‘ ll!!!WI!ll/IIIIIIIIIHIII/l/ III/IIIIIIIIIH/l(ll/11111111! 3 1293 10455 9715 {THESIS 1‘1. N «5“ RA’R Y Mmrhigan 5mm 3: bzalvefiiry i This is to certify that the dissertation entitled A Reconstituted Lesbian Family: An Ethnographic Study to Formulate a Grounded Theory of Closeting Behavior presented by Gail Iris Knapp has been accepted towards fulfillment of the requirements for _thLEh__D_ degree in JamilgLEnology Mt 114m Major professor Date 16 February 1982 MSUivnnAfn man-.1 - .- m’r ' I v‘v —‘ 042771 MSU LIBRARIES u. RETURNING MATERIALS: Wm remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. ‘me .-'-: ms ““03 % 1% A RECONSTITUTED LESBIAN FAMILY: AN ETHNOGRAPHIC STUDY TO FORMULATE A GROUNDED THEORY OF CLOSETING BEHAVIOR By Gail Iris Knapp A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family and Child Ecology 1982 ABSTRACT A RECONSTITUTED LESBIAN FAMILY: AN ETHNOGRAPHIC STUDY TO FORMULATE A GROUNDED THEORY OF CLOSETING BEHAVIOR By Gail Iris Knapp Using ethnographic methodology, this researcher studied the closeting behavior of one family that is living an alternative lifestyle. The family studied is reconstituted because it is formed by the relationship of two divorced parents and their respective children. In addition, this family is not typical of the majority culture since the parents are a lesbian couple. Homosexual families have not been extensively studied. Previous researchers have primarily used surveys to study readily identifiable gay people. Since many homosexuals present themselves to the majority culture as heterosexuals, the process of identifying samples of gay populations for study is very difficult. Further, an understanding of closeting behavior is essential for any researcher, counselor or therapist whose work deals with homosexuality. Data were collected from participant observation and ethnographic interviews following the methodology described by Spradley (1980). Observations were focused on. behavior that was related to the process of hiding or revealing the Gail Iris Knapp actual nature of the lesbian relationship. The data were categorized by verbal behavior, non-verbal behavior, and the general psychological mood. The use of artifacts and the physical environment were considered along with interpersonal communication. The grounded theory generated by this study indicates that closeting behavior is not a presence or absence condition, but exhibits a differentiation which for this family can be delimited as five Levels of behavior from uncloseted to closeted. The Levels were based on the lesbians' awareness of others' perception of their lifestyle. These Levels can be summarized as follows. 1. First Level is open and uncloseted. 2. Second Level is typified by all people maintaining a pretense that the women are heterosexual roommates although there is evidence that the outside person actually knows they are lesbians. 3. Third Level occurs when the lesbians are uncertain about the other person's knowledge. 4. Fourth Level is maintained with people who are in close contact with the lesbians, but for whom no other data exist that indicate that they are aware. 5. Fifth Level occurs when the lesbians are closeted, and attempt to present themselves as heterosexuals. To Lynn and Anoopa "With all that stuff" ACKNOWLEDGMENTS It is not possible to complete a task of this magnitude alone. In this section I would like to acknowledge the support of other members of the team that made this dissertation possible. The members of my committee have guided me throughout my doctoral studies. Dr. Eileen Earhart, Dr. Donald Grummon and Dr. Beatrice Paolucci have given me invaluable advice and direction. Without them this dissertation would have been of lower quality. Very special thanks are due to my committee chairperson, Dr. Linda Nelson. During the past few years she has been my mentor and supporter. This dissertation represents only a small part of what I learned from her. I have also been fortunate in having friends who assisted me in the preparation of this manuscript. Dr. Sue Wisenberg and Mr. James Drummond read earlier drafts and made many useful suggestions. The late Dr. Meredith Mead was a great help in the formulation of the original -proposal. Special thanks are also extended to Ms. Lynn Thigpen whose help with the writing, word processing and graphics was invaluable. This research would not have been possible without the assistance of the "Montes/Weaver" family who allowed me to invade their home. I appreciate their trust, coOperation and new found friendship. Perhaps in the future a time will come when this thesis is outdated, and people will not have to live in closets. TABLE OF CONTENTS LIST OF FIGURES.. ......... . ..................... . .......... CHAPTER I-INTRODUCTION ............ . ................... ..... CHAPTER II-REVIEW OF THE LITERATURE ..................... ... Sampling Problems in Research with Homosexuals ......... Content Areas of Research on Homosexuality ..... . ....... The Homosexual Family ........... . ...................... Research on Closeting Behavior ......................... The Reconstituted Family ............................... CHAPTER III-METHODOLOGY ..................................... Advantages of Ethnographic Methodology ................. Techniques for Obtaining a Sample ...................... Selection of Family Used in the Current Study.......... Research Procedure ..................................... Limitations of Observations ............................ CHAPTER IV-DESCRIPTION OF FAMILY ............................ Kinship Chart .......................................... Process of Formation of Family Used for this Study ..... Evidence that these People Consider Themselves a Family. CHAPTER V-DESCRIPTION OF GROUNDED THEORY .................... General Statement of Procedure and Theory .............. Definitions ............................................ Examples of Behavior Typical of Each Level ............. First Level ....................................... Second Level ...................................... Third Level ....................................... vi 1o 16. 21 26 29 3o 33 34 39 414 145 45 247 52 57 57 58 62 62 67 Fourth Level .......................... ..... ..... .. Fifth Level ....................................... Summary of the Five Levels ............................. An Example of Movement from Fifth Level to First Level. CHAPTER VI-DISCUSSION ....................................... REFERENCE LIST 75 8O 85 87 93 LIST OF FIGURES Figure 1-Kinship Chart .......................... . .......... A6 Figure 2-Floorplan of House ................................ 50 Figure 3-Summary Describing Evidence of Levels ............... 86 CHAPTER I - INTRODUCTION A stereotypical American concept of family is the nuclear which includes a mother, a father, and their children. However, family researchers and scholars recognize that there are various family forms. This research will examine one alternative family that deviates from the expected structure. First, this research is concerned with the reconstituted family. This type of family is formed after one or two previous nuclear families have dissolved, and one spouse from each remarries to form a new family. This new household contains two adults (parents) and one or two sets of children. Secondly, this research is concerned with a special case of the reconstituted family where the two adults are both lesbians. For purposes of this research, the lesbian family is defined as two women who have made a long-term commitment to each other that is analogous to heterosexual marriage. Also included are their children for whom they have primary custody. Kantor and Lehr (1975) have viewed the family as a distance regulating mechanism. The reconstituted lesbian family must discover ways to establish boundaries and allow information flow both within the family and between itself and other systems of the larger society. Some of these systems are of concern to any family, but several of them are unique to the reconstituted family (e.g. the kinship system of the former spouse), and at least one is unique to the homosexual family (the gay subculture). The lesbian reconstituted family is a significant concern to society today. One need only read a newspaper or watch television to see that the "invisible" minority is no longer quite so invisible. Popular books about gay people abound. There are hotly contested court battles over the ability of lesbian mothers and gay fathers to function properly as parents. There is a great paucity of academic literature about homosexuality, and that which exists has tended to focus on the question of mental health and on sexuality. In the literature about homosexuality, it has become almost necessary for the researcher to mention that the sample was not random since it is well known that many homosexuals present themselves as heterosexuals. This is called "being in the closet" by many gay publications. In the current project the researcher studied one reconstituted lesbian family in depth. The method of participant observation as described by Spradley (1980) was used. The researcher began without a preconceived hypothesis or theory, and allowed the data collected to generate a grounded theory. However, the researcher did focus her observations on closeting behavior. This variable, closeting behavior, has been a significant factor in the inability of researchers to locate unbiased samples of gay populations. If gay peOple do attempt to present themselves as heterosexuals for a portion or majority of their life, then this must be a significant factor in understanding their lifestyle. Indeed, it must influence most other aSpects of their life to some extent. In that case, it becomes an important variable for anyone interested in understanding the totality of the lesbian experience. A need for research on nonclinical, gay families has been expressed many times in the literature. The Task Force on Homosexuality of the National Institute of Mental Health (1969) indicated that there was a need for research on family dynamics as well as many other basic areas. Further, Morin (1977) has argued that previous research done with gay peOple has concentrated on a clinical model, and that ..research on lesbians and gay men should give higher priority to questions relevant to the wide variety of homosexual life-styles. Specifically, research is needed on the dynamics of gay relationships... (p. 636) By studying one lesbian couple with children from the previous marriages of each adult, this researcher hopes to help fill this gap. Gay mothers are often very reluctant to be open about their homosexuality since they are also faced with the threat of a custody challenge. This threat may come from the children's father, her own parents or ex-in-laws, or even the state itself could declare her an unfit mother. Therefore, many large studies of lesbians, (e.g. Tanner, 1978), have used only childless couples. The use of grounded theory as described by Glaser and Strauss (1967) provides an opportunity to explore the cultural system of a lesbian family while being relatively free of preconceived biases. This type of research is less likely to impose the values, morals or theoretical conceptions of the researcher or the majority culture. When a researcher begins a study with a predetermined hypothesis, data are collected that are designed to support or refute that hypothesis. The direction that the research will take is determined before data collection begins. In the grounded theory methodology, only the general focus of the research is decided initially. The data that are collected in the early stages of the process direct later observations. However, this research is not without limitations. First, in order to get depth of understanding, the researcher limited the sample size to one family. This necessarily results in a lack of generalizability to other families. However, using the Glaser and Strauss concept of "slices of data," this research will provide one slice. Other slices are available from questionnaire and interview studies, and these can be incorporated into a fuller understanding of the whole lesbian culture. U‘l Secondly, an attempt was made to find a relatively closeted family. A fully closeted family is, by definition, unavailable for study; therefore, this family must represent the best available approximation. It is hoped that this family is typical of such families; however, there are no data available to support this contention. Thirdly, it was not possible to gain access to all parts of the family system. Observation time was limited to seeing the family members in the household. It was not possible to gain entrance to the children's school, the mothers' jobs, or certain family events. Indeed, gaining such access would itself have been a violation of the very premise upon which this work is based. This family was selected because the lesbians wish to conceal their lesbianism under certain circumstances. Being able to do fieldwork in these settings could have jepordized their public facade, and this was not acceptable to them. In addition, the researcher attempted to make the family as comfortable as possible with the researcher's presence. However, there is no available means of discovering whether the observation process itself influenced their behavior. Finally, there were two different types of time limitations. The study was completed during a one year time period from initial contact with the family to final data collection. Actual participant observation time occurred during a three and one half month period. Information about other parts of the family life cycle were only available through interviews. In addition, the researcher was only able to observe several days each week, and for only certain hours on those days. Clearly, anything that happened during the times the researcher was not present could not be observed. This research will present a grounded theory of closeting behavior based on observations made in one reconstituted lesbian household. It is expected that it will provide some insight into the closeting processes of this family type. CHAPTER II - REVIEW OF THE LITERATURE In research using conventional verification methodology, the review of literature provides a basis for the research question and hypotheses. The develOpment of grounded theory requires that the data generate conclusions through inductive processes rather than the deductive reasoning that typifies other methodologies. However, previous research is still important as a source of the original research question, and it can also be used to identify the focus of field observations. The literature on homosexuality has primarily been concerned with sexual behavior of gay people, mainly males, or attempts to search for the cause of homosexuality. This research has been done with clinical pOpulations or, more recently, with members of gay organizations. In the past decade, numerous articles and essays have been written about aspects of gay life that reflect the total lifestyle of these people (Simon and Gagnon, 1967; Riddle, 1977; Brown, 1976; and Berzon, 1978), but actual research in these other areas is scanty (Morin, 1977). Many of these authors contended that homosexuals are more similar than dissimilar from heterosexuals in areas of their lives other than sexual preference. The following sections present a brief overview of the type of research that has been done in the past and the recommendations for further research suggested by scholars interested in this lifestyle. Some of the terminology used in the literature on homosexuality needs to be defined. The word, homosexual, is used to describe either a man or woman whose primary or exclusive sexual orientation is towards a person of the same sex. This was the definition used most commonly in the literature (e.g. Moses, 1978). The word lesbian refers to a female homosexual. Many of the essays that have' been written in the past decade used the term gay to refer to homosexual people. This literature seemed to be more concerned with the total lifestyle of the gay person (e.g. Laner, 1977; Jay and Young, 1979 and Riddle, 1977). Sampling Problems in Research with Homosexuals Because it is difficult or impossible to detect gay people without their consent, and since societal pressures cause many gay people to be unwilling to be identified, all researchers desiring to use gay subjects must grapple with the inevitable biasing of their sample. Such biased samples may result in conclusions that contradict other studies where the bias is different, or can result in confounding variables so that invalid conclusions are drawn. For example, the use of patients in psychotherapy has resulted in some researchers concluding that homosexuals all have psychological problems (Spector, 1978). Researchers are now aware of the need to have nonclinical samples in order to get a more accurate picture of gay life; however, most researchers today face the problem of finding a suitable sample. Much of the research reported in the literature of the 1970's has used either interview or survey techniques (Tanner, 1978; Jay and Young, 1979; Jensen, 1974; and Laner, 1977). Rosen (1974) used both interviews and an adjective checklist with his sample of 26 lesbians from a gay organization in San Francisco. He concluded that lesbianism represents a way of life rather than an illness. These researchers used subjects that were available. That is, most subjects were obtained from gay bars or gay organizations. Hogan, Fox, and Kirchner (1977) used a large sample of 205 lesbians who answered a questionnaire. These women were all members of homophile organizations in Illinois. These researchers discovered that an unusually high percentage reported having had no siblings. Hogan, Fox, and Kirchner described two other studies with similar nonclinical samples that also found high percentages of only children. They concluded that possibly they have propitiously happened upon the cause of lesbianism. These data are in conflict with the demographics reported by other researchers. Jay and Young (1977) found that almost all of 962 lesbians they questioned answered that they had one or more siblings. Although Rosen (197A) 10 included several questions about siblings on his questionnaire, he did not report finding anything significantly related to these answers. In the table presented listing individual descriptions of causal factors for each of the 27 lesbians, being an only child was named once. Yet Hogan, et al. (1977) referred to the Rosen data as further support for their findings. This researcher wonders whether the use of only lesbians that were members of organizations provided a bias. It is possible that heterosexual women who join analogous organizations are more likely to be only children. Or perhaps lesbians without siblings are more likely to be less closeted, and therefore more likely to join such a group. It is this researcher's Opinion that this type of biasing can result in possibly invalid generalizations. Researchers now recognize the problem of selecting gay samples from clinical settings. The selection of members of homophile organizations might be presenting a different type of bias. Content Areas of Research on Homosexuality Previous research on homosexuality has concentrated on questions about etiology and sexual behavior. Weinberg (1973) summarized the several theoretical orientations of researchers searching for the cause of homosexual behavior. He Concluded that psychoanalytic and behaviorist theories 11 have directed much of this effort and have resulted in therapies designed to change the homosexual into a heterosexual. Secondly, much research has been done on the sexual behavior of gays. If homosexuality is defined as a disorder, then research is aimed at finding the cause in order to cure it or prevent it. Bieber (1962) used a psychoanalytic model to conclude that his all clinical sample manifested pathological relationships with overprotective mothers. Ibrahim (1975) interviewed 31 gay people. His sample included 27 gay men and four lesbians from the San Francisco area. They were interviewed in person by the researcher for at least one hour each. He presented brief vignettes about several of them and concluded that 29 of his subjects did not perceive their family as normal. No heterosexual controls were presented. Ibrahim concluded that there was a correlation between an unhappy home in childhood and later homosexuality. 'The interview questions were open-ended and were not fully presented in the paper. However, one questiOn that was used asked the homosexual to explain, "...the circumstances that led to your being a lesbian?"(p. 280) This question presumes that there exist clearly identifiable causes of homosexuality, and that an adult lesbian would be able to isolate and name the cause in her case. This researcher doubts whether a heterosexual would have been able to answer a similar question, and therefore 12 wonders whether Ibrahim's conclusion was warranted from these data. One of the classic collections of essays in the field is Judd Marmor's Sexual Inversion (1965). This book presents a series of essays purporting to explain the causation of homosexuality from a wide diversity of viewpoints. Hormonal and genetic bases are considered as well as a variety of psychological theories and therapies. This body of research has led to several books and essays that claim that homosexual samples obtained from clinical populations are a biased sample of all gay peOple. Weinberg (1973) has written a book on homophobia, fear of homosexuality, which he considered a serious psychological problem for the whole society. He concluded that homephobia has directed much of the research and psychotherapy concerning gay people in the past. Spector (1978) has written an essay describing the process used by gay rights activists to convince the American Psychological Association that they were using biased samples to conclude that homosexuality was abnormal. The APA decided that homosexuality was to be considered deviant, not pathological. Researchers (Weinberg, 1973 and Morin, 1977) are now discovering that the question of causation is more complex than simple psychoanalytic explanations imply. Shavelson, Biaggio, Cross and Lehman (1980) were unable to find consistent patterns in the 13 regression analysis comparing heterosexual women and lesbians with respect to their family backgrounds. The sample consisted of 26 self-identified lesbians and 26 self-identified heterosexual women in Moscow, Idaho and Pullman, Washington. All of the heterosexuals and two-thirds of the lesbians were college students in the area. The women were given two inventories to complete: one was concerned with their relationship with their parents; the other was about sex role behavior. In addition, the women were interviewed. These researchers found that the gay women were similar to the straight women. Simon and Gagnon (1967) also concluded that researchers have concentrated on the deviant behavior of gay people while ignoring the majority of their lifestyle. These researchers concluded that lesbians behave in culturally determined feminine ways that are similar to heterosexual women. Findings of this kind provide evidence that homosexuals are not different from heterosexuals in aspects of their lives other than sexual preference. During the last two decades many voices have been raised challenging the assumption that homosexuality is inherently pathological, (Simon and Gagnon, 1968; Turnage and Logan, 1975; and Task Force on Homosexuality: Final Report, 1969). (These researchers all suggest that since changing the homosexual into a heterosexual should not be the major goal of the helping professions, there is less 19 need for research on etiology than for research about how gay peeple and the gay culture function. The second most pOpular area for research on gays has been sexual behavior. In the 1970's there were several quite extensive survey and interview studies of gay people that concentrated primarily on sexuality. Masters and Johnson (1979) did laboratory studies of sexual behavior as well as interviews to compare homosexual sexuality with heterosexual sexuality. Their study group was selected during a four year screening and interviewing process to obtain a final sample of 94 male homosexuals and 82 lesbians from the St. Louis area. As a result of extensive clinical and laboratory study, they drew conclusions about the sexual behavior and self satisfaction of gays and recommended treatments for homosexual dysfunction. Bell and Weinberg (1978) interviewed more than 1000 homosexuals recruited primarily in bars, public baths and gay meeting places in the San Franscico Bay area. This is the Kinsey Institute study, and while it dealt with general demographic data, it was primarily concerned with gay male sexuality. Another questionnaire study of sexual behavior was done by Jay and Young (1979). This lengthy questionnaire included 108 fixed choice questions, and 26 open ended questions. It was completed by 962 lesbians and 4,329 gay men. Jay and Young used slightly different questionnaires for men and women. Although the survey was primarily concerned with sexual 15 behavior, it also asked many questions about background, political beliefs and personal values. Jay and Young discovered that the gay people in their sample represented a wide diversity of backgrounds, lifestyles and sexual behaviors. This was also the conclusion of the Bell and Weinberg study. These researchers decided to title their book Homosexualities in order to emphasize that they did not believe that homosexuality represented a single lifestyle. Delph (1978) used participant observation techniques to study gay male public sex in bathrooms, parks, and bars. His observations were made by playing the role of a gay man. He loitered in those areas identified as gay male sexual meeting places in New York City. His book described the ways gay men used to identify each other and initiate casual sexual contacts. There is relatively little research on any other aspects of gay life. Few reports were found related to pairs of males or females living as family units. Most of the research has been done with males. Morin has noted that only 18 per cent of research reported in Psychological Abstracts between 1967 and 197A used lesbian subjects. The Task Force on Homosexuality of the National Institute of Mental Health (1969) included a strong statement about the need for more research on gay women as well as on different aspects of the gay experience. 16 The Homosexual Family In the 1970's some researchers studied gay people as members of families. Some of this research was concerned with the aspect of coupling. That is, these researchers were interested in discovering whether homosexuals formed relationships analogous to heterosexual marriages. Several studies have been concerned with choice of partners. Laner (1977) compared self-identified gay and nongay peeple's responses on a questionnaire about qualities desired in a life partner. She collected questionnaires from 48 primarily assumed to be heterosexual men and an primarily assumed to be heterosexual women who were students in a Marriage and Family class in an Arizona university. These were compared to 69 homosexual men and 19 lesbians who returned their surveys from a mailing to members of a campus gay organization. The samples did not overlap. Laner concluded that the priorities for a permanent partner were not different for gay or heterosexual men or women. Cotton (1975) compared gay men and lesbians' values concerning choice of partners and sexual fidelity. He interviewed 30 lesbian women in New York City and compared their expressed values to those of a matched gay male group from a previous study. Information about how the sample was identified was not provided. Cotton concluded that lesbians L__— 17 were more likely than gay men to have stable monogamous relationships with others of a similar socio-economic status. There has been some research done on sex roles in gay relationships. Gay pOpulations are extremely heterogeneous, though, and it is therefore possible to find studies which reach diametrically opposing conclusions based on the nature of the sample. For example, Jensen (1974) found that lesbian couples engage in "butch" and "femme" role playing that mimics the masculine and feminine sex roles of straight society. Jensen informally observed 102 subjects, but only interviewed 34 women. Her data for the 17 couples presented indicated stereotyped role playing that was much greater for the women living in Salt Lake City and Denver than for the few women in the sample from Los Angeles. Jensen did notice a correSpondence to the sex role behavior of heterosexuals in those cities. Conversely, Sang (1977) reported to the annual convention of the American Psychological Association that her conclusions from reviewing the literature' and making her own informal observations were that lesbian couples were relatively free of role-playing. Thus, different researchers have reached diametrically opposing conclusions with different samples. Some research has considered the lesbian family as an alternative family form. Much of this research is aimed at the helping professions. Hall (1978) used brief case 18 histories to present some of the typical problems faced by lesbian families. DiBella (1979) described a project to provide family therapy for gays. DiBella noted that therapists tended to feel uncomfortable with the prospect and to focus on the sexual aspects of the relationships rather than on the interpersonal problems. DiBella complains that one problem is "...the absence of a reference model for an 'ideal' homosexual relationship"(p. A3). Studies of lesbians have considered primarily women who are childless. This is not unexpected since the threat of loss of custody is a serious concern for the lesbian mother, (Stevens, 1978; Riddle, 1977). Therefore, she is more likely to be closeted and less likely to be available to researchers. Since lesbians are women, it is not surprising that many are mothers. Their existence is manifested in the lesbian custody cases that appear sporadically in the courts. Since many (if not most) lesbian mothers choose to hide their gay identity, there is only speculation about their actual numbers. Simon and Gagnon (1967) have reported that lesbians are more similar than different from straight women. They explain that it is usual for researchers studying a deviant group to concentrate on the deviance. However, it is important to remember that lesbians are, after all, women and have a great deal in common with the majority of women. 19 Riddle (1977) has described the arguments against allowing the lesbian mother to maintain custody of her children. The basic fear of the society is that the child will be harmed from continued contact with a lesbian parent. Specifically, the courts fear that lesbians are promiscuous and may recruit or sexually molest the child. They fear the lack of heterosexual role models and worry about the effects on the child of growing up in a home that is different from socially accepted standards. Riddle attempted to answer these charges, but found that there was a serious lack of objective research on these topics. She concluded that there was no evidence that lesbians were less able to function adequately as mothers, but there was also very little evidence that lesbians were as capable as straight women. The pain of a lost custody suit based on lesbianism was described powerfully in Gibson's 1977 account of the Mary Jo Risher case. This same trauma was described in more formal fashion by Hitchens (1979). There is additional evidence about lesbian mothers to support the Simon and Gagnon claim that they are not very dissimilar from straight women. Mucklow and Phelan (1979) used adjective checklists to compare lesbian and straight mothers. They found few differences between their responses that measure self concept and maternal attitude. Pagelow (1980) compared lesbian and single mothers on questionnaire responses, interviews, and some observations 20 in public places. Her sample included 20 self identified lesbian mothers from a lesbian organization and 23 assumed heterosexual women who were members of an organization for single mothers. Pagelow commented that it was very difficult to obtain the cooperation of the lesbian mothers. She stated, "The apparent sensitivity of lesbian mothers to scrutiny and their avoidance of self-revelation may indicate perceived Oppression"(p. 192). On other dimensions she found that the lesbians seemed to have a great deal in common with the other single mothers. Interestingly, Pagelow identified the lesbians as single parents even though she reports that 45 percent of them live with a lover. Only four percent of the heterosexual mothers lived with a lover. Since lesbians do not have the Option to marry, some of these women might define themselves as members of a reconstituted family. These lesbians would not have all of the same experiences as a single mother. The other major difference she found was fear of having her private life paraded before the public. This fear centered on loss of custody, loss of job, and difficulty in finding housing. The major societal fear for the children is that somehow they will also become homosexual. This has been refuted by Green (1978) who studied 37 children of gay or transsexual parents. Based on clinical evaluations, he concluded that all of the boys and girls appeared to be 21 heterosexually oriented at the time and had appropriate gender role identities. This sample included 21 lesbian mothers and their children. All of them were involved in some type of custody action. Thus all of these mothers were "out of the closet." Research on Closeting Behavior It can be concluded that an important aspect of contemporary gay life is the question of closeting; that is, to what extent is the person interested in being known as a homosexual by the public. In addition, it has been shown that finding gay samples for research can be problematic. The closeting behavior of gay peOple may be the the source of part of this problem. Very little research was found that centered on this topic. However, many books and articles have been written for public consumption in this area. For example, Howard Brown (1976) wrote an autobiographical account of his own public announcement when he was a chief public health officer in New York City. His book includes information about his life as a closeted homosexual, his experiences after he "came out," and many anecdotes about the experiences of other gays. Berzon (1978) wrote an essay for a collection of 22 material by lesbians that encourages lesbian mothers to be open with their children. This essay is a "how to do it" for lesbian mothers, but it is based only on the experiences and opinions of the author. In the academic literature, research on this tepic is notably missing. Many research studies only allude to closeting. It is common for the researcher to note that random samples of gays are impossible to obtain and that generalizability of any findings is always unknown. Several researchers, however, have devoted part of their effort toward a study of closeting behavior. Schafer's (1976) study consisted of questionnaire responses by 150 lesbians contacted in bars and gay organizations in West Germany. The study dealt with many aspects of sexuality and social situations including several questions about closeting behavior. Schafer found that this sample of women selected members of their families and friends to whom they wished to reveal their lesbianism. About half her sample thought that lesbians should "confess Openly to their sexual orientation." However, fewer than 1/2 the mothers and 1/3 of the fathers were reported by the daughters to know about their daughters' homosexuality. Schafer presented tables about who knows, who suspects and any problems that might have resulted from peOple knowing. Schafer concluded that it is necessary for each lesbian to solve the problems derived from this orientation in one of 23 three ways. ...first, to conform, in other words, to repress one's own homosexual needs and to live in abstinence or to practice heterosexuality devoid of all pleasure; second to mask one's self, i.e. to lead a double-life by secretly practicing homosexuality while denying this to the outside world; third, to ignore the sanctions of the environment, in other words, to practice homosexuality and openly profess to it (p. 60). Schafer found “that while her sample eXpressed the attitude that the third choice was the most desirable, they described their actual behavior in terms of the second alternative. "The most common mode of adaptation is doubtlessly: defensive secretiveness and only selective openness where the person feels that no risk is involved"(p. 61). Hall (1978) also included closeting behavior in her report. She presented short vignettes about problems facing gay people and explained the need for the social sciences to recognize the lesbian family as an existing form. Hall described the problem that the lesbian mother faces in deciding whether or not to tell her children. She concluded, "They may reveal their sexual orientation to friends and colleagues long before telling close family members"(p. 38A). Tuller (1978) obtained his sample of five lesbian couples and ten gay male couples by asking gay contacts to distribute surveys to their friends. These friends were instructed to distribute surveys to their friends. This technique was used to attempt to reach gays that were not part of the more visible community. Tuller noted that it 2A was more difficult to find gay couples than single gay people for research since couples were less likely to frequent gay bars or gatherings. Thus, couples may be considered more closeted to the general society and the researcher since they were clearly less often available for study. Chafetz et al. (1974) did an interview study with 51 gay women. These women were all in Houston, Texas and either members of a homophile organization or friends of members. The researchers' major concern was stressful situations and problems that society causes for lesbians. They found that 75 percent of their sample reported feeling pressure to "act like a straight female on some occasion-in their jobs (56 percent), with their families (25 percent), and when they attended social events with straight people"(p. 719). Some studies of closeting behavior concentrate on a dichotomy between being totally closeted versus completely public. Many agree that "coming out" is a process, and that that process begins with self definition as gay. (Cass, 1979; Lee, 1977; and Albro and Tully, 1979). Albro and Tully (1979) surveyed 91 lesbians to compare how they functioned in the homosexual and heterosexual cultures. These researchers found that in general the women felt isolated from the larger society, and found friends, emotional support, and companionship primarily in the gay community. These researchers also asked the respondents to 25 rate their level of openness on a scale from one (hidden) to seven (complete openness about sexual orientation). They found a positive correlation between level of openness for women who belonged to gay organizations (mean score=4.AA) and women who did not belong (mean score=3.52). In addition, they found a negative correlation between age and level of openness. The researchers concluded that it might be difficult to find unbiased samples of older lesbians. One research study that used closeting as a major part of the research question was done by Alice Moses for her doctoral dissertation. Moses (1978) used a questionnaire to study identity management in lesbian women. She distributed 300 questionnaires (82 were returned) using the "snowball" method. Questionnaires were given to lesbians she knew who were asked to give them to their friends. The friends distributed them to their friends, and so on. The hope was that this method would reach women who would not ordinarily be located by researchers. Moses based her hypotheses upon a quite complicated theory of deviance and passing behavior. She asserted that based on the writings of Glaser and Strauss, Scheff, and Garfinkel, (Moses, p. 8-11), one can conclude that there are several different contexts (open, closed, su5picious, and pretense) and the concept of metaidentity. "Metaidentity involves my perception of my view of your view of me"(p. 11). By combining these concepts one . can 26 describe the process of selecting management behaviors. Many of these behaviors are based on the need to pass as a nondeviant, i.e. heterosexual. Moses found that her reSpondents did not feel they were identifiable as lesbians by the straight society. Moses found that the most important variable that made being deviant a problem for her respondents was not how likely they thought it was that others would identify them, but rather how concerned the respondent was about this possibility. In addition, she was unable to find a relationship between heterosexual experiences and the frequency of use of passing behaviors or difficulty in handling interactions with heterosexuals. Although in her explanation of the theoretical basis for her hypotheses Moses recognized several awareness contexts, her actual survey questions only asked the respondents to identify the people who know or suspect that the respondent is gay. Thus she was not able to draw any conclusions about the awareness context of pretense. The Reconstituted Family The study of family now includes the study of alternative family forms. It is also recognized by professionals working in the field of family study that a family is more than the sum of the individual people that 27 comprise it. In an essay, Kent (1980) has decried the lack of research on the reconstituted family. She identified various aSpects of the family system that are unique to families created by remarriage and pointed out that these are often lacking cultural expectations. For example, she described the process of formation of the family that requires that individuals accept new roles in relation to each other. The result can be the labeling of individuals based on family and cultural myths, the process of adjusting expectations of each other and of the larger kinship systems, and boundary problems caused by family members being unwilling to discuss or handle certain categories of feelings about the former spouse or step-parent. Although many of these same processes apply to the lesbian reconstituted family, Kent did not allow for this in her definition of the reconstituted family. Her definition requires a remarriage of a male and female. Marriage is an unattainable goal for lesbians in the contemporary legal and social systems. Yet the lesbian family may face these same pressures in a magnified way. If the heterosexual reconstituted family has difficulty defining itself with the assistance of societal rituals, it appears to be more difficult for the lesbian couple who do not have this help. In addition, these people may define themselves as a single family unit and behave as such with selected outsiders, but be 28 simultaneously presenting themselves as two separate family units to the society. It has been shown in this chapter that one important aspect of gay lifestyle is behavior relating to hiding and revealing the actual nature of the family. It has also been shown that there has been little research done about this process, and that many researchers have suggested that more data need to be collected that reflect the wide variety of gay experience. Since the peOple that are most readily ' available for research are those that are identifiable (i.e. clinical patients, members of gay organizations and parents in custody cases), most research has been done with uncloseted homosexuals. This researcher, therefore, decided to study the most closeted reconstituted lesbian family that could be found. CHAPTER III - METHODOLOGY Spradley (1980) has described a method of doing ethnographic research. In the present study, Spradley's technique was followed and many of his definitions used. The techniques for data collection were participant observation, observation and ethnographic interview in order to understand the process of hiding or revealing a lesbian relationship. Ethnographic research differs from experimental or survey research design. Much social science research follows a linear pattern from research problem to hypothesis to data collection and finally confirmation or nonconfirmation of this hypothesis. In ethnographic methodology, the process is cyclical with an initial statement of the research area which is followed by fieldwork in that culture which generates new research questions. The process is continuous and is only concluded at an arbitrary point where it appears that categories and generalizations have emerged that can be formulated in a theoretical manner. This is the method of generating "grounded theory" as described by Glaser and Strauss (1967). 29 30 Advantages of Ethnographic Methodology This researcher explored the cultural system of a reconstituted lesbian family that is a family type considered deviant by the larger culture with which it pretends to conform. As indicated in the review of literature section, very little is already known about this family form. Some data have been collected, but research already published has been done primarily by structured interview or questionnaire methodology. While this methodology provides much useful data, it is always possible that the researcher is overlooking pertinent areas that have not already been identified. Since the questionnaire or interview uses predetermined questions, it is important for the researcher using this methodology to have a very clear idea about the kinds of data that are relevant to the research question before beginning to collect data. Since the ethnographer changes the course of data collection continuously, this approach is sensitive to identifying the kinds of information that are important to the members of the culture being studied. Glaser and Strauss (1967) have discussed the advantages of generating grounded theory in order to identify categories that are inherently useful since they are derived directly from the data. Thus, in producing a grounded 31 theory, the researcher begins with only a general subject matter and a group that has been identified as worthy of research. The researcher then observes in that setting, i.e. does fieldwork, and allows the data to provide direction for further observations. Therefore, the process is not linear as it is in research designed to test deductive theory; it is circular, instead. As data are collected, they provide categories and contrasts that lead to more focused observations. These observations lead to more generalizations, and the process continues. Theory generated in this manner has several advantages. Perhaps most importantly, it helps the researcher understand the culture being explored _from the perSpective of the people who live in it. Thus, the research is less likely to suffer from the distortions of the researcher's culture, which is what Spradley calls being "culture bound." Further, this technique allows the data to identify areas of primary concern. Morin (1977) has decried what he calls "heterosexual bias" in research that has been done with homosexuals. His primary complaints are not concerned with the quality of research, but rather with the areas selected for research. Morin did a statistical study of all empirical research on all aspects of homosexuality described in Psychological Abstracts. He found that the majority of research questions concerned diagnosis (16 percent), causes (30 percent) and adjustment (27 percent). These topics all 32 reflect a medical model orientation of the researcher. Morin recommended less research on these topics and more research on topics such as "the dynamics of gay relationships, the development of a positive gay identity, the positive and negative variables associated with self-disclosure to significant others...and conflict involving gay civil liberties"(Morin, p.636). Rehers (1978) wrote a response to the Morin article in which he complains that Morin is advocating a bias in the opposite direction. That is, Rehers believes that Morin's recommendation for further research makes exactly the same mistake that was made previously in the opposite direction: the mistake of letting the personal values of the investigator prejudge the interpretation of eXperimental studies on homosexual behavior before the data are actually collected.(p. 510) The generation of grounded theory helps reduce the fervor of such arguments. Since the researcher does not begin with a preset collection of questions or a preset theory to test, the direction of the final research is tied directly to the actual concerns and behavior of the peOple being studied. 33 Techniques for Obtaining a Sample As described in the second chapter, a major problem for researchers wishing to study gay lifestyles is obtaining a representative sample of the population. It is very difficult to find a random sample of gay people. Many gay peOple are so reluctant to be identified that they do not subscribe to homosexually oriented publications, join gay organizations, or have any other identifications that could make them accessible to a researcher. This is recognized as a concern by peOple studying gay lifestyles, and it has become almost inevitable that every research study include a disclaimer about the biases in the sample. Even very large surveys like the sexuality study by the Kinsey Institute (Bell and Weinberg, 1978) have been geographically limited to areas where the social climate was more tolerant of gay lifestyles. Historically, researchers have used clinical populations with all the problems of biasing that implies. Recognizing the need for nonclinical samples, researchers in the 1970's have gone to gay bars, gay organizations, subscription lists for gay publications, or private parties to find subjects. While this might result in samples that are more typical of the population in terms of not being in therapy, it is likely to result in samples 34 that are more willing to be visible to the society as homosexuals than the population as a whole would be. Another technique for attempting to study the more hidden homosexual was called "snowballing" by Moses (1978). This technique consists of gaining the confidence of several gay people and enlisting their aid in contacting their friends. It is hoped that in this way friend will reach friend, and eventually, some people will be reached whose only contact with the gay community is through this friendship network. Several questionnaires and interviews have been conducted using this methodology. Of course even this method fails to reach those gay peOple who have no connection at all with other gays. At this point it is only possible to speculate about their number. Selection of Family Used in the Current Study The present study used snowballing in order to find a single family that fit several criteria. 1. They are self-defined as a lesbian family. That is, the women consider themselves lesbians and have made a lifetime commitment to each other that is analogous to a heterosexual marriage. This research is concerned with families, and it is important that the people being studied define themselves in this way. Although there are definitions of family that require legal and/or biological 35 relationships between family members, for this research such a definition would have been inappropriate. 2. They fulfill the definition of a reconstituted family. Both of the women have had previous heterosexual marriages. This research is concerned with the lesbian family as an example of a reconstituted family. Therefore, previous heterosexual relationships that resulted in children would provide the closest approximation to the definition of a reconstituted family presented by Kent (1980). 3. Both of the women have children and presently have primary custody of those children. It would be desirable if children of both sexes are represented. It is expected that information about closeting behavior in relation to lesbian mothers will be obtained. Children also provide more potential opportunities for data to appear since questions about how much to tell the children are of concern to many lesbian mothers. If children of both sexes are present, data can be obtained about differing reactions of both mothers and children that are related to the children's sex. 4. They have shared living space, finances, and goals for a sufficient amount of time that they can be considered to be past the period of initial adjustments. 5. They are relatively covert. That is, they are typical of many lesbian families in not being active in political or social gay organizations. Because they attempt 36 to appear heterosexual to the society in general, they are not the type of people readily available for most lesbian research. In order to find a family that fit these criteria, the researcher explained to several contacts in the lesbian community that she was doing a dissertation on lesbian families and was looking for a such family, adults and children, that would be willing to participate. It was further explained that the identity of the people would be fully protected, and that a closeted family was important since they would provide a better picture of what it was really like to be lesbian. Several families were identified in this manner. These families were checked in relation to the criteria and were discarded because children were not present in the home. One of the women that contacted the researcher had been a student in her Women's Studies class at the local community college several years before. She said that since she already knew the researcher and trusted her not to divulge her identity, she and her lover were willing to participate. An appointment was made at their home in June, 1980. At that time the researcher explained in general terms what would be happening, and asked several questions relating to the criteria for the study. It was found that this family conformed to the criteria. The researcher explained that she would be writing 37 notes in a notebook while watching what was happening in the household. She further explained that the study was about how lesbian families functioned, and that she was really only looking for a specific aspect of their lifestyle. However, she said that she could not be more specific at that time since if the family knew what area was being studied it could influence their behavior. The couple asked many questions about whether they would be "good enough." They were assured that the study was descriptive and that no evaluation of quality would be made. In addition, the researcher explained that after the data were collected, they would be able to look at the results. In fact, they were told that they would be asked if the information was accurate before the final report was written. The researcher asked the couple about the use of tape recordings. They explained that they would not object to tapes as long as they were not made generally available, and as long as they could request that any tapes be erased after the study was finished. The researcher explained that the tapes were only going to be used to ensure the accuracy of written notes. In addition, the researcher explained that she would try not to interfere in their lives, but that she would be willing to participate in activities if they wished. She also asked whether they would allow photographs. This was completely unacceptable to them, and the idea was 38 immediately discarded. The initial interview, which was in the basement family room of the home, was outside the hearing of the children. This interview was interrupted several times by children knocking on the door with questions for the couple. After the interview, the researcher went upstairs and was introduced to the children. Nothing was said to the children at that time, since the women wanted to talk with them about it privately. They said that they did not want the presence of the researcher to influence the children's questions about the study. However, the couple agreed to tell the children that it would be a research study about families. Since this family appeared to fit all the criteria, and since they were satisfied with the arrangements, it was determined that they would be the entire sample for this study. In return for their cooperation, the researcher agreed to provide the parents with a copy of the finished report and to conceal the actual nature of the study from the children and all others unless the couple gave their permission, and, of course, to conceal their identities. It was not possible for the women to sign a consent form since doing so would be a violation of the agreement to conceal their identities. However, the researcher gave them a letter signed by her agreeing to these conditions. This procedure was found acceptable by the Michigan State 39 University committee that reviews all research prOposals that use human subjects. The committee determined on March 2, 1981 that the researcher had taken sufficient precautions to protect her subjects from being harmed by participating in this study. Research Procedure While the researcher was awaiting approval of the research proposal she became an occasional visitor at the family's home. She established herself as a friend who was not an unusual person to see at the home. In addition, the researcher was able to gain a general familiarity with the family while making informal observations about their lifestyle in general. Thus, when formal research began the initial discomfort associated with a stranger in the home was already past. Formal observations were made between March 15 and June 13, 1981. All observations, with two exceptions, were made in the family's home. The very first observation day was a Sunday when all of the children were with their respective fathers. On that day, the researcher took the couple to a restaurant and explained exactly what would be happening and answered their questions. The researcher explained that she was studying lesbian lifestyles, but that it was necessary not to be any more specific about the nature of the research 40 in order not to influence their behavior. The couple was willing to agree to this as long as they were assured that there would be full disclosure before the study was in final written form. The researcher again tried to alleviate their fear that they would not be a good example. The couple was concerned that they might have problems that they did not wish to have generalized to all gays. They expressed much concern about presenting a negative image to the nongay world. Once again the researcher explained that the study was completely descriptive, and that she was only interested in finding out about how they lived. She assured them that she was not going to conclude that they were bad mothers or that they were in any way unhealthy; these were expressed as major concerns by the women. The second exception occurred at the end of the study, when the researcher accompanied the whole family to a restaurant for lunch. This was suggested by the researcher who invited them as a concluding activity and an expression of appreciation for their assistance. During the three months of observations, the researcher attempted to spend as much time as possible with the family. Two or three days per week, she would arrive at the house in the afternoon after one of the women had returned from work and the children were home from school. She would remain on most days until after the children were in bed, and on several occasions remained with the women until 10 or 11 at L11 night. The children spent alternate weekends with their fathers on an irregular schedule, so observations were made all day on weekends when the children were home, and several times when the children were away. On some weekends the children of one of the women were present while those of the other were with their father. Since the purpose of ethnography is "to understand another way of life from the native point of view," (Spradley, p. 3), this research was conducted by observing an actual reconstituted lesbian family during their normal day to day activities. In a further attempt to understand the processes and perceptions of family members, the researcher joined in family activities whenever appropriate. Since the goal of this research is descriptive, no attempt was made to intervene or direct any activity. On many occasions, the researcher asked questions. Primarily, these questions were intended to clarify observations or to fill in information about behavior that could not be directly observed. Since members of the family were aware of the researcher's actual role, she was able to sit unobtrusively and write notes or make tape recordings. In later chapters, these data are identified by a reference to fieldnotes or tape with a date. The tapes were always recorded in the home. On several occasions, the fieldnotes were augmented from memory immediately after the event. 42 When peOple were present that the family did not wish to know about the study, the researcher simply put away the notebook and assumed the role of visiting friend. It was necessary to miss an extended family birthday party for one of the children since neither the researcher nor the friend role would have been appropriate. Observations were only made during a three month period in the lifecycle of this family. Clearly, there were many aspects of the process being studied that happened at other times and places. In order to obtain information about other time periods, the researcher used ethnographic interviews. In general, the researcher attempted to prompt topics by asking general questions that opened up an area for discussion. Then she would remain as passive as possible. On some occasions additional questions were asked, but they were never preplanned. The researcher allowed the previous information to lead the conversation. Questions generally were either requests for clarification or for additional information. In addition, interviews occurred spontaneously. That is, an interview time was never scheduled or planned. Rather, when a topic was presented in conversation that the researcher wished to learn more about and the time was appropriate, the interview began. On very few occasions the opportunity to explore the topic at the time was not available. In those cases the researcher waited until the first opportunity and said, 43. "Remember earlier when you were talking about...?" The interview always proceeded comfortably from there. Interviews were also used to fill in time periods when the researcher was not present. As mentioned above, the researcher was not present at the birthday party, but was able to find out about it the next day. Information about other physical settings (eg. school, work, homes of fathers) was also only available through interview data. In later chapters, information from interviews is identified with the word interview and a date. The physical realities of a home with eight people each going in his or her own direction required frequent choices about which family member to follow. Initially, an attempt was made to observe all the family members approximately equally. As the research progressed it became apparent that information about closeting behavior was most likely to be observed when at least one of the parents was present. Information about affectional patterns and friendships could be obtained by observing only the children, but at no time in the first month did the children mention or even allude to lesbianism or their mothers' relationship. Therefore, during the final two months the researcher concentrated the observations in the vicinity of the mothers. An attempt was made to observe each of the women approximately equally. uu Limitations of Observations This research is limited in several ways. First, it only includes direct observations for a three month period. Secondly, it only includes observations during the periods when the researcher was present in the family home. However, other information was obtained from interviews. Finally, the researcher focused on only one aSpect of the lifestyle of this family. Necessarily, data were collected that are not related to the topic, and probably some related data were overlooked. As Glaser and Strauss (1967) suggest, this should not be a serious problem since the data have led to the theoretical generalizations. Using their concept of the "theoretical sample" and collecting data until a particular category is "saturated" has resulted in the categories described in the next chapter. This is not an exhaustive study of many lesbian families of different types. Thus, it is possible that these data will not generalize directly to other families without modification. However, these data do exist. Therefore, this theory is an accurate description of a real process. CHAPTER IV - DESCRIPTION OF FAMILY Kinship Chart This chapter will describe the family that was observed for this study. It will begin with a diagram of the family and their relationships and then will describe the process of formation of the reconstituted family. Finally, it is necessary to present information about the women and children's self perceptions about their relationship. One of the criteria for selection of a family was self definition as a family. The final section of this chapter presents data that support this conclusion for this household. A kinship chart showing the family relationships is shown in Figure 1. All names have been changed to protect the identities of the peOple. In this chart the symbol ":" is used to indicate a marriage. The ":" is used to represent the relationship between Cindy and Virginia since these women consider themselves married. Although they were not able to obtain a legal marriage (same sex marriages are not recognized in Michigan at this time), they decided to have a religious ceremony of "holy union." In 1978, they arranged with a minister from a predominantly gay church to have this ceremony. Cindy and Virginia have photographs of the event which the researcher saw. It was attended by one A5 116 p NQDGI ht <10 522$! Kmtbwech 022:4: OFF—‘26): qu< 2 thzo; m2... 6an 2‘ 2Com mdt‘QW»: tOkthmtm EOE—xm) Wh= ”KNQEDZ. 3358 5mm 9. <= 32;. ....<© 8km: Cum: Snag. Emu—...“: «~50: Bum—K» em» 30;: . >35 . mam 2mm . <24 33: «New: «'3: 2.6: :3: 3mm: 025 550:; wmt‘oi kmmm Eamon «Now: 2mm: 0 (Q2: M>Q<40 no < K COUCH FAMILY ROOM FAMILY ROOM/PLAY AREA H FIREPLACE couc BASEMENT LISAa. SUE ; w KITCHEN BEDROOM BAT" g; 9!“ r \ . 'SALLY a. TRICIA1\/ BEDROOM TED &BEN LIVING ROOM BEDROOM \ am am MAIN FLOOR FLOORPLAN OF HOUSE FIGURE 2 51 encountered all the difficulties and problems faced by other reconstituted families. Virginia worried about whether her two children who were accustomed to having their own rooms would accept four "invaders." They both worried about how the children would handle the authority of a different parent. Indeed, most of these problems did emerge during the first year. Cindy described how angry and frustrated she would become when Lisa would refuse to listen to her and say, "I don't have to listen to you. You're not my mother!" (Interview, May 12) She said that she would then telephone Virginia at work to explain what was happening. Virginia would then talk to Lisa on the phone and tell her to listen to Cindy. At the time of the observations this family had lived together for three years. Often during the times when the observer was present either of the women would say, "You know it wasn't always easy like this." Or, "This all seems to be going so well (putting the children to bed). But you should have seen the fuss two years ago!" At the time Of the observations they defined themselves as a family. When Cindy talked about the vacations plans "we" have, it is assumed by everyone that "we" includes all eight people. When Cindy said "the" kids can have a treat tonight, everyone present realized that that includes all six children. However, if Cindy said that "my" kids are 52 going with their Dad this weekend, everyone present knew that that only included those four. Evidence that these PeOple Consider Themselves a Family During the course of the observations, data were revealed concerning the family's perception and definition of itself. Of course, the mothers originally (June, 1980, interview) indicated that they consider themselves a family. This researcher was also looking for objective measures of family system behavior while collecting the data. Since the major part of the these data are in favor of this self-perception, any exceptions are useful in understanding the grounded theory of levels of closeting described in the next chapter. The following list represents saturated observations. That is, these were not single instances, and no observation is included in this list that does not appear at least five times in the field notes. Most of these observations were so consistent that the researcher ceased making special note of them after the first month or two since they were repeating consistently. Since one of the criteria for selection of a family was self definition as a family, the researcher thought it would not be sufficient simply to assert that this was the case. Indeed, it was possible that the women would call themselves 53 a family but not manifest behavior that is usually considered fundamental to a definition of family. Included in this area are the sharing of resources, sharing of long term goals and emotional commitments to each other. The following list presents Observations that support the researcher's contention that this family Operates as a system that fulfills the definition of a reconstituted family presented by Kent (1980) with the exception of heterosexual legal marriage. 1. The children share household space and resources equally. Rooms are assigned based on age and sex (see Figure 2) not considering biological family of origin. The girls selected their own roommates. Privileges are granted by whichever parent is present. Similarly, punishments are handed out by whichever parent is present. Criteria for punishments and privileges appear to be standard. For example, all four of the girls enjoy having a friend spend the night with them. This privilege is rotated based on the presence of an empty bed because one or more of the children is spending the night with his or her father. All of the children have been given ballet lessons. Tricia decided that she did not like ballet and wanted to be active in Girl Scouts instead. The other five children all go to ballet once a week, and the whole family attends the recital every year. Cindy and Virginia had hOped to buy a backyard pool 54 this summer. But the cost of braces for Sue, Sally and Lisa's teeth have made that impossible. 2. Cindy and Virginia pool all their resources. Virginia has a clerical job in a local social service agency, and Cindy drives a school bus for a suburban school system. This means that they are not supporting six children in Opulence. Cindy is the financial planner, and Virginia has relegated all tasks involving allocating of money to her, including money for Lisa and Ben. 3. Child care is shared based on availability. Virginia's job requires that she be at work from 10 a.m. to 6 p.m. or from 12 noon until 8 p.m. depending on the day of the week. Therefore, Virginia is responsible for getting the children off to school in the morning. Cindy's job begins at 6 a.m. and finishes at 5 p.m. (with a long break in the middle of the day). A babysitter was hired to meet the children at home at 3:30 and stay with them until Cindy returns from work at 5. During school vacations Cindy does not work so she stays home with the children. 4. Affectional patterns among the children are based on sex and age, not on family of origin. When the children play together Ted and Ben are inseparable. It is very unusual to see one of them without the other. Often they are playing with other children in the family or with children from the neighborhood, but they are always together. When this finding was shared with Cindy and 55 Virginia during the last week Of the study, they laughed. After thinking about it and discussing it together, they agreed that that was probably the case at this time, but they quickly asserted that earlier the two boys fought constantly. Indeed, the verbal fighting was Observed on a small scale on several occasions by the researcher, but Cindy and Virginia said that a year ago it was constant and often physical. The researcher also Observed the relationship between the girls. On one occasion she talked with Virginia about this. Virginia complained, "Lisa is the leader. She always gets things done for her by Sue. I wish Sue would learn that she isn't Lisa's private servant." The researcher had already observed the girls playing a game outside. Lisa is always the decision maker about which game to play. They always vote. Lisa leads the vote. And what Lisa wants seems to be the foregone conclusion. Lisa is clearly the oldest of the girls, and although she is only six months older than the twins, she is more mature (post puberty) and is already in the Jr. High School while the other children are still in elementary school.(Field notes, April 6) Thus, it can be concluded that these eight people share most of the things that are usually identified as necessary for a group of peOple to be defined as a reconstituted family. Indeed, it would be difficult for an outsider to identify which children are biologically related, and the 56 only Obvious sign of biological motherhood is that the children always call their own mother "Mom" and the other parent by her first name. Cindy and Virginia have demonstrated that they share a mutual commitment that is analogous to heterosexual marriage. CHAPTER V - DESCRIPTION OF GROUNDED THEORY General Statement of Procedure and Theory During the three month period of observations, the researcher recorded notes on tape and in notebooks. The procedure for analyzing and categorizing notes described by Spradley (1980) was followed by this researcher. First, the notes were analyzed for examples of general categories. That is, the researcher was looking for data that consistently reappeared. As these general categories were identified, it was possible to use them to narrow the observations to more focused areas. For example, it rapidly became clear that. certain topics were discussed only when the children were not present. This led the researcher to focus her observations on which tOpics were acceptable or unacceptable with each of the people who were present. This led to a further refinement. The researcher observed that different words were selected for similar events or artifacts. The researcher then attempted to categorize the peOple according to the words and topics. Simultaneously, other dimensions of behavior were being similarly recorded and analyzed. This process resulted in the formulation of the grounded theory presented below. 57 58 Definitions The grounded theory generated by this study contains a five level delimitation of closeting behavior. First the five levels will be defined, and then the data supporting this conclusion will be presented. The theory was Obtained directly from actual Observations in this family's home. A general summary of all the levels identified will be presented first. The levels are based on the lesbians' awareness of others' perception of their lifestyle. 1. First Level represents fully open and uncloseted relationships. People at First Level are aware of the lesbian relationship, and there is open communication about it. The couple does not attempt to censor information that would indicate that they are lesbians, and the couple is willing to behave in an Openly affectionate manner in the presence of these people. 2. The Second Level involves relationships in which lesbianism is never mentioned nor discussed openly because the couple feels certain that others would be uncomfortable if lesbianism was ever named. However, other people say or do things that make it clear that they know about the lifestyle. To identify that a person is at this level, there must be some concrete example of their knowledge. This person then engages in his or her own closeting behavior. That is, 59 the person selects Second Level language and nonverbal behavior. Thus it becomes a tacit agreement between the lesbian couple and the other that together they will pretend that the other does not know what he/she clearly does know. 3. The Third Level involves people whom the gay person believes are probably aware of their lifestyle, but they are not certain. PeOple at this Level have done or said things that can be interpreted as understanding the actual nature of the relationship. However, these are always vague or general enough so that the conclusion that the other person really knows is not certain. If the lesbians decide to keep the relationship at this Level, they are careful not to say or do anything that would confirm the other's awareness. 4. The Fourth Level includes relationships with others whom the gay peOple suspect might be aware, but they have no concrete evidence to that effect. This Level is Often typified by the gay peOple thinking, "Given what knows about me, how could possibly not know?" Fourth Level typically involves peOple with whom the couple would prefer a Fifth Level relationship, but with whom they are forced by circumstances to maintain a close contact. They su5pect that the information those people possess could be sufficient for them to have figured out the truth. 5. At the Fifth Level the gay peOple are confident 60 that the others are not aware of their lifestyle. The lesbians attempt to present themselves as heterosexuals to these people. In order to maintain this Level, the lesbians remain as distant and impersonal as possible in these relationships. The following sections provide specific examples of how the data were used to generate the grounded theory. The observations were categorized according to verbal behavior, nonverbal behavior, and mood. The researcher also considered artifacts and their use. In particular, verbal and nonverbal behavior that centered on certain objects was found to be different with peOple at different Levels. Verbal behavior refers to actual choice of words and expressions. The researcher observed that the couple selected different words according to whether the other person was aware of their lesbianism. In addition, other people also selected language that was appropriate for the people at the level of least awareness of those present. That is, if a neighbor stopped in to visit, Cindy would immediately stop calling Virginia, "Honey," as she had been doing just before the arrival of the neighbor. Nonverbal behavior refers to all actions that were not spoken. Primarily this means touching, kissing, and movement in the physical environment. The researcher observed that Cindy and Virginia would use different quantity and quality of touching that was related to the 61 level of awareness of peOple present. For example, there was a complete avoidance Of even casual touching in the presence of Fifth Level people; a brief hug or kiss on the cheek if only Second Level peOple were there; and warm, affectionate hugs combined with kissing on the lips when First Level people were present. Mood refers to the general atmosphere in the room. It is difficult precisely to define mood since it represents the sum of many different variables. The major criterion used to identify mood was evidence of psychological stress. When peOple were present from more closeted levels, the women exhibited behavior that is commonly associated with nervousness. They would sometimes pause before using a word or stop to decide where to sit or change the subject of conversation when it was getting personal. On several occasions, the researcher observed a mistake, i.e., a word or topic was broached or a touch made that was inappropriate for one or more peOple present. The inapprOpriateness was made evident as it was frequently followed by an angry gesture or expression between the women. This was always done so that the outsider was not aware of it. It was either a quick look when the outsider was looking in another direction, or it was a sharp word in another room at the first Opportunity. Mood was identified primarily by the obvious changes in atmosphere when people from more closeted levels came or 62 left. When only less closeted people were present, the atmosphere was more relaxed. Conversation did not appear to be censored. The couple did not exhibit stressful behavior, but rather appeared more comfortable. Artifacts refer to physical objects. Cindy and Virginia sleep together on a double bed. They wear wedding rings. But these objects are identified differently to people at different levels. Fifth Level peOple are not shown the bedroom and are simply told that the rings are gifts. The album of pictures of their wedding is only shown to people at the First Level. It is not the Objects themselves that are important, but rather it is the verbal and nonverbal behavior associated with them. Examples of Behavior Typical of Each Level First Level There are certain words and expressions that are exclusively used when the only peOple present are people with whom they have Open communication. Cindy and Virginia both agreed that this is also the behavior that feels most "natural" and "being ourselves." They said that it is most typical of the way they behave when alone. Of course, the presence of the reSearcher to make observations meant that there is no direct observation possible of them alone. Cindy then laughed and said, "There are some things we do 63 when alone that we don't do when you are here!" (Notes, June 5) Indeed, the above example is typical of verbal behavior at this level. Verbal behavior includes topics of conversation, labeling and describing artifacts, and the the use of certain words or phrases. One category Of verbal behavior is selection Of conversational tOpics. One example of this is sexual inferences about Cindy and Virginia's relationship (see above). These are usually done in a joking manner. For example, one evening Linda, (Virginia's sister-see Figure 1), commented that Cindy looked tired. Cindy immediately quipped, "Well I did go to bed for a couple of hours this afternoon. But I didn't sleep!" This resulted in general laughter. (Field notes, May 31, Cindy, Virginia, Linda and researcher present) Another category that is only acceptable at First Level is the wedding and related artifacts. In a conversation with Linda (Field notes, May 31), there was much conversation about the wedding and honeymoon afterwards. Linda was not invited to the wedding, though, since she was at Second Level at that time. (See excerpt of conversation with Linda on page 91.) On another day Cindy and Virginia gave an extensive description of the real significance of their wedding rings. Cindy and Virginia both showed a lesbian friend the rings 64 and explained that they had had a wedding ceremony three years earlier. (Field notes after lesbian friend's visit, March 27) Another category of speech that is only acceptable at First Level is references to lesbian couples as partners, e.g. "Susan and Jill are another couple we know." (Field notes, March 27) Very personal matters that imply a marriage-like relationship between them are also only discussed at First Level. For example, Cindy told the researcher about being unhappy when Virginia would not "fight back" during an argument but would become very quiet and withdrawn.(Taped conversation, April 16, Virginia and Cindy both present. Conversation in the basement family room after the children were in bed.) Certain words and phrases are only used at First Level. For example, gay jargon is spoken occasionally. These words are used in a positive sense by these women. ‘Their manner appears to connote a certain pride. EXpressions such as these are sometimes heard: "Those dykes are here again!" "Do you think ---- could be a dyke?" and "I don't think we have too much in common with the queers who always hang out at the bar." There are several phrases that the couple use repeatedly with each other when they are mildly annoyed. These too are reserved for First Level people. A typical exchange of this sort would be the following. Cindy, 65 Virginia, the researcher, and a lesbian friend were playing cards one evening. Cindy and Virginia were partners. Cindy was annoyed at a mistake Virginia made and the following conversation occurred: Cindy: "on, HONEY!" Virginia: "Don't get too mad. I can't afford another divorce." Cindy: "That's too bad you're stuck. You're married now!" Virginia: "I'm sorry." Whined. Cindy: "I love you." Barely whispered. (Notes, May 23) There are several terms of endearment that are constantly heard when only people at First Level are present, but not at other times. Both Cindy and Virginia call each other "honey," Cindy frequently calls Virginia, "Babe," and they sometimes will refer to each other as "my lover," "my lady" and "my partner." With respect to non-verbal behavior when others at the First Level are present, there is much touching, occasional kissing, and, rarely, very intimate hugs. The following is a typical example of First Level nonverbal behavior. Virginia returns from work. Cindy is sitting in family room on couch. They smile and say a warm hello. Children are watching television on couch and on floor. Virginia goes into bedroom area and then picks some things up from couch and carries them into kitchen. Cindy calls after her, "What are you doing?" Virginia, "Cleaning up some of this stuff." "Come on down here. You don't have to do that now." Cindy follows Virginia into kitchen. The children are all downstairs Cindy 66 gives Virginia a warm welcome home kiss, and they embrace for a moment. During that moment there has been an abrupt change in their manner towards each other. It vanishes almost as quickly as it came, and they continue to argue about whether Virginia is going to sit and have a cup of coffee or continue to clean the house.(Notes, March 10) Other examples of nonverbal behavior at this level happened Often. Cindy and Virginia would Often sit together on the couch with their arms around each other. Or they would hold hands. On two occasions there was a certain amount of horseplay between them that was quite physical. On many occasions they would steal a quick kiss. Cindy (Conversation, June 7) told many anecdotes about stealing a quick kiss when they felt confident that they would not be "caught." This even happened when peOple who were not at First Level were in the vicinity but not watching. Another type of nonverbal behavior is the manner in which the couple arranges their physical environment. Cindy and Virginia share a bedroom with a double bed. (See Figure 2.) It is located in the basement adjoining the family room, and there are no doors leading to it. It is in an alcove that is visible only from one end of the family room. It is visible, nonetheless, to anyone spending a significant amount of time in the family room. The family room is the largest and most comfortable room in the house, and it is not uncommon for Cindy and Virginia to entertain close friends there. Cindy and Virginia have only one checking account, and it has both names on it. Virginia explained to the 67 researcher that that was easier for them, and that she hoped it was not too revealing. She explained that she says nothing about it unless pressed, and she hOpes that others simply assume that this is more convenient for paying joint bills as roommates. Another indicator of the First Level that was Observed is much more difficult to identify. The mood is here defined as the emotional messages conveyed by the peOple. When the peOple present were all at First Level, the mood was very relaxed and casual. Cindy and Virginia were much more likely to laugh or joke. Their manner was less restrained. This was particularly evident when it is contrasted with the stressful atmosphere described below for Fourth and Fifth Levels. Second Level The people that Cindy and Virginia perceive to be at Second Level are those that have given irrefutable evidence that they are aware of the actual nature Of the lesbian relationship, but who have given the couple the impression that they would rather pretend that they do not know. People at this level are aware of the actual nature of the relationship between Cindy and Virginia but do not discuss it. The certainty of the knowledge of the other person comes from someone having mentioned it at some point. That is, either Cindy and/or Virginia have actually told them, or 68 the others have said or done something that clearly indicates that they know. Yet, there seems to be an unstated rule that nothing be said or done that names the relationship. This is clearly manifested by the avoidance of conversational tOpics that might include any indication that this is a gay relationship. For example, Cindy explained that she knows Frank is at the Second Level. (Interview, June 7). She said that she had had lesbian feelings before she married Frank. They were engaged, and she thought that it would be unfair to him not to tell him the truth before they married. She said, "I told him I liked women. He said that the engagement was off. He said that he couldn't take a chance on coming home and finding his wife in bed with another woman." But after six months he decided to marry her anyway. Cindy said, "He has never mentioned it again since that day. He never says a word to me. But he and his wife always invite Virginia and her kids if we go over there for a party or something." Another aspect of language is the selection of words. At the Second Level the couple change many of the words that definitely identify the nature of the relationship. They avoid gay jargon, and are disturbed by the usage of these same words by nongay peOple. The following is an excerpt from a conversation with Virginia and her sister, Linda. This conversation took place late at night, May 31, after a 69 long and casual conversation. The researcher asked whether other family members were aware of the actual nature of Cindy and Virginia's relationship. This resulted in a two hour discussion during which the researcher said very little. This excerpt includes the word "queer" which was used Often by the couple themselves in positive way. It was clear from the voice tone of Linda and from Virginia's facial grimace, that this was considered insulting here. Virginia believes that her mother and father both know. Linda concludes that her dad identified Cindy very quickly based primarily on stereotypes of lesbians. He is outspokenly anti-gay (Virginia says this is typical of many other similar prejudices), and took an immediate dislike to Cindy at first sight. Linda said that soon after Cindy had left his home the first time, he made many negative comments about the "queer." However, now he has come to know her better, and they are ,all agreed that they are really quite friendly. Linda is convinced that he really likes Cindy, and that he really thinks that Virginia is happy. Linda said, "Everyone in the family can see that Virginia is doing better and is happier now than she ever was with Bert. Dad may not like this (gestures) but he‘ likes Cindy, and he sees that Virginia is happy." The result is that Cindy is accepted as part of the family, but everyone is agreed never to name the type of relationship. In fact, Virginia's sister, Gladys, once broke this rule when not including Cindy and her children in the Christmas gift name pulling. She said that gift giving was "just for family." Virginia accepted this the first year, but thereafter refused to participate in any family function that did not include Cindy. Suddenly, Cindy became included in everything, apparently with varying levels of acceptance by different family members. There is much nonverbal behavior that is typical of Second Level. While Cindy and Virginia may touch each 70 other, it is with less obvious affection than was seen at First Level. That is, hugs are brief and need a reason. They might kiss, but it needs a reason and is invariably on the cheek. For example, one day (March 28) after eating dinner, Cindy suddenly jumped up from the table where the six children were still finishing eating and very enthusiastically said, "Thank you Virginia for cooking such a fantastic dinner!" She then kissed her briefly on the cheek. The following lengthy excerpt shows all of the processes being described. It is not possible that Sue does not know the actual nature of the relationship her mother has with Virginia. If she did not know, then she would not be embarrassed to tell the truth. When she lies, she is actually engaging in closeting behavior. That is, she is attempting to keep her friends at Fifth Level. Cindy told the researcher an anecdote about her 11 year old twin daughter Sue coming home from school and talking to Virginia about a conversation she had had with one of her friends about the sleeping arrangements at their home. Since she is telling the story second hand, she asks the researcher to ask Virginia to tell it later. After Virginia is reminded of the incident, she recalls that Sue told her about one of her friends who had asked in some surprise, "You and Lisa live together?" "Yeah, we live together." "All you kids live together?" "Yeah" (Virginia then interrupts herself to point out that this is a bizzare conversation since they have all lived together for four years, and it is inconceivable that this friend didn't already know that.) "Well, where do you sleep?" "Lisa and I share a room." Where do the others sleep? "Well, 71 Sally and Tricia share a room." "Where do Ben and Ted sleep?" "They sleep in their room." "Where does your mom sleep and Lisa's mom?" "Well, Lisa's mom sleeps in the bedroom and my mom sleeps on the couch." Virginia- "I said, 'why did you say that? You know that we sleep together. You didn't have to say that." and Sue responded with "I was embarrassed." (Note: This conversation took place between Virginia and Sue. Virginia thinks that was because she was the one who happened to be present at the time.) After making that remark, Sue immediately left the room. During the next few weeks, Virginia and Cindy both remember many examples Of the children asking questions about homosexuality. Virginia said that she simply tried to answer the questions openly and honestly and not present it in an unfavorable light. But she said that she had decided not to tell them anything about themselves Openly until they actually asked. Cindy immediately agreed. "I don't want to tell them anything that is too much for them to handle. But if they ask they are ready to hear the answer." (June 1, interview) The mood when Second Level people are present is less relaxed than with First Level. However, there does not seem to be the stress related with more closeted levels. When at the end of the observations, the researcher showed the couple these findings, Cindy immediately commented that it just didn't seem too serious to her if she "slipped and called Virginia, 'Honey' in front of Frank. After all, he would just ignore it." Third Level At this level there appears to be much talk between Cindy and Virginia about whether or not someone "knows." There is evidence that someone might know, but it isn't conclusive. In fact, it is usually open to more than one 72 interpretation. If Cindy and Virginia wished to be less closeted with the person, they could begin using words or body language that was more Open. However, the people at- this level are primarily those with whom they have regular contact; for a variety of reasons, they do not wish to be more Open with them. Language at Third Level is appropriate for two divorced women who are roommates. If they are asked questions that seem to move into more Open levels, they feel free to answer by saying that the question is too personal, (interview with Cindy, June 5). Often language used is vague. This is why it might not be clear whether or not the person really understands. Virginia is not certain whether her mother really knows- Virginia showed the researcher a letter from her mother in which she says that she is very pleased to see how well Virginia is doing, and how happy she seems. Virginia is not certain whether this is a disguised way of saying, "it's OK," or if it is really just as innocent as it appears on the surface. Cindy described an incident where she had given her brother a check drawn on the joint account when he acted surprised, Cindy immediately explained that it didn't make sense to pay so much money for two accounts when they only needed one since they pooled their money. Her brother didn't say anything else. (Interview, June 7) 73 Another person who is at the Third Level is Barb Montes. Virginia's ex-husband, Bert, is a twin. His twin brother, Bob, is divorced from Barb who maintains reasonably friendly contact with Virginia and visits Virginia's home Often. Linda and Virginia both are quite certain that most, if not all, members of the family that they associate with regularly are aware of the lesbian relationship. But the relationship is usually at Third Level. Barb. Montes once made a passing remark to Linda, "They could tell us what it is like to be queer." But Linda is not certain that this was based on actual knowledge--she suspects that it was simply a sarcastic remark since Cindy was wearing bib overalls. Virginia believes that Barb might be harboring lesbian feelings, and thus might be cultivating their friendship in an attempt to try on a new self definition. Virginia is not pleased about this, and is not interested in serving in that role. Both she and Cindy are agreed that Barb Montes should not be told openly. They are afraid that to do so would open up all sorts of potential problems for her, and perhaps for themselves, since they do not feel she can be trusted not to talk about it among other family members. (Interview with Linda and Virginia, May 31). The nonverbal behavior associated with the Third Level is aimed at maintaining an image of themselves as a family without implying that there is anything romantic about the 74 relationship. PeOple at the Third Level are expected to understand that these eight people share their lives, but they are not expected to be certain that this is a lesbian relationship. Cindy and Virginia recognize that these people may guess that they are gay, but they have decided that they do not wish them to be certain of this. The preceding example demonstrated that they expect the other person to maintain the Third Level when they have decided that the question should be left unanswered. The following example demonstrates how they maintain Barb at the Third Level. There was only one occasion when Barb was present when the researcher was there observing. (Field notes, March 28). On that day, Barb and Gladys (Virginia's sister) were already there when the researcher arrived in the early afternoon. Cindy immediately introduced the researcher as a "friend," and said that we had planned to play cards. Barb was being very rowdy and loud. Cindy was obviously unhappy about it and would mouth comments to Virginia behind Barb's back. When Barb left soon after this, Cindy immediately turned to Virginia and said, "I think you have to do something about her. She's your ex." Cindy then eXplained to the researcher and Virginia that she found Barb's behavior boorish, and she didn't want her there, but that she felt she couldn't do anything about it since she was Virginia's guest. Virginia immediately denied that she 75 wanted her there either, and she reluctantly agreed that she would have to be the one to discourage her visits. Cindy did not think that she had an appropriate role at the Third Level to allow her to take the initiative in this case. When Barb was present the mood in the room was tense. As soon as she left, both Cindy and Virginia noticeably relaxed. Both of these women agree that they have to monitor themselves when a person at Third Level is present, but this researcher concluded that they are attempting to disguise behavior or language that could be interpreted as lesbian. That is, they recognize that the. other person already possesses enough information to be guessing about the nature of their relationship, and therefore, the pressure they experience is the need to hide the proof. Fourth Level At this level Cindy and Virginia make an effort to be seen as roommates, even though they realize that Fourth Level people could easily have concluded that they are gay from external appearances. Some evidence of Fourth Level was presented earlier as part of the description of Sue's experience with her friend. While this example shows that Sue is at Second Level, it also shows that she is maintaining Fourth Level closeting with her friend. Language behavior that is used to maintain Fourth Level includes the use of words that would be expected from 76 divorced roommates. That is, the honeymoon is now a "trip"; their relationship is now called "living together"; and the other woman becomes "the woman I live with." There are no terms of endearment, and they only refer to each other by name. They use their relationship to their biological children as identifiers. For example, when Cindy went to school to pick up Ben's report card, she identified herself as "a good friend of Ben's mother" who had come because Virginia was working at that hour. Cindy believes that the teacher may be more aware since three of her children had been in her class during previous years, and also because she didn't question giving her the card.(Interview, June 7) Ted and Ben are listed as being at Fourth Level even though they often see and hear behavior that is Second and Third Levels. Cindy and Virginia several times discussed whether or not the boys really know, but their conclusion was that they are still too young to understand. (Notes, April 4) On June 1, after all the children had returned from spending a weekend with their respective fathers, Ben was noticeably upset. He explained that his grandmother was sick and in the hospital. He then announced that he was going to make a get well card to send to her. Ted said that he was going to send her one too. The researcher peeked over the boys' shoulders as they sat working diligently on the kitchen table with paper and crayons. She immediately 77 noticed that Ted was doing an exact duplicate of Ben's card. The researcher and Cindy went into the living room. In few moments Ted arrived with a worried look on his face. Cindy, "What's the matter, son?" Ted, "Can I send Ben's grandmother a card? She doesn't even know me." Cindy, "That's OK. You can send one if you want to." Ted, "Well, I'll just sign it Ben's best friend. OK?" Both Cindy and Virginia often discuss whether the children really understand. In particular, they think that the younger children are aware that they are a family, but they do not believe that they are old enough to really understand what it means to be a lesbian. This belief is qualified with a question since both of them admit that they really do not know how much the children understand. However, they are very reluctant to ask since that might create problems. Thus, they believe that the younger children are at Fourth Level, but admit that it might be Second Level since the children could know that this is a forbidden tOpic. Another example of a person at this level is the babysitter that stays with the children for about two hours each schoolday afternoon. Cindy is the person who has had almost all the contact with her since Virginia is working at that hour. The sitter was originally told by Cindy that she and her roommate needed a sitter for six children. Although this woman has never said anything that could be directly 78 identified as an indication that she is really aware, .Cindy suspects that it is possible. There are really only three reasons for Cindy's suspicions. One reason is that she has free access to the house every day. She has certainly seen the double bed. However, since that is the only free bedroom available, she might not think it is odd that they share a room. In addition, a double bed uses less space than twin beds, and the room is relatively small. Secondly, she is with the children every day, and the mothers do not know what the children might have said. Finally, one day the sitter made a comment about the work schedules of Cindy and Virginia, "It's too bad. That doesn't leave you two much time alone together." Both Cindy and Virginia are wondering whether that represents real knowledge or just a casual comment. They are not yet certain how open they wish to be or can afford to be in front of her. (Fieldnotes, April 13) The following example illustrates the uncertainty associated with the Fourth Level. Virginia wonders whether her boss at work knows. "She isn't a stupid woman. She knows I live with another woman--to share expenses and child care-- but I refuse to pretend and make up dates or anything. SO she also knows I don't see any men. She knows I was married and have children, but she reads the papers, and should know that that doesn't mean anything." (Fieldnotes, May 31) 79 This level is typified by a lack of nonverbal intimacy. Cindy and Virginia never touch at all when the children's friends are present. They will also immediately stop certain types of physical interaction when the children are present. One beautiful Sunday afternoon Cindy had been in a particularly good mood. The children were playing outside but were constantly coming in and out. Cindy started some silly horseplay with Virginia on the couch in the livingroom. Virginia responded, and soon they were both laughing and mock fighting, but whenever a child entered the room, the horseplay would stOp very casually. Although the researcher sitting in the same room was not always aware of the entrance of a child, it was apparent that both Cindy and Virginia were constantly alert to indications that a child was coming. Whenever a child entered the room he/she only saw the two women sitting next to each other on the couch apparently having a conversation. But as soon as the child had left, the horseplay would begin again. This process repeated itself four times within less than an hour.(Fieldnotes, March 29) Cindy and Virginia will usually go to the basement family room in the evening after dinner on the days when Virginia is home from work earlier. The following exchange of March 20 is typical. Cindy is sitting on one section of the large modular couch when Virginia comes downstairs. The 8O researcher is sitting on a far section of the couch. Virginia selects a seat on a third section. Cindy, "There are no kids or peOple here. Come here woman!" She pats the spot next to her. Virginia smiles at her and walks in that direction. Cindy stands up and gives her a quick kiss and pulls her on the couch next to her. This exchange shows the uncloseted behavior acceptable when only the researcher is present. It also shows Virginia still responding in Fourth Level and Cindy scolding her since it was not necessary in that situation. The previous sequence also demonstrates the changes in mood that epitomize the entrance or leaving of a person who is at the Fourth Level. The couple is very careful to behave as they believe heterosexual roommates would behave, and they are constantly monitoring their behavior. The word "peOple" was used in an emphatic all-inclusive sense to represent the society as a whole. The researcher was present, but in this case she was not part of "peOple." Fifth Level The difference between Fourth Level and Fifth Level is less one of commission than of omission. That is, at Fourth Level are people whom this couple are forced by circumstance to deal with at a personal level. At Fifth Level are people that the couple actively seeks to keep distant. The major hiding mechanism used here is avoidance.This avoidance is 81 evident in language behavior when the lesbian changes the subject or lies about an artifact. One example of the difference between Fourth and Fifth Levels can be demonstrated by the reaction to the wedding ring. At First Level, the researcher was shown the rings, and their full significance was eXplained. At Fourth Level, Cindy's niece had asked to see her ring since she thought it was pretty. Cindy gave her the ring, and suddenly realized that she might try and read the inscription inside which has both their initials and their wedding date. She said that she casually took the ring back before the niece read it.(Conversation with Virginia, April 26). However, at Fifth Level, when one of Virginia's coworkers commented about her pretty ring, Virginia just said, "thank you." The other person persisted and asked if it had any particular significance. Virginia said that she just said, "Yes, it was a gift from a special friend." And then immediately changed the subject. (Interview, June 7) Changing the subject or avoiding the situation seem to be the prevailing techniques at Fifth Level. If forced into a situation where they must deal with an uncomfortable topic, both Cindy and Virginia say that they do not want to lie. Cindy described one situation last summer when she was very tempted to lie. She had called a campground to make reservations for their vacation trip. The person asked how 82 many people there would be, and Cindy said eight. The reservations clerk said that only immediate family members could use one campsite and then asked for the relationship between the people. Cindy said that she wanted to just say that she and Virginia were sisters, but she decided to simply say that they were roommates and see what would happen. They got the reservation without difficulty, but Cindy said that if they had not, she would have simply tried another campground. (Interview, June 13). Another time they thought that lying might be the only alternative was described by Cindy. She laughed throughout the description, and the researcher concluded that this was not really considered seriously. Cindy eXplained, "We were going to have this all lesbian party for New Year's Eve. I wondered what we would do if any of our straight friends or relatives showed up. And then I thought we could tell them that all the men had gone out for beer! Fortunately, none of them showed up." (Notes, May 31) The major way they maintain Fifth Level at their jobs is by avoiding intimacy with peOple in that setting. Both Cindy and Virginia agree that they do not have friends at work, nor do they socialize with people from there. Virginia explained that she avoids talking about herself and changes the tOpic when pressed. Virginia has found that since she works with very few peOple, this is more difficult than it was at her last job in a large organization. Her 83 boss knows her better than she would like, and is perceived as at the Fourth Level, but she tries to keep her co-workers at Fifth Level. For example, one day one of the women in her Office asked her how long she had been divorced. Virginia told her that it was final two years ago. Then the woman asked her if she ever dated since she never seemed to mention anything about it. Virginia said, "No, I'm not interested in dating yet." And then she immediately changed the subject.(Interview, June 12) Virginia also said that she avoids going to office parties or gatherings since she would not want to go without Cindy, but could not bring her either. Cindy echoed the same sentiments. But she felt less pressure to become close to peOple at work since she is out on the bus by herself most of the time.(Interview, June 12) Cindy and Virginia described one exception to this rule. They had attended an all-women coffee house sponsored by a local lesbian organization. Virginia suddenly realized that she recognized one of the women there from work. This happened 18 months ago when Virginia held a different job in a large organization. Cindy explained, "There was a gay woman who worked with Virginia at her old job. She knew all about us, but she wouldn't even be seen having a coffee break with me. She said Virginia was OK because she looks straight, but she said I was too obvious."(Field notes, 84 April 9) This process also happens with family members that are not very close. One of Virginia's cousins who lives up north is getting married. The invitation came addressed to Virginia and to her children, Lisa and Ben. However, this woman is a distant relative with whom there is no day-tO-day contact. Virginia believes that they are totally unaware of her present living arrangements and have simply been told that she is divorced. Virginia and Cindy are going to go to the wedding; the children will be with their respective fathers. Virginia mentioned that her mother said that she was wondering when both of them would be going up. Virginia said that the relatives are not expecting Cindy, but that they would not be surprised that since she is divorced, she would not want to travel alone and might bring along a "friend."(Field notes, May 23) The nonverbal behavior and mood are both very formal and proper. The couple attempts to present themselves as two women who happen to be living together. They hope that those people who know that there is a roommate will assume that the arrangement is strictly for financial and security reasons. Indeed, there is virtually no show of affection. The mood is tense since the women are constantly watching their behavior. On one occasion when the paper delivery boy came to collect the money, Cindy selected a chair on the Opposite side of the room from where Virginia 85 was sitting. It was apparent that she was demonstrating that she did not sit near Virginia. The researcher never saw either Of the women even touch each other casually outside the house. If they sit on the patio on the side of the house, they invariably select chairs rather than the garden couch. Summary of the Five Levels Figure 3 is a table summarizing the data obtained. It can be seen that while there is some overlap in behaviors, different peOple experience Cindy and Virginia in different ways. People may sometimes be present when a more closeted level person is also present. In that situation, the other person is also expected to behave at a more closeted level. Since the researcher was at First Level, it was possible for her to observe examples of behavior at all five levels. Indeed, it was required that she participate in not being more Open than the level of the least aware person present. This table details the five levels identified, the people identified as operating at each level at the time the data were collected, and examples of behaviors along each dimension for each level. Of course, this list is not exhaustive, but it represents conclusions drawn as a result of repeated observations of each behavior. As will be shown 86 n NCDGE 952(le WOZWQSN 02~m~towma >2<===w 0:205 622°2<20 >n 323:2. O2.2O=OF .82 3322.» 4.285.. 2.2.2032 no 32.3.9... 2.5:. 2» no 32.3.9... 229: 25.20.... .mfiaztt wh2w2m02(22< G2awmdu wO2W22k 92.223220 uxhhbono. 02:4 tmzu622 «525 .3298}. .2932: 02.35... 02.2.2.3 O2 5252.: 3:83 02. 30.. 932.02... 2.3.2339. 8.320 m3.<2 . 2.262.353 .. 22023.32 3 <2»: 02... «a .3 us. .326 .3 399:. 3:3“... 523.... m . 2 302...... 02.223: SOS. «2.25.30 3.23 m :20 32.35.... 2.: $2302 R.» ...m? . 332.320» 36:. 952.92.. 02.22:: 4.22.2.3. 2332.... .33.. 96.30 :20... <22. 520...... «on: O... ozoowm 0. 53.3.3 20...: 2.3.. 9...: up $23.02. .20: .65. .3238 2 3.2.8 Rozean.2ooe42a Q2< ”02.2 302m.1=§ .>Q<.~ >3 .w2>0 .IWWDO .2W>Od «0 6.: WAQDOU 23¢qu «302. (023 QW>WI~ hut; GOO: CO; <2mu 4 (Gym) . 202 29> (IWO 4 matOWQ KO mudti