marriages) for Catho- lics in the Detroit area, related to education in church schools I by sex, 1962. Rates of Intermarriage Years of Education in Church Schools Catholic Men Catholic Women ~Four years or more ‘- 25.3(79)a _ ~ 15.4-(65)a Less than four years - 32.3 (31) 30.6 (3o) Totals - 27.3 (110) 21.0 (101) C312 (1 d.f.) .54 3.24 p -.46b =.07b v2 .005 .032 8Numbers in parentheses indicate the number of cases in each category. bNote-c of Table 6-6 applies here. As in the 1958 data, we here define four years of experience in Catholic schools as the minimum in which we would expect social control to operate. The number of cases in the 1962 subesample is too small for us to distinguish between the types of interfaith marriage, as we did in Table 5-12. It is interesting to see, never- theless, that the overall tendency of intermarriage is the same for both men and women as it was in Table 5-12, with only the Catholic women approaching our accepted level of statistical significance. If the three husbands and five wives who changed 22 Protestant had also been asked about their education in Catholic schools, it is likely that we would have slightly higher intermarriage rates for those without experience in Catholic schools, as predicted. 267. This statement is supported by thel958 data, which showed that those who had not attended Cathdlic schools were more likely to change their religious preference if they intermarried. The data of the 1962 survey allowed for the testing of another hypothesis concerning the influence of the church on re- ligious endogamy. It was this: 3b. Protestant-Catholic marriages will be negatively related to frequent attendance at religious services the year before marriage. The wives answered this question for themselves only, not for their husbands. There is no basis for establishing the predicted relation- ship among the Catholic wives, since only oigoo out of 101 reported attending services less than once a week during this time. Another curious thing about this distribution is that the only two who said they attended services once a month or less were BEE in a mixed marriage. Perhaps these were also among the four unmixed Catholic marriages which, as we saw earlier, were performed by a civil off- icial. There is no consistent relationship between interfaith mar- riages and church attendance for our Protestant wives either. There were just 25 out of 80 who said that they had attended re- ligious services less than once a week during the year before their marriage, and these did not have a very distinctive rate of inter- 'marriage. It was somewhat surprising to discover that 16 out of the 55 reporting at least weekly attendance were wives who had changed go Catholic. Apparently they had already adopted the Catholic practice of weekly Mass during most of the year before their marriage. 268. Close Ethnic Tiesoond Interfaith Marriago 4a. Protestant-Catholic marriages will be positively related to being an American of the second and subsequent generations. This hypothesis was tested in the 1962 data for four categories: husbands and wives of both Catholic and Protestant religious pref- erence. In none of the four was the prediction born out by steadily increasing rates of intermarriage from the first through the fourth generation. Actually, we have very few belonging to the first generation in this sample, no more than eight per cent in any of these four categories. Since there is no consistent relationship between these two variables, it does not seem desirable to show a table of the rates of intermarriage. We shall simply make a few summary statements, as we did in the case of the 1958 and 1959 data for the same hypothesis. First, it is clear that the 1962 newlweds represent a later generation than the 1958-59 surveys of the general population. Here we find the Protestants concentrated in the fourth-or-later and the third generations (not the second, as in 1958-59); where- as half of the Catholics belong to the third generation, with al- most all of the rest divided rather evenly between the second and the fourth-or-later generations. Although the number of cases is too small to be statisti- 'cally significant, it seems to be worth noting that the marriages in which one partner (probably Protestant) changed to Catholic were concentrated in the third and fourth generations. More pre- cisely, 18 of the 19 wives who made this change were of the third or fourth-and-later generations: a. v., 95 per cent fell into 269. this category, wheras 84 per cent of all the Protestant wives were in these generations. Five of the seven Protestant husbands who made the same change belonged to the fourth-or-later generae tion, while this category held a ratio of four to seven for all Protestant husbands. This is only a slight tendency, but it reminds one of Lenski's finding that: "The more Americanized the individuals, the greater the probability that mixed marriages will lead to the conversion of one partner or the other, and that reli- gious unity will eventually be established within the nuclear family."5 This tendency appears among the Protestant but not the catholic husbands and wives in the 1962 data. The Influence of Anomic Situations Since all the marriages being discussed in this chapter took place in 1961, we have no basis here for testing wartime as an anomic situation. However, we do have better information con- cerning age at marriage in this sample than we had in Chapter V. Our hopothesis stated: 5b. Protestant-Catholic marriages will be positively related to getting married at a late age, with consequent emancipation from.home. As we saw earlier, this hypothesis was found to hold for the Catho- lic-sample in Heiss' Manhattan study. John H. Burma discovered that interracial marriages were also more characteristic of someh 5Gerhard Lenski, The Religious Factor, Garden City, N. Y.: Doubleday, 1961, pp. 49-50; italicized in the original. 270. what older men and women in Los Angeles.6 Because the sample is small, the writer thought it best to use only a gross distinction for age at marriage. In the case of men, those aged 25 years and over were considered ”late mar- riages;" for women, it was those who were 23 years old or over who were so defined. Even with this simple dichotomy, our hypo- thesis was not supported. In fact, for Catholic women and for both Protestant and Catholic men, there was a slight tendency in the opposite direction, i;_g., for intermarriage to be associated with the younger age category. It does not seem likely that we would find a significantly different relationship if we chose a different cutting point for age at marriage. More than two-thirds of the husbands and wives in the 1962 sample were native Detroiters. Even without controlling for re- ligious preference, only 23 wives and 24 husbands lived longest on a farm or in a small city or town. Therefore we do not have adequate data here for testing the relationship of rural-urban migration to interfaith marriage. To fill in the picture, we ex- amined the marriages of Protestants and found that 45 of the 70 migrations originated in Michigan or in the South; among marriages of Catholics, 20 of the 61 migrations originated in Michigan, with the remainder scattered through other areas. It also appeared that the marriages in which one partner (especially the wife) changed to Catholic were concentrated among the native Detroiters. 6"Interethnic Marriages in Los Angeles, 1948-1959," paper read at the annual meetings of the American Sociological Associa— tion, Washington, D. 0., August 30, 1962. 271. TABLE 6-9.--Religious type of marriage among Catholic men in the Detroit area, 1962, related to attendance at college by them and/ or their fathers. Attendance at College Religious Type of Marriagg Unmixed Mixed All Father and/or son attended college - 39 8 47 Neither attended college - _&l_ _gg_ _Q; Totals - 8O 30 110 CH12 (1 d. f.) = 4.3 p <:.05 V2 = .039 Hence this type of marriage, which we discussed earlier, is not a consequence of migration. Another of our hypotheses which was intended to test the influence of an anomic situation was the following: 5d. Protestant-Catholic marriages will be positively related to husband's extreme upward mobility in education, compared to his father. It was thought that the parents' influence would be exerted in favor of religious endogamy, in general, but that a son who far surpassed his father in education would be less likely to accept this influence. Actually, no such relationship appears for either Catholic or Protestant husbands. In this sample we can do a better job of predicting the type of marriage if we think of soy; educa- tion in college as being associated with a low intermarriage rate. Table 6-9 shows how this is evidenced in the 1962 Catholic sample. If either the father or the son has attended college, the son is less likely to enter upon a mixed marriage. The distribution of religious types of marriage among Protestant men was in the same 272. direction as for Catholics, but it was far from being statistically significant. How should we interpret this relationship? It seems plausible that the existence of a norm of religious endogamy is expressed and understood better at the college level of education. In this context it may be regarded as a social norm (what a £31;- .giggg man ggghg to do), or it may simply be the perception of a marital risk which is operating (what an intelligent man will avoid). In any case, we have no evidence here of the sons rejecting the norms of their fathers. Since the newly-married men of the 1962 survey have not yet reached the peak of their occupational career, we shall not try to compare them with their fathers in terms of occupational mobility. We have now completed the analysis of our 1962 data. All that remains for us is to try to pull together the findings reported both in this chapter and in the preceding one. This we shall do in the next and final chapter. An evaluation of the theoretical scheme we have proposed will then be in order, as well as some suggestions for future research. CHAPTER VII SUMMARY AND CONCLUSIONS In this final chapter we shall first review our theory of social control as it relates to interfaith marriages, i;_£;9 the series of hypotheses found in Chapter III. Then we shall consider all of our Protestant data and Catholic data separately, arriving at a composite picture of those in each religious group who intermarry in Detroit. Finally, an attempt will be made to highlight what appear to be the distinctive features of methodol- ogy and the more significant findings of this study, so that fu- ture researchers may more easily build on this experience. How Our Theory of Social Control Has Fared in This Study In Tables 7-1 and 7-2 we have summarized the statistics which express our findings in this study regarding the hypotheses set down at the beginning. By examining these two tables we should be able to evaluate this theoretical approach. Perhaps the most obvious fact is that in both tables only 223 of our five hypothe- ses for anomic situations turned out to be statistically signifi- cant, XlE° occupational mobility. Even in this case the level of probability attained is due more to the large sample than to the strength of the relationship, as is evident from the samll value of V2 in each case. For the other four situations, anomia simply does not explain anything in the data as we have them. Regarding war- 273. 274. TABLE 7-l.--Protestant interfaith marriages in terms of social control: summary of findings from the Detroit Area Study, 1955, 1958, 1959, and 1962. Year(s) of Subject of Hypothesis Survey CHI2 Parent of Mixed 1958 Religions (la)a 1959 - Mother's Religious Activity (lb) 1962 2.33 Parents Were Greatest Religious Influence (1°) (1958) (8.50)e Close Relatives are (1958) (15.38) Protestant (2a) 1962 3.0 Relatives Would Discourage Change to Catholib (2b) 1958 9.3 Attendance at Services the Year Before Marriage (3b) 1962 Close Ethnic 1958, 1959 Ties (4a) 1962 Wartime Marriage (53) 1955, 1959 Late Marriage (5b) 1962 Rural-Urban 1955, 1958 Migration (5c) 1959 Upward Educational Mobility (5d) 1962 Occupational 1955, 1958 Mobility (5e) 1959 2 .1 Level of (p) Degree of (V2) Probability Relationship --b -- .10 .029d (.05) (.044) (.001) (.084) .08 .049 .005 .051 .04 .004 8In this first column, the number of the hypothesis is given in parentheses. pp. 84-85. For a full statement of each hypothesis, see bStatistics are not shown if the level of probability exceed- ed .50. was tested. The double hyphen indicates, however, that the hypothesis CCf.p.263 regarding a Protestant—Catholic difference in the cutting-point for the independent variable. 275. dFor the meaning of V2, of Table 5-6, footnote d. SThe statistics within parentheses were obtained not by a simple test of the hypothesis regarding intermarriage, but by re- lating the independent variable to the several religious types of intermarriage. time marriages and rural-urban migration, marriages entered into in these situations may have broken up in large numbers and there- fore would have been eliminated from our sample before the inter- views were taken. However, the writer is more inclined to think that.rura1-urban migration has not been an anomic situation for some decades now. Since the hypothesis regarding late marriages and upward educational mobility were tested on newlyweds, the absence of broken marriages from the sample would not account for the lack of a significant relationship. However, in both cases the sample was also small and may be non-typical. With a larger sample, the writer would be inclined to look for a non-linear re- lationship between age at marriage and interfaith marriage, those who depart most from the median age being the most likely to vio- late the social norm of religious endogamy. In the case of upward educational mobility, that sub-category of college students who live on campus, away from their less well educated parents, may still show the predicted anomic behavior. Moving up these two tables from the bottom, we find that interfaith marriages are not related to being an American of the second or a later generation for either Protestants or Catholics. In fact, we discovered no consistent relationship here, one way or the other. It is this writer's opinion that the only way to gauge the influence of this variable would be to examine each ethnic 276. TABLE 7-2.--Catholic interfaith marriage in terms of social control: summary of findings from the Detroit Area Study, 1955, 1958, and 1962.a — m Year(s) of Level of (p) Degree of (V2) Subject of Hypothesis Survey CHI2 ProbabilityA Relationship Parents of Mixed 1958 Religions (la) 1959 9.83 ..002 .025 Mother's Religious Activity (lb) 1962 4.66 .03 .046 Parents Were Greatest Religious Influence (1c) 1958 -- -- _- Close Relatives Are (1958) (19.60) (.001) (.141) Catholic (2a) 1962 4.2 .04 .044 Relatives Would Discour- age Change to Protestant (2b) 1958 7.5 .025 .055 1958 3.0 .10 .017 Education in Church (1958) (4.41) (.30) (.025) Schools (38) - Men 1962 .54 .46 .005 1958 .56 .50 .004 Education in Church (1958) (17.26) (.001) (.138) Schools (3a) - Women 1962 3.24 .07 .032 Attendance at Services the Year before Marriage (3b) 1962 -- -- -- Close Ethnic 1958, 1959 Ties (4a) 1962 -- -- -- Wartime Marriages (5a) 1955, 1959 -- -- -— Late Marriage (5b) 1962 -- -4 -- Rural-Urban 1955, 1958 Migration (5c) 1959 -- -- -- Upward Educational Mobility (5d) 1962 -- —- -- Occupational 1955, 1958 Mobility (5c) 1959 3.1 .002 .007 277. aEach of the footnotes for Table 7-1 applies to this table also. group separately, controlling for early religion and the size of the group. It is very unlikely that anyone will find data in the United States which are equal to this task. Immigration has slackened greatly since World War I. Marriages between genera- tions tend to obscure any relationship associated with a single generation. Our expedient of categorizing individuals by their closest tie with another country (even if only one grandparent) is probably the best way to handle this part of the problem, but it is not wholly satisfactory. Furthermore, the numerous inter- ethnic marriages make it difficult to associate a pattern of re- ligious intermarriage with any specific nationality group. To avoid this problem (but not to solve it), we have categorized individuals according to the original nationality on their father's side. This simply hides the interethnic marriages. 0n the whole, it is probably undesirable to try to do much with this hypothesis anymore in this country. It was not surprising to the writer that we did not find interfaith marriages to be related to the frequency of attendance at religious services during the year before marriage. We could test this only for a small sample of women, who generally show less variation in religious attendance than men do. As sociologists have occasionally observed, there is a tendency to take on the be- havioral patterns of a group with which one is beginning to ident- ify. Since we have found that more than half of the interfaith marriages of Protestants and Catholics involve conversion to one 278. religious group or the other, we should expect that these changers would take on the pattern of attendance at religious services which is characteristic of the group of destination rather than the one in which they were socialized. Therefore Protestant women who marry Catholics should have somewhat higher frequency of church attendance (if they also have converted) than those who do not intermarry. The hypothesis as stated should hold only for those whose early religion was Catholic. To test it satisfactorily, we should pinpoint rather closely the time of a change in religious preference, to know when we should expect a change in frequency of attendance at religious services. Actually, we should be able to find a better index than this to reflect the influence of the church as an agent of social control. For Catholics, education in church schools is certainly such an index. The writer was somewhat surprised not to find a stronger negative relationship between this religious education and interfaith marriages. Perhaps we should have been forewarned by Lenski's report that this variable had little relationship to attitudes toward interfaith marriage. We did find a tendency in the predicted direction for both men and women, but not a statis- tically significant one. However, we found one of our strongest relationships between education in Catholic schools (for women, at least) and BEE changing to Protestant or other in case of inter- marriage. Maintaining membership in the Roman Catholic Church is apparently the function of social control which church-affiliated schools fulfill better than they do the prevention of interfaith marriages themselves. 279. Still moving up the two tables from the bottom, we come to the hypotheses which concern relatives as agents of social control. The previous findings of the Detroit Area Study, which we reviewed in Appendix A, have led us to regard relatives as a potentially important agent of social control. However, as we noted in the earlier discussion of these findings, it does not help our theory much to discover that people in mixed marriages fafl.close to signi- ficantly more relatives who do not share their religious preference. To predict intermarriage, we should gather this information before the marriages have taken place. The closest we could come to this was in the interviews with the newlyweds. When those who had inter- married showed a modest tendency to claim more "close" relatives who were not of their faith, this probably (though not certainly) reflects the situation before marriage. Such a situation, accord- ing to our theory, would not be conducive to the enforcing of a norm of religious endogamy. So much for that index of the influ- ence of relatives. On the other hand, one of the most satisfying indices we found from the point of view of social control was the statement that relatives would discourage the respondent from changing his (her) religious preference. This is an exercise of social control, pure and simple. Although this control is per- ceived by the respondent as being directed against changing relig- ious preference, it was found to have a significant negative re- lationship to interfaith marriages for both Protestants and Catho- lics. This held for those whose marriage was still religiously mixed as well as for those in which one partner had changed relig- ioug preference. Social control is a good explanatory concept for 280. interfaith marriages when we can find a good measurement for it. Finally, in this reverse order in which we are reading Tables 7-1 and 7-2, we come to the role of parents as agents of social control for religious endogamy. Here we are talking about the two people most closely affected by the marriage choices of their children. Their interaction with the new in-law is likely to be greater than anyone else involved in the socialization of the marrying person. The newcomer will be the parent of their grandchildren. If religious beliefs and practices are important to them, they have reason to exercise whatever social control is in their power to support religious endogamy by their own children. It was rather surprising, therefore, to find that naming parents as having the greatest influence on the respondent's religious beliefs was not associated with observance of religious endogamy for either Protestants or Catholics. However, our theory of social control is supported even here by the finding that those who name their parents as having had the greatest religious influence on them are also much less likely to change their religious preference when they do intermarry. This was particularly true of Protestants. One of the better prima facie indicators we have of the exercise of social control is in the extent of the mother's religious activity as remembered by the young brides of 1962. A satisfactory negative relationship was found between this variable and interfaith marriages, stronger for Catholics than for Protestants. The mother's religious example in the life of the family evidently had some in- fluence on marriage choices. As for our hypothesis which predicted that parents who were in mixed marriages would find their children 281. following their example in significant numbers, this was establish- ed for Catholics but, surprisingly, not for Protestants. The only explanation we can offer for not finding a stronger relationship here is that we do not know how many of the parents of unmixed re- ligions were originally of mixed religious prefrences. This would obscure the relationship we were predicting. On the whole, it is clear that parents and relatives, at least by the measurements which we have used, are the most effec- tive agents of social control in the matter of religious endogamy. Romantic love, as an American theme, apparently has not removed marriage choices completely from the affairs of the family. The influence of the Church, as exerted through the Catholic schools, is more effective in maintaining membership than in preventing interfaith marriages. Ethnic groups as a whole do not seem to be agents of social control for religious endogamy. And our anomic situations were singularly ineffective predictors of interfaith marriages. Even the interpretation of the weak relationship be- tween occupational mobility and interfaith marriages is in doubt. It may be an intervening variable or a spurious relationship, or the direction of causation may actually be reversed. Finally, it must be more than coincidence that, Just as in Heiss' Manhattan study, our hypotheses (in almost every comparable instance) re- ceived more support from the Catholic data than from the Protestant data. It has been a valuable invention for our purposes to dis- tinguish the types of interfaith marriage according to whether either partner changed religious preference. In several instances, these 282. distinctions have supported our theory of social control when the’ simple dichotomy of religiously mixed 25, unmixed marriages did not show much relationship to the independent variable. Unpredicted Findings Concerning Protestant Intermarriages In our combined sample of 935 marriages involving Prot- »estants, 36 per cent were found to be interfaith marriages. (This went up to 53 per cent for our smaller, possibly atypical, sample of 97 marriages of Protestants in 1962.) When we compare this rate with what would be expected in the Detroit area if random inter- marriage occurred, we arrive at a ratio of .53 (in 1962, .78). In other words, interfaith marriages of Protestants are about half as common in fact as they would be if unrestricted in any way. This ratio becomes just .22 if we look only at the marriages of those who are Protestant at present, very close to Glick's ratio of .19 based on a national sample. When we distinguish the types of inter- faith marriages by Protestants in our combined sample, it appears that about one in three such marriages involves a change from the Protestant group, whereas about one in four involves a change £3 a Protestant faith. In the rest neither partner changed. The im- balance is not great, but it does seem noteworthy in a sample as large as this. When we shift our focus from marriages to individuals, it turns out that just 22 per cent of the 1,534 Protestants in our combined sample have married non-Protestants. Protestant women showed an intermarriage rate of 24 per cent, compared to 20 per cent for Protestant men. The women show some tendency to change their religious preference, whereas the men are more likely to be 283. in a marriage that is still religiously mixed. The same tendencies appeared in the small 1962 sample. This same study of newlyweds showed that 33 of the Protestant mixed marriages, according to our operational definition, took place at ceremonies performed by Cath- olic priests, compared to 13 which were performed by Protestant clergymen. As for Protestant denominations, those respondents who claimed no specific denomination and those who said they were Episcopalians had the higher rates of marriage with non-Protestants, whereas the Baptists had the lowest rate among the larger denomina- tions. Protestants of English background (on the father's side) showed a notably higher rate of interfaith marriages than those of German background. Protestants who are natives of the Detroit area showed a considerably higher rate of intermarriage with non-Prot- estants than did all types of Protestant migrants to Detroit. A Composite Sketch of the IntermarryinggProtestant in Detroit There is not just one set of background characteristics be- longing to Protestants who marry non-Protestants. However, we can suggest a few attributes which are modal for them, even though they do not express a true average. The Protestant who contracts an interfaith marriage is likely to be a native Detroiter of English extraction. This person is a bit more likely to be a woman and, if so, to become a Catholic in the process; she is probably a high school graduate whose father finished grade-school. She remembers her mother as being religiously active with the family "sometimes" or less often. Whether husband or wife, the Protestant probably grew up in one of the denominations of English or Scotch origin 284. (Methodist, Presbyterian, or Episcopalian). He (or she) does not feel that relatives would discourage him from changing to Catholic, and not all the relatives he feels close to are Protestant. In age at marriage and in the income and occupation of the family head, this marriage will not be notably different from Protestants who have not intermarried. Unpredicted Findings Concerning Catholic Intermarriages The Catholic rate of intermarriage in the Detroit area was five points higher, at 41 per cent, than that of the Protestants. However, when the adjustment was made for the size of the two groups in the general p0pu1ation, the Catholic ratio (.56) became more similar to that of the Protestants (.53). In the smaller sample of 1962, both the actual rate and the ratio of this to the "expected" rate were almost identical, in the case of Catholics, with those just reported from the combined sample. When Glick computed his ratio of actual to expected intermarriage for Catholics in the United States, based onthe reported present religious preference of respondents and spouses, he arrived at a figure of .26. Our comparable figure for Catholics in the Detroit area was .22, show- ing even less intermarriage in this city than is true of the nation- al population. When we distinguish the religious types of inter- marriage among Catholics of the Detroit area, it appears that thggg out of ten change their religious preference in each direction, i;_g;,‘£g or away from Catholic, whereas the other four out of ten remain in mixed marriages. With Catholics, as with Protestants, three out of five mixed marriages involve some change of religious preference by one of the partners. 285. Looking at the individual Catholics instead of at their marriages, we see that 26 per cent of the 1,333 Catholics in our combined sample are in interfaith marriages. Rather surprisingly, in view of earlier studies, there was practically no difference in the rate of intermarriage for Catholic men and women. There were some sex differences, however, among those Catholics who did intermarry. The men were noticeably more likely to marry a Catho- lic convert than were the women, this difference being balanced equally by the larger prOportion of Catholic women who changed their own religious preference and those who remained in a mixed marriage. This pattern was present to some extent also in the smaller 1962 sample. In addition, this sample showed that 36 out of the 53 mixed marriages involving Catholics were performed by a Catholic priest. In terms of their national background, Catholics of Polish descent have a low rate of interfaith marriage, whereas the Irish rate is just a little higher than that of the German and Italians. In our combined sample, Catholic intermarriages were also character- ized by a slightly higher level of income and by a professional or other white collar occupation (except sales). Small town and rural migrants to Detroit showed somewhat higher rates of interfaith mar- riage than did the Catholic natives of Detroit or migrants from other cities. Somewhat paradoxically (in view of high rates of Catholic intermarriage for higher occupations), our small 1962 sample revealed lggg intermarriage for Catholics if either father or son had attended college. 286. A Composite Sketch of the Intermarrying Catholic in Detroit As in the case of Protestants, there is no single type of Catholic who enters upon an interfaith marriage. The details we give here comprise a modal description rather than an average type. The Catholic in an interfaith marriage is likely to have been born vin Detroit of Polish parents, since these categories are prepon- derant in the Catholic population of Detroit--even though their ragga of intermarriage are relatively low. The income of the inter- married couple will be relatively high, and the husband is likely to be a skilled worker or a professional or managerial person. One of the partners in this marriage has probably changed religious preference, the direction of change being more likely to follow the husband's preference. The wedding probably took place in the Cath- olic Church. The Catholic woman who has intermarried probably would 22E remember her mother as having been religiously active "quite often" when the family was growing up. She probably has not attended a Catholic school for as long as four years, and not all of the rela- tives she feels close to are Catholic. Whether husband or wife, the intermarrying Catholic feels that relatives would discourage . him (her) from changing to Protestant. The husband of this couple is very likely to be the son of a blue collar worker and is almost as likely to be a blue collar worker himself. In time, the child- ren of this couple will probably also enter into an interfaith mar- riage. 287. What This Study Tells Future Researchers Undertaking research on the selection of marriage partners from the point of view of the norm of religious endogamy has a sociological interest which, it seems to this writer, can be match- ed by very few research topics. We are dealing here with a key decision made by almost all members of society. It is a decision which is the subject of many social norms representing value systems which sometimes conflict. Several social institutions intersect at this point, and the social control which supports them in various ways is brought to bear on this single decision.' In studying interfaith marriages in this way, one seems to ex- perience the dynamism of the entire society. What of importance have we learned from this study which can be put to use by future researchers? Probably the chief point is the new evidence of a need to know the early religious prefer- ence of married people if you wish to study the background factors associated with interfaith marriages. Since three-fifths of the mixed marriages in our combined sample would not have been dis- covered if we had known only the present religious preference of married people, it seems that a sociological researcher should never be content to base conclusions about the background factors of interfaith marriages on such inadequate information. PeOple who intermarry do change their religious preference in large numbers. We have had to assume that those who changed religious preference did so about the time of marriage (or in relation to their choice of a marriage partner). We did report some evidence to support this assumption. However, it would be better in future data-gathering to 288. ask a question or two about the relationship of the change in re- ligion (if it occurred) to the time and circumstances of the mar- riage. The time-lapse between the wedding and the day of the inter- view has been a serious handicap in testing some of the hypotheses of this study. Not only is there danger of memories being distorted, but questions sometimes have referred to a current.situation rather than to the circumstances prior to the wedding. In addition, there is the fact that some married partners break up during the years following the wedding, especially if it has been an interfaith mar- riage. Due to these factors, it would seem best in a study such as this to draw a sample from the state's marriage records, as was done for the Detroit Area Study's 1962 sample. An interview with- in a month or two after the wedding would be preferable to waiting six or eight months. Restricting the study to first marriages of whites is desirable until such time as other categories of the p0pu1ation can be sampled adequately. We have said much in the course of this thesis concerning the factor of the size of a group as related to its rate of inter— marriage. The writer would like to make clear that he realizes that there is great practical importance in this matter. A small group may be faced with extinction through intermarriage with other members of the general p0pu1ation in a given locality. There seems to be a gravitation to the patterns of behavior of the majority in any community, with the result that the children of an intermarriage will (other things being equal) probably be drawn to affiliate with the majority. Hence, in a practical sense, a high rate of intermar- 289. marriage for a small group is a threat to survival. However, it still seems to this writer that, once the foregoing has been stated in general and observed in a given locality, a sociological explan- ation for differences in rates of intermarriage is still lacking. It is important, whenever possible, to control for the influence of this factor of the size of groups before comparing rates of intermarriage. So far, the best way to do this seems to be to compute the ratio of this observed rate to the rate which would be expected from a chance distribution of marriage choices. Un- fortunately, it is necessary to know the frequency distribution of a variable in a particular p0pu1ation before one can compute this expected rate, a thing which is rarely possible with such variables as religious preference, national origin, migration his- tory, and number of generations in America. Considering the results of this study and that of Jerold Heiss in Manhattan, the concept of social control seems to be an apprOpriate one for studying the factors leading to violations of an endogamous norm. The agents of social control which seem to be most significant, as indicated by our measurements are the parents and the family or relatives. These are the individuals who are generally perceived to be most closely concerned with the marriage choices of the younger generation, even in a modern society. The church's interest, so far as our limited evidence reveals it to us, is primarily directed toward maintaining its membership and secondarily toward discouraging marriages which would be a threat to this primary concern. As for the ethnic group's interest in this matter, it seems to this writer that this factor cannot be studied 290. fruitfully in globo, but each ethnic groups should be considered, individually. The writer's attempt to use the concept of anomic situa- tions in relation to social control of interfaith marriages was not successful. We have suggested earlier the particular circum- stances in which these hypotheses might stand up under a test. The one anomic situation which did stand up in this study, occu- pational mobility, is still open to other interpretations which are perhaps equally plausible. It is in the study of this factor, however, that the relationship between socio-economic hypergamy - and religious intermarriage might be explored with some wide im-1 plications. The several religious types of intermarriage as used in this study have been of some value in expressing degrees of social control in relation to the maintenance of membership. There are probably still other ways that these categories could be studied profitably. This, then, brings our thesis to an end. In Spite of its complexities, it goes quite directly to the heart of sociological interests. Happily, its study has not only been intellectually rewarding but has thrown some light on a practical problem which is common in our society. BIBLIOGRAPHY Books Adams, Romanzo. Interracial Marriage in Hawaii. New York: Mac- millan, 1937. Anderson, Elin. We Americans: A Study of Cleavage in an American City. Cambridge: Harvard University Press, 1937. Anderson, Walfred A. Marriages and Families of University Graduates. Ithaca: Cornell University Press, 1950. Angell, Robert C. Free Societyfiand Moral Crisis. Ann Arbor: Uni- versity of Michigan Press, 1958. The Integration of American Society: A Study of Groups and Institutions. New York: McGraw-Hill, 1941. Asch, Solomon E. Studies of Independence and Conformity: l. A Minority of One Against a Unanimous Majority. Washington: American Psychological Association, 1956. Barron, Milton L. PeOple Who Intermarry: Intermarriage in a New England Industrial Community. Syracuse: Syracuse University Press, 1946. Bell, Howard M. Youth Tell Their Story. Washington: American Council on Education, 1938. Berger, Morroe, Theodore Abel, and Charles H. Page (eds.) Freedom and Control in Modern Society. New York: Van Nostrand, 1954. Bernard, Luther L. Social Control in Its Sociological Aspects. New York: Macmillan, 1939. Black, Algernon D. If I Marry Outside My Religion. Public Affairs Pamphlet No. 204. New York: Public Affairs Pamphlets, 1954. Blalock, Hubert M. Social Statistics. New York: McGrawAHill, 1960. Blood, Robert 0., Jr., and Donald M. Wolfe. Husbands and Wives: The Dynamics of Married Living. Glencoe, 111.: Free Press, 1960. 291. 292. Bloom, Leonard, Ruth Riemer, and Carol Creedon. Marriages of Japanese-Americans in Los Angeles: A Statistical Study. Berkeley: University of California Press, 1945. Bogue, Donald J. Population Growth in Standard MetrOpolitan Areas, ‘ 1900-1950: With an Explanatory Analysis of Urbanized Areas. Washington: Government Printing Office, 1953. Bossard, James H. Marriage and the Child. Philadelphia: Uni- versity of Pennsylvania Press, 1940. and Eleanor S. Boll. One Marriage, Two Faiths: Guid- ance on Interfaith Marriage. New York: Ronald Press, 1957. and Eleanor S. Boll. Ritual in Family Living: Phila- delphia: University of Pennsylvania Press, 1950. Bowman, Henry A. Marriage for Moderns, 4th ed. New York: .McGraw- Hill, 1960. Brown, Francis J., and Joseph S. Roucek (eds.) One America: The History, Contributions, and Present Problems of Our Racial and National Minorities. Revised. New York: Prentice- Hall, 1945. Brunner, Edmund de S. Immigrant Farmers and Their Children. New York: Doubleday Doran, 1929. Canadian Government. Canada Year Book, 1959. Ottawa: Bureau of Statistics, 1959. Carpenter, Niles. The Sociology of City Life. New York: Long- mans, Green, 1931. Catlin, George B. The Story of Detroit. Detroit: Evening News Association, 1923. City Plan Commission. Economic Base of Detroit. Detroit: City Plan Commission, 1944. Clemens, Alphonse H. Marriage and the Family: An Integrated Approach for Catholics. Englewood Cliffs, N. J.: Prentice- Hall, 1957. Colacci, Mario. Christian Marriage Today: A Comparison of Roman Catholic and Protestant Views with Special Reference to Mixed Marriages. Minneapolis: Augsburg, 1958. Comte, Auguste. The Positive Philosophy of Apguste Comte. Trans- lated and condensed by Harriet Martineau. New York: Calvin Blanchard, 1858. 293. Coser, Lewis A., and Bernard Rosenberg (eds.) Sociological Theory: A Book of Readings. New York: Macmillan, 1957. Dowd, Jerome. Control in Human Societies.. New York: D. Appleton- Century, 1936. Drachsler, Julius. Democracy and Assimilation. New York: Macmillan, 1920. Intermarriage in New York City: A Statistical Study of the Amalgamation of European PeOples. New York: .J. Drachsler, 1921. Durkheim, Emile. The Division of Labor. Translated by George Simpson. Glencoe, 111.: Free Press, 1947. . The Rules of Sociological Method. 8th edition, trans- lated by Sarah A. Solovay and John.H. Mueller, edited by George E. Catlin. Glencoe, 111.: Free Press, 1958. . Suicide: A Study in Sociology. Translated by John A. Spaulding and George Simpson. Edited by George Simpson. Glencoe, 111.: Free Press,.l951. Elliott, Mabel A., and Francis E. Merrill. Social Disorganization. 3d ed. New York: Harper, 1950. Ellwood, Charles A. The Psycholggy of Human Society. New York: D. Appleton-Century, 1927. English, H. B. and Ava C. A Comprehensive Dictionary of Psycho- logical and Psychoanalytic Terms. New York: Longmans, Green, 1958. Fichter, Joseph H., S.J. Southern Parish. Vol I: Dynamics of a City Church. Chicago: University of Chicago Press, 1951. Foote, Nelson N., and Leonard S. Cottrell, Jr. .Identity and Inter- personal Competence: A New Direction in Family Research. Chicago: University of Chicago Press, 1957. Freedman, Ronald, Pascal K. Whelpton, and Arthur A. Campbell. Family Planning, Sterility, and Population Growth. New York: McGraw-Hill, 1959. Gillette, John M., and James MI Reinhardt. Problems of a Changing Social Order. New York: American Book, 1942. Glick, Paul. American Families. New York: Wiley, 1957. Graeber, Isacque, and Steuart H. Britt (eds.) Jews in the Gentile World. New York: Macmillan, 1942. 294. Handlin, Oscar. The Uprooted. Boston: Little, Brown, 1951. Henry, Andrew F., and James F. Short, Jr. Suicide and Homicide. Glencoe, 111.: Free Press, 1954. Herberg, Will. Protestant - Catholic - Jew: An Essay in American Religious Sociology. New York: Doubleday, 1955. Hertzler, Joyce 0. American Social Institutions: A Sociological Analysis. Boston: Allyn and Bacon, 1961. Historical Records Survey Project, WPA. Inventopy of Church Archives of Michigan: Dearborn Churches. Detroit: Michigan Histor- ical Records Survey Project, 1940. . Directory of Churches and Religious Opganizations: Greater Detroit, 1941. Detroit: Michigan Historical Re- cords Survey Project, 1941. Hoover, Harry E. Attitudes of High-school Students toward Mixed Marriages. Washington: Catholic University Press, 1950. Johnson, F. Ernest (ed) Patterns of Faith in America Today. New York: Harper, 1957. Kahl, Joseph A. The American Class Structure. New York: Rinehart, 1957. Kapadia, Kanailal M. Marriagp and Family in India. Zdexi New York: Oxford University Press, 1958. Kelly, George A. The Catholic Family Handbook. New York: Dell, 1962. . The Catholic Marriage Manual. New York: Random House, 1958. Kornhauser, Arthur. Detroit as the People See It. Detroit: Wayne University Press, 1952. Landis, Paul H. Social Control: Social Organization and Disorgan- ization in Process. Revised. Philadelphia: Lippincott, 1956. LaPiere, Richard T. A Theopy of Social Control. New York: McGraw- Hill, 1954. Leiffer, Murray H. City and Church in Transition: A Study of the Medium-Sized City and Its Orggnized Religious Life. Chicago: Willett, Clark, 1938. 295. Lenski, Gerhard. The Religious Factor: A Sociological Study of Religion's Impact on Politics, Economics, and Family Life. Garden City, N. Y.: Doubleday, 1961. Lind, Andrew W. Hawaii's PeOple. Honolulu: University of Hawaii Press, 1955. Linton, Ralph. The Study of Man: An Introduction. Student's Edition. New York: Appleton-Century-Crofts, 1936. Lipset, Seymour, and Reinhard Bendix. Social Mobility in Indus- trial Society. Berkeley: University of California Press, 1959. Loomis, Charles. Social Systems. Princeton, N. J.: Van Nostrand, 1960. Lumley, Frederick E. The Means of Social Control. New York: Cen- tury, 1925. Lynn, Sister Annella, R. S. M. Interracial Marriages in Washington, D. C.y 1940-47. Washington: Catholic University Press, 1953. McGrath, Sister M. Evodine, O. S. F.. The Role of the Catholic College in Preparing for Marriage and Family Life. Wash- ington: Catholic University Press, 1952. MacIver, Robert M., and Charles H. Page. Society: An Introductory Analysis. New York: Rinehart, 1949. 2 Marden, Charles F. Minorities in American Society. New York: American Book, 1952. Mayer, John E. Jewish-Gentile Courtships: An Exploratory Study of a Social Process. New York: Free Press of Glencoe, Mayor's Interracial Committee.. Maps of 1940 Popplation: Racial and Nationality Group Distribution. Detroit: Mayor's Inter- racial Committee, 1950. Merton, Robert K. Social Theory and Social Structure. Revised. Glencoe, 111.: Free Press, 1957. IMihavovich, Clement 8., Gerald J. Schnepp, S. M., and John L. Thomas, S.J. A Guide to Catholic Marriage. Milwaukee: 'Bruce, 1955. thonahan, Thomas P. The Pattern of Age at Marriage in the United States. 2 vols. Philadelphia: Stephenson, 1951. 296. Nelson, Erland. Pergistence of Attitudes of College Students Four- teen Years Later. Washington: American Psychological Association, 1954. .O'Brien, John A. Happy Marriage. New York: Popular Library, 1957. Panunzio, Constantine. Major Social Institutions: An Introduction. New York:‘ Macmillan, 1939. Park, Robert E., and Ernest W. Burgess. Introduction to the Science of Sociology. Chicago: "University of Chicago Press, 1924. Parsons, Talcott. Essays in Sociological Theory. Revised. Glencoe, 111.: Free Press, 1954. , Robert F. Bales, et al. Family, Socialization, and Inter- action Process. Glencoe, 111.: Free Press, 1955. . The Social System. Glencoe, 111.: Free Press, 1951. . The Structure of Social Action. Glencoe, 111.: Free Press, 1949. , and Edward A. Shils (eds.) Toward a General Theory of Action. Cambridge: Harvard University Press, 1951. Phillips, Arthur. Survey of African Marriage and Family Life. London: Oxford University Press, 1953. Pike, James A. If You Marry Outside Your Faith. New York: Harper, 1954. Pound, Roscoe. Social Control Through Law. New Haven: Yale Uni- versity Press, 1942. Rogoff, Natalie. Recent Trends in Occupational Mobility. Glencoe, ‘ 111.: Free Press, 1953. Ross, Edward A. Social Control: A Survey of the Foundations of Order. New York: Macmillan, 1901. Rossi, Peter H. Why Families Move: A Study in the Social Psychology of Urban Residential Mobility. Glencoe, 111.: ‘Free Press, 1955. Roucek, Joseph S., et a1. Social Control. New York: D. Van Nostrand, 1947. Schmdedeler, Edgar, O. S. B. LookingyToward Marriage: A Modern Text for the High School Student. Washington: Family Life Bureau, National Catholic Welfare Conference, 1948. _ . Marriage and the Family. New York: McGraw-Hill, 1946. 297. Schnepp, Gerald J., S. M. Leakage from a Catholic Parish. Wash- ington: Catholic University Press, 1942. Schuyler, Joseph B.,S.J} Northern Parish: A Sociological and Pastoral Study. Chicago: Loyola University Press, 1960. Shibutani, Tamotsu. Society and Personality: An Interactionist Approach to Social Psychology. Englewood Cliffs, N. J.; Prentice-Hall, 1961. Simmel, Georg. The Web of Group~Affiliations. Translated by Reinhard Bendix. With his Conflict, transalted by Kurt H. Wolff. Glencoe, 111.: Free Press, 1955. Simpson, George, and Milton Yinger. Racial and Cultural Minorities: An Analysis of Prejudice and Discrimination. New York: Harper, 1958. Small, Albion W., and George E. Vincent. An Introduction to the Study of Society. New York: American Book, 1894. Smith, Rockwell C. The Church in Our Town: A Study_of the Rela- tionshipdbetween the Church and the Rural Community. Revised. New York: Abingdon Press, 1955. Smith, William C. Americans in the Making: The Natural History of the Assimilation of Immigrants. New York: D. Appleton- Century, 1939. Spencer, Herbert. Principles of Sociology. Vol. 11. London: Williams and Norgate, 1906. Stonequist, Everett V. The Marginal Man: A Study in Personality and Culture Conflict. New York: Charles Scribner's Sons, 1937. Sumner, William G. Folkways: A Study of the Sociological Impor- tance of Usages, Manners, CustomsJ Mores, and Morals. New York: Ginn, 1906. , and Albert G. Keller. The Science of Society, Vol. 111.. New Haven: Yale University Press, 1927. Thomas, John L., S.J. .The American Catholic Family. Englewood Cliffs, N. J.; Prentice-Hall, 1956. Thomas, William 1., and Florian Znaniecki. The Polish Peasant in Europe and America: Monograph of an Immigrant Gropp. Vol. 1: Primary-Group Organization. Chicago: University of Chicago Press, 1918. “Treacy, Gerald C., S.J. Love Undying: A Simplified Edition of the Engyclical Casti Connubii, on Christian Marriege, by Pope Pius X1. New York: Paulist Press, 1944. 298. Tufts, James H. America's Social Morality. New York: Henry Holt, 1933. U. S. Bureau of the Census. United States Census of Population and Housing: »l960. Census Tracts,yDetroit S. M. S. A. Final . Report PHC(l)-40. Washington: Government Printing Office, 1962. . United States Census of Population: 1960, Detailed Characteristics,dMichigan. Final Report PC(1)-24D. Wash- ington: Government Printing Office, 1962. . United States Census of Population: 1960, General Social and Economic Characteristics, Michigan. Final Report PC(1)- 24C. Washington: Government Printing Office, 1962. Vidich, Arthur J., and Joseph Bensman. Small Town in Mass Society: Class, Power, and Religion in a Rural Community. Princeton: Princeton University Press, 1958. Ward, Lester F. Dynamic Sociology, or Applied Social Science. New York: D. Appleton, 1883. Warner, W. Lloyd, and Leo Srole. The Social Systems of American Ethnic Groups. New Haven: Yale University Press, 1945. Wells, George R. Individualipy and Social Restraint. New York: D. Appleton, 1929. Wessel, Bessie B. An Ethnic Survey of Woonsockep, R. 1. Chicago: University of Chicago Press, 1931. Westoff, Charles F., Robert G. Potter, Jr., Philip C. Sagi, and Elliot G. Mishler. Family Growth in MetrOpolitan America. Princeton: Princeton University Press, 1961. Williams, Robin M., Jr. American Society: A Sociological Inter- pretation. New York: Alfred A. Knopf, 1959. Winch, Robert F. Mate-Selection: A Study of Complementary Needs. New York: Harper, 1958. Woods, Sister Frances J. C.D.P.. Cultural Values of American Ethnic Groups. New York: Harper, 1956. Woofter, Thomas J. Races and Ethnic Groups in American Life. New York: McGraw-Hill, 1933. liynn, John G. How Christian Parents Face Family Problems. Phila- delphia: Westminster Press, 1955. Ykaung, Kimball. Sociology: Cincinnati: American Book, 1942. 299. Zahn, Gordon C. German Catholics and Hitler's Wars: A Study in Social Control. New York: Sheed and Ward, 1962. Articles Abbott, Walter M., S.J. "Two Council Fathers: Cardinal Leger and Cardinal Cushing," America, 108 (1963), pp. 864-867. Abrams, Ray H. "Residential Pr0pinquity as a Factor in Marriage Selection: Fifty Year Trends in Philadelphia," American Sociological Review, 8 (1943), pp. 288-294. American Sociological Society, Publications. "Social Control," Papers and Proceedings of the American Sociological Society, 12 (1917), entire issue. Anderson, C. Arnold. "Our Present Knowledge of Assortative Mating," Rural Sociology, 3 (1938), pp. 296—302. Annella, Sister M. "Some Aspects of Interracial Marriages in Wash- ington, D. C.," Journal of Negro Education, 25 (1956), pp. 380-391. Antonovsky, Aaron. "Toward a Refinement of the 'Marginal Man' Con- cept," Social Forces, 35 (1956), pp. 57-62. Ayoub, Millicent R. "Parallel Cousin Marriage and Endogamy: A Study in Sociometry," Southwestern Journal of Anthropology, 15 (1959), pp. 266-275. Baber, Ray E. "Some Mate Selection Standards of College Students and Their Parents,” Journal of Social Hygiene, 22 (1936), pp. 115-125. "A Study of 325 Mixed Marriages," American Sociological Review, 2 (1937), pp. 705-716. Bakal, Carl. "The Risks You Take in Interfaith Marriage," Good Housekeeping, 149 (1959), pp. 62-63. Barron, Milton L. "The Incidence of Jewish Intermarriage in Europe and America," American Sociological Review, 11 (1946), pp. 6—13. "Research on Intermarriage: A Survey of Accomplishments and Prospects," American Journal of Sociology, 57 (1951), pp. 249“2550 Becker, Howard. "Normative Reactions to Normlessness," Sociological Review, 25 (1960), pp. 803-810. American Beesley, Thomas Q. ”What It Means to Marry a Protestant," The Forum, 82 (1929), pp. 226-230. 300. Beigel, Hugo G. "Body Height in Mate Selection," Journal of Social Psychology, 39 (1954), pp. 257-268. Bell, Wendell. "Anomie, Social Isolation, and the Class Structure," Sociometry, 20 (1957), pp. 105-115. Bend, Emil. "Marriage Offers in a Yiddish Newspaper--1935 and.1950," American Journal of Sociology, 58 (1952), pp. 60-66. Bernard, Luther L. "The Conflict Between Primary Group Attitudes and Derivative Group Ideals in Modern Society,” American Journal of Sociology, 41 (1936), pp. 611- 623. Besanceney, Paul H., S.J. "Unbroken Protestant-Catholic Marriages Among Whites in the Detroit Area," American Catholic Sociological Review, 23 (1962), pp. 3-20. Bettelheim, Bruno, and Morris Janowitz. "Ethnic Tolerance: A Function of Social and Personal Control," American Journal of Sociology, 55 (1949), pp. 137-145. Biegel, Hugo G. "Romantic Love," American Sociological Review, 16 (1951), pp. 326-334. Biesanz, John. "Inter-American Marriages on the Isthmus of Panama," Social Forces, 29 (1950), pp. 159-163. Blumberg, Leonard, and Robert R. Bell. "Urban.Migration and Kin- ship Ties," Social Problems, 6 (1959), pp. 328-333. Bond, Horace M. "Two Racial Islands In Alabama," American Journal of Sociology, 36 (1931), pp. 552-560. Borke, Helene, and Margaret B. McFarland. "Effect of Tradition on Husband Choice and Role as Wife: A Case Report," Journal of Clinical Psychology, 15 (1959), pp. 128-132. Bossard, James H. "The Age Factor in Marriage: A Philadelphia Study, 1931," American Journal of Sociology, 38 (1933), pp. 536-549. "Ecological Areas and Marriage Rates," American Journal of Sociology, 44 (1938), pp. 70-85. . "Marrying Late in Life," Social Forces, 29 (1951), pp. 405-408. , and Harold C. Letts. "Mixed Marriages Involving Lutherans," Marriage and Family Livipg, 18 (1956), pp. 308-310. . "Nationality and Nativity as Factors in Marriage," Amer- ican Sociological Review, 4 (1939), pp. 792-798. 301. "Residential Pr0pinquity as a Factor in Marriage Selec- tion," American Journal of Sociology, 38 (1932), pp. 219- 224 , Eleanor S. B011, and Winagene P. Sanger. "Some Neglected Areas in Family Life Study," Annals of the American Academy of Political and Social Sciences, 272 (1950), pp. 74-84. Bowerman, Charles E. "Age Relationships at Marriage, by Marital Status and Age at Marriage," Marriage and Family Living, 18 (1956), pp. 231-233. , and Barbara R. Day. "A Test of the Theory of Complemen- tary Needs as Applied to Couples During Courtship," Ameri- can Sociological Review, 21 (1956), pp. 602-605. Bowman, Claude C. "Uncomplimentary Remarks on Complementary Needs," American Sociological Review, 20 (1955), p. 466. Bressler, Marvin, and William.M. Kephart. "Marriage and Family Patterns of an Academic Group," Marriage and Family Livipg, 16 (1954), pp. 121-127. Brittain, Clay V. "Adolescent Choices and Parent-Peer Cross-Pres- sures," American Sociological Review, 28 (1963), pp. 385- 391. Bronfenbrenner, Urie. "Socialization and Social Class Through Space and Time," in Eleanor E. Maccoby, Theodore M. Newcomb, and Eugene L. Hartley (eds), Readings in Social Psychology, 3d ed., New York: Henry Holt, 1958, pp. 400- 425. Broom, Leonard, and Eshref Shevky. "Mexicans in the United States, a Problem in Social Differentiation," Sociology and Social Research, 36 (1952), pp. 150-158. Brown, James S. ."Social Class, Intermarriage, and Church Membership in a Kentucky Community," American Journal of Sociology, 57 (1951), pp. 232-242. Browning, Charles J., Malcolm F. Farmer, H. David Kirk, and G. Duncan Mitchell. "On the Meaning of Alienation," American Sociological Review, 26 (1961), pp. 780-781. Bugelski, B. R. "Assimilation Through Intermarriage," Social Forces, 40 (1961), pp. 148-153. Burchinal, Lee G., and Loren E. Chancellor. "Catholics, Urbanism, and Mixed-Catholic Marriage Rates," Social Problems, 9 (1962), pp. 359-365. 302. , William F. Kenkel, and Loren E. Chancellor, "Comparison of State and Diocese-Reported Marriage Data for Iowa, 1953-57," American Catholic Sociological Review, 23 (1962), pp. 21-29. "Membership Groups and Attitudes Toward Cross-Religious Dating and Marriage," Marriage and Family Living, 22 (1960), pp. 248-253. , and William K. Kenkel. "Religious Identification and Occupational Status of Iowa Grooms, 1953-1957," American Sociological Review, 27 (1962), pp. 526-532. Burgess, Ernest W. "The Family in a Changing Society," American Journal of Sociology, 53 (1948), pp. 417-422. , and Paul Wallin. "Homogamy in Personality Characteris- tics," Journal of Abnormal and Social Psychology, 39 (1944), pp. 475-481. , and Paul Wallin. "Homogamy in Social Characteristics," American Journal of Sociology, 49 (1943), pp. 109-124. Burma, John H. "Research Note on the Measurement of Interracial Marriage," American Journal of Sociology, 57 (1952), pp. 587-589 I Campisi, P. J. "Ethnic Family Patterns: The Italian Family in the United States," American Journal of Sociology, 53 (1948), pp. 443-449. Cartwright, Dorwin, and Alvin Zander. "Group Pressures and Group Standards: Introduction," in their Group Dynamics: Re- search and Theory, 2d ed., Evanston, 111.: Row, Peterson, 1960, pp. 165-188. Cary-Lundberg, Isabel, and Elwin H. Powell. "On Durkheim, Suicide, and Anomie," "Rejoinder to Dr. Cary-Lundberg," American Sociological Review, 24 (1959), pp. 250-253. Catapusan, B. T. "Filipino Intermarriage Problems in the United States," Sociology and Social Research, 22 (1938), pp. 265-272. Catholic Husband and a Protestant Wife. "1 Love You, But ..... ," Sign, 31 (1952), pp. 12-15, 74-76. Catton, William R. "The Functions and Dysfunctions of Ethnocentrism: A Theory," Social Problems, 8 (1960-61), pp. 201-211. Cavan, Ruth. "Regional Family Patterns: The Middle Western Family," American Journal of Sociology, 53 (1948), pp. 430-431. 303. Centers, Richard. "Marital Selection and Occupational Strata," American Journal of Sociology, 54 (1948), pp. 530-535. . "Occupational Mobility of Urban Occupational Strata," in Sigmund Nosow and William H. Form (eds.), Man, Work, and Socieiy, New York: Basic Books, 1962, pp. 336-343. Chambliss, Rollin, "Contributions of the Vital Statistics of Finland to the Study of Factors That Induce Marriage," American Sociological Review, 22 (1957), pp. 38-48. Chancellor, Loren E., and Thomas P. Monahan. "Religious Prefer- ence and Interreligious Mixtures in Marriages and Divorces in Iowa," American Journal of Sociology, 61 (1955), pp. 233-239. Chapin, F. Stuart. "The Protestant Church in an Urban Environment," in Paul K. Hatt and Albert J. Reiss, Jr. (eds.), Cities and Society: The Revised Reader in Urban Sociology, Glencoe, 111.: ‘Free Press, 1957, pp. 505-515. Cheng, C. K., and Douglas S. Yamamura. "Interracial Marriage and Divorce in Hawaii," Social Forces, 36 (1957), pp. 77-84. Child, Irvin L. "Socialization," in Gardner Lindsey (ed.), Hand- book of Social Psychology, Reading, Mass.: Addison - Wesley, 1954, II, pp. 655-692. Christensen, Harold T. "Student Views on Mate Selection," Marriage and Family Living, 9 (1947), pp. 85-88. Christopherson, Victor A., and James Walters. "Responses of Prot- estants, Catholics, and Jews Concerning Marriage and Family Life," Sociology and Social Research, 43 (1958), pp. 16-22. Cizon, Francis H. "Interethnic and Inter-religious Marriage Patterns in Parish X," American Catholic Sociological Review, 15 (1954), pp. 244-255. Clarke, Alfred C. "An Examination of the Operation of Residential Pr0pinquity as a Factor in Mate Selection," American Socio- logical Review, 17 (1952), pp. 17-22. Cloward, Richard A. "Illegitimate Means, Anomie, and Deviant Be- havior," American Sociological Review, 24 (1959), pp. 164- 176. Coakley, Thomas F. "Mixed Marriages, Their Causes, Their Effects, Their Prevention," Lumen Vitae, 4 (1949), pp. 455-460. Cook, Robert (ed.) "Spotlight on Marriage," P0pulation Bulletin, 17 (1961), pp. 61-79. 304. Coser, Rose L. "Insulation from Observability and Types of Social Conformity," American Sociological Review, 26 (1961), pp. 28-390 Cox, Oliver C. "Sex Ratio and Marital Status Among Negroes," American Sociological Revigg, 5 (1940), pp. 937-947. Curtis, Richard F. "Occupational Mobility and Church Participation," Social Forces, 38 (1960), pp. 315-319. David, Jacques. "Marriages mixtes et mariages religiousement homo- génes en Suisse," Lumen Vitae, 4 (1949), pp. 463-474. Davie, Maurice R., and Ruby Jo Reeves. "Pr0pinquity Before Mar- riage," American Journal of Sociology, 44 (1939), pp. 510- 517. » Davis, Kingsley, "Intermarriage in Caste Societies," American Anthropologist, 43 (1941), pp. 376-395. Davis, Thurston N. "The Loneliness of Man," America, 104 (1960), pp. 550-552. Dean, Dwight, and Jon A. Reeves. "Anomie: A Comparison.of a Cath- olic and a Protestant Sample," Sociometry, 25 (1962), pp. 209-212. . "Alienation: Its Meaning and Measurement," American Sociological Review, 26 (1961), pp. 753-758. DePorte, J. V. "Marriages in the State of New York with Special Reference to Nativity," Human Biology, 3 (1931), pp. 376- 396. Detroit Board of Education. The Detroit Educational Bulletin, Research Bulletin No. 9 (October, 1925), entire issue. Dinitz, Simon, Franklin Banks, and Benjamin Pasamanick. "Mate Selection and Social Class: Changes During the Past Century,‘ Marriage and Family Living, 22 (1960), pp. 348-351. Dohrenwend, Bruce P. "Durkheim's Egoism, Altruism, Anomie, and Fatalism," American Sociological Review, 24 (1959),.pp. 466-473. Douglass, Truman B. "Ecological Changes and the Church," Annals of the American Academy_of Political and Social Sciences, 332 (1960), pp. 80-88 Dreikurs, Rudolf. "The Choice of a Mate,” International Journal of Individual Psychology, 1 (1935), pp. 99-112. 305. Dubius. "New Light on Mixed Marriages," Ecclegiaptigal ngigw, 82 (1930), pp. 412-417. ‘ Duncan, Otis D., et_§l, "The Factor of Age in.Marriage," American Journal of Sociology, 39 (1934), pp. 469-482. Dunn, Edward S., S. J. "Mixed Marriages in 1947, " Ecclesiastical Review, 102 (1949), p. 34. Editors. "Editorial," Social Compass, 9 (1962), pp. 441-444. Ellis, Albert, and Robert M" Beechley. "Assortative Mating in the Parents of Child Guidance Clinic Patients," American Sociological Review, 14 (1949), pp. 678-679. , and Ruth R. Doorbar. "Recent Trends in Sex, Marriage, and Family Research, " Marriage and Family Research, 14 (1952), pp. 338- 340. Ellsworth, John 8., Jr. "The Relation of Population Density to Residential Pr0pinquity as a Factor in Marriage Selection, American Sociological Review, 13 (1948), pp. 444-448. Engelman, Uriah Z. "Intermarriage Among Jews in Switzerland, 1880- 1920," American Journal of Sociology, 34 (1928), pp. 516- 523. Erikson, Kai T. "Notes on the Sociology of Deviance," Social Prob- lems, 9 (1962), pp. 307-314. Femminella, Francis X. "The Impact of Italian Migration and Ameri- can Catholicism," American Catholic Sociological Review, Festinger, Leon, Stanley Schachter, and Kurt Back. "The Operation .of Group Standards," in Group Dynamics, Dorwin Cartwright and Alvin Zander (eds.L Evanston, 111.: Row, Peterson, 1960, pp. 241-259. Fishman,.Joshua A. "Childhood Indoctrination for Minority-Group Membership," Daedalus, 90 (1961), pp. 329-349. Freedman, Ronald. "The Detroit Area Study: A Training and Re- search Laboratory in the Community," American Journal of Sociology, 59 (1953), pp. 30-33. , and-Deborah Freedman. "Farm Boy in the City," in Ronald Freedman et al., Principles of Sociology, revised, New York: Henry Holt, 1956, pp. 462-471. Freeman, Howard E., and Gene G. Kassebaum. "Exogamous Dating in a Southern City," Jewish Social Studies, 18 (1956), pp. 55-60. 306. Freeman, Linton. ”Homogamy in Interethnic Mate Selection," Sociology and Social Research, 39 (1955), pp. 369-377. Gilbert, William H., Jr. "Memorandum Concerning the Characteris- tics of the Larger Mixed Blood Racial Islands of the Eastern United States," SOCial Forces, 24 (1946), pp. 438- 447. ‘ Glick, Paul C., and Emanuel Landau. "Age as a Factor in Marriage," American Sociological Review, 15 (1950), pp. 517-529. "Intermarriage and Fertility Patterns Among Persons in Major Religious Groups," Eugenics Quarterly, 7 (1960), pp. 31-38. , and Hugh Carter. "Marriage Patterns and Educational Level," American Sociological Review, 23 (1958), pp. 294- 300. Goldberg, David, and Harry Sharp. "Some Characteristics of Detroit Area Jewish and Non-Jewish Adults," in Marshall Sklare (ed.) The Jews: Social Patterns of an American Group, Glencoe, 111.: Free Press, 1958, pp. 107-118. Golden, Joseph. "Characteristics of the Negro-White Intermarried . in Philadelphia," American Sociological Review, 18 (1953), pp. 177-183. . "Facilitating Factors in Negro-White Intermarriage," Phylon, 6 (1945), pp. 273-284. . "Patterns of Negro-White Intermarriage," American Socio- logical Review, 19 (1954), pp. 144-47. "Social Control of Negro-White Intermarriage," Social Forces, 36 (1958), pp. 267-269. Golovensky, David 1. "The Marginal Man Concept: An Analysis and Critique," Social Forces, 30 (1952), pp. 333-339. Goode, William J. "The Theoretical Importance of Love," American Sociological Review, 24 (1959), pp. 38-47.‘ Grayson, Kirk. "The Filipinos," Annals of the American Academy of Political and Social Sciences, 223 (1942), pp. 45-48. Greeley, Andrew M. "Areas of Research on Religion and Social Or- ganization)‘ American Catholic Sociological Review, 23 (1962), pp. 99-112. . "Some Aspects of Interaction Between Religious Groups in an Upper Middle Class Roman Catholic Parish," Social Compass, 9 (1962), pp. 39-61. 307. Greenacre, Phyllis. "Child Wife as Ideal: Sociological Consider- ation," American Journal of Orthopsychiatpy. 17 (1947), pp. 167-171. Greenberg, Joseph H. "Sex Distribution and Marital Status," Sociology and Social Research, 33 (1949), pp. 368-372. Gurvitch, Georges. ”Social Control," in Georges Gurvitch and Wilbert E. Moore (eds.). Twentieth Century Sociology, New York: PhilOSOphical Library, 1945, pp. 267-296. Hajda, Jan. "Alienation and Integration of Student Intellectuals," American Sociological Review, 26 (1961), pp. 758-777. Hajnal, John. "Analysis of Changes in the Marriage Pattern by Economic Groups," American Sociological Review, 19 (1954), pp. 295-302. "Differential Changes in Marriage Patterns," American Sociological Review, 19 (1954), pp. 148-154. Hamilton, Maxwell. "We Wouldn't Marry Each Other Again," Catholic Digest, 19 (1955), pp. 8-11. Handlin, Oscar. "Historical Perspectives on the American Ethnic Group," Daedalus, 90 (1961), pp. 220-232. Harris, Daniel. "Age and Occupational Factors in the Residential Pr0pinquity of Marriage Partners," Journal of Social Psychology, 6 (1935), pp. 257-261. Harris, J. Arthur. "Assortative Mating in Man," POpular Science Monthly, 80 (1912), pp. 476-492. Harte, Thomas J., C. SS. R. "Trends in Mate Selection in a Tri- Racial Isolate," Social Forces, 37 (1959), pp. 215-221. Heer, David M. "The Marital Status of Second-Generation Americans," American Sociological Review, 26 (1961), pp. 233-241. "The Trend of Interfaith Marriages in Canada: 1922-1957," American Sociological Review, 27 (1962), pp. 245-250. Heiss, Jerold S. ”Premarital Characteristics of the Religiously Intermarried in an Urban Area," American Sociological Re- view, 25 (1960), pp. 47-55. Herskovits, Melville J. "Social Selection in a Mixed Population," Proceedings of the National Academy of Sciences, 12 (1926), pp. 587-5930 Hewitt, Lester E. "Students Perceptions of Traits Desired in Them- selves as Dating and Marriage Partners." Marriage and Family Living, 20 (1958), pp. 344-349. 308. Hill, G. W., and James D. Tarver. "Marriage and Divorce Trends in Wisconsin, 1915-45," Milbank Memorial Fund Qparterly, 30 (1952), pp. 5-17. Hill, Reuben. "Campus Values in Mate Selection," Journal of Home Economics, 37 (1945), pp. 554-558. "A Critique of Contemporary Marriage and Family Re- search," Social Forces, 33 (1955), pp. 268-277. "Marriage and Family Research," Eugenics Quarterly, 1 (1954), pp. 58-65. ”Review of Current Research on Marriage and the Family," American Sociological Review, 16 (1951), pp. 694-701. Himes, Joseph 8., Jr. "Mate Selection Among Negro College Students,‘ Sociology and Social Research, 33 (1949), pp. 204-211. "Value Consensus in Mate Selection Among Negroes," Mar- riage and Family Living, 14 (1952), pp. 317-321. "A Value Profile in Mate Selection Among Negroes," Mar- riage and Family Living, 16 (1954), pp. 244-247. Hobart, Charles W. ”The Incidence of Romanticism During Courtship," Social Forces, 36 (1958), pp. 362-367. Hollingshead, August B. "Age Relationships and Marriage," Ameri- can Sociological Review, 16 (1951), pp. 492-499. "Class and Kinship in a Middle Western Community," American Sociological Review, 14 (1949), pp. 469-475. . "Class Differences in Family Stability," in Reinhard Bendix and Seymour M. Lipset (eds.), Class, Status, and Power, Glencoe, 111.: Free Press, 1953, pp. 284-292. "The Concept of Social Control," American Sociologi- cal Review, 6 (1941), pp. 217-224. "Cultural Factors in the Selection of Marriage Mates," American Sociological Review, 15 (1950), pp. 619-627. Hughes, Everett C. "Institutions," in Alfred M. Lee (ed.), Principles of Sociology, revised, New York: Barnes and Noble, 1959, pp. 225-284. Humphreys, Norman D. "The Changing Structure of the Detroit Mexican Family: An Index of Acculturation," American Sociological Review, 9 (1944), pp. 622-626. Hunt, Chester L. "Cotabato: Melting Pot of the Philippines," Phillippine Social Sciences and Humanities Review, 19 (1954), pp. 40-72. 309. , and Richard W. Coller. "Intermarriage and Cultural Change: ‘A Study of Philippine-American Marriages,"-Socia1 Forces, 35 (1957), pp. 223-230. Hunt, Rolfe L. "Religion and Education," Annals of the American Academy of Political and Social Sciences, 332 (1960), pp. 89-100. Hunt, Thomas C. "Occupational Status and Marriage Selection," American Sociological Review, 5 (1940), pp. 495-504. Inamine, Otome, Phyllis Kon, Yan Quai Law, and Marjorie Okamoto. "The Effect of War in Inter-Racial Marriage in Hawaii," Social Process in Hawaii, 10 (1945), pp. 103-109. Janks, A. E. "Ethnic Census in Minneapolis," American Journal of Sociology, 17 (1912), pp. 776-782. Johnson, Hildegard B. "Intermarriages Between German Pioneers and Other Nationalities in Minnesota in 1860 and 1870," American Journal of Sociology, 51 (1946), pp. 299-304. Johnson, Roswell H. "Mate Selection," Proceedinge of the Second International Congress of Eugenics, 1 (1921), pp. 416-425. Jones, Harold E. "Homogamy in Intellectual Abilities," American Journal of Sociology, 35 (1929), pp. 369-382. Jones, Robert C. "Ethnic Family Patterns: The Mexican Family in the United States," American Journal of Sociology, 53 (1948), pp. 450-452. Kallman, Franz J., and Eva Bondy. "Applicability of the Twin Study Method in the Analysis of Variations in Mate Selection and Marital Adjustment," American Journal of Human Genetics, 4 (1952), pp. 209-222. Katz, Alvin M., and Reuben Hill. "Residential Pr0pinquity and Marital Selection: A Review of Theory, Method, and Fact," Marriage and Family Living, 20 (1958), pp. 27-35. Kelly, E. Lowell. "A Preliminary Report on Psychological Factors in Assortative Mating," Psychological Bulletin, 34 (1957), p. 749. "Psychological Factors in Assortative Mating," Psycho- logical Bulletin, 37 (1940), p. 473. Kennedy, Ruby Jo R. "Premarital Residential Pr0pinquity and Ethnic Endogamy," American Journal of Sociology, 48 (1943), pp. 580-584 a 310. . "Single or Triple Melting Pot? Intermarriage in New Haven, 1870-1950," American Journal of Sociology, 58 (1952), pp. 56-59. "Single or Triple Melting Pot? Intermarriage Trends in New Haven, 1870-1940," American Journal of Sociology, 49 (1944), pp. 331-339. Kent, Donald P. "Subjective Factors in Mate Selection: An Ex- ploratory Study," Sociology and Social Research, 35 (1951), pp. 391-398. Kerckhoff, Alan C., and Thomas C. McCormick. "Marginal Status and Marginal Personality," Social Forces, 34 (1955), pp. 48-55. . "Notes and Comments on the Meaning of Residential Pro- pinquity as a Factor in.Mate Selection," Social Forces, 34 (1956), pp. 207-213. , and Keith E. Davis. "Value Consensus and Need Comple- mentarity in Mate Selection," American Sociological Review, 27 (1962), pp. 295-303. Kernodle, Wayne. "Some Implications of the Homogamy-Complementar- ity Needs Theories of Mate Selection for Sociological Re- search," Social Forces, 38 (1959), pp. 145-152. Kimber, J. A. Morris. "Intersectional and Intrasectional Marriages in a Southern Bible College," Marriage and Family Living, 18 (1956), pp. 151-153. Kirkpatrick, Clifford, and Theodore Caplow. "Courtship in a Group of Minnesota Students," American Journal of Sociology, 51 (1945), pp. 114-125. . "A Statistical Investigation of the Psychoanalytic Theory of Mate Selection," Journal of Abnormal and Social Psychology, 32 (1937), pp. 427-430. . "Student Attitudes Towards Marriage and Sex," Journal of Educational Sociology, 9 (1936), pp. 545-555. Kiser, Clyde V. "Current Mating and Fertility Patterns and Their Demographic Significance," Eugenics Quarterly, 6 (1959), pp. 65-82. Kbhn, Melvin L. "Social Class and Parental Authority," American Sociological Review, 24 (1959), pp. 352-366. Kolehmainen, John 1. "A Study of Marriage in a Finnish Community," American Journal of Sociology, 42 (1936), pp. 371-382. 311. Koller, Marvin R. "Residential Pr0pinquity of White Mates at Mar- riage in Relation to Age and Occupation of Males, Columbus, Ohio," American Sociological Review, 13 (1948), pp. 613-616. Ktsanes, Thomas. "Mate Selection on the Basis of Personality Type: A Study Utilizing an Empirical Typology of Personality," American Sociological Review, 20 (1955), pp. 547-555. Landis, Benson Y. "Trends in Church Membership in the United States," Annals of the American Academy of Political and Social Sciences, 332 (1960), pp. 1-8. Landis, Judson T. "Marriages of Mixed and Non-Mixed Religious Faith," American Sociological Review, 14 (1949), pp. 401- 407. "Religiousness, Family Relationships, and Family Values in Protestant, Catholic, and Jewish Families," Marriage and Family Living, 22 (1960), pp. 341-347. Landis, Paul H., and Katherine H. Day. "Education,as a Factor in Mate Selection," American Sociological Review, 10 (1945), pp. 558-560. . "Rural-Urban Migration and the Marriage Rate," American Sociological Review, 11 (1946), pp. 155-158. Langhorne, M. C., and Paul F. Secord. "Variations in Marital Needs with Age, Sex, Marital Status, and Regional Location," Journal of Social Psychology, 41 (1955), pp. 19-37. Lauquier, Helen C. "Cultural Change Among Three Generations of Greeks," American Catholic Sociological Review, 22 (1961), pp. 223-232. Lauriat, Patience. "Marriage and Fertility Patterns of College Graduates," Eugenics Quarterly, 6 (1959), pp. 171-179. Lee, Alfred M. "Socialization of the Individual," in Alfred M. Lee (ed.), Principles of Sociology, revised, New York: Barnes and Noble, 1959, pp. 285-341. - Lee, Dorothy. "Individual Autonomy and Social Structure," in Bartlett H. Stoodley (ed.), Society and Self: A Reader in Social Psychology, New York: Free Press of Glencoe, 1962, pp. 223-233. Lee, Rose H. "The Marginal Man: A Re-Evaluation and Indices of Marginality," Journal of Human Relations, 4 (1956), pp. 27- 39. Leiffer, Murray H. "Interfaith Marriages and Their Effect on the Religious Training of Children," Lumen Vitae, 4 (1949), pp. 442-452. 312. . "Mixed Marriages and Church Loyalties,” The Christian Century, 66 (1949), pp. 78-80. . "Mixed Marriages and the Children,” The Christian Century, 66 (1949), pp. 106-109. Lenski, Gerhard E. "Social Correlates of Religious Interest," American Sociological Review, 18 (1953), pp. 533-544. Leslie, Gerald R., and Arthur H. Richardson. "Family Versus Campus Influences in Relation to Mate Selection," Social Problems, 4 (1956), pp. 117-121. Lieberson, Stanley. "The Old-New Distinction and Immigrants in Australia," American Sociological Review, 28 (1963), pp. 550-565. Lind, Andrew W. ”Attitudes Toward Interracial Marriage in Kona, Hawaii," Sociel Process in Hawaii, 4 (1938), pp. 79-83. Lipset, Seymour Ma "Social Mobility and Urbanization," in Paul K. Hatt and Albert J. Reiss, Jr. (eds.), Cities and Socieiy: The Revised Reader in Urban Sociology, Glencoe, 111.: Free Press, 1957, pp. 458-466. Litwak, Eugene. "Geographic Mobility and Extended Family Cohesion," American Sociological Review, 25 (1960), pp. 385-394. "Group Pressure and Family Breakup: A Study of German Communities," American Journal of Sociology, 61 (1956), pp. 345-354. . "Occupational Mobility and Extended Family Cohesion,” American Sociological Review, 25 (1960), pp. 9-21. . "The Use of Extended Family Groups in the Achievement of Social Goals: Some Policy Implications," Social Problems, 7 (1959-60), pp. 177-187. Liu, William T., and Frank Fahey. "Delinquency, Self Esteem, and Social Control: A Retreductive Analysis," American Cath- olic Sociological Review, 24 (1963), pp. 3-12. Locke, Harvey J., Georges Sabagh, and Mary M. Thomas. "Interfaith Marriages," Social Problemo, 4 (1957), pp. 329-333. Lowrie, Samuel H. "Dating Theories and Student Responses," Ameri- can Sociologicalmgeyiey, 16 (1951), pp. 334-340. . "Racial and National Intermarriage in a Brazilian City," American Journal of Sociology, 44 (1939), pp. 685-698. 313. McClusky, Howard Y., and Alvin Zander. "Residential Pr0pinquity and Marriage in Branch County, Michigan," Social Forces, 19 (1940), pp. 79-81. McCormick, Thomas C., and Boyd E. Macrory. "Group Values in Mate Selection in a Sample of College Girls," Social Forces, 22 (1944), pp. 315-317. McGinnis, Robert. "Campus Values in Mate Selection: A Repeat Study," Social Fotoee, 36 (1958), pp. 368-373. Mack, Raymond W., Raymond J. Murphy, and Seymour Yellin. "The Protestant Ethic, Level of Aspiration, and Social Mobility: An Empirical Test," American Sociological Review, 21 (1956), pp. 295-300. McKain, Walter C., Jr., and C. Arnold Anderson. "Assortative Mating in Prosperity and Depression," Sociology and Social Research, 21 (1937), pp. 411-418. Maier, Joseph, and William Spinrad. "Comparison of Religious Be- liefs and Practices of Jewish, Catholic, and Protestant Students," Phylon, 18 (1958), pp. 355-360. Marches, J. R., and G. Turbeville. "The Effect of Residential Pr0pinquity on Marriage Selection," American Journal of Sociology, 58 (1953), pp. 592-595. Marcson, Simon. "Intermarriage and Generational Status," Phylon, 12 (1951), pp. 357-363. "Predicting Intermarriages," Sociology and Social Re- search, 37 (1953), pp. 151-156. . "A Theory of Intermarriage and Assimilation," Social Forces, 29 (1950), pp. 75-78. Marvin, Donald M. "Occupational Pr0pinquity as a Factor in Marriage Selection," dournal of the American Statistical Association, 16 (1918), pp. 131-156. Mayer, Albert J., and Harry Sharp. ”Religious Preference and World- ly Success," American Sociological Review, 27 (1962), pp. 218-227. Meier, Dorothy L., and Wendell Bell. "Anomia and Differential Access to the Achievement of Life Goals," American Sociolo- gicaldReview, 24 (1959), pp. 189-202. Merton, Robert K. "Conformity, Deviation, and Opportunity-Structure," American Sociological Review, 24 (1959), pp. 177-189. "Intermarriage and the Social Structure: Fact and Theory," Psychiatry, 4 (1941), pp. 361-374. 314. Middleton, Russell, and Snell Putney. "Religion, Normative Stand- ards, and Behavior," Sociometry, 25 (1962), pp. 141-152. Mills, Theodore M. "Equilibrium and the Processes of Deviance and Control," American Sociological Review, 24 (1959), pp. 671- 679. Mitchell, Donald. "Residential Pr0pinquity and Marriage in Carver and Scott Counties, Minnesota," Social Forces, 20 (1941), pp. 256-259. Mitchell, Roy. "An Ethnic Distance Study in Buffalo," Sociology and Social Research, 40 (1955), pp. 35-40. Mizruchi, Ephraim H., and Robert Perucci. "Norm Qualities and Differential Effects of Deviant Behavior: An Exploratory Analysis," American Sociological Review, 27 (1962), pp. 391-399. "Social Structure and Anomia in a Small City," Ameri- can Sociological Review, 25 (1960), pp. 645-654. Monahan, Thomas P. "One Hundred Years of Marriages in Massachusetts," American Journal of Sociology, 56 (1951), pp. 534-545. , and Loren E. Chancellor. "Statistical Aspects of Mar- riage and Divorce by Religious Denomination in Iowa," Eugenics Quarterly, 2 (1955), pp. 162-173. Mowrer, Ernest R. "Social Crises and Social Disorganization," Amer- ican Sociological Review, 15 (1950), pp. 60-66. , and Harriet Mowrer. "The Social Psychology of Marriage," American Sociological Review, 16 (1951), pp. 27-36. Mui, Lorna H. ”Social Structure and Anomia,” American Sociological Review, 26 (1961), pp. 275-277. National Opinion Research Center. "Jobs and Occupations: A Popu- lar Evaluation," in Reinhard Bendix and Seymour M. Lipset (eds.), Class, Status, and Power: A Reader in Social Stratification, Glencoe, 111.: Free Press, 1953, pp. 411- 426. Neal, Arthur G., and Salomon Rettig. "Dimensions of Alienation among Manual and Non-Manual Workers," American Sociological Revlgfl; 28 (1963), pp. 599-608. Nelson, Lowry. "Intermarriage among Nationality Groups in a Rural Area of Minnesota," American Journal of Sociology, 48 (1943), pp. 585-592. Nettler, Gwynn. "A Measure of Alienation," American Sociological Reyieg, 22 (1957), pp. 670-677. 315. Newcomb, Theodore M. "Recent Changes in Attitudes Towards Sex and Marriage," American Sociological Review, 2 (1937), pp. 659- 667. Nimkoff, Meyer F. "Occupational Factors and Marriage," American Journal of Sociology, 49 (1943), pp. 248-254. Notestein, Frank W. "Differential Age at Marriage According to Social Class," American Journal of Sociology, 37 (1931), pp. 22-48. Nurge, Ethel. ”Factors Operative in Mate Selection in a Phil- ippine Village,: Eugenics Qparterly, 5 (1958), pp. 162-168. Obenhaus, Victor, W. Widick Schroeder, and Charles D. England. "Church Participation Related to Social Class and Type of Center," Rural Sociology, 23 (1958), pp. 298-308. Ogburn, William F. "Recent Changes in.Marriage," American Journal of Sociology, 41 (1935), pp. 285-298. Opler, Marvin K. "Woman's Social Status and the Forms of Marriage," American Journal of Sociology, 49 (1943), pp. 125-146. Orzack, Louis H. "Preference and Prejudice Patterns Among Rural and Urban Schoolmates," Rural Sociology, 21 (1956), pp. 29-33. Ozaki, Shigeo. "Student Attitudes on Interracial Marriage," Social Process in Hawaii, 6 (1940), pp. 23-28. Panunzio, Constantine. "Intermarriage in Los Angeles, 1924-33," American Journal of Sociology, 47 (1942), pp. 690-701. Pearson, Karl. "Assortative Mating in Man," Biometrika, 2 (1903), pp 0 481-498 0 Perplexus. "Some Revelations of a Recent Parish Census," Ecclesi- astical Review, 82 (1930), pp. 312-314. Pfeffer, Leo. ”Issues That Divide," Journal of Social Issues, 12 (1956), pp. 21-39. POpe, Liston. "Religion and the Class Structure," in Reinhard Bendix and Seymour M. Lipset (eds.), Class, Status, and Power: A Reader in Social Stratification, Glencoe, 111.: Free Press, 1953, pp. 316-323. Popenoe, Paul. "Assortative Mating for Occupational Level," Journal of Social Psychology, 8 (1937), pp. 270-274. "Mate Selection," American Sociological ReviQK, 2 (1937), pp. 735-743. 316. . "Recent Trends in American Marriages," Eugenical News, 32 (1947), pp. 9-11. Powell, Elwin H. "Occupation, Status, and Suicide: Toward a Re- definition of Anomie," American Sociological Review, 23 (1958), pp. 131-139. Price, C. A., and J. Zubrzycki. "The Use of Inter-Marriage Sta- tistics as an Index of Assimilation," POpulation Studies, 16 (1962), pp. 58-69. Price, Edward T. "A Geographic Analysis of White-Negro-Indian Racial Mixtures in the Eastern United States," Annals of the Association of American Geographers, 43 (1953), pp. 138-155. Prince, Alfred J. "Attitudes of College Students Toward Inter- Faith Marriages," Marriage and Fami1y_living, 19 (1957), p. 120, abstract from The Coordinator, 5 (1956), pp. 11-23. . "Factors in Mate Selection," The Family Life Coordina- tor, 10 (1961), pp. 55-58. "A Study of 194 Cross—Religion Marriages," The Family Life Coordinator, 11 (1962), pp. 3-7. Punke, Howard H. "Marriage Rate Among Women Teachers," American Sociological Review, 5 (1940), 505-511. Putney, Snell, and Russell Middleton. "Rebellion, Conformity, and Parental Religious Ideologies," Sociometry, 24 (1961), pp. 125-135. Rainer, John D., and I. Lester Firschein. "Mating and Fertility Patterns in Families with Early Total Deafness," Eugenics Qoarterly, 6 (1959), pp. 117-127. Rankin, Lois. "Detroit Nationality Groups," Michigan History Magazine, 23 (1939), pp. 129-211. Reiss, Albert J., Jr. ”Change in the Occupational Structure of the United States, 1910 to 1950," in Paul K. Hatt and Albert J. Reiss, Jr. (eds.), Cities and Society: The Revised Reader in Urban Sociology, Glencoe, 111.: Free Press, 1957, pp. 424-431. Resnik, Reuben R. "Some Sociological Aspects of Intermarriage of Jew and Non-Jew,” Social Forces, 12 (1933), pp. 94-102. Reuss, Carl F. "Research Findings on the Effects of Modern-Day Religion on Family Living," Marriage and Family Living, 16 (1954), pp. 221-225. 317. Risdon, Randall. "A Study of Interracial Marriages Based on Data for Los Angeles County,” Sociology and Social Research, 39 (1954), pp. 92-95. 7 Rogers, Everett M., and A. Eugene Havens. "Prestige Rating and Mate Selection on a College Campus," Marriage and Family Living, 22 (1960), pp. 55-59. Rosen, Bernard C. ”The Reference Group Approach to the Parental Factor in Attitude and Behavior Formation," Social Forces, 34 (1955). pp. 137-144. Rosow, Irving. "Issues in the Concept of Need-Complementarity," Sociometry, 20 (1957), pp. 216-233. Rossi, Peter H. and Alice S. "Some Effects of Parochial School Education in America," Daedalus, 90 (1961), pp. 300-328. Roucek, Joseph S. "The Development and Status of Social Control in American Sociology," American Catholic Sociological Review, 20 (1959), pp. 107-123. Ruchames, Louis. "Race, Marriage, and Abolition in Massachusetts," Journal of Negro Hietory, 40 (1955), pp. 250-273. Samenfink, J. Anthony. "A Study of Some Aspects of Marital Be- havior as Related to Religious Control," Marriage and Family Living, 20 (1958), pp. 163-169. Schachter, Stanley. "Deviation, Rejection, and Communication," in Dorwin Cartwright and Alvin Zander (eds.), Group Dynamics, revised, Evanston, 111.: Row, Peterson, 1960, pp. 260-285. Schellenberg, James A. "Homogamy in Personal Values and the Field of Eligibles," Social Forces, 39 (1960), pp. 157-162. , and Lawrence S. Bee. "A Re-Examination of the Theory of Complementary Needs in Mate Selection," Marriage and Family Living, 22 (1960), pp. 227-232. Schiller, Belle. "A Quantitative Analysis of Marriage Selection in a Small Group,” doppnoi of Social Psychology, 3 (1932), pp. 297-319. Schooley, Mary. "Personality Resemblances among Married Couples," Journal of Abnormal and Social Psychology, 31 (1936), pp. 340-347. Schnepp, Gerald J., S.M., and Louis A. Roberts. "Residential Pr0pinquity and Mate Selection on a Parish Basis," Ameri- pan Journal of Sociology, 58 (1952), pp. 45-50. "Three Mixed Marriage Questions Answered," Catholic World, 156 (1942), pp. 203-207. 318. Schwartz, Shepard. "Mate-Selection among New York City's Chinese Males," American Journal of Sociology, 56 (1951), pp. 562- 568. Seeman, Melvin. "On the Meaning of Alienation," American Sociolo- gical Review, 24 (1959), pp. 783-791. Sharp, Harry. "Graduate Training Through the Detroit Area Study," American Sociological Review, 26 (1961), pp..110-114. , and Morris Axelrod. "Mutual Aid Among Relatives in an Urban Population," in Ronald Freedman, et al. (eds.), Principles of Sociology: A Text with Readings, revised, New York: Henry Holt, 1956, pp. 433-439. , and Allan Feldt. "Some Factors in a Probability Sample Survey of a MetrOpolitan Community," American Sociological Review, 24 (1959), pp. 650-661. Simpson, Richard L., and H. Max Miller. "Social Status and Anomia," Social Problems, 10 (1963), pp. 256-264. Sirjamaki, John. "Culture Configurations in the American Family," American Journal of Sociology, 53 (1948), pp. 464-470. Slotkin, J. S. "Jewish-Gentile Intermarriage in Chicago," Ameri- can Sociological Review, 7 (1942), pp. 34-39. Smith, Eleanor, and J. Greenberg Monanane. "Courtship Values in a Youth Sample," American Sociological Review, 18 (1953), pp. 635-640. Smith, Ernest A. "Dating and Courtship in Pioneer College," Sociology and Social Research, 40 (1955), pp. 92-98. Smith, Mapheus. "Similarities of Marriage Partners in Intelligence," American Sociological Review, 6 (1941), pp. 697-701. Smith, William M., Jr. "Rating and Dating: A Re-study," Marriage end Family Living, 14 (1952), pp. 312-317. Srole, Leo. "Social Integration and Certain Corollaries: An Ex- ploratory Study," American Sociological Review, 21 (1956), pp. 711-712. Staley, Austin J., O. S. B. "Racial Democracy in Brazilian Marriage," American Catholic Sociological Review, 21 (1960), pp. 146-164. Staudt, Virginia M. "Attitudes of College Students Toward Marriage and Related Problems. 11. Age, Educational, Familial, and F Economic Factors in Marriage," Journal of Psychology, 34 (1952), pp. 95-106. _I_ U 319. Stern, Bernhard J. "Intermarriage," in Edwin R. Seligman (ed.), Eneyclopedia of the Social Sciences, VIII, pp. 151-154. Stoodley, Bartlett H. "Normative Attitudes of Filipino Youth Compared with German and American Youth," in Bartlett H. Stoodley (ed.), Society and Self: A Reader in Social Psychology, New York: Free Press of Glencoe, pp. 203-215. Strauss, Anselm. ”The Ideal and the Chosen Mate," American Journal of Sociology, 52 (1946), pp. 204-208. "The Influence of Parent-Images upon Marital Choice," American Sociological Review, 11 (1946), pp. 554-559. "Personality Needs and Marital Choice," Social Forces, 25 (1947), pp. 332-335. Sundal, A. Philip, and Thomas C. McCormick. ”Age at Marriage and Mate Selection: Madison, Wis., 1937-1943," American Sociological Review, 16 (1951), pp. 37-48. Sussman, Marvin B. "The Isolated Nuclear Family: Fact or Fiction," Social Problems, 6 (1959), pp. 333-340. , and H. C. Yeager. "Mate Selection Among Negro and White College Students,” Sociology and Social Research, 35 (1950), pp. 46-49. "Parental Participation in Mate Selection and Its Effects upon Family Continuity," Social Forces, 32 (1953), pp. 76-81. Tarver, James D. "Trend in Age at Marriage of Wisconsin Men and Women, 1909-1940," American Sociological Review, 16 (1951), pp. 246-247. Thomas, John L., S.J. "The Factor of Religion in the Selection of Marriage Mates," American Sociological Review, 16 (1951), pp. 487-491. "Out-Group Marriage Patterns of Some Selected Ethnic Groups," American Catholic Sociological Review, 15 (1954), pp. 9-18. ' "Religious Training in the Roman Catholic Family," Ameri- can Journal of Sociology, 57 (1951), pp. 178-183. Traynor, Victor J., S.J. "Urban and Rural Mixed Marriages," Social Order, 6 (1956), pp. 154-158. U. S. Bureau of the Census. "Religion Reported by the Civilian Population of the United States: March, 1957," Current POpulation Reports: Population Characteristics, Series P-20, No. 79 (1958). 320. Useem, John and Ruth H. "Minority-Group Pattern in Prairie Society," American Journal of Sociology, 50 (1945), pp..377-385. Vail, James P., and Virginia Ma Staudt. "Attitudes of College Students toward Marriage and Related Problems. 1. Dating and Mate Selection," Journal of Psychology, 30 (1950), pp. 171-182. - Van den Berghe, Pierre L. "Hypergamy, Hypergenation, and Mis- cengenation," Human Relations, 13 (1960), pp. 83-91. Van Leeuwen, Bernard, 0. F. M. "Le mariagemixte, facteur de déchristianisation en Hollande," Lumen Vitae, 4 (1949), pp. 425-438. Vernon, Glenn M4 "Interfaith Marriages--Bias in Professional Pub- lications," Religious Education, 55 (1960), pp. 261-264. Veroff, Joseph, Sheila Feld, and Gerald Gurin. "Achievement Moti- vation and Religious Background," American Sociological Review, 27 (1962), pp. 205-217. Vogel, Ezra. "The Go-Between in a Deve10ping Society: The Case of the Japanese Marriage Arranger," Human Organization, 20 (1961), pp. 112-120. Wallace, Karl M. "An Experiment in Scientific Matchmaking," Mar- riage and Family Living, 21 (1959), pp. 342-348. Weatherby, Ulysses G. "Race and Marriage," American Journal of Sociology, 15 (1910), pp. 433-453. Wessel, Bessie B. "Comparative Rates of Intermarriage among Dif- ferent Nationalities in the United States," Eugenical News, 15 (1930), pp. 105-107. . "The Ethnic Survey of New London, Conn., 1938-44, A Resurvey after Some Twenty Years," American Journal of Soci- ology, 50 (1944), pp. 85-98. "What It Means to Marry a Catholic," The Forum, 81 (1929), pp. 339- 345. Williams, Robin M., Jr. "Religion, Value-Orientations, and Inter- group Conflict," Journal of Social Issues, 12 (1956), pp. 12-20. Winch, Robert F. Thomas and Virginia Dtsanes, "Empirical Elabor- ation of the Theory of Complementary Needs in Mate-Selec- tion," Journal of Abnormal and Social Psychology, 51 (1955). pp. 508-513. 321. . "Interrelations between Certain Social Background and Parent-Son Factors in a Study of Courtship among College Men," American Sociological Review, 11 (1946), pp. 333-341. "Personality Characteristics of Engaged and Married Couples," American Journal of Sociology, 46 (1941), pp. 686-697. , Thomas and Virginia Ktsanes. "The Theory of Complemen- tary Needs in Mate-Selection: An Analytic and Descriptive Study," American Sociological Review, 19 (1954), pp. 241- 249. . "The Theory of Complementary Needs in Mate-Selection: Final Results on the Test of the General Hypothesis," American Sociological Review, 20 (1955), pp. 552-555. "The Theory of Complementary Needs in Mate-Selection: A Test of One Kind of Complementariness," American Socio- logical Review, 20 (1955), pp. 52-56. Wirth, Louis, and H. Goldhamer. "The Hybrid and the Problem of Miscegenation," in Otto Klineberg (ed.), Characteristics of the American Negro, New York: Harper, 1944, pp. 249- 369. Woywood, Stanislaus, 0. F. M. "Marriage Between Catholics and Non-Catholics," Homiletic and Pastoral Review, 40 (1940), pp. 409-419. "Marriage Impediment of Mixed Religion," Homiletic and Pastoral Review, 40 (1940), pp. 633-643. "Marriage of Catholics to Unbaptized Persons," Homiletic and Pastoral Review, 40 (1940), pp. 756-766. Yamamoto, George K. "Some Patterns of Mate Selection Among Naichi and Okinawans on Oahu," Social Process in Hawaii, 21 (1957), pp. 42-49. Yamamura, Douglas S., and Janet H. Higa. "Dating Preferences of University of Hawaii Students," Social Process in Hawaii, 20 (1956) 9 pp. 4-150 , and Raymond E. Sakumoto. "Inter-Ethnic Friendship and Dating Practices," Social Process in Hawaii, 19 (1955), pp. 35-44. Yinger, J. Milton. "Social Forces Involved in Group Identification and Withdrawal," Daedalus, 90 (1961), pp. 247-262. Young, Pauline V. "Social Problems in the Education of the Immi- grant Child," American Sociological Review, 1 (1936), pp. 419-429. 322. Znaniecki, Florian. "Social Organization and Institutions," in Georges Gurvitch and Wilbert E. Moore (eds.), Twentieth Century Sociology, New York: Philosophical Library, 1945, pp. 172-217. Zubrack, Fred. "Adolescent Values in Marital Choice," Marriage and Family Living, 21 (1959), pp. 77-78. Unpublished Manuscripts Burma, John H. ”Interrethnic Marriages in Los Angeles, 1948-1959," Paper read at meetings of the American Sociological Associa- tion, Washington, D. C., August 30, 1962. (Mimeographed) Detroit Commission on Community Relations. "Racial and Ethnic Characteristics of the POpulation: Detroit Metropolitan Area," A document based on the United States Census for the Detroit Standard Metropolitan Area, 1940, 1950, and 1960. (Multilithed) Freedman, Ronald, and David Goldberg. "Family Growth in the Detroit Area: The prospectus for the research program of the Detroit Area Study for 1961-62," Project 882, Detroit Area Study, Survey Research Center, University of Michigan, September, 1961. (Mimeographed) Heiss, Jerold S. "Interfaith.Marriage in an Urban Area." Unpub- lished Ph. D, dissertation, Indiana University, Blooming- ton, Indiana, 1958. Klietsch, Ronald G. "Pr0pinquity: Distance and Direction Cor- relates of Marriage Frequencies." Community Data: Stearns County, 1900-1960, Division of Rural Sociology-~Institute of Agriculture, University of Minnesota, 1961. (Duplicated) "Pr0pinquity: Distance and Direction in Mate Selection." Stearns County, Minnesota, University of Minnesota Miscell- aneous Bulletin No. 1079, May, 1961. (Duplicated) Kolivosky, Michael E. "Intermarriage Between Two Divergent Ethnic Groups as an Index of Assimilation." Unpublished Ph. D. dissertation, Michigan State University, E. Lansing, Mich- igan, 1953. Schellenberg, James A. "Models of Mate Selection: Rational, Ir- rational, and Non-Rational? Paper read at the meetings of the Ohio Valley Sociological Society, E. Lansing, Michigan, May 4, 1962. (Typewritten) Sharp, Harry. "The Detroit Area Study: A Review of Some Recent Research," Detroit Area Study, Report No. 1704, University of Michigan, Ann Arbor, Michigan, 1960. (Mimeographed) 323. "Family Income in Greater Detroit: {1951-1959." Detroit Area Study, Report No. 1681, University of Mich- igan, Ann Arbor, Michigan, 1960. (Mimeographed) "The Nationality Background of Detroit Area Residents," Detroit Area Study, Report No. 1203, University of Michigan, Ann Arbor, Michigan, 1957. (Mimeographed) APPENDIX A THE HISTORY AND DEMOGRAPHY OF THE DETROIT AREA In the proposal submitted to the Department of Sociology and Anthropology of Michigan State University at the outset of this study the following paragraphs were included: . The writer proposes initially to put on paper his limited knowledge (not being a native of Detroit) of the his- tory and movement of peoples in the Detroit area. Next, he will specify his predictions for the Detroit area based on this limited knowledge and the hypotheses stated in Section III of this proposal. Finally, he proposes to make a modest study of the history and demographic materials concerning the Detroit area, trying to use this new information to improve his predictions made without the benefit of a community study. The proposed "modest study" of the Detroit area "com- munity" would include a good history, U. 8. Census reports on the economy and population characteristics of the Detroit Metropolitan Area, relevant analyses of Detroit Area Study data, gross records of the movements of ethnic groups into and with- in this area, some published expressions of the norms of these groups relative to intermarriage. ' This part of the research project is not central to the design. However, it may be regarded as an illustration of the following hypothesis: If a study of documents on a community's growth is added to empirically-based generalizations from other areas, then the prediction of community relations such as inter- marriage patterns and factors will be notably improved. This appendix to the study is an attempt to carry out this part of the study. Let us begin with the writer's preliminary impressions, as recorded on December 31, 1962. Personal Impressions of Detroit's History It seems to me that Detroit was established as a trad- ing center and fort by the French in the late seventeenth or 324. 325. early eighteenth century. It occupied a strategic position on the straits between Lake Erie-and Lake Huron, providing a land route from the settlements of the St. Lawrence to the Ohio and Mississippi valleys as well as a center for the local fur trade. The French from Canada were the only white settlers until the British took control during the French and Indian War. ' It is my impression that Detroit grew slowly during the eighteenth and nineteenth centuries, attracting moderate num- bers of Anglo-Saxons, Irish, and Germans for/its commerce and industry (including horse-drawn carriages). At the start of the twentieth century its Sports car enthusiasts began to be- come auto magnates. Detroit became hungry for laborers. The old stock immigrant groups increased their numbers and were joined by other nationalities from Eastern and Southern Europe. Italians, Slovaks, Russian Jews, and especially Poles are here in conspicuous numbers; also settlements of Maltese and Greeks. Scandinavians, Iberians, and Orientals seem to have very small representations. World War II brought another labor shortage, with consequent importing of populations (Negro and white) from the border states of this country or those farther south, plus a fair number of Mexican-Americans. Detroit has had its share of intergroup strife, notably the Henry Ford and Father Coughlin anti-Semitic propaganda and the more recent race riots. Economically, Detroit has not been very well balanced. It is a banking center, but its industry is very strongly tied to the automobile. Consequently, Detroit has felt the depres- sion as well as the boom years more than most cities. The seasonal and relatively unpredictable production of automobiles has also contributed toward labor strife, especially in the depression years when the Congress of Industrial Organizations (CIO) was organizing the whole auto industry. Detroit as a whole, like most large cities, gives its political allegiance to the Democratic Party, which in turn is closely tied to the United Auto Workers (UAW). Labor-manage- ment disputes, as well as urban-rural conflicts, are-much in evidence in Detroit. Its newspapers follow the understandable pattern of supporting business, Republican, and urban interests. The city itself is governed by a mayor and city council, who are elected by popular vote-on a non-partisan ticket (I believe). The suburbs, also included in the Detroit area, are more typi- cally Republican and conservative.. The school systems in the Detroit area, both public and private, seem to enjoy good reputations. The principal institutions of higher education in the area are Wayne State University (formerly a community college) and the University of Detroit, a Catholic college. The University of Michigan in Ann Arbor and Eastern Michigan in Ypsilanti also are accessible. As in the city itself, engineering is a strong division in several of these institutions. The automobile also seems to dominate the traditions and values of the Detroit area. The city prides itself on its industrial capacity and is generally oriented toward change 326. and growth. Fine arts which do not fit this context have not flourished. Apart from ethnic traditions, the more humanistic values and ideologies seem to receive rather slight attention. Religious institutions, however, are much in evidence; Conventions of religious groups are frequently held in Detroit.- How might we expect interfaith marriages to be affected? If we restrict our attention to the white Christians, it seemm that the Poles and French Canadians would be the Catholic groups least likely to intermarry religiously. White Southern Prot- estants have probably immigrated too recently to engage in much intermarriage as yet. However, Protestants and Catholics who migrate from other cities (not farms) are probably upwardly mobile and more likely to be involved in interfaith marriages (at least on the part of Catholics). With these impressions on record, let us now see what knowledge can be added by a community study such as indicated, and whether this knowledge has any notable bearing on the prediction of inter- faith marriages. Geography and Early History1 It was Detroit's strategic location on the water, rather than the land, routes which led to its founding in 1701 by Antoine Cadillac and two hundred French Canadians. King Louis XIV, through his minister Count Pontchartrain, ordered a fort to be established here on the complaint of the beaver fur traders at Mackinac. It seems that the British traders, operating from the vicinity of the present Albany, New York, had to be prevented from competing for Indian trade by offering the cheaper English rum in larger quanti- ties than the French could profitably supply of their brandy. When, in 1758, Francois Picote de Bellestre, the last French commandant at Detroit, surrendered to the British under Major Rogers, 1Most of the facts in this and the following section are drawn from George B. Catlin, The Story of Detroit, Detroit: Even- ing News Association, 1923. This was recommended to the writer by a librarian of the Burton Historical Collection at the Detroit Public Library as "the best single volume history of Detroit." 327. this outpost still contained only 300 dwellings and about 2,000 inhabitants. Although the American Revolution was successful on the Atlantic seaboard, the British refused for some time to surrender Detroit and other western outposts to the Colonies because pre-War American debts to British creditors were still outstanding. Conse- quent to the Canada Act of 1791, Detroit was, in effect, part of the territory of Upper Canada and actually sent its two representa- tives to the first session of the Canadian legislature at Newark on September 17, 1792. The American flag, with its fifteen stars, did not fly over Detroit until July 11, 1796, when Col. John Francis Hamtramck took peaceful possession. From a population of 2,200, which included 178 Negro and Indian slaves, the number of residents soon dwindled to about 500, due to the exodus of British settlers to the Canadian side of the river. The low point of Detroit's vitality was reached when a fire completely leveled the town in 1805. Gen. William Hull, as Governor of Michigan Territory, and Judge Augustus B. Woodward went to Washington in that year and obtained permission to lay out a new city. Their plan was more-or- less followed, and each resident over seventeen years of age was given a lot of not less than 5,000 square feet. For a short time during the War of 1812 the British again held possession of Detroit. Even in 1816 it was still only a town of 850 inhabitants. The Flow of Immigration into the Detroit Area When in 1825 the Erie Canal was opened, a new era began for Detroit. Passenger traffic between the port of New York City 328. and the Great Lakes grew rapidly, and settlers in the eastern states began to migrate westward in swarms. What they found at Detroit was still largely a French settlement, as is evidenced by the suc- cessful candidacy of Father Gabriel Richard in 1824 as Michigan's representative to Congress. However, most business leaders in the community even then were of British stock. Equally significant for Detroit's growth was the early development of railroads in Michigan. Railroad lines were com- pleted to Toledo in 1837, to Chicago in 1847, and, most important, to the East through Canada in 1854. The Irish, many of whom had helped build the Erie Canal, were recruited in the 1830's to do railroad work. The Germans came soon after, in such numbers that they totaled nearly 35,000 by 1874, or about one third of the pop- ulation.2 But already the completion of Detroit was in the process of another change due to the arrival of increasing numbers of Poles, until they became the largest single national group with about 300,000 in the Polish colony in 1930.3 Starting somewhat later than the Polish, the Italians also immigrated in large numbers, until first and second generation Italians in the Detroit area totaled 67,711 in 1930.4 Russia should also be named as one of the largest contributors to Detroit's population during this period, 2Detroit Free Press, December 8, 1874. All references to newspapers in this study were made available through the courtesy of the Burton Historical Collection at the Detroit Public Library. 3Lois Rankin, "Detroit Nationality Groups," Michigan History Magazine, 23 (1939), p. 177. 41bid., p. 154. there being 50,507 from that country by 1930.. 329. It was estimated that fully half of these were Russian Jews, about 15,000 were Ukrainians and 10,000 were "Slavic Russians."5 groups in a metropolitan area like Detroit. It is difficult to estimate the relative size of ethnic The U. 8. Census determines the country of origin for the foreign born and native born of foreign or mixed parents, but this information is not often gathered in one place for a series of years. Political boundaries change, and countries may sometimes be grouped together in statis- tics. Besides, the various nationalities have all had their dis- tinctive patterns of immigration, relative to peak years and dura- tion of influx. However, for what it is worth, the following gives the ranks of the fifteen largest foreign-born groups in the Detroit area for selected years: gégy 1925a 1940b: 1950b 1960b ‘1. Poland Canada Canada Canada 2. Canada Poland Poland Poland ~3. German Great Britain Great Britain Great Britain 4. Russia Germany Italy Italy 5. Italy Italy Germany Germany 6. England U. S. S. R. U. S. S. R. U. S. S. R. 7. Hungary Hungary Hungary Asia 8. Austria Austria Asia Hungary 9. Scotland Asia Austria Yugoslavia 10. Belgium Yugoslavia Ireland Austria 11. Ireland Greece Yugoslavia Greece 12. Rumania Ireland Greece Czechoslovakia 13. Yugoslavia Czechoslovakia Czechoslovakia Ireland 14. Lithuania Mexico Mexico Mexico 15.. Greece So. & Cent. Am. So. & Cent. Am. So. & Cent. Am. tional Bulletin, Research Bulletin #9 (October, 1925), p. 14. 3Adapted from Detroit Board of Education, The Detroit Educa- This is based on the population of the city of Detroit only. 5Ibid., p. 190. 330. bAdapted from."Racial and Ethnic Characteristics of the Population, Detroit Metropolitan Area," Unpublished document pre- pared by the Research Division, Detroit Commission on Comunity Relations. This is based on data from the U. S. Census for the Detroit Standard Metropolitan Area. It can be noted from the above that Canada and Great Britain continue to be important sending nations for Detroit in recent years, as they were in the early history of the city. Two other ethnic groups, not foreign, must be mentioned. Negroes have been present in large numbers in Detroit since the 1830's; during both WOrld Wars they migrated to Detroit in very large numbers. Similarly, during and after World War II, whites from the Border States moved into the area by the thousands. In 1957, it was estimated that the southern white population in Detroit numbered ' about-200,000.6 We must not leave this section-without considering the growth of population in the Detroit Metropolitan Area as a whole. Rapid urbanization has been taking place everywhere in the United States during this century. In the years 1900-1950, the popula- tion of the Nation nearly doubled in size, but the population of the country's 162 Standard Metropolitan Areas became 3% times its 1900 size.7 Detroit experienced its most rapid growth in the dec- ades 1910-20 (112.7%) and 1920-30 (66.7%).3 In the-two year period at the start of war production, April 1, 1940, to May 1, 1942, ‘Detroit.showed the largest population gain of any metropolitan 6Detroit Free Press, August 26, 1957. 7Donald J. Bogue, Population Growth in Standard Metro- politan Areas, 1900-1959, Washington: Government Printing Office, 1953, p. vii. 81bid., p. 63. 331. area in the country, adding 336,000, compared to 231,000 for Washington, 149,000 for Chicago, and 131,000 for Los Angeles.9 But perhaps the most striking indication of Detroit's phenomenal growth is the following. Of the fourteen Standard Metropolitan Areas with a population over one million in 1950, the highest gaggp of growth, 1900-1950, were: Los Angeles 2,199.0% Detroit 606.6% San Francisco - 312.7% Cleveland 218.0% Chicago 163.6%10 No city can touch Los Angeles in break-neck expansion, but neither does Detroit have a close rival for second place. Development of Detroit's Industry Surprisingly enough, considering Detroit's northern loca- tion, the first industry of notable size in the city was the pro- cessing of tobacco, particularly cigars and chewing tobacco. For over fifty years this activity increased, until in 1922 the cigar busi- ness employed 12,000 persons.11 Another surprise to this writer was the discovery that Detroit was in the 1920's the chief center of the stove industry in America. Starting with J. Dwyer & Co. in 1861, five companies became engaged in stove making and by 1923 were employing 5,000 men in Detroit.12 9Detroit News, December, 1942. 10Bogue, op. cit., pp. 62-69. 11Catlin, pp, cit., p. 471. 121bid., pp. 472-477. 332. Two events farther north were of special significance for Detroit's economic growth. In the 1840's, Douglass Houghton and William A. Burt surveyed the vast wealth of iron and copper in the Upper Peninsula of Michigan. In 1855, through the promotion of William.Burt's son_John and other Detroiters and supported by a land grant from Congress, a ship canal was completed at Sault St. Marie to make it possible for shipping to navigate the eighteen foOt fall between the level of Lake Superior and Lake Huron.13 Eber B. Ward was a pioneer in the steel industry and, in 1864, manufactured at Wyandotte (near Detroit) the first Bessemer steel made in America. However, the panic of 1873 made his associates in Detroit timid for a while; so he and others deve10ped this industry chiefly in the neighborhood of Chicago.14 The promotion of railways began in Michigan before there was.any rail communication with the East. It was therefore neces- sary to build railroad cars (modeled on the old stagecoach) in Detroit. Already in 1898, the Detroit plants doing this work em- I ployed more than 9,000 men.15 Another large industry during those days was the manufacture of pharmaceuticals, in which three com- panies employed a total of over 5,000 people by 1923.16 However, as everyone knows, it was the automobile whose development and manufacture made Detroit one of the most populous 13Ibid., pp. 493-494. 14Ibid., pp. 498-499. 151bid., p. 653. 16Ibid., pp. 655-656. 333. centers in the United States. The Oldsmobile of 1901, made in Detroit, was the first of its automobiles to sell by the thousands. It was in Detroit also that Henry M. Leland developed the Cadillac and later the Lincoln. After John and Horace Dodge had contributed their talents to the making of the first successful Olds and the early Fords, they formed their own company in Detroit. Of course, the most successful auto manufacturer was Henry Ford. In his com- pany's first year of operation 1,708 cars were produced, but after twelve years he had already passed the one million mark.17 To judge from Rankin's frequent mention of the fact, it was Ford's dramatic announcement in 1913 of a wage scale of $5. a day which attracted recent European immigrants from other parts of the coun- try to Detroit.18 In 1925, the Ford Motor Co. employed 51,533 persons in Highland Park and 70,682 in the River Rouge and Dearborn area.19 If the making of automobiles (together with armaments dur- ing the two World Wars) has made Detroit the third city in America in manufacturing employment, it has also given the city's leaders some headaches. There were 29.5 million automobiles in use in America in 194L,and 45.9 per cent of the men who made them worked in the Detroit area. For Detroiters this means that unless automobiles are sold, the economic structure of the area suffers drastically. 'In 1940, the automobile industry directly provided two-thirds of the manufacturing employment and one-third of all employ- ment in the area. When to this total are added the industries ‘17Ibid., p. 718. 18Rankin, op. cit., pp. 134, 141, 164, 175. 19Detroit News, January 10, 1926. 334. supplying the automobile industry and the industries, trades and services supplying the auto workers with secondary goods and services, it is found that the automobile industry directly and indirectlg'is the source of 80 per cent of all employment in the area.2 ' As a consequence of this concentration, Detroit has experi- enced severe growing pains in the peak years of auto sales and in wartime, followed by extreme unemployment in the Great Depression and in the transitions from wartime to peacetime production. In addition, the annual re-tooling for the production of new models has created much seasonal unemployment, a condition which has been largely corrected in recent years. However, due largely to the strains just mentioned, labor-management relations in the Detroit area have not always been ideal. The City's Churches One subject very pertinent to this study is the develop- ment of religious groups in the Detroit area.- During the first century of Detroit's existence as a fort and trading post, St. Anne's Roman Catholic Church, built by the French colony in 1701, was the only church maintained there. Shortly after'the great fire of 1805, this condition was remedied. The Congregationalist Rev. David Bacon and his wife had taught school in Detroit.before that, and several itinerant Methodist ministers visited Detroit; but it was not until 1818 that the Presbyterian Rev. Gideon Lanning promoted the building at River Rouge of the first Protestant 20City Plan Commission, Economic Base ofiDetroit, Detroit: City Play Commission, 1944, p. 7. 335. Church in Michigan.21 The Methodists built their first church in Detroit in 1821, the Episcopalians in 1828, the Presbyterians in 1825, and the Baptists in 1835, although in each instance the con- gregation was organized several years prior to the erection of the church. According to the city directory of Detroit in 1846: Church founding and building had brought into exist- ence seventeen societies and fifteen churches...Ste. Anne's, ...Holy Trinity,...St. Mary's,...St. Peter's...St. Paul's Episcopal,...the First Presbyterian,...the First Methodist,... the Schotch Presbyterian,...the First Congregational,...the’ Second Methodist,...the Colored Methodist,...the First Baptist, ...the Colored American Baptist,...the German Lutheran,...the Bethel or Mariners' Church,...In addition there was a German Evapgelical Lutheran Society...and a society of the Disciples In 1941, a directory of churches in the Detroit area was published, with this summarizing statement: This ecclesiastical directory of Greater Detroit is a listing of approximately 90 specific church denominations, operating within the Greater Detroit area, controlling some- 1,300 churches and institutions, and in addition to these groups there are approximately 100 independent or non-affiliated churches.23 Asia rough index of the activities of the various churches in the Detroit area, the number oprages used to list the churches and institutions of each group of denominations is as follows (in order): 21Catlin, op. cit., p. 220. 221bid., p. 423. 23Historical Records Survey Project, WPA, Directory of Churches and Religious Organization : Greater Detroit;_lgfll,Detroit; Michigan Historical Records Survey Project, December, 1941. 336. Roman Catholic Church - 36 pages Churches of God‘ - 3 Lutheran Bodies - 18 " Evangelical & Reformed Ch.- 3 Baptist Bodies - 16 ” Latter-Day Saints - 3 Methodist Bodies - ll " and 2 pages each, for: Presbyterian Bodies - 7 " Adventists, Assemblies 9f Christian Scientist - God, Church of God in Christ, (incl. practitioners) - 7 " Churches of Christ (non-affil- Prot. Episcopal Church - 6 " iated), Jewish Bodies - 5 " Disciples of Christ, Eastern (almost all Orthodox) Orthodox Churches, Salvation Army, Spiritualists. Unfortunately, as a departure from the proposal for this study, set down at the beginning of this Appendix, the writer did not find it possible to search out published expressions of the norms of these groups in Detroit relative to interfaith marriages. Some Earlier Findings of the'Detroit"Aréa Stpgy Since the University of Michigan has been conducting an annual survey by interviewing residents in the Detroit area since 1951, countless facts have been discovered and analyzed over the years. Those which seem to be pertinent to this study will be summarized here. First the books, then shorter reports. Daniel R. Miller and Guy E. Swanson asked many interesting questions about child-training by Detroit mothers in their study conducted in 1952-53. Although the evidence presented is not strong, the answer they give to two questions is of possible rele- vance to the present study: Are Protestants more likely than Roman Catholics to emphasize the teaching of self-control and personal accountability in training their children? The answer to both of these ques- tions is generally positive within the entrepreneurial middle and lower classes, but ambiguous within the bureaucratic 337. setting.24 On this basis, if we assume an equally strong norm among Prot- estants and Catholics against interfaith marriages, we might expect Pratestants to be somewhat more successful in obtaining compliance. However, the connection seems somewhat tenuous. In the 1955 Detroit Area Study, directed by Robert 0. Blood, Jr., and Donald M. Wolfe, it was found that relatives lived nearby and were important to most of the housewives interviewed. "Only a very small minority of Detroit area couples (11 per cent) have no relatives living in the metropolis." Over half of the housewives said they had at least one relative living in the same "neighborhood" as themselves.25 Furthermore, about half of the couples were related to six or more other Detroit area families. Three-quarters of them had siblings in the Detroit area. Visiting relatives at least once a week was the mode, whereas only one in four reported that she never attended large family gatherings.26 Four out of five considered relatives more "important" than friends.27 The obvious conclusion to be drawn from all of this for the present study is that relatives are potentially important agents of social control in Detroit. The 1955 study also found, contrary to a common impression, 24The Chan in Americgn Parent: A Study in the_Qetroit Area, New York: John Wiley & Sons, 1958, p. 169. 25A Social Profile of Detroit: 1955, Ann Arbor University of Michigan, 1956, p. 21. 25Ibid., p. 23. 271bid., p. 25. 338. that the patriarchial family is not characteristic of immigrants or of Catholics in the Detroit area.28 Perhaps this notion has been exaggerated in the past. It seems, rather, to be the husband's high status in the community which gives him a position of power in the decision-making of the family.29 The findings of this study relative to religious homogamy were discussed above in the chapter on Social Control. The 1958 Detroit Area Study, conducted by Gerhard Lenski, is rich in findings relevant to the hypotheses of the present study.. In communal solidarity and strength, for instance, white Protestants in Detroit were found to be closely similar to Catho- lics. Respectively, 86 and 84 per cent reported their spouse to be of the same faith, 73 and 70 per cent said their spouse was raised in the same faith as themselves, 76 and 79 per cent stated that all or nearly all of their relatives were of the same faith, whereas 38 and 44 per cent said the same about all or nearly all of their close friends.30 These facts would seem to imply that we should not expect to find differences in Detroit in the degree to which these two religious groups hold or enforce a religiously endogamous norm. 28Robert 0. Blood, Jr., and Donald M, Wolfe, Husbands and Wiveg; “The Dynamics of Married Living, Glencoe, 111.: Free Press, 1960, p. 25. 29Ibid., p. 36. 30Gerhard Lenski, The Religious Fgctor: A Sociological Studyiof Religion's Impact on Politics, Economics, and Familnyife, Garden City, N. Y.: Doubleday, 1961, pp. 35-36. 339. In contrast to Will'Herberg's thesis, Lenski also finds' thatzamong white Protestants and Catholics in Detroit the second generation did not attend weekly religious services in fewer numr bers than the first generation of Americans; in fact, second genera- tion Catholics were more active by this criterion. For both groups the third generation increased its religious participation.‘ More- over, Southern-born white Protestant migrants to Detroit are no more active than first or second generation Americans, even though they have been in this country for six or more generations. The increased religious activity of the third and later generations of non-Southern whites holds even when controlling for the class posi- tion of socio-religious groups. But more frequent church attend- ance is also associated with being in the middle class rather than in the working class and with having more years of education.31 Since church attendance is Lenski's criterion for associational vitality, perhaps we should not look for these factors to be re- lated to a communal aspect like religious endogamy. To find a close relationship with the same factors would weaken his distinction; between the community and association dimensions of a religious group. Compared to the Jewish and Negro Protestant groups, Lenski's data showed both white Protestant and Catholic groups to be far less concentrated residentially in the twelve sections of the Detroit area. They were also more evenly divided across economic 311bid., pp. 41-46. 340. classes than the other two groups, with Protestants more frequently belonging to the middle class and Catholics to the working‘clas‘s.32 Both of these factors would be conducive to higher intermarriage rates between the two groups. In studying the relationship of church attendance (asso- ciational involvement) to intergenerational occupational mobility, Lenski found only 19 Protestants and 27 Catholics who were upwardly mobile. However, his conclusions are interesting: On the basis of these data it appears that involve- ment in the Catholic church does not have the same consequences as involvement in the white Protestant churches. At best it seems to be irrelevant to mobility, and at worst, something of a hindrance. The relationship between upward mobility and current church attendance among Catholics therefore appears to be a result of the influence of mobility on attendance rather than the reverse, as appears to be the case among white Prot- estants. The fact that upwardly mobile sons and daughters of working-class and farm parents are more faithful in their at- tendance at Mass seems to reflect conformity, or overconformr ity, to the standards of their new social eers rather than the influence of childhood socialization.3 Since the number of cases is so small, we may find that this is a special sub-sample of people who have changed their religious persuasion from that of their parents when they married, thus per- haps adopting the mode of church attendance of their new religious group. It is not altogether clear from the text that Lenski was talking about parents and children who belonged to the same reli- gious group during their married life. A test of the hypothesis on occupational mobility in our present study should present a good comparison with the conclusions Lenski draws above. In fact, 321bid., pp. 72-73. 33Ibid., p. 105. 341. the plausibility of the interpretation proposed for the present study is supported by the following observation of Lenski's study, when we remember that "communal involvement" is measured largely by religious endogamy: When we examined the interrelations between vertical mobility and the involvement of white Protestants and Catho- lics in their subcommunities, we found that a high degree of communal involvement was consistently linked with a low rate of vertical mobility. Those who were highly involved in their socio-religious sub-communities were less likely either to have fallen in the class system than were those who were more marginal to the group. This pattern was especially marked for Catholics.34 This is quite consistent with our hypothesis that upwardly mobile persons do not feel bound by a religiously endogamous norm, Lenski found the same pattern to be true for the downwardly mobile in both religious groups. Citing the 1952 Detroit Area Study, as well as his own, Lenski found Catholics to be more closely tied to their kin group by the criteria of being less likely to have no relatives living in the Detroit area, of being more likely than Protestants to have weekly visits with relatives rather than neighbors, and of being more likely to name relatives rather than friends as having greatest influence on religious beliefs.35 In fact, total separation from the kin group was especially pronounced among middle-blass white Protestants. Therefore we should expect relatives to be less effective as agents of social control in the matter of religiousi endogamy among Protestants. The fact that, at all economic levels, 341bid., p. 110. 35Ibid., pp. 195-197. 342. white Protestants more often than Catholics said that it was most important for a child to learn "to think for himself" rather than "to obey" also points in the same direction.36 What about Catholics who attend Catholic schools? we have hypothesized that the Church is here acting as an agent of social control and that Catholics in such schools are less likely to con- tract marriages with Protestants. Before taking note of Lenski's relevant finding, let us put on record the observations of one of Detroit's historians: Unlike the Poles, the Italians have established few parochial schools....However, the Italians have nothing to compare with the Polish "Home." Most of their societies have only male membership,...With the drift from the church and the lack of cohesion in their social life, the restraining influ- ence of religion and group 0 iniOn no longer function among Italians of a certain class. 7 Lenski's facts are very similar: The one really substantial difference associated with the type of education Catholics received appeared when we con- sidered nationality background. Although 46 per cent of the Catholics of northwest European background (N'95) reported a Catholic education, only 28 per cent of the Catholics with a south or east European background (N’122) reported this.33 Since class and income are not related to this, he is inclined to attribute the difference to a survival of the Latin anti-clerical tradition. Perhaps. It may also be that, since the State and benefices support the Church in southern Europe, the Italians have no tradition of personally supporting their church or school. In 36Ibid., pp. 200-201. 37Rankin, op. cit., pp. 159-160. 38Lenski, op. cit., p. 241. 343. any case, we must be alerted to this difference when we examine the influence of Catholic schooling on interfaith marriage rates. We can now turn to several unpublished reports written by Harry Sharp, Director of the Detroit Area Study, 1954-61. He has drawn together some findings which are likely to have relevance to the hypotheses of the present study. For instance, his analysis of income differentials in recent years in Detroit gives us an important context for marriage patterns. He finds that greater Detroit is still a high income community, though declining rela- tive to the rest of the nation, with low income families becoming somewhat more numerous. Differences between the incomes of blue- collar and white-collar workers remain steady, with the median for semi-skilled factory workers being just 60 per cent of the income of professionals, managers, and proprietors. However, unemploy- ment in the 1958 recession "was much more frequently faced by blue-collar workers than by other labor force members." Moreover, ’the relationship of income to education level is becoming stronger, the median income advantage of college over seventh grade education increasing from $2,500, in 1951 to $5,400, in 1959.39 How do such things relate to interfaith marriages? They reinforce our impres- sion that the workers of Detroit, especially at lower levels, experience considerable economic strains, and may therefore be anomic relative to social norms such as endogamy. The experience 39Harry Sharp, "Family Income in Greater Detroit: 1951- 1959," Detroit Area Study, Report #1681, Ann Arbor: University of Michigan, 1960. (Mimeographed.) 344. of educational and occupational mobility is perhaps more likely now than formerly to produce an anomic outlook. Migration is, of course, very pertinent to marriage.patterns. Using the data of the 1952 survey, Sharp notes that, "Of all persons 21 years old or older now living in greater Detroit, only about three out of every ten are native-born residents of the community.40 Only 15 per cent of the population are foreign-born; another 25 per cent belong to the second generation. It turns out that adults of Polish background are more likely to be "natives" of the Detroit area than are adults of any other extraction. As of 1952, most migrants to Detroit who were not-foreign-born came from other come munities of the North. Two out of five Negro Detroiters were born on a farm in the South, but only about one in twenty migrant whites comes from a southern farm. However, migration from.the South seems to be increasing relatively. As in other urban centers, internal migration in the Detroit area has been pronounced. In 1950, more than one-third of the total population of greater Detroit lived within six miles of the center; ten years later the proportion was closer to one-fifth. Consequent to these and other residential moves, "approximately one half of those adults who attend a given Detroit area church have been going there no longer than five years."41 This does 4OHarry'Sharp, "The Detroit Area Study:' A Review. of Some Recent Research, "'Detroit Area Study, Report #1704, Ann Arbor: University of Michigan, 1960. (Mimeographed) 41Ibid., pp. 6-7. 345. not sound like the average parish or congregation can expect to be a center for social control relative to religious endogamy. What is the religious composition of the Detroit area's white population? Almost one half of the adults are Protestants; a slightly smaller proportion are Catholics; about 4 per cent are Jewish; 3 per cent express no preference. About 40 per cent of the whites are either Lutherans, Methodists, Calvinists (primarily- Presbyterians), or Baptists, in nearly equal proportions. Episco- palians add another 4 per cent, and smaller Protestant "sects"‘ include about 3 per cent.42 Since expgcted rates of intermarriage are higher for smaller groups, we should keep these proportiOns in mind when examining interfaith marriage rates in Detroit. Ethnicity is certain to be a complicating factor in trying to predict rates of interfaith marriages. There are at least fifty different ethnic groups in the Detroit area, with most of them having more than 10,000 representatives in the community.43' The groups show different tendencies toward residential concentration. Some, like the Poles, Hungarians, Jews, and Italians, clung to- gether in colonies. Others, like the Canadians, British, Irish, and Germans, were mostly dispersed from.the beginninga44l Follow- ing World'War I, there was a break-up of nearly all of the foreign colonies in the city, hastened by the rapid migration of Negroes. 421bid., pp. 7-8. 43Harry Sharp, "The Nationality Background of Detroit Area Residents," Detroit Area Study, Report #1203, Ann Arbor: University of Michigan, 1957. (Mimeographed) 44Rankin, op. cit., passim. 346. into Detroit at that time.45 'However, even in 1940, heavy ethnic concentrations were characteristic of Negroes, Polish, Italians, Russians, and Belgians, but much less true of Asians, Irish, Swedes, Greeks, Austrians, Germans, and Y’ugoslavs.46 we may assume, there- fore, that ethnic ties are still of considerable importance in some groups, even though we occasionally have_seen a statement such as, "Children of Italian parentage seem to repudiate the language, religion, and customs of their fathers more often than do the children of other foreign groups."47 For an example of . ethnic ties, as recently as 1941, Archbishop Mooney responded to severe dissension among Polish Catholics by stating the principle that, "Priests who labor in our Polish parishes today must be - bilingual."43 ‘Even in 1957, there were.approximately 200 churches and synagogues in the Detroit area which were attended chiefly by persons of a given nationality background.49 we should expect these differences to diminish in time, since among Detroiters under thirty-five years of age there is now little distinction in educational level between the various ethnic groups. New Expecpapions for Rates of-Prpgestant-Catholip ";E§£££sx§8'ffifpetr01t Now we should be ready to conSider whether all of these 451bid., p. 155. 46Mayor's Interracial Committee, Maps of I940 Population: Racial and Nationality Group Distribution, Detroit: ‘Mayor's Inters racial Committee, 1950. " ’ 47Rankin, op. cit., p. 161. 43Detroit News, April 24, 1941.1 49Sharp,"The Nationality Background of Detroit Area Residents," op. Cit., p. 4. 347. findings regarding the Detroit area "community" give us reason to reverse, omit, or modify any of the hypotheses of this study which were set down on the basis of sociological theory and the generali- zation of earlier research. In general, although we cannot compare Detroit with other communities through the data of this study, we have two reasons for expecting the situation here to be more anomic regarding so- cial norms than elsewhere:. 1) Detroit has grown more rapidly than other cities, bringing in more migrants; 2) the city's dependence on the auto industry has meant seasonal unemployment and great fluctuations in depression and prosperity. On the other hand, since we learned that only three out of ten adult Detroiters were native to the area in 1952, perhaps most of the marriages we shall be looking at were not contracted in the Detroit area. rIn this case the results should not be much different from other communities. Looking at the specific hypotheses proposed for this study, Blood's study has told us that we should not think in terms of patriarchal families. Our hypotheses concerning parents as agents of social control should apply to mothers as much as to fathers. For our second specific hypothesis and its sub-hypotheses, we have found in Blood's study that the family or relatives are, indeed, potentially strong agents of social control. However, Lenski's data seem to tell us to expect this to be more true of Catholics than of Protestants in Detroit. Since attendance at Catholic schools has been shown to be more characteristic of those having a northwest European national origin than of the south European groups, we should ideally control 348. for national origin when testing the relationship of Catholic school attendance to interfaith marriages. However, the number of cases is likely to be too small to accomplish this. Regarding the ethnic group as an agent of social control, we can use Lenski's distinction between the community and associa- tion aspects of a religious group to tell us to expect increased interfaith marriages not only in the second, but also in the third and later, generations. Even though attendance at church increases (as he finds) with Americanization, communalfeeling'may become weaker. For the second generation, it has been theorized that the parent-child strains may weaken social control. For later generations, the emphasis is here being placed on the tendency to enter more into the larger community, weakening the sub-community ties through AmericanizatiOn. Furthermore, as a result of our study of the Detroit area, we have more knowledge now of individual ethnic groups, in case it becomes possible to use our data to look at them individually or in categories. Sharp found that only five per cent of the white immigrants to Detroit came from a southern farm. This suggests that we may not have enough cases in.our data to teat‘the hypothesis that migration from.a farm to Detroit will be associated with inter- faith marriages--as a manifestation of anomie. Perhaps we can achieve some test of it by including those who have migrated from a small town. Lenski's "communal involvement" was partly measured by intermarriage. Since he-found that this involvement was less strong among those who had experienced occupational mobility, either 349. upward or downward, we should also include the downwardly mobile in our hypothesis of association with interfaith marriage. How- ever, if the relationship holds up, our explanation will flow from the anomic situation. In summary, through our study of the Detroit area, we have been led to expand two of our hypotheses in response to Lenski's findings and concepts. We are now more aware of difficulties to be met in testing the hypotheses concerning Catholic schools and rural migration. But we have added reason to expect our hypotheses about social control by parents and family to be borne out. Beyond this, our study of the Detroit "community" has given us a context for understanding and interpreting what we do find in our data. APPENDIX B DETAILED MARRIAGE CHOICES AMONG PROTESTANTS AND CATHOLICS, INCLUDING THE POSSIBLE CHANGES IN RELIGIOUS PREFERENCEa Year of Detroit Area Survey Religion of Parppers 1955 1958 1959 All Years , Husband always Catholic 178 104 168 450 Wife always Catholic (11)b Wife formerly Protestant, 18 17 22 57 now Catholic (12) 30 17 ll 58 Wife always Protestant (13) Wife formerly Catholic, 1 1 now Protestant (14) Wife formerly Catholic, now other than Catholic or Protestant (15) Wife formerly Protestant, now other than Catholic or Protestant (16) Wife formerly other, 1 10 6 17 now Catholic (17) Wife formerly other, now Protestant (18) 3 1 2 6 Wife always other (19) .(230)(149) (210) (589) Subtotal aThis table categorizes the marriage choices only of those in the population who have been Protestants or Catholics. Hence each marriage will include at least one who was a member of either of these groups. The "other" on which these choices sometimes fall may include Jew, Orthodox, a person with no religious preference, or one whose preference was simply not ascertained. It is assumed, for the 1955 survey, that the respondent's former religious preference solved before the interview have been lost from the sample. The number in parentheses after each category is a code number, for easy reference. 350. 351. 1955 1958 1959 All Years Husband formerly Protestant, now Catholic 17 5 ll 33 Wife always Catholic (21) Wife formerly Protestant, 2 l 3 now Catholic (22) Wife always Protestant (23) Wife formerly Catholic, now Protestant (24) Wife formerly Catholic, now other than Catholic or Protestant (25) Wife formerly Protestant, now other than Catholic or Protestant (26) Wife formerly other, 1 1 now Catholic (27) Wife formerly other, now Protestant (28) Wife always other (29) (19) (6) (12) (37) Subtotal Husband always Protestant 17 20 21 58 Wife always Catholic (31) Wife formerly Protestant, 3 3 now Catholic (32) 229 137 195 561 Wife always Protestant (33) Wife formerly Catholic, 18 10 9 37 now Protestant (34) Wife formerly Catholic, now other 2 2 than Catholic or Protestant (35) Wife formerly Protestant, now other than Catholic or Protestant (36) Wife formerly other, 1 1 1 3 now Catholic (37) Wife formerly other, 4 2 3 9 now Protestant (38) 3 3 2 8 Wife always other (39) (277) (I73) (231) (681) Subtotal 352. 1955 1958 1959 All Years Husband formerly,Catholic, now Protestant Wife always Catholic (41) Wife formerly Protestant, now Catholic (42) 12 12 8 32 Wife always Protestant (43) Wife formerly Catholic, 6 5 2 13 now Protestant (44) Wife formerly Catholic, now other than Catholic or Protestant (45) Wife formerly Protestant, now other than Catholic or Protestant (46) Wife formerly other, now Catholic (47) Wife formerly other, 1 1 now Protestant (48) 1 1 Wife always other (49) (19) (18) (10) (47) Subtotal Hu§band formerly Catholig, now pther than Catholic or Protestant 3 l l 5 Wife always Catholic (51) Wife formerly Protestant, l 1 now Catholic (52) 5 4 2 11 Wife always Protestant (53) Wife formerly Catholic, 1 1 now Protestant (54) Wife formerly Catholic, now other 1 1 than Catholic or Protestant (55) Wife formerly Protestant, now other than Catholic or Protestant (56) Wife formerly other, now Catholic (57) Wife formerly other, now Protestant (58) 2 1 3 Wife always other (59) (12) (6) (4) (22) Subtotal 353. 1955 1958 1959 All Years Husband formerly Protestant, now other than Catholic or Protestant 3 3 6 Wife always Catholic (61) Wife formerly Protestant, now Catholic (62) 2 2 2 6 Wife always Protestant (63) Wife formerly Catholic, now Protestant (64) Wife formerly Catholic, now other than Catholic or Protestant (65) Wife formerly Protestant, now other 2 2 than Catholic or Protestant (66) Wife formerly other, now Catholic (67) Wife formerly other, 1 1 now Protestant (68) 2 2 Wife always other (69) (9) (2) (6) (l7) Subtotal Husband formerly other than Catholic or Protestant, now Catholic 3 15 8 26 Wife always Catholic (71) Wife formerly Protestant, now Catholic (72) 1 1 Wife always Protestant (73) Wife formerly Catholic, now 2 2 Protestant (74) Wife formerly Catholic, now other than Catholic or Protestant (75) Wife formerly Protestant, now other than Catholic or Protestant (76) «J (4) (l ) (8) (29) Subtotal 354. 1.2121228me Husband formerly other, now Protestant l 1 Wife always Catholic (81) Wife formerly Protestant, now Catholic (82) 5 10 7 22 Wife always Protestant (83) Wife formerly Catholic, 1 1 now Protestant (84) Wife formerly Catholic, now other than Catholic or Protestant (85) Wife formerly Protestant, now other than Catholic or Protestant (86) (5) (ll) (8) (24) Subtotal Husband alggys other than Catholip or Protestant 3 4 7 Wife always Catholic (91) Wife formerly Protestant, l 1 now Catholic (92) 3 4 2 9 Wife always Protestant (93) Wife formerly Catholic, now Protestant (94) Wife formerly Catholic, now other 3 3 6 than Catholic or Protestant (95) Wife formerly Protestant, now other 3 3 than Catholic or Protestant (96) (12) (9) (5) (26) Subtotal 587 391 494 1472 Grand totals APPENDIX C INTERVIEW QUESTIONS USED IN THIS ANALYSIS OF THE DETROIT AREA STUDY SURVEYS OF 1955, 1958, 1959, AND 1962 In each survey the sex and race of the respondent were determined by observation. Before each interview began, all per- sons in the dwelling unit were listed as to their age, sex, marital status, and relationship to the head of the household. Other ques- tions which we planned to use for this analysis are recorded here with the numbering, style of typing, and directions for the inter- viewer which appeared in the interview schedule for each survey. In 1955 and 1962 the respondent was always the wife. Q. 2 Q. 3. Q. 76. Q. 77. Q. 79. Detroit Area Study #828, in 1955 We would like to know how close to you your relatives and your husband's relatives live. Do any of them live here in the Detroit area? (IF NEEDED) The "Detroit area" is any place in Wayne, Macomb or Oakland Counties. Which ones live in the Detroit area? Is this your first marriage? (IF NO) Did your first marriage end by death or divorce? How long have you been married (to your present husband)? CENSUS DATA: Q. 6. Q. 6a. Q. 6b. Q. 61:. Q. 6d. How long have you lived in the Detroit area? Where did you live most of your life before youcame here? Where were you born? Have YCU ever lived on a farm? (IF YES) 'Where? (STATE) 355. 6e. 7. Q. 7a. Q. 8 Q 8a. Q 12 Q. 12a. Q 14. Q. 14a. Q 15 Q. 16. Q. 16a. 356. Between what ages? What was the highest grade of school you completed? (IF ATTENDED COLLEGE) How many years of college did you complete? What is your religious preference? (IF PROTESTANT) What religious denomination is that? What was your total family income in 1954, considering all sources such as rents, profits, wages, interest and so on? How much of that was the income of the head of the family? What was your mother's religious preference while you were growing up? (IF PROTESTANT) What religious denomination is that? The forefathers of all Americans came from outside the United States originally. What is the original nation- ality of your family on your father's side? (IF WIFE BORN IN U. S.) Was your father born in the United States? IF YES) Was your father's father born in the United States? Now we would like to ask a few questions about your husband. 6. 7. 20920209201020 {3 a: 6a. 6b. 6c. 6d. 6e. 7a. How long has he lived in the Detroit area? Where did he live most of his life before he came here? Where was he born? Has he ever lived on a farm? (IF YES) Where? Between what ages? What was the highest grade of school he completed? (IF ATTENDED COLLEGE) How many years of college did he complete? What is your husband's religious preference? (IF PROTESTANT) What religious denomination is that? 13. 14. 14a. 15. 16. 16a. 17. 34. 34a. 34c. 34d. 34e. 34f. 35. 37. 57. 357. What is his occupation: (IF NEEDED) That is, what sort of work does he do? What was your husband's mother's religious preference while he was growing up? (IF PROTESTANT) What religious denomination is that? What is the original nationality of your husband's family on his father's side? (IF HUSBAND BORN IN U. S.) Was your husband's father born in the United States? (IF YES) Was your husband's grandfather on his father's side born in the United States? What was your husband's father's usual occupation while your husband was growing up? Detroit Area Study #849, in 1958 One of the first things we are interested in is the kind of work Detroiters do. What is your job at the present time? (BE SPECIFIC) Do you have a religious preference? That is, are you either Protestant, Catholic, Jewish, or something else? (IF PROTESTANT) What specific denomination is that, if any? Have you always been a (specific denomination)? (IF NO) What was your preference previously? What were your reasons for changing? When did you make this change? What is (IF DEAD: was) the religious preference of your father? (GET SPECIFIC PROTESTANT OR JEWISH DENOMI- NATION) What is (IF DEAD: was) the religious preference of your mother? (GET-SPECIFIC PROTESTANT OR JEWISH DENOMINATION) Which of the people or things on this card (SHOW CARD V) has had the greatest-influence-on your religious beliefs? 358. CARD V Friends Teachers Husband, wife or children Parents Ministers, priests, or rabbis Books TV or radio \zo‘th-‘th-I PROTESTANT RESPONDENTS ONLY: Q. 71. Of those relatives you feel really close to, what pro- portion are Protestants? (CARD VIII) 1. All of them 2. Nearly all of them 3. More than half of them 4. About half of them 5. Less than half of them 6. None of them Q. 72. As a general rule, do you think it is wiser.for Prot- estants to marry other Protestants, or not? Q. 75. If for some reason you decided to become a Catholic some day, do you think any of your friends or relatives would try to discourage you, or not? Q. 75a. (IF WOULD NOT TRY) Do you think any of your friends or relatives would feel at all unhappy or disturbed about it, or not? CATHOLIC RESPONDENTS ONLY: Q. 83. Of those relatives you feel really close to, what pro- portion are Catholics? ‘(CARD VIII) Q. 84. As a general rule, do you think it is wiser for Catho- lics to marry other Catholics, or not? Q. 87. If for some reason you decided to become a Protestant some day, do you think any of your friends or relatives would try to discourage you, or not? Q. 87a. (IF WOULD NOT TRY) Do you think any of your friends or relatives would feel at all unhappy or disturbed about it, or not? RESUME QUESTIONS FOR ALL RESPONDENT”: Q. 130. Is this your first marriage? Q. 130a. Q. 131. Q. 132. Q. 142. Q. 145. Q. 145a. Q. 146. Q. 146a. Q. 147. Q. 147a. Q. l47b. Q. 147C. Q. 152. Q. 153. Q. 154. Q. 155. 359. Did your first marriage end by death or divorce? Is this your wife's (husband's) first marriage? How long have you been married to your (present) wife (husband)? What is (was) your husband's (wife's) job? (BE SPECIFIC) Did he (she) get any of his (her) education in parochial schools or other schools run by his (her) church? (IF YES) How many years did he (she) attend church schools? What is your husband's (wife's) religious preference? (IF PROT.) What denomination is that? Did your husband (wife) ever change his (her) religious preference? (IF YES) What was he (she) previously? Why did he (she) change? When was that approximately? What is the original nationality of his (her) family on his (her) father's side? Was his (her) father born in the U. 8.? What is the original nationality of his (her) family on his (her) mother's side? Was his (her) mother born in the U. S.? CENSUS DATA: Q. 6. Q. 7. Q. 7a. Q. 8. Q. 8a. Q. 8b. What was the highest grade of school you completed? Did you get any of your education in parochial schools or other schools run by your church? (IF YES) How many years did you attend church schools? How long have you lived in the Detroit area? Where were you born? Have you ever lived on a farm? 53 P9910 8c. 8d. 23. 24. 25. 26. 27. 28. 28a. 29. 29a. 141. 141a. 142. 142a. 143. 3600 In what state or country? Between what ages? What was your father's usual occupation while you were growing up? (BE SPECIFIC) What is the original nationality of your family on your father's side? Was your father born in the United States? What is the original nationality of your family on your mother's side? Was your mother born in the United States? Were any of your grandparents born in the U. 8.? (IF YES) How many were born in the U. 3.? What was your total family income in 1957, considering all sources such as rents, profits, wages, interest, and so on? (CARD X) -. O-$999. $1000-1999. $2000-2999. $3000-3999. $4000~4999. $5000-5999. $6000-6999. $7000-7999. $8000-9999. $10,000-14,999. $15,000 or more Q‘OCDNO‘LHJ—‘le—‘O How much of your total family income was the income of the head of the family? Detroit Area Study #855, in 1959 Do you have a religious preference? This is, are you either Protestant, Catholic, Jewish, or something else? (IF PRCT.) What specific denomination is that, if any? Have you always been a (SPECIFIC PREFERENCE)? (IF NO) What other preferences have you had? What is, or was, your father's usual religious denomi- national preference? (BE SPECIFIC: E.GL, BAPTIST, ORTHODOX JEW, ETC.) Q. 144. 361. What is, or was, your mother's usual religious denomis national preference? (BE SPECIFIC). CENSUS DATA: Q. 5. Q. 6. Q. 6a. Q. 6d. Q. 6e. Q. 6f. Q. 9. Q. 9d. Q. 9e. Q. 10. Q. 14. Q. 15. Q. 16. Q. 17. Q. 18. Q. 19. Q. 20. Q. 21. Q. 23. Q. 23a. Q 25. Q. 25a. What was the highest grade of school you completed? How long have you lived in the Detroit area? (IF NOT ALL LIFE) Where were you born? Have you ever lived on a farm? (IF YES) In what state or country? Between what ages? (IF MARRIED) Is this your first marriage? (IF NO) Did your first marriage end by death or divorce? How many years have you been married ( altogether)? What is your occupation? (BE SPECIFIC). What is the original nationality of your family on your father's side? Was your father born in the United States? Was your mother born in the United States? How many of your grandparents were born in the United States? (IF MARRIED) What is the original nationality of your (spouse's) family on his (her) father's Side? Was his (her) father born in the United States? Was his (her) mother born in the United States? How many of his (her) grandparents were born in the United States? What was your father's usual occupation while you were growing up? How many years of school did he complete? What was your (spouse's) father's usual occupation while he (she) was growing up? How many years of school did your (spouse's) father complete? 362. What was your total family income in 1958, considering all sources such as rents, profits, wages, interest and so on? (CARD 17). How much of your total family income was the income of the head of the family? CENSUS DATA ABOUT OTHER ADULTS IN THE DWELLING UNIT (Spouse): Q. 10. Q 42. Q. 42a. Q. 43. Q. 43a. Q. 4 Q. 4a. Q. 5 Q. 7 Q. .7a. Q. 8. Q. 9 Q 11. Q 12. What is his (her) occupation? (BE SPECIFIC). What is his (her) religious preference? (IF PROT.) What denomination is that? (ASK IF THIS PERSON IS R'S SP USE) Has his (her) reli- gious preference ever been different from what it is now? (IF YES) What other preferences has he (she) had? Detroit Area Study #882, in 1962 (Zero Parity, i. ea, No Children How many brothers and sisters did you have altogether when you were growing up? Do you have any married brothers or sisters now living in the Detroit area? Does your mother live in the Detroit area? How many brothers and sisters did your husband have al- together when he was growing up? Does your husband have married brothers or sisters now living in the Detroit area? Does your husband's Mother live in the Detroit area? Do any members of your husband's or your immediate family live in this neighborhood? When you and your husband get together with relatives, do you see your relatives more or your husband's? Do your relatives or your husband's relatives have large family gatherings in which you participate on birthdays, anniversaries, holidays, or any other time? Would you say that you participate: 3. Once/month or more 2. Several times a year OOPO O 14. 15. 15c. 21. 81. 81a. 81b. 82. 82a. 82b. 83. 363. 1. Less often 0. Never When you think of the people you know really well, and feel close to--people you can be Yourself with and dis- cuss your problems with--are most of them friends or relatives? Now, of all your friends and relatives, would you think of the three married women neargyour own age that you feel closest to and know best? Would you just give me the first name of each? It's so we won't get them.mixed up. Let's start with the one you see most often. What is (NAME'S) religious preference--that is--is she Catholic, Protestant, or Jewish? In what month and year were you married? Were you married by a Protestant clergyman, Catholic priest, Rabbi, or somebody else? What kind of job would you like your husband to have ten years from now? What is your husband's main occupation? (BE SPECIFIC). (IF UNEMPLOYED: What is his usual occupation?) (IF STUDENT: What kind of job will he probably take when he finishes school?) Now, thinking of your mother at the time when you were growing up, how often did she talk about religious matters, or participate in religious activities with the family? Would it be quite ofteni sometimes, almost never, or never? And now, I want to ask you about your religious prefer- ences--that is, are you Protestant, Catholic, Jewish or something else? Have you always been (PROTESTANT, CATHOLIC, JEWISH)? (IF NO) When did you make this change? 19 (YEAR). What is your husband's religious preference? Has your husband always been (CATHOLIC, PROTESTANT, JEWISH)? (IF NO) When did he make this change? 19 (YEAR). In the year before you were married, how often did you usually attend religious services? Q 89. Q 90. Q. 97. 364. (IF R IS PROTESTANT) 0f the relatives you feel close to, about how many are Protestants? Would you say all, nearly all, more than half, or less than half? What specific Protestant denomination do you consider yourself to be, if any? (IF R IS CATHOLIC) 0f the relatives you feel close to, about how many are Catholic? Would you say all, nearly all, more than half, or less than half? CENSUS DATA FOR ALL RESPONDENT": Q. 1. Q la Q. 1c Q. 2 Q. 3 Q 3a. Q. 4 Q 4a Q. 5 Q. 7. Q 7a. Q 7b. Q 7c. Q. 8 Q So. Now, just a few questions about your background and we'll be through. How long have you lived in the Detroit area? (IF NOT ALL LIFE) In what state were you born? Before your marriage was the place you lived longest a large city of 50,000 or more; the suburbs of such a large city; a small city or town; or on a farm? How many years of school did you finish? (IF R IS CATHOLIC) Did you get all, some, or none of your education in Catholic schools (or colleges)? (IF SOME) For how many years did you go to Catholic schools? Were either of your parents born outside the United States? (IF NO) Were any of your four grandparents born out- side of the United States? How many years of school did you; father finish? Now, about your husband...how long has he lived in the Detroit area? (IF NOT NTIRE LIFE) In what state was he born? in what state did he live longest before your marriage? Before your marriage was the place he lived longest a large city of 50,000 or more; the suburbs of such a large city; a small city or town; or on a farm? How many grades of school did your husband finish? (IF STILL IN SCHOOL) How many years of school do you expect him to complete? 10. 13. 13a. 14. 15. O-OD‘ID 365. (IF HUSBAND IS CATHOLIC) Did he get all, some, or none of his education at Catholic schools (or colleges)? (IF SOME) For how many years did he go to Catholic schools? About what was your total family income last year-- 1961--for you and your husband-—including all sources such as wages, profits, interest, and so on? (SHOW CARD 9). CARD 9 Under $1,000. $1,000-1,999. $2,000-2,999. $3,000-3,999. $4,000-4,999. i. $8,000-8,999. $5,000-5,999. j. $9,000-9,999. $6,000-6,999. k. $10,000-14,999. $7,000-7,999. 1. $15,000 or more 3‘00th How much of the total family income was earned by your husband? (SHOW CARD 9) (RESPONSE LETTER) Were either of your husband's parents born outside the United States? (IF NO) Were any of his four grandparents born outside the United States? How many grades of school did his father have? What was the main occupation of your husband's father (STEPFATHER) While your husband was growing up? (6 to 16) (BE SPECIFIC).