Lovers, legal strangers, and parents: Negotiating parental and sexual identity in family law

[T]he concepts of homosexuality and adoption are so inherently mutually exclusive and inconsistent, if not hostile, that the legislature never considered it necessary to enact an express ineligibility provision .... Homosexuality negates procreation. Announced homosexuality defeats the goals of adoption.

-In the Matter of Adoption of Charles B., 1988

To suggest that adoption petitions may not be filed by unmarried partners of the same or opposite sex because the legislature has only expressed a desire for these adoptions to occur in the traditional nuclear family constellation of the 1930's ignores the reality of what is happening in the population.

-Matter of Adoption of Camilla, 1994

Although Domestic Relations Law does not explicitly define the term "parent," we are of the view. that the petitioner does not come within the meaning of that term.

-Alison D. v Virginia M., 1990

An oft-citied theme in law and society research is the interplay between the social world and everyday life, on the one hand, and the law and its institutions, on the other, in reciprocally constructing meaning (Ewick & Silbey 1998; McCann 1994; Phillips & Grattet 2000; Sarat & Kearns 1993; Yngvesson 1993).Just as the social world necessarily influences the making and interpretation of law, the law has the ability to influence, whether coercively or subtly, our social existence. One particularly potent way in which this happens is through the power of the law to create and shape meanings, definitions, and identities in ways that are consequential for the everyday lives of citizens. Many researchers have discussed, for example, the highly consequential and potentially devastating personal impact of being labeled as a "criminal" by the legal system (Becker 1963; Braithwaite 1989; Goffman 1963; Lemert 1951). The effects of acquiring such an identity come not only in the form of structural limitations (such as the inability to vote) and institutional requirements (such as requiring one to "register" or stay in contact with a parole officer) but also in more subtle and personal forms-what has alternatively been called "labeling" (Lemert 1951) or "shaming" (Braithwaite 1989). The process of being cast as a "criminal" and inheriting that identity is an extreme and quite visible example of how the law may shape and impose identities.

This process also happens in more subtle ways, under other conditions and in other legal forums. Feminist scholars have noted repeatedly the power of legal institutions to define, delimit, and constrain women's activities and identities in both the public and the private spheres.1 Groups seeking assistance from the law in asserting their political and legal rights have also been subject to the law's definitional powers (see Espeland 1994; Merry 1990). Espeland (1994) notes, for example, how the law can simultaneously represent a group's interest (in her case, the Yavapai community of Native Americans) and impose on the group an identity that differs from the group's self-constructed identity.

In many cases, judges are in the position to decide who are legitimate legal actors and who are not-thus, defining some people or groups as appropriate legal subjects and imposing on others a status of legal nonexistence (those who are deemed to have "no standing" or whose problems are deemed to be outside the realm of the legal authority) (see, e.g., Merry 1990). IMAGE FORMULA 13

In the area of family law, the potential for meaning making and identity formation is particularly potent, because of its notoriously indeterminate nature and vague standards.2 Family courts routinely engage in the process of defining and evaluating many concepts whose meanings are often taken for granted-but in reality, are anything but static-such as "family," "parent," and "harm." In her auto-ethnographic account of open adoptions, Yngvesson (1997) discusses the process of being defined as either a "real" or an "illegal" parent. The adoptive family, like other nontraditional family forms, is particularly vulnerable, Yngvesson argues, to the powers of the law to disrupt and redefine the identities of its members. Because it does not fit within the archetypal framework of "family" assumed by the law-consisting of one parent of each gender, who are married, and whose children are biologically related to them-an element of instability is introduced when such familial entities encounter the law.

As perhaps the most hotly contested of alternative family forms, this instability and vulnerability is particularly present in families headed by one or more gay or lesbian parent. It is becoming increasingly apparent that the homosexuality of one or more parents in a custody suit introduces an added element of contestation (Armanno 1973; Lin 1999). Given the relatively new social realities of the "gayby boom" (a term commonly given to describe the increase in planned gay and lesbian headed families as a result of new reproductive technologies) and the expanding public and legal acceptance of alternative sexualities, courts are increasingly being forced to deal with the intersection of sexuality and family law in a way that forces them to question-or at least temporarily suspend-their standard operating assumptions, definitions, and identifications, which would generally assume "homosexual parent" to be a contradiction in terms. The presence of these nontraditional family forms in the courtroom disrupts the standard dichotomous and exclusive categories on which family law and its understandings of sexuality and parenthood are based. In such cases, where the actors do not fit into the presumed traditional gendered family roles, judges must either create new meaning and new legal categories or stretch the existing categories, forcing nontraditional families into traditional categories-to the exclusion of those for whom the judges cannot find a fit (Harrison 1995). These categorizations, in turn, are crucial in the process of shaping and defining the identity of the legal actors involved.

In this study, I analyze the adjudication of gay and lesbian parents' child custody cases as a site of meaning making and identity formation. In this context "identity" refers not only to an IMAGE FORMULA 17

individual's own self-image but also to how one is constructed as a subject in law and represented in legal discourse. As Espeland (1994:1150) insightfully notes, "The relationship between what is often considered the exemplar of the 'public' sphere-law-and what we might suppose is our most 'private' realm-our conceptions of self-may seem like a study in oppositions, but like many oppositions, the one often informs (if not requires) the other."

Thus, one's identity is composed in interaction and is necessarily constrained and impacted by the law's categories and definitional processes. This is particularly true in an arena such as child custody, where frequently the focus is on defining and categorizing, through legal findings and processes, the facets of the self that are often thought to be the most personal-family ties and sexuality.

In this article, I show how judges over the past 50 years have engaged in discursive processes that create legal and social meaning and, either explicitly or implicitly, shape and redefine the identities of gay and lesbian parents and would-be parents. This investigation explicates the definitional processes engaged in by judges in any legal forum in which there are written decisions. At stake are answers to such questions as, How do courts go about defining key concepts whose meanings are often taken for granted-such as "parent," "family," and "homosexual"? How do these definitions differ from those of the litigants before them, and how do they vary across decisions? Do they recognize the fluidity of these terms and their meanings? Are some identities and meanings prioritized over or predicated on others? How are these meanings negotiated, shaped, and settled over time? How are the presumed mutually exclusive statuses of homosexuality and parenthood brought together or reconciled? The specific focus here is on sexual and familial identity formation in the context of gay and lesbian parents' family court cases; yet the findings are relevant generally to the constitutive analysis of identity formation and add to existing law and society scholarship on meaning-making processes in the law (Ewick & Silbey 1998; McCann 1996; Phillips & Grattet 2000).

In addition to its theoretical applications, the empirical contribution of this work to the study of gay and lesbian parents' legal treatment and status is significant. This study follows a long tradition of legal and social scientific scholarship that has been instrumental in rendering visible the existence of gay-, lesbian-, and bisexual-headed households (Benkov 1994; Lewin 1993; Hunter & Polikoff 1976; Rivera 1987), explaining the problems faced by these parents in family court (Seidel 1989; Sella 1991; Sheppard 1985), refuting myths that portray them as inadequate parents (Bigner & Jacobsen 1989; Falk 1989; Green et al. 1986; Patterson 1995; Patterson & D'Augelli 1998; Stacey & Biblarz 2001), and putting forth innovative legal strategies aimed at over-- coming these problems (Burks 1994; Hitchens & Price 1978-79; Peltz 1995; Polikoff 1990; Wray 1997). For instance, evidence of the oft-cited "gayby boom" (Pressley & Andrews, cited in Lin 1999) of the past decade is manifest in this study, in that over half of the cases occur in the 1990s-even though they begin in 1952. The current project moves beyond these important foundational studies, however, to ask new questions about the impact of law on gay- and lesbian-headed families over the past 50 years. Most importantly, this work emphasizes the need for both micro-- and macro-level analyses of the power of judicial narrative and its impact on lesbian, gay, bisexual, and transgender (hereafter LGBT) parents and their children by looking to the narratives themselves, both individually and as a discursive whole.

In the pages that follow, I begin with a brief over-view of the legal context of child custody and visitation, followed by a discussion of my methodology. Next, I offer a conceptual background for the study, locating it at the intersection of family, law, and sexuality and discussing the complexity of that intersection. Here, I comment on the social history of "family" in gay and lesbian communities and activism surrounding gay and lesbian family issues, the resistance to legal and social recognition of gay and lesbian families, and the legal formulations and incarnations of family and parenthood.

I then begin my analysis of identity formation with a discussion of common assumptions about the etiology and nature of sexual and familial identity generally, in the eyes of the law, and in the context of child custody decisionmaking. This discussion serves as a precursor to the crux of my analysis, an in-depth examination of the discursive processes of identity negotiation and its attendant forms, tensions, and outcomes: in particular, the imposition and negation of identities by the court, and the negotiations and compromises that result when these identities collide with or contradict the litigants' representations of self. These sections are followed by an analysis of the notion of "commitment" to identity-how different and presumably incompatible identities are prioritized or hierarchically arranged and performed.

Throughout these analytical sections, several themes loom large-in particular the notion of "homosexual recruitment," accusations of deviance, and varying definitional framings of family and parents. I draw on specific cases throughout to demonstrate how law defines family and its members, how parties are either included or excluded from the legal process based on these definitions, and how sexual and familial identities are shaped and reshaped in law. I conclude by arguing that there are specific mechanisms, including discursive compromises and "passing" gestures, that allow for the reconciliation in law of previously incompatible identities, and assert that the negotiation of identity in court should be framed as a power struggle wherein the stakes are self-determination and self-definition.

The Legal Context: A Brief Note on Custody and Family Law

As noted previously, family law has often been named as perhaps the most indeterminate and varying area of law practiced in the United States; for that reason, as well as the variance in the specifics of family law from jurisdiction to jurisdiction due to a tradition of federalism in this area of law, it is difficult to summarize the contents of family law across all 50 states (Charlow 1994; Guggenheim 1994; Parker 1994; Polikoff 2000; Schneider 1991). A few commonly accepted standards do exist, however.

The previously mentioned "best interest of the child" standard, however vague, is the guiding principle in determining custody between two legally recognized parents. Likewise, the best interest standard plays a role in deciding whether to allow the adoption of a child by a single parent or a married couple. However, for this standard to be applied in adoption cases, both the child and the parent(s) have to be determined eligible-and the requirements for eligibility vary significantly from state to state. In determining visitation claims, the standard is quite different. For a legal parent to be denied visitation, one must find that such visitation would be detrimental to the child. For this reason, one requires a much higher standard of proof to deny a parent visitation than one does to deny him or her custody. Likewise, it is more difficult to change custody once it has been awarded-to do this, it must be shown that there has been a significant change in circumstances since the time of the original custody award. Finally, the standard of review for custody cases is as vague as the standard for determining custody-for a decision to be overturned on appeal, it must be shown that there was an "abuse of discretion" by the trial judge. What constitutes an abuse of discretion, of course, varies significantly not only from state to state but also from judge to judge.

Method of Analysis

The analysis that follows is based on a data set of 235 appellate court decisions from 1952 to 1999.(3) These decisions were gathered using Westlaw and LEXIS legal research software, using the key words "child," "custody," "homosexual," "bisexual," and "lesbian."4 I sifted through the cases to exclude only those in IMAGE FORMULA 27

which child custody or visitation is not at stake, or in which at least one of the parties is not (allegedly) homosexual, bisexual, or transsexual.5 I included all recorded appellate custody and visitation cases in which one or both of the parties' homosexuality or bisexuality was raised as an issue for discussion. These cases contain three general categories of claims: divorce cases involving a heterosexual marriage in which one party subsequently comes out as gay or lesbian and there is an ensuing custody or visitation battle; adoption cases in which a single gay man, a lesbian, or a same-sex couple wishes to adopt a child, or in which the nonbiological parent in a same-sex parenting dyad wishes to formally adopt; and "lesbian split" cases, in which a same-sex couple ends their relationship and the nonbiological mother seeks custody or visitation rights.

The data set includes all cases that were considered on appeal, whether or not the decisions were ultimately overturned. These cases range in appellate court levels and come from 44 states, as well as the District of Columbia and the U.S. Court of Appeals, 4th Circuit. For each case, I completed a descriptive summary, including year, level of court, and state or jurisdiction, as well as a history, a description of the parties involved, a brief summary of the key issue(s), the outcome of the case, and other pertinent information. One hundred and fifty cases involved a divorce or split between a man and a woman, one of whom came out as gay or lesbian; 27 involved the breakup of a same-sex couple; and 32 were adoption cases.6 Analytically, the focus was on both the outcomes of the cases and the rationales given in the judicial narrative for these outcomes. Therefore, in subsequent close readings of the cases I looked for, summarized, and coded (when necessary) key analytic dimensions of the cases, including such factors as the judges' definitions of "parent," "family," "harm," and "best interest," any references to deviance or morality in reference to the parents' sexualities, the legal arguments IMAGE FORMULA 30

put forth by the litigants, and the nature of any expert testimony given.7

The inclusion in the analysis of all known appellate custody cases involving mention of one or more gay, lesbian, or bisexual parents in this study ensures empirical comprehensiveness and allows for a historical perspective. Yet, it is important to address the rationale and consequences of using only appellate custody cases in the archival analysis. Since decisions in trial courts generally are not recorded, were I to attempt to gather them, I would not be able to give an accurate account (or count) of all custody cases in which the parents' sexuality is at issue. In particular, a comprehensive analysis of a historical nature would be nearly impossible. The content of an appellate decision also differs from those at the trial level in that the purpose of the former is to evaluate issues of procedure, not fact (although the facts are at least rehashed in almost all appellate decisions). Additionally, those who pursue their cases to the appellate courts tend to have more resources than those who do not and therefore are a self-- selected group, to some extent. By using only the population of appellate cases, however, I focused on cases that were highly conIMAGE FORMULA 36

tested and whose outcomes became legal precedent. The latter is particularly important, as the focus here is on judicial decision-- making as a mechanism by which legal meanings, and ultimately social meanings, are made and changed. Furthermore, while the standard for review of child custody cases-"abuse of discretion"-is thought to be high, it is also quite vague and allows for multiple interpretations (Benkov 1994; Rubenstein 1993). That my search revealed 235 appellate cases is evidence that although the majority of trial-level custody cases may not be reviewed at the appellate level, enough are to allow me a suitable pool of decisions to analyze. Finally, while the use of appellate decisions may be problematic if one were attempting to predict the outcomes of custody cases, in this study the focus is on the language of the law and the framing of legal definitions and narratives-- thus, the use of appellate cases is appropriate. Indeed, appellate decisions have been used successfully in several similar studies of family law (Daly 1995; Fleming 1997; Lin 1999; Mason & Quirk 1997; Sheppard 1985), as well as in the most recent sociolegal work on judicial language and decisionmaking (Phillips & Grattet 2000).

The use of such records of judicial decisions in the study of legal narratives is a hallmark of the constitutive approach to the analysis of legal language and that which it signifies. McCann (1996:477) comments that "the concern for exploring citizen meaning making activities and interactions places a premium on qualitative techniques such as . . . archival investigations," and that "[l]egal language importantly shapes how citizens understand, explain, and negotiate . . . in practical social interaction." With specific regard to the study of gay and lesbian experience with the law, Eskridge (cited by Lin 1999) remarks that "legal scholarship is inevitably narrative." These narratives not only reveal the perspectivity and biases of the storyteller (most often the judges), but narratives can also "have tremendous informational value, which can have great impact in debunking erroneous and negative stereotypes [about gay men and lesbians]" (pp. 748-49). MacDougall (2000:16), in his discussion of judicial treatment of gay men and lesbians in Canada, comments on the importance of legal language: "Words and meaning are everything in law.... Applying a particular word is not just a matter of deducing the appropriate label from a given state of affairs; the application of that word can also induce certain consequences or assumptions about what those consequences will be." Such analyses of legal narratives are best accomplished through close reading and content analysis of the judicial decisions (Johnson 1987; Kort 1962; Marvell 1978; Mercer 1998).8 IMAGE FORMULA 39

Content analysis, defined by Holsti (1969:14) as "any technique for making inferences by objectively and systematically identifying specified characteristics of messages," enables the researcher to distill and organize themes in the documents to be analyzed, as well as to emphasize critical reflection in the construction of categories and interpretation of text (Mercer 1998). These elements, in particular the latter, are crucial in analyzing the construction of identity in legal narrative and action. As Ewick & Silbey (1995:202) note, the study of legal narratives reveals how they may "function in mediating action and constituting identities."

Conceptual Background: At the Intersection of Sexuality, Family, and Law

The intersection of non-normative sexualities and the family is particularly troublesome for the law because homosexuality and the family have traditionally been thought of as mutually exclusive institutions (Lewin 1993; Lin 1999). While Fineman (1995) discusses the law as being formed around and aimed at preserving the "sexual family" as its normative and empirical base, this sexual family is distinctly heterosexual. The law and its institutions are neither doctrinally nor ideologically equipped to deal with the possibility of a family whose heads of household look less like "Ozzie and Harriet" and more like "Rozzie and Harriet" (Sullivan 1996). The legal (if not social) existence of same-- sex parenting couples is in large part a recent phenomenon-- only two such dyads are documented in my data set as having had a case heard at the appellate level before 1990. But the law's ill preparedness for dealing logistically with the breakup of a family in which there are two mothers, for instance, is far from merely a problem of novelty. There is an ideological and highly moralistic component to the courts' traditional reticence in recognizing gay men and lesbians' roles as parents, and it goes beyond the courts' ambivalence (at best) toward homosexuality in general.

The confluence of homosexuality with the family, one of the most revered and protected institutions in American life, taps into a deep-seated, underlying collective fear (Dalton & Bielby 2000; Fineman 1993; Sella 1991). This fear is rooted in two culturally embedded systemic beliefs. The first is that homosexuals, more than just aberrations from the norm, are overly sexualized and promiscuous to the point of depravity (Kinsman 1996; Lin 1999). This stereotype, although empirically groundless, is in some ways not surprising, given that the lesbian, gay, bisexual, and transgender (LGBT) community is the only minority group largely defined by its sexual practice.9 This belief encourages the fear that the presence of gay men and lesbians in the family will corrupt it as an institution (Lin 1999). The second factor is based on the widespread disbelief that LGBT individuals would be interested in or capable of nurturing children and having a family. Gay men and lesbians are not seen as possessing the attributes needed to be a parent-so the two images have often been irreconcilable in the eyes of many, including the court (Stacey 1996).

But long before the "gayby boom" of the 1980s and 1990s, families were being built in gay and lesbian communities (Nardi 1997; Weston 1991). These "families of choice" not only served as support networks for gay men and lesbians who had been ostracized by their biological families,10 but they challenged the traditional notion of family and the concept that people have to be linked genetically in order to be considered a family (Stack 1974). Thus, the creation of these alternative family networks was both functional and subversive. They demonstrated the vicissitude of the family form and, at the same time, located a new space at the intersection of family, law, and sexuality. While these families most often did not come to the attention of the law, they showed that the taken-for-granted definitions of family offered by the law were not adequately inclusive (Say & Kowalewski 1998; Sullivan 1999; Weston 1991).

When gay and lesbian family issues did finally come to the attention of judges and legislators, it was largely within a context of activism on the part of lesbians, who had long been excluded from the major currents of claimsmaking in both the women's movement and the gay rights movement (Polikoff 2000). While the early gay rights movement focused on issues that most affected gay men, such as sodomy laws, it was not until the mid-- 1970s that women in the movement began attaining visibility in the area most affecting them-child custody challenges based on their sexual orientation. Although there is no way to know how many children were being raised by lesbians before this, and cases appeared sporadically in the appellate courts as early as 1952, the first activist group dedicated to this cause, the Lesbian Mothers National Defense Fund, appeared in 1974; and the major gay legal rights organization in the United States, Lambda Legal Defense and Education Fund (hereafter Lambda), first became involved in a lesbian-mother custody case in 1977. By 1990, Lambda and the American Civil Liberties Union had filed "friend of the court" briefs or had provided representation in at least 15 appellate cases. IMAGE FORMULA 45

In the late 1970s and the 1980s, reproductive technology became widely available for lesbians to choose motherhood outside of the heterosexual marriage context-the demographic phenomenon previously noted as the "gayby boom." Again, there is no way to accurately count the number of such families; but in 1985 the first second-parent adoptions for lesbian couples who had conceived via alternative insemination were granted in Alaska and Oregon (Benkov 1994; Polikoff 2000).

In the legal arena, other divergent family forms, such as those involving stepparents or extended-kin relations, have necessitated the development of legal formulations to define pseudo-parental figures and other interested parties for custody purposes. Some of these legal formulations, such as de facto parenthood, confer a parent-like status to a person who has assumed the day-to-day duties of a parent such that he or she can participate in court proceedings but cannot generally obtain custody and is not recognized as a legal parent (Sella 1991). Other legal forms, such as equitable parenthood and in loco parentis, create parental rights analogous to those of a biological parent. The former is based on a mutually acknowledged parent-child relationship and is often used in situations where a father has proceeded on the assumption that he is the child's natural parent but ultimately finds out that this is not the case (Sella 1991). The latter creates parental rights for someone voluntarily providing support or care and is often used to confer legal status to stepparents (Polikoff 1990). In some cases, stepparents are allowed to adopt a child to formalize the relationship. While generally courts and legislatures will only recognize one parent of either sex, these adoptions are allowed based on the "stepparent exception" to the rule, which would otherwise require that the biological parent of the same sex have his or her parental status terminated. In addition, contractual parenting agreements are sometimes entered into in an attempt to formally recognize non-- biological parents.

All of these legal statuses and definitions were originally formulated to accommodate heterosexual extensions of the family, such as stepparents and grandparents; but lesbian and gay co-- parents have since begun using them to assert their parental rights and identities in court. Trial courts in some states began allowing second-parent adoptions to gay and lesbian co-parents based on the use of parental contracts or analogies to stepparent adoptions as early as 1985; however, this right was not explicitly institutionalized in case law until 1993 (Logue 2001).11 Likewise, some gay and lesbian non-biological parents had attained visitaIMAGE FORMULA 49

tion rights as de facto parents or had started using the doctrine of in loco parentis at the trial level as early as 1984; yet, such arguments were not successful in appellate decisions until 1996.(12) Often, the courts were reticent to apply these labels to same-sex partners or former partners, finding that they did not qualify because of their lack of a blood or marriage relationship. Thus, gay and lesbian non-biological parents were often excluded from the legal process and were not identified as parents or even as pseudo parents, but as "third parties" or "legal strangers" who were not allowed to have an interest in the custody proceedings.

The Etiology of Identity: Essentialism, Transmission, and Labeling

Innately embedded in the study of identity-sexual and otherwise-are questions about its etiology. Certainly in the case of family identity, in particular non-biological kinship ties, the law's powers to regulate, create, and deny identity are evident (Lewin 1993; Stacey 1996; Yngvesson 1997). Gay and lesbian parents encounter this reality when they attempt to legally solidify their parental identity by adopting a child whom, in many cases, they may have been raising since birth.13 Yet, as Yngvesson (1997) discusses, parenthood (in particular, motherhood) is also a culturally articulated and naturalized status whose roots are thought to be outside the law (see also Lewin 1993).

Although this discourse of "natural" and fundamental parental identity has most often been rooted in biology, there is increasingly a question as to whether biology alone is sufficient to identify a person as a parent (Benkov 1994; Stacey 1996). In the well-known case of Thomas S. v. Robin Y. (1993), the judge rejected a sperm donor's paternity claim, arguing that he had not acted as a parent throughout the child's life, as had her two lesbian mothers: "To Ry [the child], a parent is a person who a child depends on to care for her needs. To Ry, Thomas S. has never been a parent since he never took care of her on a daily basis.... In her family, there has been no father."14 This case and others like it suggested that parental identity cannot be biologically essentialized; rather, the acquisition of such an identity must be deliberate and active, and there is a distinctly performative aspect to it. IMAGE FORMULA 53

The same distinction is useful in discussing the nature of sexual identity. Homosexuality has most often been defined, in law and in public discourse, by the sexual act(s) most associated with it. In ruling on its criminal status, the Supreme Court referred to sodomy as "the behavior that defines the class" of homosexuals (Bowers v. Hardwick, 1986). Yet, scholars have shown that this rendering of act and identity as mutually defining is problematic for a number of reasons-in particular because of the reality that heterosexuals also engage in sodomy and not all self-identified homosexuals do (Halley 1997). Subsequent decisions such as that of the U.S. Court of Appeals in Watkins v. United States Army (1988), in an effort to remain consistent with Bowers, made a distinction between act and identity by finding that the status of homosexuality could be protected under equal protection doctrine, even if the conduct associated with it was not (Eskridge & Hunter 1997). Such developments and contradictions have prompted many in academe to adopt what Halley (1997:91) has called a "personhood definition" of sexual identity, in which the classes of homosexual and heterosexual are defined by the form of personality they may share rather than the sexual behavior in which they engage. Yet this distinction is not universally accepted, and often these two facets-behavior and personhood-are conflated, particularly in discussions of how one "becomes" homosexual.

The dispute over the etiology of sexual identity is complex and highly contested. The familiar "nature versus nurture" debate is politicized by the important social and legal consequences its answers have for LGBT communities, individuals, and their families. At stake in particular are two widely significant questions: first, is homosexuality an "immutable trait" that can be protected as a suspect status under equal protection doctrine and civil rights laws; and second, is homosexuality socially learned or otherwise communicable, such that it may be transmitted from parents to their children? With regard to the first question, the LGBT community has been hesitant to take a side in the debate because of the potentially troubling implications of either conclusion. Scholars of sexuality have often rejected the essentialist/ constructionist dichotomy as poorly conceived and overly constraining (Halley 1997; MacDougall 2000). The point has also been made that sexuality cannot be analogized to race or gender in terms of immutability because it must be made visible, therefore, it may be seen as mutable, at least in the performative sense (Eskridge 2001).

The issue is no more resolved in the legal arena. Courts have often been put in the position of debating how one's sexual identity is formed. As early as 1957, the Wolfenden Report, sponsored by the British government, argued that homosexuality was a "'state or condition' that [could not] come under the purview of criminal law," while maintaining that a purely biological explanation for its etiology would wrongly absolve these men of "responsibility" for their actions (Kinsman 1996).15 The implication of the argument that one is "born" into his or her sexual identity not only simplifies and essentializes an irrefutably complex and contingent social and personal part of human existence, it also raises the specter of a potential genetic "cause" of alternative sexualities-itself a deeply troubling step toward possible eugenics implications. Yet to argue that sexuality is entirely a socially learned or constructed facet of personality may have the legal consequence of foreclosing on the possibility of adding sexual orientation as a suspect status to be protected constitutionally-and, as suggested above by the Wolfenden Report, may invite personal criticisms and unrealistic expectations that individuals should be able to "control" or change their sexual identities at will. The rejection of sexual orientation as an immutable and therefore protected status-and the consequences of such a rejection in terms of civil rights-was apparent in a number of high-profile cases, including the landmark decisions of Bowers v. Hardwick (1986) and Padula v. Webster (1987).16

The second question posed above-whether homosexuality may be learned or transmitted from parent to child-is equally complex, problematic, and unresolved for both courts and scholars. In 42 (18%) of the 235 cases analyzed here, the judges discussed the possibility of a child's sexual orientation being affected by that of his or her parent(s). The process this transmission might entail has sometimes been left unarticulated by courts-homosexuality has been analogized as a contagious disease, which may infect children by some unspecified means.17 Often in the legal arena, however, judges and opposing litigants have asserted more specifically that this transmission of sexuality may happen in a passive or an active way: Children might model the behaviors of their parents and learn their sexual identity, or parents may actively "recruit" their children by making conscious efforts to "convert" them to homosexuality (Lin 1999). The primary difference between these explanations has been the difference in the intentionality-and therefore culpability-of the parents in contributing to their child's possible alternative sexuality. The second explanation-the "conversion" model-envisaged homosexuality as a sort of cult recruiting membership. This model was evident in Hertzler v. Hertzler (1995), in which a IMAGE FORMULA 58

mother was accused of "immersing" her children in homosexuality: "[T]he record is ... replete with [the mother] Pamela's intensive and unrelenting efforts to immerse the children in her alternative lifestyle, seemingly to the point of indoctrination." Similarly, in J.P. v. P.W. (1989), a gay father's visitation rights were restricted based on the perception that he had "advocated" his "lifestyle" to his children. Many parents and advocates have commented that the expectations that a gay or lesbian parent-- who has no doubt experienced significant stigma, exclusion, ridicule, legal difficulty, and possibly violence as a result of his or her sexual orientation-would want her or his children to experience the same type of treatment are misdirected, if not absurd (Ali 1989; Lin 1999).

The first explanation proposed above-the learning model-- was less likely to implicate the parent in such "cult-like" behavior, but nevertheless assumed that he or she was ultimately responsible for a child's sexual identity as a role model. In Bennett v. O'Rourke (1985), for example, the court revoked custody from a lesbian mother based on the belief that her child would "model" her sexual behavior and become a lesbian:

In light of the fact that here the homosexual parent and the minor child are both female, we consider this factor particularly important because of the increased chance of role-modeling .... Common sense dictates that a child should not be exposed to such an unhealthy attitude, especially without the consistent presence of a father to help counteract its ill effects.

In response to such claims, a number of developmental and social psychologists have presented longitudinal evidence both in court and in scholarly forums that refutes the hypothesis that a child is likely to "catch" homosexuality from a parent (Ball & Pea 1998; Green 1982; Green et al. 1986; Patterson 1995). In addition, they have pointed out the flawed logic of assuming gay children must derive from gay parents-since most of the parents who were studied or who appeared in court came from heterosexual parents (Armanno 1973; Hitchens & Price 1978-79; Lin 1999). Any evidence of differences between these two types of families generally revealed differences in traits that would be seen as advantageous to the children of gay or lesbian parents, such as open-mindedness and self-assuredness (Bigner & Jacobsen 1989; Stacey 1996; Stacey & Biblarz 2001).

Some gay and lesbian litigants have successfully used this research to counter claims in court that their children will become homosexual if they gain custody or visitation rights. As early as 1985, judges in S.N.E. v. R.L.B. (1985), relying on the advice of 16 expert witnesses, rejected the notion that a child would "catch" homosexuality from his lesbian mother. Often, such evidence has not been considered or accepted as relevant in judicial decisions, and the discourse of homosexuality as a contagion has continued to be quite common in court. As late as October 1999, judges in the case of Eldridge v. Eldridge reflected this line of reasoning in restricting a lesbian mother's visitation rights based on "sexual orientation and behavior modeling issues."

Beyond the debate over the origin and causes of sexual orientations, however, the process of self identifying-or of being identified-as homosexual is by no means a simple one. MacDougall (2000:17) notes the impact of the various identity labels applied to gay men and lesbians by the law:

What flows from the use of the term 'homosexual,' in particular what assumptions are made, what stereotypes are applied once the label is assigned, is ... significant .... [J]udges have been reasonably content to allow the consequences to be determined by stereotypes, on the whole negative and marginalized stereotypes, the content of which has not been defined by homosexuals.

This labeling process is more than a passive acceptance of extant stereotypes regarding homosexuality, however. When the law affixes a particular label, depiction, or status, it asserts discursive control over alternate sexual identities. In other words, the law must have a hand in defining identity in order to regulate non-- normative sexualities (Brown 1995; Gavigan 1998).

Far from recognizing its fluidity, actors in the legal arena have tended to essentialize or stabilize sexual identity (Halley 1997). Self-identification and the reading or misreading of one's sexual identity by others are two of the ways in which this essentializing process has occurred. In the custody cases of gay and lesbian parents, both of these processes have been evident. Perhaps the most blatant examples of this were those cases in which the judges, in their written decisions, explicitly denied or contradicted a person's own sexual self-image as articulated in court. The social and legal production of sexual identity were laid bare here, and jurists were literally charged with answering the blunt question: homosexual or not?18

Beyond the basic question of whether someone was in actuality (or in perception) gay or lesbian, sexual identity has also been interpreted and negotiated in more normative terms as part of a discourse of "deviance." This discussion of sexuality as deviance took place in 37, or 16%, of all of the cases examined. To be labeled in the courtroom as homosexual, some have suggested, is to be labeled as "other" (MacDougall 2000; Seidman 1997). MacDougall (2000:20) notes, "Historically, the effect of the label `homosexual' and its related terms has been ... to render the activIMAGE FORMULA 67

ity or the people to whom the label is applied beyond the pale of acceptability. The use of the label homosexual and its various historical and modern equivalents has been, for the most part, an exercise in marginalization."

In recent legal history in particular, the "otherizing" of homosexuality has taken a different and distinctly more judgmental tone than that of other marginalized identities, such as racial minorities, women, and so on (Lin 1999). More than just being identified as non-normative, gay and lesbian identities alternatively have been constructed in court as either pathological or immoral-or as having "pernicious" effects on their children.19 Countless legal and social science scholars have enumerated and presented evidence to counter the multiple ways in which gay and lesbian parents have been assumed to be deviant-by virtue of either excessive promiscuity, a penchant for molesting children, psychological instability, or a desire to "convert" children to homosexuality (Kinsman 1996; Lin 1999; O'Toole 1989).

Even though the medical community has not officially considered homosexuality to be a psychological disorder since 1974, this discourse of deviance has been present in decisions throughout the past 50 years (Bayer 1987). Judges have defined the sexual identity of gay and lesbian parents as "sexual disorientation"20 (emphasis added), as an "abnormalcy,"21 an "illness,"22 or a "psychological disturbance."23 The practice of identifying gay men and lesbians as deviant has of course been consequential in determining whether they are fit parents. In a functionalist sense, however, it has also served to illustrate what a "good parent" should be:

The practice of the social labeling of persons as deviant operates . . . as a mechanism of social control. . . [I]t helps to provide a clear-cut publicized and recognizable threshold between permissible and impermissible behaviour .... The creation of a specialized, despised and punishable role of homosexual keeps the bulk of society pure. (McIntosh 1968:101)

Sexual identity has not been the only type of identity that has been shaped and negotiated in gay and lesbian parents' child custody cases. Their familial identity-as parent, spouse, and so on-has also been at issue in these cases. In 33% of all of the cases (70 total), there was substantial discussion of how "parent" and "family" are defined. The question of what constitutes a parent generally in legal and social terms has been addressed previously in the context of defining the family and its members in legal proceedings; but the constitution of or the negation of IMAGE FORMULA 71

one's familial identity has also occurred on a more micro level in the everyday interactions of the family members themselves, between the judges and the litigants, and, ultimately, in the courts' rendering of individual written decisions (Sarat & Kearns 1993). Familial identity is also unique in that it is a necessarily relational status or construct. Even in cases where a judge does not consider an individual to be a parent, the individual must still be named in a way that identifies his or her relationship to the other members of the family-whether it is as the "lover" of the other parent, the caretaker of the child, or even as a "third party." Thus, the law may negate a person's familial identity by rendering it invisible, without actually erasing it from existence or memory (Yoshino 2000).

Negotiating Parental and Sexual Identity in Family Court

The Legal Imposition and Negation of Identity

Undoubtedly, one cannot assume that a court ruling would change one's self-identification as a gay or lesbian parent. However, as this analysis will show, the court imposes identities on the litigants before it, either affirmatively or through negation, in both a symbolic and a practical sense. This theme has persisted in cases throughout the data set, despite the passage of time. The most striking example of such an imposition of identity was evident in the case of Guinan v. Guinan (1984) in which the court's judicial narrative assumed the mother involved in the litigation was a lesbian, despite her denial. Fifteen years later, such an imposition of identity was still evident in the case of D.L. v. R.B.L. (1999), in which the court assumed a father was bisexual because he liked to spend time with his male friends and because of his "attachment to more feminine-type articles of furniture." Sometimes, as in D.L. v. R.B.L., the imposition of a homosexual identity on a person who did not claim such an identity for him- or herself was based on what the court perceived to be gender non-- normative behavior or preferences. In most cases, however, it was based on the friendships, associations, and affiliations of the person-or on rumors and pure conjecture on the part of relatives, friends, and former spouses. In other words, the imposition of homosexual identity by the court was likely to be the result of a person's friendship with a gay man or lesbian-or with anyone of the same sex with whom he or she spent a significant amount of time-in combination with the active imagination of other individuals and legal actors.

Such was the case In re Mara (1956), in which the mother was assumed to be a lesbian based on "evidence of female homosexuality" between some of her friends and her roommate. By imposing a sexual identity in this way, the law was able to accomplish three things: impute sexual meaning to any aspect of a person's life, regardless of the person's own experiential reality or personal imagination; reify social stereotypes about how gay men and lesbians act and how they relate to others-even in their absence; and redefine one of the most intimate facets of a person's identity against her or his will.

In 37 cases (16% of all cases), the court challenged parents who presented their own sexual identifies as healthy or unproblematic, imposing a label of "deviant" or "mentally ill" in their decision. In at least one case the court threatened to impose further visitation restrictions on a father if he continued to teach his children that his homosexuality was not immoral: "If the father persists in his vehement espousal to the child of the 'desirability' of his chosen lifestyle ... the authorities would support even greater restrictions upon his rights of visitation" (J.L.P.(H.) v. D.J.P. 1982). In 81 cases (34.5% of all cases), the parents submitted social scientific data and/or expert testimony aimed at supporting their self-representations as healthy and non-deviant. In the Matter of J.S. & C. (1974), for example, was one of the first cases of record in which a gay or lesbian parent introduced expert testimony based on the recent decision of the American Psychiatric Association (APA) to remove homosexuality from their Diagnostic and Statistical Manual of mental disorders (DSM), directly contradicting the assertion that homosexuality was an illness per se, or that the gay father in question was not psychologically stable. Thus, despite the court's efforts to impose a deviant identity, parents were sometimes able to counteract this framing of their sexual identity.

When the type of identity being reshaped or imposed by the court was family related, the court could discursively and literally dictate or change the character of one's relationship with his or her partner and child In all, there were 16 cases involving proposed second-parent adoptions and 27 cases involving one lesbian parent against another. In 20 of the 27 cases involving a split between two lesbian parents, the judges rendered the non-biological mother a non-parent by asserting the claim that she had no natural or legal link to the child.

Several scholars have commented on the tendency for courts not to accept the notion that a child may have two parents of the same sex (Burks 1994; Polikoff 1990; Hunter & Polikoff 1976; Sella 1991) Most legal formulations of parenthood and definitions of the term "family" have made it difficult, if not impossible, to recognize the parental status of non-biological lesbian mothers on equal footing with their partners or former partners, the biological mothers-even after the biological mother's death. In two cases, McGuffin v. Overton (1995) and Matter of Guardianship of Astonn H. (1995), the surviving mother in a lesbian parenting dyad fought to define herself as a parent in order to retain custody of a child after the biological mother's death.

More typical, however, were cases such as In re Custody of H.S.H.-K. (1995), in which a non-biological mother faced her former partner in a custody and visitation battle and enumerated the traditional parental duties she had fulfilled during the child's upbringing; or cases such as In re Adoption of Baby Z. (1996), a second-parent adoption case in which both lesbian mothers sought to present the non-biological mother as an equal participant in deciding to have a child and in the childrearing process. Thus, someone who for years identified herself and was identified by others as a mother could, at the end of a custody trial, emerge as a "third party" or even a "legal stranger." Someone who had a solid parental bond with her child could be renamed as a "lover" or a "friend" of the "real" mother; as a result, her identity was defined only in relation to her partner and not by virtue of her connection to her child.

In Alison D. v. Virginia M. (1990), for example, the non-biological mother was named in court as "[the] woman who had a live-in relationship with the child's mother"-despite the fact that she had raised the children since birth and the children bore her last name and referred to her as "mommy." Similarly, in the case of West v. Superior Court (1997), the judges denied the non-biological mother standing and confirmed the biological mother's contention that her former partner did not have the right to "drag" her-the child's only "natural" mother-to court. In the case of Kathleen C. v. Lisa W. (1999), the judge held that the non-biological mother might have been considered a de facto parent were she still living with her former partner but that she had subsequently lost her parental status once she was no longer cohabiting with the child's biological mother: "[A]lthough appellant exhibited the characteristics of a de facto parent during her relationship with respondent, absent any legislative or case authority granting a nonparent visitation rights over the objection of the biological parent ... we cannot grant those rights here (emphasis added)."

Paradoxically, however, judges in 13 cases denied a second mother an adoption or other recognition of being a non-biological same-sex parent based on the lack of an official, legal, or institutionally defined relationship with her partner. As mentioned previously, most states have a rule stipulating that the court can only recognize one parent of each gender. This rule often leads to an exercise in absurdity; for instance, in the case of a lesbian couple, the non-biological mother can only be recognized as a parent if the biological mother's parental rights are terminated. Since the only way to circumvent this rule is via the "stepparent exception" (whereby in the event of a parent's remarriage, his or her new spouse can be legally recognized as a parent), the issue becomes one of whether the relationship between the two parents can be defined as a marriage. In most such cases, although not all, the judges were not willing to define same-sex couples as spouses for the purpose of allowing a ruling analogous to the stepparent exception.24 In Interest of Angel Lace M. (1994), for example, the court found that

[s]ince Wisconsin does not recognize same-sex marriages, [the] woman who cohabitated with the child's mother and who shared equally in raising [the] child was not [the] child's stepparent, even though the woman and the child's mother symbolically solemnized their commitment to each other by partaking in [a] marriage-like ceremony (emphasis added).

Thus, while in certain instances, judges redefined a family member to exist only by virtue of her or his romantic relationship, they were generally not willing to recognize this relationship as a permanent or official bond. Again, the effect of this practice was threefold: it removed the second parent's identity as a parent, it denied her or him any possible contention for custody, and it reinscribed existing stereotypes about the impermanence of homosexual relationships.

The Negotiation of Identity: Reciprocal Processes and Discursive Compromises

It should not be surprising that the self-representations of litigants often collided or were directly contradicted by the judge-- imposed identities they acquired in court. The tensions, struggles, and negotiations that resulted from these contradictions not only illustrate in vivid form the process of reciprocal meaning making and construction of identity that is the object of constitutive theories of law and society (see Harrington & Yngvesson 1990; McCann 1996), they also mark a new terrain in which mutually exclusive or contradictory images can be reconciled.

This interplay was apparent in moments of negotiation in which the impact of both the law and the individual were visible in the eventual settling of meanings and identities. An instructive example is the case of Karin T. v. Michael T. (1985). Michael T., a transvestite who was genetically female, married Karin T. and assumed the role of her husband. Karin was subsequently donor-- inseminated so the two could have a child. Upon the child's birth, Michael signed the child's birth certificate in the space provided for "father" and assumed responsibility as the child's father. When the couple broke up, Karin attempted to nullify the marriage and argued that Michael should not be recognized legally as the child's father since Michael was not actually a man. IMAGE FORMULA 86

Instead, however, the court adopted a constructionist approach and found that, since Michael T. had signed the birth certificate and had acted as both husband and father for the duration of the family's existence, (s)he should be considered a legal parent to the child, "in view of [the] agreement to which respondent affixed her signature as father which stated that the children produced by artificial insemination were respondent's own legitimate children." This finding seemed a remarkably progressive gesture for a court of the mid-1980s-but because of the unique circumstances surrounding this case, its impact on other custody contests between same-sex parents was limited. In fact, the appellate courts would not recognize two parents of the same sex again until 1992 (Matter of Adoption of Evan).

In the previously cited case of Kathleen C. v. Lisa W., although ultimately the court ruled that Kathleen C. did not have standing to assert parental rights, elements of negotiation and arbitration of meaning were present in the judges' rationale. While it may be practically or emotionally difficult for a couple to continue to live together after a rather contentious split, as was the case here, the fact that the court considered Kathleen, a non-biological lesbian mother, to be a de facto parent prior to the end of her relationship with Lisa-the biological mother-was evidence of the inroads nontraditional families have made in asserting their legal existence. In this instance, Kathleen's self-identification as a mother was confronted with the law's ideal of the heterosexual family, and a sort of discursive compromise resulted whereby she was recognized-although fleetingly and retrospectively-as a parental figure. In another case, In re Price v. Price (1997), the court issued a different sort of compromise. Here, in response to the allegation that the children may be harmed by exposure to their lesbian mother's sex life if she retained partial custody, the court decided to restrict the sexual activity of both parents when the children were residing with them:

We make no moral judgments concerning Mother's lifestyle as it applies to her and her friends. However, we cannot disregard a parent's activity that may have an impact in the developmental stage of a child's life. We feel it is unacceptable to subject children to any course of conduct that might signify approval of anv illicit conduct whether it be between homosexuals or heterosexuals.

Instead of accepting wholesale the mother's assertion of self in which her sexual identity was unproblematic, the court located a conciliatory gesture by which it could find the mother and father to be equally suitable parents, yet still define her lesbian existence as not entirely acceptable.

The negotiation of identity in court has often been a function of the law's ability to protect a group, serving its interests and conferring on it certain rights while simultaneously infringing on its powers of self-definition and self-determination (Brown 1995; Espeland 1994). One common theme in the denial of a parent's custody rights has been the assertion that a gay or lesbian parent might taint the sexual development of his or her child and effectively "turn" the child homosexual. In the 1980s, when social scientific studies began to emerge that refuted this hypothesis, their impact on negotiations of sexual identity in court was complex. Some judges chose to ignore or deny the validity of these findings, but others accepted and even embraced this evidence in defense of gay and lesbian parents' custodial rights. In one well-cited case, S.N.E. v. R.L.B. (1985), the judges rejected the claim that contact with their mother would turn the children gay or lesbian, asserting "there is no suggestion that this [the mother's lesbianism] has or is likely to affect the child adversely. The record contains evidence showing that the child's development to date has been excellent ... and that there is no increased likelihood that a male child raised by a lesbian would be homosexual."

Yet, this judicial strategy-denial of the gayness-as-contagion hypothesis-can be a double-edged sword. Although its manifest intent and immediate effect has been to allow gay men and lesbians to retain or gain custody and adoption rights in their individual cases, the latent effect has been a reification of the belief that a homosexual identity is inherently problematic. Implicit in the defensive claim that exposure to a gay or lesbian parent will not influence a child to become gay or lesbian is the assumption that such a result is undesirable and in fact would constitute "harm" to the child.25 Thus, although S.N.E. v. R.L.B. was lauded as a victory for gay and lesbian parents' rights, it was premised on the fact that the judges were able to neutralize the perceived threat of what was considered an adverse effect on the child's sexual identity. Similarly, in Matter of Adoption of Child by J.M.G. (1993), the judges allowed a second-parent adoption by refuting the proposition that having two lesbian parents would be harmful to the child's "normal sexual development" (emphasis added). By premising the decision in this way, the court found a way to affirm the person's self-asserted parental identity but still retain discursive control over his/her sexual identity, defining it as abnormal. These decisions demonstrated the ability of the law to represent a marginal group's interests while concurrently rewriting the group's identity (Espeland 1994).

The negotiation of identity can be seen in historical trends as well. A historical look at judicial decisions over the past 50 years reveals a gradual and cumulative shaping of the positioning and portrayal of gay and lesbian parents in law. Just as the evolution IMAGE FORMULA 95

of nearly any type of law can be seen as an incremental "settling" of meaning (Phillips & Grattet 2000), the eventual shaping of certain distinct facets of sexual and familial identity is apparent in these cases. One example of this process is the discursive treatment of same-sex relationships.

Particularly in pre-1990s custody decisions, knowledge of the existence of a same-sex partner for anyone involved in a child custody suit primarily functioned as "ammunition" for the opposing party in the custody battle,26 or at the very least, as fodder for gossip. In Wolff v. Wolff (1984), for example, a father was rumored to be bisexual and to have slept with a male babysitter, an assertion that he adamantly denied. Evidence of these relationships was most likely, as in Mr. Wolff's case, to be presented in an accusatory form, putting the object of the accusation in a defensive stance. Faced with such allegations in court, a gay or lesbian parent would be forced either to reveal the relationship if one existed-with the potential that this revelation would effectively foreclose her or his chances for custody-or deny its existence. In one early case, Immerman v. Immerman (1959), the court revoked custody from a mother who was not previously identified as a lesbian because it learned that her former husband had entered her home unannounced and found her engaging in sexual activity with another woman. In another case, Henry v. Henry (1988), a lesbian mother was forced to move out of the home she shared with her partner and to relinquish contact with her as a condition of partial custody rights. Despite the gaps in time, these cases nonetheless remain anchored in the thematic context of the pejorative framing of homosexuality and same-sex relationships among parents in custody disputes

This dynamic shifted, however, in the late 1980s, as a tacit recognition of same-sex relationships became more common. By the 1990s, the relational identity of same-sex partners had evolved in the eyes of the law from a uniformly deviant status to a conditionally accepted social phenomenon. This change is evident in the contrast between the cases cited above and later cases-particularly those in which both members of a same-sex couple were litigants. In one second-parent adoption involving a lesbian couple, for example, the decision acknowledged that "courts have long construed statutes to meet the changing needs of our growing society .... [and] the concept of 'family' has expanded," and it granted the adoption (Matter of Adoption of Camilla, 1994). In J.A.L. v. E.P.H. (1996), the judges overturned a IMAGE FORMULA 99

lower court's decision that a non-biological mother did not have standing to pursue custody of her children after her relationship with the biological mother ended, and found that she stood in loco parentis based on her membership in the "nontraditional family." While such a contrast in case outcomes may have the effect of eclipsing 50 years of back-and-forth give-and-take struggle and negotiation, it should be clear that this was a tentative, gradual, and hard-fought change. Through this process, the possibility of a normalized same-sex "domestic partner" identity, in many cases analogous to a spousal identity, was created in the family courts.

Similarly, the declining impact in court of homosexuality as an accusation of impropriety marks a contemporaneous negotiation of and change in the normative evaluation of gay and lesbian parents' sexual identity over time. Many judges fall far short of embracing the homosexuality of a parent, or even treating it as a non-issue in custody matters, but they are no longer as likely to reference it as per se evidence of pathology or moral unfitness.27 This is quite obviously a change that is evident not only in family court. A myriad of legal, social, representational, and epistemological factors too numerous to list are clearly implicated in the changing perceptions and normative treatment of homosexuality. Nevertheless, a more subtle feature of this progression and its impact on the shaping of gay and lesbian parents' identity is evinced in the narratives of these cases.

In recent years, openly LGBT parents began to bring custody cases into the courtroom in far greater numbers-either forcing increased acceptance of their sexual identity in court, or as a function of it. Whereas in earlier cases courts forced gay and lesbian parents into a posture of defensiveness, denial, or apology when confronted with their sexuality, over time courts have expanded and renegotiated their range of possible responses to and representations of these litigants.

Even in the face of overtly hostile judges, parents in the 1980s and 1990s were more likely to assert their sexual identities in court and in their social lives (as evidenced in court) without apology or reticence, despite a possible negative outcome. In the Louisiana case of Scott v. Scott (1995), for example, the mother and her partner were open about their relationship, both at IMAGE FORMULA 103

home with the children and in court; consequentially, the court revoked the mother's custody and gave it to her ex-husband.

Such cases resulted in a conflicting dialogue in which, eventually, the court began to accept litigants' self-asserted sexual identities, but did so only conditionally. This acquiescence was conditioned on a parent not overtly acting out or making known his or her sexual orientation-what was often, in the judicial narratives, called "flaunting." In one case, a lesbian mother was accused of "flagrantly flaunting her relationship with [her partner] in the presence of the minor child" (Dailey v. Dailey, 1981). Conversely, in the case of M.A.B. v. R.B. (1986), a gay father was awarded custody of one of his children for not having "flaunted" his homosexuality and because, "[his] behavior has been discreet, not flamboyant." Thus, a discursive space emerged as a result of the reciprocal processes of conflict, tension, and gradual acquiescence, in which a "don't ask-don't tell" approach was taken with regard to sexual identity and a discourse of "outness," or of commitment to identity, became central.

Commitment to Identity: Prioritizing and "Passing"

As discussed above, gay and lesbian parents' commitment to their identity has been a common narrative in child custody cases. This narrative has been applied, although in different ways, to both sexual identity and parental identity. The central question concerning one's commitment to his or her sexual identity is whether this identity is an ever-present constant or whether one can (or should) effectively mute it at times. In 105 cases of the 235 that are recorded, for instance, the court criticized a gay or lesbian parent for associating with others who were openly gay or lesbian. In the judicial narrative of 29 cases of the total 235, the court assumed that parents should be able to separate their behavior from their sexual identity and not "act on" their homosexuality. This question is important in many child custody cases, in which courts have presumed that a parent's homosexuality-or children's knowledge of it-has negative effects on children. These effects, previously enumerated in part, include a conversion of the children to homosexuality (Falk 1989; Hitchens & Price 1978-79; Stacey & Biblarz 2001), the social stigma associated with having a gay parent (Cox 1994); Seidel 1994; O'Toole 1989), and a disruption of the child's normal development or gender identity (O'Toole 1989; Patterson 1995; Sedgwick 1997). The law has evinced a sometimes tacit and sometimes well-articulated expectation that it was a homosexual parent's duty to shield his or her child from the evidence and manifestations of her or his sexuality. Consequently, courts have often reprimanded parents for not closeting their sexuality, or for being too involved in the homosexual community.

Such was the reasoning of In the Matter of J.S. & C. (1974), in which a gay father who had held a position as director of the National Gay Task Force28 was denied visitation with his children, at least in part because of his gay rights activism and the prospect that his children might be exposed to these activities: "[The] defendant's total involvement with and dedication to furthering homosexuality has created an environment exposure to which in anything more than a minimal amount would be harmful to the children." Despite the passage of time, a similar rationale was used to deny custody to a gay father in Marlow v. Marlow (1998) because of his involvement in gay and lesbian church groups and with the organization Parents and Friends of Lesbian and Gay Persons (PFLAG). Alternatively, those who concealed their homosexuality and promised to continue to do so were praised in court and rewarded with custody or visitation rights-as in the otherwise quite progressive decision of M.A.B. v. R.B. (1986), where a custody award to a gay father was premised on the finding that he was not open about his sexual orientation in public or with the children.

The willingness and ability of a parent to conceal, or at least to not overtly display, her or his homosexuality is implicated in the question of how committed one is to her or his sexual identity. Sexual identity is unique in that it is not a visible trait, but must be rendered visible in a "coming out" process (Dubin 1998; Rich 1980; Troiden 1998). Evaluations of such visibility are implicit in judicial narratives of a parent's commitment to sexual identity. These narratives involve questions about how sure litigants are about their sexual orientation, how open, or "out," they are, and how integrated they are in their lifestyle, in the gay and lesbian community, and in queer politics. For instance, in Jacobson v. Jacobson (1981), the judges contended that a lesbian mother's sexual orientation was "beyond her control," but acting on it was not, and she was expected not to express her sexual identity verbally, socially, or sexually. The judges in In the Matter of J.S. & C. (1974) referred to the involvement of the father in the gay rights movement as an "obsessive preoccupation."

In other cases, parents were portrayed as flaunting their alternative sexualities by virtue of their gender non-normative appearance, behaviors, or preferences. Again, this theme was evident across time. In Newsome v. Newsome (1979), judges chastised a mother for keeping Ms. magazine in her home, as they saw this as an open sign of radical feminist lesbianism and not gender-- appropriate. In another case, Ward v. Ward (1996), judges cited the female child's poor hygiene habits and "prefer[ence] to wear men's cologne" as evidence of disturbed gender identity and IMAGE FORMULA 110

harm from having lived with her lesbian mother. The court therefore awarded custody to her father, who had previously been convicted of murdering his first wife over a custody dispute. In most of these cases there was a sense that the law was encouraging-or even demanding-that parents engage in the normalizing strategy of "passing" as heterosexual or asexual (Calavita 2000; Goffman 1963).29 Through this strategy, homosexuality could be rendered compatible with parenthood in the eyes of the law.

In the judicial narratives, litigants' commitment to their own parental identity has often been inextricably linked to their sexual identity. The courts' presumption of the incompatibility of homosexuality and parenthood, discussed above, has also manifested in their imposition of a hierarchy of identities, in which judges assumed that gay and lesbian parents valued their sexuality over their families. Although their parental identity brought them to family court, their sexual identity was constituted as a "master status," eclipsing other facets of their lives and persons (Goffman 1963; Lin 1999; Say & Kowalewski 1998).

Lin discusses the stereotyped perception of homosexuals present in family court decisions, which "illustrate[s] the reductive nature of these narratives ... [and] establishes the sexuality of lesbians and gays as their most prominent and defining characteristic" (1999:758).30 In order to gain or retain custody of their children, courts required that gay and lesbian parents prove that their sexual identities and relationships were not and would never be prioritized. In Roe v. Roe (1985), for example, a gay father was accused of "choosing his own sexual gratification" over his child because he openly affirmed his gay identity. Lesbian mothers, in particular, have persistently been put in the position of "choosing" between motherhood and lesbianism. In the cases of Hall v. Hall (1980) and Bottoms v. Bottoms (1996), for instance, the court chastised each lesbian mother for being involved in a same-sex relationship and, ultimately, removed her child from her custody. In Hall, the court found that "the interests of the children could well be subverted by [the mother's] relationship, which was clearly the chief priority in her life"; in Bottoms, the court criticized the mother because she "felt her individual rights [to live with a companion] were as important as her child's." This IMAGE FORMULA 114

theme was evident in earlier cases as well. In Towend v. Towend (1976), after extensive questioning of the mother and her partner regarding their sexual activities, a judge stated that he was "struck by the primacy that ... the two lesbians ... give to multiple organisms (sic). They mean more to them apparently than the children."

Conversely, parents who were willing to forego any same-sex romantic partnership or even social recognition of their sexual identity were found to be appropriately dedicated to the parental role. Yet even this was sometimes not enough to prove commitment to their family identity: In one case, a gay father's award of partial custody was premised on his not being too "out"-but at the same time he was called untrustworthy for having hid his homosexuality from family members (In Interest of R.E.W., 1996).

In cases involving adoptive parents, the onus of proving their appropriate commitment to parenthood as a status was doubly challenging because of the court's predisposition to recognize and privilege biological family ties. While biological parents generally can only have their parental status revoked by a judge if they are proven unfit or abusive, adoptive parents must affirmatively prove their fitness and their ability to forge a bond with the child. For gay and lesbian parents, several structural obstacles, including state laws barring adoption or foster parenthood by gay and lesbian individuals, as well as a lack of legal provisions for dual-parent adoption by same-sex couples, have often exacerbated this difficulty.31 In one case, Matter of Appeal in Pima County juvenile Action (1986), the court barred a bisexual man from adopting a child based on the criminal status of sodomy in Arizona at that time: "It would be anomalous for the state on the one hand to declare homosexual conduct unlawful and on the other create a parent after that proscribed model, in effect approving that standard, inimical to the natural family, as head of a state-created family."

However, even when these impediments did not exist at the state level, states have shown significant resistance to adoption by gay and lesbian individuals or couples. In a case where an abused and abandoned girl was placed by the Department of Social Services in the care of her adult gay brother, for instance, the court barred the adoption and removed the child based on their finding that the home was unsuitable because of her brother's homosexuality (State ex rel. Human Services Department, 1988). Moreover, the contention that gay men or lesbians would be willing or able to prioritize a child's needs over their own sociosexual activity and involvement-when they do not have a pre-ordained obligation to do so-was often the subject of significant doubt in judiIMAGE FORMULA 118

cial narratives. A 1988 decision denying an adoption petition by a gay man reflected the sentiment of many judges: "The so-called `gay lifestyle' is patently incompatible with the manifest spirit, purpose and goals of adoption" (In the Matter of Adoption of Charles B, 1988).

Conclusion

The myriad ways in which identity is represented, imposed, negotiated, produced, and reproduced in these child custody cases reveals both diversity and commonality in the sociolegal existence of gay and lesbian parents. This set of cases confirms what has been suggested repeatedly in the past: the family and its constituent membership is by no means a stable or easily definable institution. Despite courts' efforts to legally operationalize the status of "parent," this analysis substantiates the claim that such efforts to "define down" parental identity do not necessarily reflect the social and personal realities of those they are meant to describe. Nevertheless, as one can see from the cases reviewed here, the law's attempts to impose its own definition of "parent," "legal stranger," or "third party" show that courts are not immune to the reality of the diverse family forms that exist, and at key moments they have even recognized this diversity and its implications for legal parenthood and social policy.

At the same time, this analysis has revealed some common threads in the judicial treatment and framing of gay and lesbian parenthood, including elements of coercion, resistance, and negotiation. Most of the judicial narratives studied-across types of custody cases and combinations of parties-have exhibited, either explicitly or subtly, courts' efforts to legally and discursively control and inhibit alternative sexualities. In their extreme form, these efforts included revoking outright custody of a child of a gay or lesbian parent, impugning a parent's moral character, or defining his or her sexual identity as deviant. In their less overt form, judicial narratives manifested these efforts in their expressions of conditional support for gay and lesbian parenthood: custody and visitation rights conditioned on a parent's effectual disavowal or abandonment of her or his sexual identity and partnership, assurances that "harm" will not result in the form of "turning" a child gay or lesbian, and recognition of the parental rights of same-sex partners only with the proviso that this recognition does not constitute support for same-sex marriage rights. That parents were commonly rewarded for not being "too gay"-- for remaining appropriately closeted-resulted in a Pyrrhic victory analogous to psychiatrists' declassification of homosexuality as a pathology, as long as the individual otherwise exhibited "gender appropriate" behavior (Sedgwick 1997). This research also evidences efforts of a discursive negotiation-a decades-long dialogue between LGBT parents and the law-marked by assertions, concessions, and impositions. In the course of this dialogue, as is revealed in this analysis, the one-time unimaginable intersection of family, law, and alternative sexualities brings to bear a number of issues that not only impact the composition of the family but also common understandings of the origin, character, and mutability (or immutability) of sexual and familial identity. The discursive negotiations evident in these custody decisions might best be seen as power struggles, where the powers at stake are those of self-determination and self-definition.

Power, as a relational phenomenon, has been described as "transformative capacity" (Giddens, cited in Yngvesson 1993; see also Foucault 1980). When the law imposed identities on litigants-as sexual deviants, unfit parents, or legal strangers-it exercised discursive control over the shaping of their identities and senses of self; a process that began with the law's capacity to define one as in or out of the domain of legitimate legal claims-- makers-what has often been called the "gatekeeping" function in law (Baum 1978; Yngvesson 1993; Yoshino 2000). If the non-- biological mother in a lesbian parenting dyad was defined in law as a "third party" or a "legal stranger" by virtue of her lack of blood relation, she did not have legal standing to seek custody or adoption. She was then effectively removed from legal existence-reduced by the law to a non-entity in the family, despite her social and personal experience. Similarly, a sperm donor could be defined in to fatherhood despite his lack of a social role in the family. These examples demonstrate what Espeland (1994:1150) has described as "a more subtle form of power ... the potential ... to construct the particular type of subject who is allowed to have an interest."

These judicial narratives, beyond determining who may be considered an interested legal actor, demonstrate how the law can coercively construct a person's identity in ways that conflict with one's sense of self, thus denying one the "epistemic authority" to define him- or herself (Halley 1997). Such an exercise of discursive control has an equally powerful impact, whether aimed at a redefinition of one's sexual or familial identity. Constructing a gay parent as a "legal stranger" or a lesbian as a "deviant" denies the individual certain material and ideological rights. In this way, certain rights can been seen as constitutive of identity-and vice versa (Sarat & Kearns 1995). The right of a particular person or couple to marry, for example, imputes both a potential identity-"spouse"-and a set of privileges, including child custody privileges. The denial of this right also constitutes a denial of one's freedom to self-define and narrows the range of available identities a person may designate for him- or herself. Power and identity, then, are inextricably linked; as Espeland (1994:1176) notes, "[T]he categories that impinge on identity are perhaps among the more potent and least stable."

The cases analyzed here elucidate the power of the law to impose definitions and identities on the actors before it. Yet, the construction of identity in law is not a one-way process. Gay and lesbian parents have put forth self-images that are often contradictory to those imposed by the court-and in doing so, have asserted their own expressions of self-hood and power. At the very least, by virtue of their presence in the appellate courts, these parents forced the law to confront and document-if not formally recognize-the existence of alternative family forms. They impacted legal discourse by necessitating the convergence of two institutions previously thought of as divergent-alternative sexualities and family-in judicial narratives. They resisted the law's closeting tendencies by affirming their sexual identities in court, and they brought their personal realities to bear by naming themselves as parents, despite their legal exclusion from the family.

Thus, the social world of the parents and the legal world of the judicial narratives inevitably impinge on each other (Yngvesson 1993). This two-way process is the essence of the negotiation of identity-how legal identities emerge, are contested, and are eventually settled and prioritized. This process is indicative of how meaning is made in law, in both a symbolic and a literal sense.

It is important to note, however, the implications of this process and its constituent parts for individuals and for social justice more generally. The imposition of identity has not only legal and discursive but also social and psychological effects that can be harmful. By imposing deviant or sexual identities that draw on and reify harmful stereotypes, judicial narratives perpetuate not only legal but also social injustices that have long been directed at LGBT communities and individuals. They also may negatively impact a person's own self-image such that the representation imposed from without-a presumption of psychological instability, for example-begins to be internalized (MacDougall 2000).

Social injustice and harm may also result in more subtle ways than the external imposition of identity. Feminists have long warned of the dangers of "dismantling the master's house with his own tools" (Bar On 1993). The legal contests and negotiations of meaning that lesbian and gay parents engage in to gain custody rights and to define themselves as parents may also have unintended consequences for the framing of LGBT identity more generally. Proving oneself as a "fit parent" in court, for instance, may involve setting oneself apart from other gay men and lesbians-those who "flaunt" their sexuality or live with their partners. Such a gesture may win an individual parent custody rights and defy stereotypes imputed to him or her, while simultaneously marking a distinction between "good" and "deserving" lesbian and gay parents and those who are "bad" or "undeserving" because of the way they choose to embody or commit to their sexual identities (Robson 1995). The result, what Robson (1995) calls the "legal domestication of lesbian [and gay] existence," may reify the gendered stereotypes that are ultimately invoked to define gay and lesbian parents as problematic. Thus, we are reminded that legal rights have the potential to enable one's expression of his or her own identity, but at other times they may facilitate the regulation and domination of this or similar identities (Brown 1995; Espeland 1994).

An understanding of the hegemonic potential of identity construction in court, however, is only one component of this analysis. In identifying and articulating the complex processes involved in the negotiation of identity in family court, a new constitutive terrain is revealed in which identities previously thought to be mutually exclusive-"gay/lesbian" and "parent"-are rendered compatible. This is accomplished through a variety of mechanisms of discursive compromise: the construction of an "identity hierarchy" where priorities and commitments to differing statuses are articulated; "passing" gestures, which require the strategic muting of certain identity traits; or the provisional and temporary conferrals of an otherwise unattainable status. Through these processes, gay and lesbian parents are eventually constituted as legally viable subjects rather than as cultural contradictions. The same mechanisms of discursive compromise might be found at work in the merging of other seemingly contradictory identities-"Islamic feminist" (Kasravi 2001), "moral criminal" (Katz 1988), or "elite deviant" (Simon 1999).

The negotiation of identity in law is both inevitable and productive. While the power of law to impose labels and coercively shape identity is explicit in the experiences of gay and lesbian parents in family court, so are their efforts to self-identify, to make their existence and their families visible, and to force the law to confront new social realities. The discursive interplay among these forces, regardless of their independent successes and failures, results in multiple iterations and reformations of existing concepts, institutions, and identities, and allows for the emergence of new meanings, new identities, and new sites of contest and compromise. Judith Stacey, in her powerful book In the Name of the Family, discusses the transformative potential of queer families as something that could benefit the entire institution of family-if we as a society could only "get used to it" (1996:139). The negotiation of identity in law, comprised of the struggles, tensions, and creativity of sociolegal actors, is at once a potent and subtle articulation of power, and a gradual procession toward getting used to it.