New York University Law Review
79 N.Y.U.L. Rev. 1134
LENGTH: 61398 words
ARTICLE: DISCRIMINATION BY PROXY AND THE FUTURE OF TITLE VII
NAME: Camille Gear Rich *
* Copyright (c) 2004 by Camille Gear Rich. Many thanks to the numerous people who provided useful critique and support during the writing of this Article. My sincerest thanks to Dean Anthony Kronman of the Yale Law School for his invaluable advice and friendship, and to the Yale Law School for its financial support during the germinal phase of this Article. Also thanks to Professors Vicki Shultz, Reva Siegel, and Jack Balkin for their inspiration and guidance, and special thanks to my husband Stephen M. Rich for the loving debates that no doubt enriched my efforts. This Article is dedicated to the memory of Hayley S. Thomas.
HIGHLIGHT: Courts interpreting Title VII have long treated race and ethnicity as biological, morphological concepts and discrimination as a reaction to a set of biologically fixed traits. Meanwhile, they have rejected claims concerning discrimination based on voluntarily chosen physical traits or "performed" behaviors and that communicate racial or ethnic identity. Yet race and ethnicity are effectively produced - that is, they do not exist until one is socially acknowledged as possessing socially coded racial or ethnic markers, whether they are fixed physical features, voluntary appearance choices, or behaviors. This Article argues that it is error to distinguish between Title VII cases concerning morphological as opposed to voluntary racially or ethnically marked features, as the discriminator's motives and the effects of her behavior are the same. Moreover, the morphological model of race/ethnicity is fundamentally contradicted by contemporary biological and sociological studies on race, discrimination studies, and identity performance theories, which indicate that individuals actively work to "perform" racial and ethnic status regardless of, and sometimes in spite of, their morphological traits. Drawing on these studies, this Article shows that courts must hear discrimination claims based on voluntary features if they are to provide a more credible analysis of modern forms of discrimination.
After almost four decades of enforcement, one would expect that workers would better understand the scope of Title VII's race and national origin discrimination protections. However, all too often, workers are surprised to discover that the voluntary behaviors that they perceive to be an essential part of their racial or ethnic identity are treated by courts as a marginal concern, 2 This mistake in perception often has tragic consequences.
The case of McBride v. Lawstaf 3 is a classic example. In that case, Corrine McBride, an employee at the Lawstaf temporary employment agency, challenged the agency's policy of not referring persons who wore all-braided hairstyles to temporary jobs. McBride [*1137] informed Lawstaf that she believed that its grooming policy served as a cover for discrimination against blacks and, unless it abandoned the policy, she would report the company to the Equal Employment Opportunity Commission (EEOC). 4 Lawstaf instead terminated McBride and subsequently prevailed on the Title VII retaliatory discharge claim McBride brought against the company. 5
To the legal scholar, this result is unsurprising, as it has long been established that Title VII does not prohibit discrimination based on "voluntary" or "performed" aspects of racial or ethnic identity. Therefore, Lawstaf legally could institute a policy rejecting applicants who wear braids and terminate McBride for challenging this policy. 6 A layperson who values these kinds of practices would be dismayed by the court's ruling in the McBride case. In her mind, it seems clear that Lawstaf's rule prohibiting all-braided hairstyles functionally screens out large numbers of blacks from the agency's employment pool and, on its face, explicitly articulates animosity towards blacks. It is "discrimination by proxy." To the layperson, it seems wholly irrelevant that the conduct at issue is "voluntary." 7
[*1138] Sometimes a statute's failure to address a layperson's understanding of a particular injury raises little concern, as the statute reflects Congress's measured consideration of the costs and benefits that are appropriate in addressing a particular injury. 8 In cases like McBride, however, the law's disconnect with the layperson's understanding of her injury is more disturbing because the decision turns on a judicially constructed definition of race that has never faced congressional scrutiny. Indeed, a review of the legislative history of Title VII shows that Congress has never indicated that race or national origin should be defined under the statute in a manner that categorically bars all claims concerning voluntary aspects of racial or ethnic identity. 9 These judicially constructed definitions are also a source of concern because they contradict prevailing sociological scholarship and biological studies on race and ethnicity 10 and fundamentally contradict [*1139] prevailing scholarship on the cognitive processes that inform discrimination. 11
In light of these concerns, this Article argues that courts should abandon the current definitions of race and ethnicity under Title VII that exempt from protection "voluntary" aspects of racial and ethnic identities - what I call "race/ethnicity performance." Race/ethnicity performance is defined as any behavior or voluntarily displayed attribute which, by accident or design, communicates racial or ethnic identity or status. 12 It covers racially and ethnically coded indicia such as [*1140] hairstyles and other aesthetic choices, as well as dialect, language choice, and accent. 13
In order to apprehend fully the need for race/ethnicity performance discrimination protections, one first must understand how the judiciary's focus on so called "biological" race and ethnicity allows employers to engage in race or national origin-based discrimination without triggering Title VII's protections. Courts have held that an employer only will be held liable under Title VII when she sanctions an employee because the employee involuntarily displays a biological, visible or palpable characteristic associated with a disfavored racial or ethnic group. 14 The unspoken corollary proposition is that employers therefore have virtually unfettered authority to select or penalize workers based on appearance and behavior, provided that it cannot be shown that the employer's preferences are linked to some involuntary biological and visible race/ethnicity-associated feature. 15 This regime imposes two costs on minority employees. First, it permits employers, through grooming codes or other rules, to discriminate against workers by proxy, disproportionately screening out or penalizing workers from disfavored racial/ethnic groups based on aesthetics 16 or [*1141] behaviors statistically correlated with these groups. 17 Second, and equally important, it devalues the psychological and dignitary interests that employees have in race/ethnicity performance. 18 Under this regime, workers engaged in race/ethnicity performance have no protection from workplace rules that are intended by design to communicate and reinforce their employer's antipathy for their individual racial/ethnic group.
The question thus becomes: Why do we continue to employ a regime that allows discrimination based on race/ethnicity performance? Why do we allow employers to discriminate based on race or ethnicity as long as their behavior does not implicate so-called biological or immutable characteristics? This Article shows that the courts' focus on the biological/voluntary distinction is fundamentally unprincipled and illogical, as the discriminatory animus in cases involving so-called biological racial or ethnic traits and voluntary, performed racial or ethnic traits operates identically. 19 In these two kinds of cases, the employer discriminates against the employee because she has triggered a cultural code associated with a low-status race or ethnic group. In both types of cases, the employer sanctions the employee [*1142] because of a fear of racial or ethnic presence: The employee's appearance reminds the employer of the employee's minority status and her potential to disrupt the current cultural hegemony of the workplace. 20
This Article suggests that the reason for the courts' emphasis on biology is more a consequence of American history, culture, and politics rather than any logical or scientific proposition. 21 It also notes that the discrimination paradigm currently in use is informed by an assimilationist world view that is the product of a specific historical moment, during which Americans were encouraged to believe that status equality merely required us to tolerate biological, visible markers of race and ethnicity, but that we could and should require, in less formal ways, that voluntary racial and ethnic markers be surrendered. 22 This paradigm is also based on a paternalistic logic which posits that only the truly assimilated will be able to participate fully in civil life, and that employers can be trusted to facilitate that process. However, assimilationist and paternalist approaches to discrimination issues have been challenged deeply in contemporary society; as such, an antidiscrimination regime that uncritically adopts this kind of approach seems disturbingly outdated, 23 and it will be blind to a large swath of behavior we now recognize as discrimination.
Therefore, this Article argues that there is an urgent need to redefine Title VII's definition of race and ethnicity to include both biological, visible racial/ethnic features and performed features associated with racial and ethnic identity. Part I of the Article uses Gordon Allport's seminal work, The Nature of Prejudice, 24 to describe the cognitive processes that undergird discrimination. I refer to these processes, collectively, as racial or ethnic ascription. Section A explores morphology-based ascription (the primary means by which subjects are racialized and ethnically categorized in society). It then notes that morphology-based ascription does not operate solely as a consequence of biology, revealing that the racial and ethnic assignment process is, in fact, an enterprise informed by socially derived [*1143] knowledge. Section B explores performance-based ascription and shows that the same ascription process that informs discrimination triggered by morphology also triggers discrimination in cases involving "voluntary" race/ethnicity-associated behavior. Section C further argues that, since the same cognitive process is at issue in cases involving involuntary morphological racial or ethnic traits and in cases involving "voluntary" race/ethnicity performance, Title VII should offer racialized and ethnic subjects protection for both of these triggers for discrimination.
Part II builds on the previous discussion about the cognitive logic of discrimination and examines the sociological justifications for protecting voluntary race/ethnicity-associated behavior under Title VII. It relies on identity performance theory and group psychology models to explain why Title VII should prohibit discrimination based on voluntary aspects of racial and ethnic identity. Section A uses Judith Butler's "performativity" model 25 to explain the psychological and dignitary value that race/ethnicity-associated behavior has for employees. Section B draws insights from group psychology to explain why race/ethnicity performance tends to disturb members of outgroups. 26 The Section discusses both group position theory and aversive racism to describe the dynamics of contemporary discrimination. Section C explores how these theories assist us in understanding the competing stakes at issue in race/ethnicity performance cases, including the employer's and the employee's interest in freedom of expression, the employee's dignitary concerns, and society's interest in protecting against conditions that encourage racial and ethnic segregation and stratification.
Part III examines the rhetorical and doctrinal tools that courts have developed to justify the rule exempting race/ethnicity performance from Title VII's protection. Section A explains how courts in the seminal cases on this issue severed race and national origin into two parts, creating a distinction between involuntary, biological features and voluntary aspects of each protected status or identity category. Section B shows that the rationales supplied for the distinctions between biological and voluntary features resulted from courts' use of undertheorized analogies between the interests raised in gender performance cases and the distinct interests at stake in race and ethnic identity performance cases. 27 Section B then shows that the [*1144] race/ethnicity performance framework allows the development of narratives that better account for the various competing interests in race and national origin discrimination cases: the worker's autonomy, dignity, and social justice concerns; the employer's expressive and financial concerns; and society's interest in maximizing equality and opportunities for interracial and cross-ethnic interaction. Section C explores the repercussions triggered by certain ill-conceived judicial opinions equating gender and race/ethnicity performance and demonstrates how, with more thoroughly considered parallels between these components of identity, courts could yield more productive insights about identity performance generally. Section C argues that our current approach relies on flat parallels that do not take into account the different weight autonomy, dignity, and social stratification concerns play in race and national origin discrimination cases as compared with gender cases.
Finally, Part IV addresses the legal, economic, and political concerns most likely to be invoked in opposition to the race/ethnicity performance model. Section A examines legal challenges, including statutory construction arguments, special rights claims, and concerns about enhancing judges' power to codify racial and ethnic identity. Section B addresses market-based challenges, namely the concern that race/ethnicity performance protections will interfere with employers' ability to market their products, discipline their workers, and increase their potential liability for workplace discrimination. Section C addresses the primary political concern about the model, namely that race/ethnicity performance protections encourage separatist attitudes in an era in which Americans need to be more focused on becoming a more cohesive community. 28 Part IV demonstrates that rather than encouraging divisiveness, race/ethnicity performance protections help ensure that the workplace is a space for mutual recognition and respect regardless of race or ethnicity. These protections create the context for a free exchange of race/ethnicity-associated traditions, the [*1145] necessary precursor to creating a diverse but cohesive American community.
One of These Things is Not Like the Other: Learning to Recognize Racial and Ethnic Difference
Section A of this Part begins with a description of morphology-based racial and ethnic ascription. Next, it deconstructs the morphology-based racial/ethnic ascription process, showing that this kind of ascription is not a natural reaction to objective biological facts but, rather, is a learned response based on socially derived knowledge about how to identify particular races and ethnic groups. Section B explores performance-based ascription, demonstrating its similarities to the morphological ascription process, as well as noting certain differences which further establish the need for antidiscrimination protections for performative features of protected identities. Section C explores the reasons courts likely have been loath to recognize performance-based discrimination, despite its obvious parallels to morphology-based discrimination and suggests ways to address these concerns. Specifically, this Section shows that judges' failure to account for the role politics and culture play in triggering morphology-based racial and ethnic ascription has made them unnecessarily wary about proscribing discrimination that is triggered by race/ethnicity-associated behaviors - what I call performance-based racial and ethnic ascription.
A. Morphology-Based Ascription
1. A Primer on Morphology-Based Racial and Ethnic Ascription
The cognitive process that people rely on to ascertain the race or ethnicity of another person is referred to here as racial or ethnic ascription. Racial/ethnic ascription works by triggering racial or ethnic associations when one sees another person display certain traits. 29 The most common form of racial and ethnic ascription is morphology-based ascription. This kind of ascription occurs when a subject interprets another person's visible, physical features to correlate with a set of features she identifies with a certain race or ethnic group. [*1146] These features include skin color, hair texture, and nose or eye shape. 30 The subject learns to correlate these traits with one of three or four racial categories and, in some cases, an ethnic subgroup. 31 Stated alternatively, morphology-based racial/ethnic ascription operates by those "common sense" cognitive rules that cause a person to conclude automatically that chocolate skin tones signify that a person is African, that olive skin tones indicate Latin origin, that certain eye shapes are Asian or Caucasian, or that particular nose structures are Caucasian or African. 32 This cognitive process is such a well-entrenched part of social interaction that it typically functions unnoticed. This is evidenced by the fact that most Americans believe that they can, upon review of a person's physical traits, easily identify the person's race or ethnicity. 33
Most people assume that race/ethnicity-associated morphology provides an objective, unchanging basis for ascription. But as a person's range of human interaction increases, she is confronted with evidence that demonstrates the limits and inconstancy of morphological markers, 34 and her ability to decisively link specific morphological [*1147] features with particular races or ethnic groups grows less sure. For example, an individual may encounter persons whose morphological features simply are indeterminate 35 or whose morphology could associate them with any one of a number of races or ethnic groups. 36
Additionally, with more experience, an individual learns that the paradigmatic traits of her own race or ethnic group are shared by other groups. For example, a black worker will discover that the range of brown skin tones she has learned to interpret as indicating that a person is black also might signal an Asian background, in the form of Pakistani or Indian heritage, a Latino background for Cape Verdean or Dominican persons, or a Middle Eastern heritage. Alternatively, she learns that the paradigmatic morphology associated with other races or ethnic groups appears in her own racial or ethnic group. For example, the Irish American worker identifying as white learns that Jews, Italians, and Greeks, who also identify as white, have nose shapes and hair textures that sometimes correlate with persons of African heritage. Collectively, these experiences counsel that there are no definitive morphological markers for a particular race or ethnicity. They also demonstrate that the racial and ethnic constructs that each person employs in daily life are shaped by her individual life experience.
Ironically, although we are surrounded by all the information necessary to deconstruct race and ethnicity, most of us selectively process this evidence and maintain the view that there are stable, morphologically distinct races and ethnic groups. Indeed, some evidence even suggests that once an individual learns to correlate certain morphology with racial and ethnic difference, she is even more inclined to interpret ambiguous physical features as being correlated with established racial and ethnic categories. 37 Despite our ingrained reliance [*1148] on these constructs, most people can recall some experience in which morphology has failed to provide a clear basis for identifying another person's race or ethnicity, even if we ultimately decided that the person belonged within an established racial or ethnic group. Yet even after these experiences showing their failures and limitations, Americans continue to use racial and ethnic constructs to assist them in social interactions. Given their obdurate, central role in American life, one must assume that race and ethnicity provide some social value; otherwise, these constructs would not endure. The question is: What helpful role do these constructs play in social life?
The first and easiest explanation for the resilience of racial and ethnic constructs is that they enable people to draw quick, generalizable assumptions about the tastes, interests, or beliefs of an individual and whether she will be amenable to certain kinds of contact. 38 Sometimes these generalizations are benign. Although the individual who is racially or ethnically categorized may be offended by the attitudes attributed to her, the consequences of the mistake are minimal, as they can be rebutted through further interaction. Other morphology-based ascriptive attributions pose more serious problems, such as racist or ethnically-biased generalizations about an individual's physical or intellectual potential. These capacity-based generalizations are more disturbing because, first, they tend to decrease opportunities for interaction and, second, they serve as a basis to deny individuals social benefits. 39 Antidiscrimination law is concerned only with this second, capacity-based group of race and ethnicity-based generalizations.
The second reason for our reliance on racial and ethnic categories stems from our intellectual and legal history. The belief in stable, morphologically distinct racial groups is a remnant of biological theories of race and ethnicity that were hegemonic in the scientific community in America until the early twentieth century. 40 During that [*1149] period, eugenics provided a powerful scientific rationale for creating distinctions between the races, specifically between African slaves and white slaveholders, 41 as well as between whites eligible for citizenship and non-white "others" granted contingent license to live and work in the United States. 42 Eugenics also grounded ethnic difference; ethnic groupings appeared and disappeared as these distinctions were used as rationales to explain the allegedly different capacities of subgroups within racial categories. 43 Although eugenics has receded from center stage, our history has been indelibly shaped and marked by racial and ethnic constructs; therefore, we continue to refer to these constructs to understand our history and the connection of past struggles to contemporary disputes.
This Article argues that it is critical to tap into the insights provided by earlier morphological debates in order to have a complete understanding of contemporary racial and ethnic constructs and contemporary discrimination. Indeed, far from being mere historical artifacts, these earlier fights about morphology hold useful lessons for understanding some of the trends in contemporary debates about race, ethnicity, and discrimination. The following subsection explores some of these lessons, including how past fights about racial and ethnic morphology can provide a basis for theorizing about how morphological fights will play out in the future and their effect on the distribution of social benefits.
2. Understanding the Politics of Morphology-Based Ascription
Because race and ethnicity have played such a central role in American political and social life, the manipulation of morphological race and ethnic group constructs during each historical period provides [*1150] a snapshot of the social tensions of the era. Additionally, these early cases demonstrate courts' recognition of the fluid nature of racial and ethnic morphological descriptions and their ability to change the rules of the game when established morphologic descriptions no longer served their intended purposes. For example, a review of racial determination cases from the eighteenth and nineteenth centuries shows that morphological race and ethnic group descriptions often shifted to accommodate the political needs of the time. 44 Specifically, morphological descriptions of racial groups were continually manipulated to maintain the separate tiers of rights accorded to white citizens, black slaves, and other immigrants. 45 Ethnicity within racial groups also was variously highlighted or ignored as whites determined how to market slaves 46 and how to make arguments about distinguishing between waves of "non-white" immigrants. 47 Legislators and judges during this period acted as though they simply were translating eugenics' "scientific" premises into rational, bright line legal rules; however, a review of these cases shows that they only selectively referred to scientific materials and relied more heavily on common sense and social exigencies to define the races.
Court decisions of this era sometimes highlight the contemporaneous social tensions in stark terms. For example, in the nineteenth century, courts policing the divide between black and white persons faced serious challenges in enforcing morphology-based race distinctions. Much of the problem stemmed from the fact that the paradigmatic physical traits associated with races failed to faithfully track bloodlines and genes. 48 Consequently, legislators and courts applying slave codes, hypodescent rules, 49 and immigration laws were forced to treat biracial and multiracial persons who displayed paradigmatic white morphology as members of subordinated minority groups. 50 [*1151] Also, as various state legislatures and courts adopted different standards of purity for their bloodline rules, the range of features legally recognized as white differed from case to case and jurisdiction to jurisdiction. 51
In some cases, the court would openly reveal its frustration with the inconstancy of these morphological paradigms, particularly when presented with skin color claims. Indeed, during the mid-nineteenth century, white skin served as the primary morphological feature distinguishing whites from blacks. However, courts recognized that this feature could be misleading, as this hallmark trait also appeared within various other racial and ethnic categories. It was clear to them that white skin was not a feature exclusive to European Americans. The Michigan Supreme Court commented on this problem in People v. Dean, recognizing that "there are white men as dark as mulattoes, and ... pure blooded Albino Africans as white as the whitest Saxons." 52 In light of this fact, the court found it difficult to assert that white skin should be treated as a presumptive marker of white identity. Because of these concerns, it resorted to a default position, using blood line evidence to definitively resolve the question of the [*1152] plaintiff's racial standing. 53 Dean is an important case for our purposes because the court explicitly acknowledges that morphology-based descriptions of the races provide no clear answers and are based on contingent, selective interpretations of physical evidence.
The symbolic ambiguity of white skin became more of a problem for courts as the number of mulatto slaves possessing this culturally-loaded morphological feature increased. The social and political imperatives of the era were clear: White slave owners who had sexual relationships with black slaves needed ways to ensure that their mixed race offspring remained a part of the slave class. 54 This well-established rule enabled them to easily increase their property holdings and also to avert potential family conflict. Cognizant of this pressure, the Virginia Supreme Court maneuvered in Hudgins v. Wrights 55 to increase the importance of physical characteristics other than white skin in race determination cases. The Hudgins court explained that "nature has stampt upon the African and his descendents two characteristic marks, besides the difference of complexion, which often remain visible long after the characteristic distinction of colour either disappears or becomes doubtful; a flat nose and woolly head of hair." 56 With this rhetorical move, the Hudgins court reordered the hierarchy of morphological traits used to denote "whiteness" and, in doing so, defeated hopeful mixed race plaintiffs' claims that white skin would be treated as a presumptive marker of "white" identity.
The Hudgins decision is notable for making it clear that courts could amend or retract morphological rules for identifying the races if and when the occasion required. As in Hudgins, courts that wanted to deny multiracial plaintiffs the privileges of whiteness could simply choose to emphasize additional morphological bases upon which to deny mixed race plaintiffs' claims. The Hudgins decision is also notable in that it attempts to naturalize and sanitize the obvious political component of its decision by citing the "Laws of Nature" to justify [*1153] its decision to increase the morphological burden of proof for mixed race litigants.
America's early immigration cases provide additional illustration of how courts in our early history openly acknowledged the fluid nature of morphological race definitions and shaped them to deal with the political imperatives of the moment. Early citizenship litigation focused on maintaining whiteness as an exclusive, limited category because only those persons recognized as white were granted the full rights of citizenship. 57 Although a number of scientific authorities provided guidance on how to categorize persons within racial groupings, a review of citizenship litigation between 1790 and 1870 shows that the courts only selectively referred to these materials. Instead, they manipulated the evidence and interpreted the requirements of whiteness with the singular goal of complying with social expectations.
For example, in In re Ah Yup, 59 the court was asked to determine if Ah Yup, a Chinese man, legally should be recognized as white. The court here, again, admits the inconstancy of morphological racial paradigms, explaining that the term "white person" refers to an "indefinite description of a class of persons," as "none can be said to be literally white, and those called white may be found of every shade from the lightest blonde to the most swarthy brunette." 60 Consequently, the court concluded that the term "white person" should be interpreted in a manner consistent with the "well settled meaning in common popular speech," which indicated that it "would intend a person of the Caucasian race." 61 Here, again, one sees that the significance of white skin (arguably the primary social marker of whiteness) could be [*1154] cabined if recognition of the feature would defy social expectations about members of this racial category.
Similarly, in Ex Parte Shahid, 62 the court expressed doubt about the validity of morphological inquiry in race determination cases. In Shahid, the court was asked to determine whether a "walnut"-colored Syrian man should be legally recognized as a "white person." 63 Characterizing the morphological inquiry as a thankless task, the court questioned whether it was possible to determine "what degree of colorization ... constituted a white person as against a colored person" and indicated that it was loathe to take "responsibility by ocular inspection of determining the shades of different colorization where the dividing line [stood] between white and colored." 64 After an exegesis on various cases regarding the proper dividing lines between the races, the court rejected these authorities, opting instead to resolve the question of the litigant's racial status based on its own statutory interpretation, which in this case was simply the "common" or "natural" definition of "white person" for the Congress that authored the citizenship statutes. 65 The court ruled that because the Congress that created the laws in question presumed that the term "white" only referred to white persons of European origin, Shahid's claim must be denied. 66
These examples from the racial determination cases between the eighteenth and nineteenth centuries show that even when our laws were expressly conditioned on the view that there were stable, morphologically distinct races, courts manipulated morphological paradigms, scientific data, and blood line evidence with the ultimate goal of producing race definitions that matched the social expectations of their time. 67 These cases confirm that the morphological inquiry about race, and even ethnicity, has always been fluid and always has [*1155] been informed by social expectations. 68 They also reveal that courts, more than any other branch of government, often were expected to create and rationalize the so-called "objective" morphological definitions that were articulated in each period, with minimal guidance from the legislature. 69
Michael Omi and Howard Winant's theory of "racial formation" identifies the social pressures that caused courts to produce these changing morphological race definitions and explains the role courts played in these definitions' continuing evolution. 70 Omi and Winant posit that any given definition of race (or, by extension, ethnicity) is the product of a "racial formation [which] refers to the process by which social, economic and political forces determine the content and importance of racial [or ethnic] categories." 71 Winant further explains that the process for creating definitions for races and ethnic groups is a continuing social enterprise, which he calls "racial signification." 72 This process is "inherently discursive, ... variable, conflictual, and contested at every level of society." 73 Consequently, we should expect fluctuations in these concepts' definitions, as the meaning of race and ethnicity will vary in different societies and over historical time. 74 Also, in order for a racial or ethnic construct to gain hold, he explains, it must be responsive to and reflective of the political and social exigencies of the period. 75 Therefore, we should expect to see several competing definitions of racial and ethnic groups being offered by different segments of society at any particular time. These various definitions can come from many different segments of society, including "elites, popular movements, state agencies, cultural and religious organizations, and intellectuals of all types [who] develop racial projects, which interpret and reinterpret the meaning of race." 76
[*1156] Omi and Winant's work suggests that the seeming inconsistency that judges displayed in the previously discussed racial determination cases should be viewed as a roadmap for understanding the social pressures that informed discussions of race during the period in which these decisions were rendered. Their analysis helps us understand that judges, as legitimators of the existing social order, were put in the position of attempting to unify and explain contested and conflicting racial and ethnic group definitions. In the process, they revealed their own social and political affiliations.
A number of legal historians have already documented how American racial formations shaped morphological race definitions during the eighteenth, nineteenth, and early twentieth centuries; however, more attention should be devoted to examining how current political tensions inform contemporary debates about morphological racial and ethnic definitions. Although prior legal debates about race concerned the morphology necessary to establish one's whiteness and gain the privileges attached to that status, 77 current debates center on policing the boundaries of minority identities in order to ensure that the antidiscrimination benefits extended to these groups are not squandered. 78 For example, we still can see these definitional problems concerning race/ethnicity-associated morphology in Title VII claims brought by racially ambiguous plaintiffs, 79 as well as debates about the rights of Native American tribes. 80 These fights also arise more informally when individuals raise "authenticity" challenges [*1157] to persons who seek to claim benefits reserved for minority groups. 81 These contemporary political disputes confirm that concerns about morphological definitions of racial and ethnic categories are not limited to any particular historical period.
Several preliminary insights about contemporary disputes about race/ethnicity-associated morphology can be drawn based on the trends observed in earlier morphology debates. These trends collectively suggest that there will be more, rather than less, strain on morphological definitions in the future. First, we must recognize that in the absence of any legislative directive to do otherwise, courts will continue to try to resolve racial determination cases based on morphological evidence. Second, as legal and material benefits accrue to a particular morphologically defined race or ethnic group, there will be an increase in the number of persons willing to claim that racial or ethnic identity. Third, higher numbers of interracial unions will lead to higher numbers of interracial offspring, and more challenges about whether a person seeking to claim certain benefits should actually be recognized as a "minority." These trends indicate that the morphology debates are far from over and emphasize the importance of continuing to track fights over race/ethnicity-associated morphology.
Despite the inherent instability of morphological racial/ethnic paradigms, most Americans have no reservations about disclosing that they can easily identify the morphology they associate with specific races or ethnic groups, 82 although they feel less comfortable revealing that they ascribe certain cultural, behavioral, or aesthetic preferences with certain races or ethnic groups. 83 Both sets of frameworks, however, play a key role in identifying the racial and ethnic identity of others. The use of these less discussed "performative" frameworks and the discomfort associated with revealing them are further examined in the Section that follows.
B. Race/Ethnicity Performance-Based Ascription
In the same way that people develop a lexicon of morphological traits that they use to identify a person's race or ethnicity, they also develop a taxonomy of voluntary behaviors that signify race and ethnicity. That is, in the same way that we associate skin color, eye color, or nose shape with particular races or ethnic groups, we also maintain beliefs about the dialects, aesthetics, and mannerisms that signal one's race or ethnic status. For example, many people associate a certain accent with Italian Americans and, based on this voluntary marker, 84 can identify the race or ethnicity of the speaker without difficulty. Similarly, many people recognize that all-braided hairstyles and dreadlocks are part of the cultural legacy of blacks and may assume that a person wearing one of these hairstyles is African American or of West Indian ancestry. These generalizations extend to clothing as well. Saris, bindis and pashminas are associated with Southeast Asian women, despite the fact that these items have been remarketed by the American fashion industry for the women generally. 85
However, similar to morphology-based ascription, once a person has acquired a broader cultural experience, she has the ability to question the associations she makes between particular practices and certain races or ethnic groups. For example, a person may learn that what she perceives to be an Italian American accent is not shared by most Italian Americans, but is actually an accent generally associated with persons from a certain area of Brooklyn. Similarly, a person may associate a certain dialect with blacks, only to discover that the speech pattern is actually a dialect that is prevalent generally among working class Southerners, regardless of race. Even cultural race/ethnicity-associated behavior is susceptible to this kind of re-contextualization. For example, a person may associate enjoyment of rap music with black urban youth, based on the disproportionate number of black artists, but find this belief challenged when she is exposed to research [*1159] studies that show that this music frequently is purchased in large numbers by whites and other races. 86 Importantly, however, similar to the treatment of race/ethnicity-associated morphology, individuals will tend to continue to believe in the correlation of racially and ethnically marked practices with their perceived community of origin, despite the receipt of contradictory information.
Additionally, like morphology-based ascription, only some performance-based ascription constitutes discrimination. Again, as with morphological features, there is nothing inherently wrong with a person interpreting voluntary features, such as hairstyle or accent, as relevant evidence of another person's race or ethnic identity. Moreover, there is little hope or clear justification for training people not to do so. A problem only develops when this ascriptive process is combined with an impulse to devalue a person because she appears to belong to a particular race or ethnic group. By way of example, there is nothing inherently problematic with assuming that a morphologically ambiguous individual with a certain accent is of Pakistani or Indian heritage. The problem results when these race/ethnicity-associated voluntary behaviors cause a person to be subject to stigma. If an employer assumes that a person with a Pakistani or Indian accent will not be as intellectually sharp as a white employee, Title VII should be concerned with this situation.
There are certain clear parallels between the morphological and performative ascription processes; however, to truly understand the particularities of performance-based discrimination, it is necessary to focus on the dynamics that distinguish it from morphology-based discrimination as well.
First, the race/ethnicity performance model posited here provides that certain racially or ethnically marked behaviors, traits, and styles will continue to be associated with their communities of origin even while, in varying degrees, they may be recognized and appreciated by outgroup members. 87 The important point, however, which has not previously been acknowledged, is that when voluntary practices are circulated beyond the community that engendered them, they often remain inflected by the stigma associated with their creators. This insight brings logical clarity to some of the more unusual discrimination [*1160] scenarios that have developed in recent years. Consider, for example, that we intuitively understand that a person from a low-status race or ethnic group can trigger sanctions when she engages in behavior associated with her race or ethnic group. The race performance analysis, however, suggests that members of high status racial groups may also be sanctioned for engaging in these same low-status ethnically or racially marked practices for discriminatory reasons. Importantly, this is not true of low-status morphological racial traits which, when present in a person who is otherwise apparently a member of a high status racial category, tend to be rendered invisible. Low-status performative traits (whether behavioral or aesthetic) prove much harder to ignore and more often will serve as a basis for sanction.
These propositions are made clearer in the following example. Consider a scenario in which a white Anglo male, who is a fan of Bob Marley, decides to assemble his blond hair into dreadlocks to demonstrate his affinity and respect for the musician. When he appears at work the following day wearing the hairstyle, his coworkers as well as his employer sanction him, telling him he looks too ethnic, or indicating that the hairstyle is unhygienic and unappealing. The negative comments made about his hairstyle are functionally identical to those made to African Americans wearing the hairstyle, and the displeasure of his racial group is no less than it would be if he were African American. Indeed, their hostility may be even more acute if they perceive this change in style to indicate the young man's larger, more general acceptance or celebration of West Indian culture. In contrast, if the same blond white male inherited a broad nose from an African ancestor, this feature is far more likely to be ignored, and rendered invisible by his other morphological features. Racially coded morphological features have far less disruptive symbolic value than any racially coded practices one engages in.
Once we acknowledge the special stigma attached to race/ethnicity-associated practices, this understanding opens a new field of antidiscrimination problems. It suggests that battles about race/ethnicity performance are not "merely" about personal freedom; rather, they are cultural group status contests. When read as part of a larger "status war," the discrimination in cases involving whites who adopt low-status race/ethnicity performance behaviors by members of their own racial group must be understood as an effort to maintain the discriminator's ethnic or racial group's standing in the racial/ethnic hierarchy of the workplace and prevent the perceived lower-status [*1161] group's practices from being considered of equal value and disseminated. 88
Stated more simply, higher-status outgroup members who engage in ethnically and racially marked practices associated with low-status groups will be sanctioned for that behavior because it poses an even greater threat of cultural transmission than when performed by its perceived progenitors. In the traditional discrimination case, the discriminator targets the minority individual engaging in race/ethnicity performance out of concern that the minority individual continues to identify with or celebrate her minority identity. In the non-traditional case, the discriminator targets the non-minority individual engaged in low-status ethnically and racially marked practices because it demonstrates that a subordinate culture is being transmitted to higher-status outgroups. In both cases, however, the discriminator holds the same negative racial/ethnic animus: Her devaluing beliefs about the low social value of the community that engendered a particular practice trigger her to sanction persons engaged in this practice.
The second feature that distinguishes race/ethnicity performance from the morphology cases is the issue of agency; some distinction must be made on this basis to perform a precise analysis. 89 To that end, I distinguish between discrimination cases that involve passive race/ethnicity performance (cases in which the subject is unaware that her practices function as racial/ethnic signifiers) and discrimination based on active race/ethnicity performance (cases in which the subject consciously has chosen to engage in racially or ethnically marked practices because of their symbolic value).
A few examples make this distinction clear. An employer is discriminating against an employee based on passive race/ethnicity performance when, during a phone interview, he denies an otherwise qualified applicant a job because the caller unknowingly speaks in a dialect associated with working class black communities 91 The "active" race/ethnicity performance scenario is when the same caller emphasizes her racially coded speaking style to ensure the employer knows she is a minority. Typically, in both of these circumstances, the employer does not mention race or ethnicity as the basis for his decision to deny the applicant the job opportunity. He simply indicates that the caller did not have the "polish" he normally looks for in employees. 92 The employer, however, is motivated by racial/ethnic animus in both cases; he reacts negatively to the speaker's voice because it triggers negative, stereotypical assumptions about the applicant's race or ethnic group - a reaction which the law forbids when the animus concerns a physical trait. The discriminator's thinking in these cases is identical to the thinking that would inform his actions in a morphology-based discrimination case: The caller has triggered a cultural code associated with a low-status race or ethnic group and, as a consequence, is denied an opportunity.
Discrimination based on "passive" race/ethnicity performance should deeply offend our notions of fairness because the target is unaware that she is manipulating racial or ethnic signifiers and is denied a valuable opportunity before she is given a choice to abandon or suppress the offending characteristic. 93 Indeed, these cases typically concern the most marginalized, vulnerable minority workers: poor individuals with few cross-cultural contacts who, prior to employment, [*1163] spend the majority of their time in highly residentially segregated minority communities. 94 Also, in many cases, the unknowingly marked job seeker faces an employer who lives in a racially or ethnically segregated neighborhood with limited opportunity for cross-cultural contact; such an employer is quite likely to find even passive race/ethnicity performance behavior foreign and threatening. This innocent job seeker suffers, not because of conscious action on her part, but because of our society's failure to combat the problem of racial and ethnic residential segregation.
Also, workers who engage in passive race/ethnicity performance should appear sympathetic even for strong proponents of assimilation, as it denies these workers access to a diverse work environment - the precondition for an exchange of cultural perspectives. This kind of discrimination tends to relegate the most ethnically and racially marked workers to low-status and low-paying jobs, offered by employers who disregard these features because these workers are anonymous (e.g., factory workers) and the employers do not have to interact with them, or jobs where the workers have no contact with the general public and therefore have no bearing on the public image of the employers' business. For these reasons, discrimination based on passive race/ethnicity performance has the greatest potential for creating socially stratifying and segregating effects now that morphology-based employment discrimination has largely abated.
Apart from the issues of access and opportunity, discrimination based on passive race/ethnicity performance should also offend our belief in the value of personal dignity. Why should a person be required to shed passively acquired racially or ethnically marked mannerisms when they have no bearing on her potential performance of the job at issue? Indeed, once a heavily-marked job seeker is denied an opportunity because of these passive traits and behaviors, she faces an important decision. Now that she is aware that her community's practices are undesirable, she must decide whether to shed these attributes, a decision that may be experienced as a truly traumatic betrayal of her concept of self. The question is: Is this how we want to reward people who are willing to leave ethnic and racial enclaves [*1164] and socialize in a wider cultural context? Many of these workers simply will withdraw and lose interest in further cross-cultural interaction. Many may feel a need to emphasize racial/ethnic pride as a result of this dignitary injury. It should offend our basic notions of fairness to leave these individuals at the mercy of an employer's subjective views about the relative value of different ethnic communities. Indeed, after two decades of identity politics, it seems unfair to tell this worker that she must assimilate in order to fairly compete in society.
Discrimination based on "active" race/ethnicity performance raises equality concerns as well, although of a different order. The purest examples of these cases involve workers who fit the morphological profile of their claimed racial group, but also engage in practices intended to affirm their identities. Discrimination based on active race/ethnicity performance occurs, for example, when a white employer suddenly begins to give his Indian employee less favorable reviews and smaller bonuses after she starts wearing a sari and bindi. 95 The employer always has known that the employee was Indian; however, something about her sari and bindi appears unprofessional to him and, on that basis, he sanctions the employee for her actions. Similarly, active race/ethnicity performance discrimination occurs when white and black coworkers begin sabotaging a Latino employee's workstation when the employee starts speaking Spanish with Latino coworkers. 96 Although the Latino employee's coworkers long have assumed, based on her morphology, that she is Latino, the problems begin when they suspect that she is insulting them in Spanish for the amusement of other Spanish-speaking coworkers. 97 Importantly, in both of these examples, the victim-employee's morphology is not the primary trigger for discrimination. Rather, the discrimination [*1165] problem arises because of active race/ethnicity performance. 98
The previous two examples explored scenarios in which a person morphologically categorized as a low-status ethnic or racial subject is sanctioned for race/ethnicity performance associated with her group. Some race/ethnicity performance cases, however, will concern persons with ambiguous morphology who find that when they adopt racially or ethnically marked behavior they become targets of discrimination. For example, a person with ambiguous racial morphology may work for her employer without incident for years but find that she is subject to adverse action shortly after she is overheard speaking in Spanish to family members during a personal call. In these circumstances, it is the employee's voluntary racially or ethnically marked behavior that functionally "outs" her in the workplace. Similarly, a light-skinned employee with straight hair may be regarded as white by her coworkers for years but be "outed" and subject to stigma when she wears a headwrap or kente cloth to work. In both of the aforementioned scenarios, the employee's morphology is irrelevant. The discrimination is triggered by active race/ethnicity performance. Again, the cognitive process behind the discrimination in these cases is the same as those concerning racial and ethnic morphology: The individual triggers a cultural code associated with a low-status group, causing the individual to be sanctioned. 99
Some will feel more reluctant to offer antidiscrimination protections for active race/ethnicity performance precisely because it involves questions of agency and demonstrates the target's unwillingness to shed racially and ethnically coded distinctive behavior in an effort to fit in at the workplace. Indeed, under this logic, the undoubtedly expressive nature of this conduct suggests that individuals [*1166] engaged in this behavior seek out some positive social benefit by celebrating this aspect of their identity and, therefore, should be willing to bear sanctions triggered by negative reactions to that identity. This tempting line of argument, however, is rendered inadequate once we consider the cognitive logic and the motivations of the discriminator. Again, just as in the morphology-based discrimination cases, the discriminator's negative reaction to this coded behavior has nothing to do with its inherent value, and everything to do with a negative predisposition to its community of origin. When the discriminator's response is analyzed, it seems clear that active race/ethnicity performance is equally entitled to protection.
C. Dangerous Minds: Judges' Role in Recognizing Parallels Between Morphology-Based and Performance-Based Racial/Ethnic Ascription
The previous discussion revealed that the same basic cognitive process informs discriminatory behavior in both morphology-based and performance-based discrimination cases. The next logical question is: Why have courts failed to recognize these parallels? Part of the answer is that some courts na<um i>vely have regarded morphological race as a biological given instead of as a cultural construction. 100 Consequently, they have not considered how the inquiry into morphology-based discrimination requires them to probe the alleged discriminator's personal racial and ethnic lexicon. However, each discrimination case requires the court to become familiar with the frameworks that the alleged discriminator employs to identify racial/ethnic morphological difference, the consequences of recognizing this difference, and the stereotypes that she attributes to particular groups.
Given the widespread currency of the scholarship demonstrating the constructed nature of race, many courts likely understand at this point that morphology-triggered discrimination cases do not involve a purely objective inquiry. However, they would likely insist that the landscape of morphological markers is far more stable than those involving racially and ethnically coded behavioral and aesthetic choices. These judges distinguish the morphology cases from performance-based [*1167] discrimination cases by arguing that the performance cases require them to engage in a rapidly changing arena of political and cultural concerns, areas that are beyond the court's traditional purview. 101 However, as this Part shows, the morphological paradigms for racial and ethnic difference are constantly shifting and are inherently political; they never have been (or ever will be) stable and fixed. Rather, both types of discrimination cases (morphology-based and performance-based) involve a similar subjective inquiry, to identify what symbols or cues trigger racial or ethnic recognition and discrimination in the employee's workplace. Therefore, a court's refusal to examine performance-based discrimination cases because of their political underpinnings seems indefensible.
Some who are persuaded about the similarities between these kinds of discrimination cases still will worry that courts are ill-equipped to negotiate the political questions necessary to identify the voluntary race/ethnicity-associated behaviors that trigger discrimination. They rightly note that, in modern antidiscrimination cases, courts are rarely required to inquire into race/ethnicity-associated morphological markers (and the social and political realities that inform these symbols) and therefore have sidestepped the difficulties presented by the earlier race determination cases. 102 In contrast, if race/ethnicity performance is accepted by the courts as a basis for discrimination claims, there will be significant, bitterly contested fights in Title VII cases simply to identify which behaviors should be counted as race/ethnicity performance. While these concerns should not be dismissed, they overestimate potential problems. The small amount of litigation about the morphological markers of race and ethnicity points to the fact that there is a shared consensus or "common sense" as to the racial/ethnic significance of many morphological features. Similarly, there is likely some consensus about the most common traits, mannerisms, and behaviors that constitute race/ethnicity performance. Indeed, employers typically do not defend against allegations of discrimination by challenging morphology-based racial or ethnic categorizations, out of the recognition that the court's "common sense" will, in most cases, dictate that such a defense is meritless. 103
[*1168] Similarly, in race/ethnicity performance cases, this "common sense" will, to some degree, dissuade defendant-employers from requiring a court to engage in extended analysis regarding whether a race/ethnicity-associated practice is generally socially recognized. Plaintiffs, of course, should be required to present some evidence to tie the claimed practice to their identity, and courts may validly inquire into the quality of these proffers. Courts also worried about "political correctness" concerns may be reluctant to tie certain practices to particular racial or ethnic groups, despite the plaintiff's request, out of fear that they are creating a negative legal portrait of a given racial or ethnic community. Their focus, however, should be on the individual litigant's proffer, with the knowledge that any conclusions they draw about the relationship of the practice is intended for a limited purpose and not properly fodder for general social commentary. However, the issue of establishing the significance of a performative behavior is more likely to heavily burden plaintiffs from marginal racial/ethnic groups that do not have a social profile in a particular community. In these cases, as well, the recommendation that we rely on the judge's racial lexicon or "common sense" validly triggers concern.
Concerns about judges' use of "common sense" in antidiscrimination cases are based on the lessons taught by the Critical Legal Studies (CLS) movement. CLS scholars 104 have previously shown that judges' use of "common sense" in their analyses often meant that they incorporated racist or sexist premises into their decisions. 105 More recently, however, several CLS scholars have adopted the pragmatic [*1169] view that one cannot fairly expect judges to conduct a judicial inquiry that is not informed by their subjective views and social knowledge. 106 These scholars contend that judges simply cannot avoid using "common sense" in their opinions and, moreover, that decisions involving these common sense principles often push limits and challenge other understandings, thereby advancing the antidiscrimination project and the quest for equality. 107
Robert Post, the scholar perhaps most identified with this view, characterizes the judicial inquiry as part of a dialectic relationship between community norms and antidiscrimination goals. 108 He recognizes that judges will "apply antidiscrimination law in ways that implicate it in the very practices it seeks to modify," but argues that this is necessary for community norms and antidiscrimination goals to remain in dialogue and for judicial opinions to be defensible to their readers." 109 Similarly, Katharine Bartlett explains that we do not necessarily need judges to step beyond "common sense" or community expectations in advancing equality, as "community norms limit legal alternatives while also defining the terms required for equality to exist." 110 She argues that any equality initiative must be stated in terms that resonate with common sense or it will fail to gain hold. In Bartlett's opinion, "The law limits the permissible effect of community norms while defining higher ideals to which the community might aspire. Both operate simultaneously and reciprocally as cause and effect, as carriers of ... stereotypes and propellants for change, as provisional givens and targets for reform." 111 Stated more simply, Bartlett argues that current social relations allow us to envision what might change in order to achieve social equality.
This more hopeful attitude about how the judiciary functions may seem overly optimistic, given judges' prior attitude towards the race/ethnicity performance cases, whereby their "common sense" told them that performative features of racial and ethnic identities were of marginal import. 112 While they are mindful of these concerns, some [*1170] CLS scholars are trying to develop models that address both the threat and potential of "common sense" in a way that will prevent judges from caricaturing races and ethnic groups or subordinating these groups' interests to the assimilationist pressures of our time. Post advises that, as an initial step, we must think clearly about the ways in which we hope to reconstruct social practices that inform constructions of race, gender, and national origin. He suggests that "principles be articulated that will guide and direct the transformation of [these] social practices." 113 Similarly, Bartlett advises that "courts should approach challenges to practices grounded in community norms by attempting to identify the cultural meanings underlying them and determining to what extent they impose burdens that disadvantage" one group in relation to another. 114 She argues that this inquiry may "reveal status distinctions buried in ... norms that have become normalized by their cultural familiarity." 115
If we apply these insights to help us understand how judges should approach cases concerning race/ethnicity performance discrimination, several lessons are clear. First and foremost, when judges resort to "common sense" in their decisions, they must scrutinize their potentially assimilationist views to determine whether their perspective tends to subordinate one race or ethnic group to others. Also, they must remain vigilant to the possibility that they hold certain common sense beliefs about cultural conformity which have become naturalized by their cultural familiarity, but may in fact fundamentally offend the idea of equality. 116
In short, in urging the use of "common sense," my goal is for judges to employ a more actively engaged, self-critical form of "common sense" 117 in race/ethnicity performance discrimination cases that is informed by notions of equality rather than unarticulated knee- [*1171] jerk responses produced by their upbringing. Judges should expect that there will be cases in which they will encounter identity practices that are unfamiliar or beyond the limits of their racial and ethnic ascription frameworks, and, in these circumstances, they must educate themselves about these new practices and carefully consider the plaintiff's proffer as to the significance of these practices. Also, they must be prepared to re-examine their "common sense" notions about assimilation if they are to identify those circumstances when an employer is employing assimilationist justifications for policies that harass minority workers. If judges can mobilize this modernized, enlightened version of common sense, they are far more likely to be able to apprehend those circumstances where race/ethnicity performance discrimination is at issue. 118 This is perhaps the most honest, pragmatic way to move forward, for there is little hope of creating the fluid, responsive inquiry necessary to support the inquiry into performance-based discrimination if we do not rely on judicial discretion.
II The Psychology of Race/Ethnicity Performance: Motivations and Reactions
Part II offers insights from social psychology and identity performance theory to better explain people's interest in and reaction to race/ethnicity performance. Section A provides background on the Nigrescence 120 the seminal theories in social psychology on racial identity development, to provide a basis for understanding what role race and ethnicity play in identity development. This Article then shows how Judith Butler's performativity model, 121 a model which originally was applied to theories about gender identity, provides a superior framework for understanding racial and ethnic identity formation issues, as it is more generalizable across races and ethnic groups than traditional theories [*1172] and is more consistent with individuals' lived experiences. Additionally, it better correlates with quantitative psychological studies on racial/ethnic identity development.
Section B employs Herbert Blumer's "group position" theory 122 and John Dovidio and Samuel Gaertner's theory of "aversive racism" 123 to explain why persons opposed to race/ethnicity performance, consciously or unconsciously, perceive these behaviors as assaults on the existing racial and ethnic status hierarchies in the workplace and, by extension, on the value and importance of their own chosen racial and ethnic identities. Section C then illustrates how the above insights about race/ethnicity performance 124 will allow courts to better understand the motivations and competing equity claims of employers and employees in Title VII cases.
A. Understanding Racial Identity: The Role of Performativity
Since the advent of racial identity development studies, scholars have primarily focused on African Americans' identity development experiences, leaving the identity construction challenges faced by [*1173] whites, Latinos, Asians, and other groups undertheorized. 125 This original emphasis on African Americans likely stemmed from the fact that several major American civil rights movements highlighted certain models of African American identity, 126 and, as a consequence, a mistaken consensus developed that American political conflicts were largely a struggle between African Americans and whites. 127
[*1174] The most famous of these racial identity development models are the Nigrescence models, specifically the one created by William Cross, 128 and the Black Identity Development (BID) model created by Bailey Jackson. 129 Influenced strongly by the 1960s civil rights struggles that raged during the decade prior to their formation, each identity model treats minority racial identity formation as a liberatory sequence during which the individual transitions from a non-identified or negatively racially identified person with low self-esteem to a strongly racially identified subject who is better socially adjusted and resilient in discriminatory circumstances. 130 Both theories posit that if an individual is hindered in her attempts to move through this liberatory sequence, she will continue to suffer from low self-esteem. 131
The Nigrescence and BID models were widely accepted and applied for several decades. BID even was used to describe the experiences of racial and ethnic minorities other than African Americans. 132 However, during the 1990s, these models were criticized on a number of grounds. First, some charged that these progression frameworks failed to comport with individual minority actors' life [*1175] experiences, as individuals do not undergo a simple progress of increasing racial/ethnic identification. Instead, minorities often show a varied reliance on racial/ethnic identity during different phases of their lives 133 and mobilize different versions of a racial/ethnic identity in a given time period, depending on their context. 134 Additionally, critics argued that these progression models did not account adequately for the varied importance or "salience" of racial/ethnic identity among individuals, as some persons choose to structure their identities based on some other socially salient belief or feature. 135 Also, the progression models suggested that the emotionally healthy African American eventually would gravitate towards Afrocentricity; in reality, "black" communities exhibit a wide array of race-associated practices, many of which have no correlation to Afrocentrism. 136 Also, Cross's and Jackson's models offered no insight into why certain identity practices develop, or how individuals choose between various race/ethnicity-associated behaviors.
Additionally, the Nigrescence and BID models were focused specifically on African American racial identity, making it appear that racial identity development issues are primarily of minority concern. As such, these models offered no insight as to the important question of white ethnic minorities' experiences of racial/ethnic identity development or, for that matter, the identity development processes of any group whose racial/ethnic identity is not premised on the need for liberation from white cultural hegemony. 137 Mindful of these criticisms, [*1176] Cross and Jackson more recently have tried to refine the early progression models. 138
The limitations of the currently hegemonic racial identity models suggest that this area of identity studies might be better served by theorizing a fundamentally new paradigm for understanding racial and ethnic identity development, one which at the start is designed to be generalizable across groups. Judith Butler's model of performativity 139 appears uniquely suited to assist us in developing this model, as it provides for a more fluid understanding of a racial or ethnic subject's need to express racial and ethnic belonging, accounts for variations in race/ethnicity-associated behaviors, and recognizes the varying reasons for the different intensity of race/ethnic identification between both individuals and racial and ethnic groups. 140
Butler introduced the concept of performativity in her seminal book Gender Trouble, 141 and further elaborated on the theory in Bodies That Matter. 142 She posits that speech acts and behavioral gestures should be understood as an individual's attempt to articulate her belonging to a social identity category. 143 Butler's primary goal in Gender Trouble was to challenge our understanding of gender's relationship to sex in order to show that, although sex is treated as a biological fact and gender as a cultural artifact, sex is in many ways just as socially constituted as gender. 144 However, in the course of this discussion, Butler argues that a social actor (or "subject") can be understood only as attempting to locate herself within matrices of discourse about various identity categories, 145 and is driven to compulsively repeat those speech acts and symbolic behaviors associated with a certain category in order to establish her place as a social being. 146 As [*1177] she explains, there is no gender identity behind expressions of gender; that identity is performatively constituted by the very expressions that are treated as its results. 147 These premises, based on gender identity, are equally applicable to our understanding of racial and ethnic identity.
Butler's description of subjectivity is complex and requires further explanation. Butler summarizes her view as follows:
Language is not an exterior medium or instrument into which I pour a self and from which I glean a reflection of that self... . The substantive "I" only appears as such through a signifying practice that seeks to conceal its own workings and to naturalize its effects. Further, to qualify as a substantive identity is an arduous task, for such appearances are rule-generated identities, ones which rely on the consistent and repeated invocation of rules that condition and restrict culturally intelligible practices of identity. Indeed, to understand identity as a practice, and as a signifying practice, is to understand culturally intelligible subjects as the resulting effects of a rule-bound discourse that inserts itself in the pervasive and mundane signifying acts of ... life... . There is no self that is prior to the convergence or who maintains "integrity" prior to its entrance into this conflicted cultural field. There is only a taking up of the tools where they lie, where the very "taking up" is enabled by the tool lying there. 148
While this excerpt cannot provide a comprehensive account of performativity, it captures the four propositions most relevant to our discussion of the development of racial or ethnic identity.
The first proposition, which is perhaps the hardest to accept, is that one does not possess a racial/ethnic identity prior to articulating one's place in a paradigm of racial and ethnic difference. 149 Butler's insight, that the concept of an "I" only appears as a consequence of a signifying practice, when applied to a discussion of racial/ethnic identity suggests that that identity can only be claimed through speech acts [*1178] and behaviors - "performative" acts. Therefore, part of the process of constituting oneself as a social actor requires the acceptance and recognition of racial/ethnic codes and markings and the mobilization of these codes to ensure that other actors read them in the manner that ensures one is placed in the desired race or ethnic group. Language and behavior play a key role in this process.
As applied to the daily challenge of asserting a personal identity, the proposition can be instrumentalized as follows. Part of the process of socialization requires that one take up the codes of social meaning about races and ethnic groups and use them as tools in one's individual identity development project. Therefore, as an individual learns about racial and ethnic difference, she constructs her self-concept in relation to these codes. However, the individual has no racial or ethnic identity prior to this socialization, prior to her decision to take up and mobilize codes of racial and ethnic meaning. 150 Typically, this moment of decision is invisible or transparent; it is not experienced as a moment of conscious choice. However, as described by multiracial or racially ambiguous persons, the moment of choice can be palpable and even traumatic, and its consequences are long standing. 151
By treating race and ethnicity as the effect of a signifying practice, enacted through speech and behavior, performativity does not deny what some would call the material realities of race and ethnicity - the importance of physical or morphological characteristics in the process of racial and ethnic ascription. 152 Performativity certainly recognizes the importance of the physical body in identity development. Physicality, in large degree, may limit what racial or ethnic identity an individual can perform successfully. 153 Performativity recognizes this fact, but it refuses to treat race and ethnicity as a [*1179] mere function of race/ethnicity-associated morphology. 154 Rather, it suggests that while certain physical traits may suggest a particular racial or ethnic identity or interfere with the performance of one's chosen identity category, physicality is not entirely determinative of the issue. For example, some people actively perform racial or ethnic identities in an attempt to cancel out the contrary symbolic effect of their morphology, 155 and are successful in doing so. Also, groups of people with the same racial or ethnic morphology often engage in bitter contests over the proper performance of that racial or ethnic identity, sanctioning persons whom they perceive to engage in inadequate race/ethnicity performance. 156 These disputes demonstrate that even individuals who possess strong racially and ethnically marked morphological features regard race/ethnicity performance as important. 157 Therefore, although morphology plays a role in racial and ethnic identity development, performance plays an equally important role in the process.
The second premise of performativity that this Article explores is drawn from Butler's observation that identity maintenance is an "arduous task"; she suggests that repetition of symbolic acts plays a key role in claiming a certain racial or ethnic identity. 158 These symbolic acts and behaviors may take the form of aesthetic choices, such as hair and clothing styles, speech patterns, or use of dialect. Because of the arduous nature of the process, individuals may develop practices that either permanently or semi-permanently mark the body, providing a stable basis for public recognition of their chosen racial/ethnic identity. 159 The compulsion to engage in these acts appears natural and invisible, and it inserts itself into the most mundane, everyday aspects of life. I would further suggest that when an individual is faced with a challenge either to her hold on a racial or ethnic identity or to the relative status ranking of her race or ethnic group, she will feel an even stronger drive to engage in acts that reaffirm her chosen identity. This proposition is illustrated in cases where an [*1180] employee engages in a range of racially or ethnically marked practices, even after having been punished for others, in an attempt to find a kind of racial/ethnic identity performance that will be tolerated by her employer. 160
David Kertzer counsels that we should not underestimate the great psychological significance that symbolic performances have for the individual in everyday life. He provides another set of justifications for respecting the need for group-affiliated performances, or personal "rituals," 161 explaining that they allow us to cope with the stresses of a constantly changing world by "building confidence in our sense of self by providing us with a sense of continuity." 162 As he explains, each symbolic performance sends an important psychological message - "I am the same person today as I was twenty years ago and as I will be ten years from now" - and "gives us confidence that the world in which we live today is the same world we lived in before and the same world we will have to cope with in the future." 163 For individuals who have chosen to enact a particular racial or ethnic identity, race/ethnicity-associated practices provide certain assurances about their group position and importance in the world, even though they know that certain material or personal realities will not remain the same.
The third proposition of racial/ethnic performance is that, in order to be effective, racial/ethnic performance must correspond to those symbolic representations that are culturally intelligible as being representative of or correlated to a particular racial or ethnic identity. 164 As Butler explains, the deployable materials for individual expression are "rule-generated identities, ones which rely on the consistent and repeated invocation of rules that condition and restrict culturally intelligible practices of identity." 165 Therefore, in order to serve their purpose, race/ethnicity performances must have a symbolic meaning both within the performing individual's ascriptive [*1181] frameworks and for the community to which she seeks to present this racial or ethnic identity. 166 For example, a morphologically indeterminate person claiming a West Indian identity would be more likely to emphasize a West Indian accent in her discussions with Caribbean Americans and African Americans to "perform" her ethnic identity than to wear the colors of her country's flag. Both actions are symbolic and are intended to send the same message; however, the accent is a more generally recognized means of activating references for a West Indian identity. This is not to suggest that individuals will never generate idiosyncratic practices intended to function as race/ethnicity performance. 167 Through pastiche, they occasionally will. However, these idiosyncratic practices are not a focus of this Article because when an individual is punished for engaging in this behavior, it is less likely that the sanction is based on discrimination because the audience for the "performance" often does not understand their meaning.
The fourth proposition, which is captured more subtly in Butler's discussion, is that the performance of racial and ethnic identity provides a person with a sense of agency. 168 There is no "self" before one's attempt to assert one's existence by seizing the identity categories offered by language. The claiming of this position is what allows one to articulate ideas and act on other issues. 169 Stated alternatively, when an individual claims her identity as a racialized person and takes up the social codes for expressing racial difference, she is given access to a variety of ways to express her views about social groups and a variety of options for cultural expression. She can then locate herself within existing discussions on a variety of social issues. Of course, a [*1182] person's decision to take up and use racial and ethnic codes in her identity project does not determine automatically what her view will be on any particular issue. However, by taking up these codes, one opens up a set of options and, in exchange, is given the materials with which one can ensure that one consistently elicits the desired responses from ingroup and outgroup members.
This proposition is explored in more detail in Kertzer's analysis of the symbolic and communicative importance of group-associated practices. 170 He explains that group-associated practices provide an individual with three benefits: They (1) allow a person to affirm her connection to her cultural past or the historic struggles of her group; (2) cultivate feelings of solidarity in order that she may understand her "place in the world"; and (3) assist her to view herself as performing distinct social roles. 171 These propositions apply equally in cases of race/ethnicity performance. For example, an African American person's decision to wear cornrows provides all three of the psychological benefits described above: The braids allow her to (1) affirm her connection to a larger African or African American community; (2) express solidarity with that community on a daily basis; and (3) actualize and valorize her connection to a history of black opposition and oppression - a heroic opportunity perhaps rarely afforded in everyday life. Indeed, challenging her employer's prohibition against wearing the hairstyle or enduring the negative reaction of coworkers may bring her additional psychological benefits.
Intuitively, the next question in our investigation becomes: How does the racial or ethnic subject locate and choose performative behaviors? Butler's theory is not designed to address this question, but certain conclusions can be drawn from social psychology in the area of racial and ethnic identity formation. First, social psychologists indicate that racial and ethnic identity, in the first instance, is shaped by the family. 172 Therefore, this unit should be seen as the primary inculcator to race/ethnicity performance behaviors. Second, the individual's [*1183] wider community also provides her with a choice of race/ethnicity performance options. 173 Importantly, the options available will vary based on the social conditions that the individual contends with and the physical resources at her disposal. Therefore, race/ethnicity performance will look different in integrated as opposed to highly segregated communities, urban as opposed to rural areas, and religious as opposed to secular communities. Race/ethnicity performance options also are offered in music and other media representations of race and ethnicity, whether they are behaviors associated with popular heroes or provided by negative media accounts of racial and ethnic subjects as victims or criminals. 174
Butler and Kertzer's insights provide the basis for three general observations about the motivations and concerns of persons engaged in race/ethnicity performance, as well as a way for courts to predict and understand individuals' reactions to attempts to police this behavior. First, they suggest that an individual's hold on an identity is at many times tentative and unsure, and it is likely to be most at risk when an incident occurs that challenges an individual's connection with that identity or ranks the identity as low-status. Translated to the employment context, I believe that courts reviewing Title VII cases may see scenarios in which plaintiffs' race/ethnicity performance practices actually increase when employers prohibit such behavior because they are forcing minorities to question the status ranking of their group and to defend that group. The worker who challenges her employer's prohibition is attempting to address a dignitary harm. She simultaneously may experience the rule as traumatic and disorienting as she attempts to maintain her hold on her chosen racial or ethnic identity. 175
Second, different performative behaviors actually may be part of the same race/ethnicity performance project. Because there are multiple sources for generating ideas about racial/ethnic identity, individuals may vacillate between different identity performance options until they temporarily find a collection of performative behaviors that are suited to their environment, circumstances, and personal tastes. 176 Therefore, rather than looking for an employee to engage consistently [*1184] in a single kind of race/ethnicity performance, courts should consider whether the employee is switching behaviors because she is trying to identify the set of behaviors that suit her lifestyle needs or, importantly, because she is trying to find a form of race/ethnicity performance that comports with the employer's workplace rules and dress code. Without this insight, courts might interpret an employee's shifts between race/ethnicity performance behaviors as willful recalcitrance when they actually may show that the employee is attempting to comply with her employer's rules.
The third observation is that persons engaged in the same race/ethnicity performance behaviors may offer very different justifications for why the behavior has symbolic importance to them. Symbolic practices are able to wed large and diverse groups under a single heading precisely because the behaviors have multiple meanings and can signify each simultaneously. 177 As a result, they allow groups to "build political solidarity in the absence of consensus" based on a mythic conception of unity that can be used for other purposes. 178 This proposition, as it pertains to our discussion, suggests that courts should recognize that members of a racial or ethnic group may agree that certain practices, aesthetics, or behaviors are important for that community but differ as to the justifications for their importance. Despite these multiple meanings, individuals can and will mobilize based on a particular symbolic practice when they feel outgroup members are challenging their way of life. This spirit of cooperation eventually may lead them to mobilize based on other issues. 179
This dynamic is observed easily in cases involving black persons' interests in wearing dreadlocks. Some people wearing the hairstyle indicate that their interest in doing so stems from their belief in Nazarite precepts and the religious beliefs of the "original" Black Jews; others justify the practice as part of Rastafarianism; and still others argue that dreadlocks are a secular celebration of African American [*1185] beauty and culture. 180 Dreadlocks function as symbols for all of these perspectives and the potentially associated social and political trends they represent within black communities.
The example above illustrates an essential point: The fact that a practice does not have a single symbolic meaning to its community of origin is separate from and irrelevant to the fact that it functions as a disturbing racial or ethnic signifier to those who would discriminate against a subordinate race or ethnic group. Stated alternatively, the multiple symbolic meanings associated with a race/ethnicity-associated practice do not muddy the reality that this practice functions as a trigger for discrimination by outgroups. Therefore, although courts should look to precedent to identify practices that have been recognized as expressive of a given racial/ethnic identity, they should not expect that the justification linking the practice with a given identity will remain the same in all geographic areas or over time. Even when these justifications differ, the only relevant inquiry is whether the alleged discriminator sanctioned the subject engaged in the behavior because she recognized it as racially/ethnically coded behavior and because of her antipathy for a particular racial/ethnic group.
Given these insights about the importance of racial and ethic identity practices, the question becomes: Do we want to preserve an antidiscrimination regime that only protects against discrimination based on physical features when, in many circumstances, minority workers are equally targeted because of voluntary race/ethnicity-associated behavior? Consider the repercussions: By preserving a rule that protects minority workers only from sanctions triggered by morphological difference, Title VII creates a dynamic that sabotages the minority workers it most wants to encourage - those raised in ethnic or racial enclaves who, despite their fears about cultural unfamiliarity, are willing to try interracial workplace experiences. It seems fundamentally wrong to subject workers who are willing to face this challenge to the whims of an employer spurred by invidious discriminatory motives. Sadly, many poor minorities who attempt to interview for jobs at white-collar businesses are dealt an unforgettable dignitary blow, as they learn about "discrimination by proxy" and the vulnerability they face because they display voluntary racially and ethnically marked features. These workers are painfully reminded of [*1186] their difference and the low status of their racial or ethnic group. 181 Additionally, they learn that employers can and will discriminate against them based on these voluntary differences with impunity. These are the lessons taught by a regime that ignores, and even subsidizes "discrimination by proxy," as Title VII, at present, fails to provide these workers with a remedy. 182
B. Understanding Prejudice: Outgroup Reactions to Race/Ethnicity Performance Behavior
Our understanding of an individual's interest in race/ethnicity performance would be incomplete without an explanation of the effect of this behavior on outgroup members. The following discussion traces outgroups' reactions to race/ethnicity performance, relying heavily on recent applications of "group position theory" as interpreted by Lawrence Bobo, 183 in addition to Samuel Gaertner and John Dovidio's studies on "aversive racism." 184
Before engaging in a more detailed treatment of individual prejudice models, it is helpful to have some background on the history and development of prejudice studies in the United States. John Dovidio explains that prejudice studies can be separated into three waves. 185 The first wave in the late 1950s was dominated by psychologists who posited that prejudice was an individual psychopathology. 186 This pathology model, however, proved unworkable, as it failed to explain why, despite legal initiatives and reeducation efforts, Americans continued to engage in prejudiced behavior. Faced with the reality that discrimination was a continuing feature of social life, prejudice studies gave birth to a "second wave," a period in which scholars treated prejudice as a normal cognitive social process, influenced by both internal personality variables and external pressures. 187 These scholars examined prejudice as a social structural phenomenon - that is, a force in dialectic relationship with social institutions. Specifically, they recognized that social institutions encouraged the development and use of racial/ethnic constructs by citizens, and, in turn, these institutions found that they were required to use these constructs in order to match their citizens' understandings of their own [*1187] identities. 188 At the end of the second wave and moving into the third, scholars recognized that because of the moral hegemony of antiracist ideas, prejudiced persons now experience a cognitive split: They claim to have egalitarian views but still feel compelled to engage in prejudiced behavior. 189 The third, or current wave, therefore, builds upon this important recognition and explores the conscious and unconscious impulses for prejudiced reactions. 190 Third wave scholars also examine more closely the justifications manufactured by persons with these cognitive splits, exploring the reasons and rationales that prejudiced individuals offer, as they find ways to characterize what is clearly discriminatory behavior in a way that does not challenge their professed egalitarian views. 191 Additionally, these studies inquire into the exchange between discriminator and target, examining how and when the target perceives these more subtle forms of discrimination and the target's feelings of antipathy and sense of agency. 192
The group position and aversive racism theories straddle the divide between the second and third waves. Sociologist Lawrence Bobo's work is particularly astute, marrying the priorities of each wave successfully. Bobo's theory of group position relies heavily on a second wave precept - that discrimination is a consequence of social structure. 193 Group position theory posits that racial prejudice is not a collection of irrational feelings and stereotypes about race but rather "[is] best understood as a general attitude or orientation involving normative ideas about where one's own group should stand in the social order vis-a-vis an outgroup." 194 Group position theory, however, posits that this anxiety about group status manifests itself in two forms. High-status group members will be concerned about: (1) their group's relative status-group ranking at a particular time, as compared with perceived subordinate groups, and (2) articulating their own values in such a way that the group maintains cohesion and excludes unwanted others. 195 With regard to this second component, this [*1188] impulse may take a variety of forms, as "dominant group members must make an affectively important distinction between themselves and [perceived] subordinate group members ... linked to ideas about the traits, capabilities and likely behaviors of subordinate group members." 196 The critical distinction between group position theory and more classic models of discrimination is that dominant group members "are not merely saying that the racial minority group is different or lesser, or even simply venting an affective hostility or resentment, but rather (and perhaps centrally) claiming that their relative status is [potentially] significantly diminished by this difference." 197
Group position theory was originally created to study whites' relationships to minority groups; however, Bobo decisively moves group position theory into the third wave when he shows that the premises of the theory are equally helpful in understanding "how members of a subordinate group come to view members of a dominant group." 198 Group position theory thus can be applied to "relations among and between racial minority groups in a multiethnic social setting." 199 Bobo identifies the four foundational beliefs that trigger individuals to engage in group position thinking: (1) "a feeling of superiority on the part of their group"; 200 (2) "a belief that the [*1189] subordinate group is intrinsically different and alien"; (3) "a sense of proprietary claim over certain rights, statuses, and resources"; and (4) the "perception of threat from members of a subordinate group who harbor a desire for a greater share of dominant group members' prerogatives." 201 Members of different races and ethnic groups will exhibit these same anxieties, both when they enjoy cultural hegemony, and when they perceive that their group's position has lost ground in relation to another perceived subordinate minority group. 202 Therefore, in the employment setting, we can expect to see conflicts about race/ethnicity performance between whites and subordinate minority groups, as well as between two relatively low-status minority groups, one of which appears to enjoy cultural hegemony or preferred status in a particular workplace.
One of Bobo's most valuable insights is his discussion of dominant group members' "sense of proprietary claim over ... rights, statuses, and resources." 203 This provides a basis for understanding the stakes of the cultural or group status battles that occur in race/ethnicity performance cases. Disputes about status can be over material, tangible goods, such as jobs, property, businesses, or political opportunities, or they may concern intangible goods such as prestige or the power to create an atmosphere or "area of intimacy." 204 This insight suggests that employers' or coworkers' desire to sanction an employee for race/ethnicity performance may be a response to the concern that the minority worker's behavior may in some way alter the cultural climate of the workplace, raising the status of her group, or destroying the intimacy enjoyed by relatively higher-status group members whose cultural perspective dominates the workplace. The event that triggers group status anxieties need not be an actual threat; anxiety can be triggered even by a perceived threat to the status rankings of dominant group members. 205 Therefore, although the race/ethnicity performance behavior of a single employee may present little danger of changing the overall cultural environment of the worksite, [*1190] if it is perceived as such, it will trigger complaints and sanctions from outgroup members.
A few examples make this point clear. Consider a scenario in which white employees in a majority white workforce notice that a small number of Latino employees speak Spanish during breaks. White employees may perceive these incidents as threats to their group's status because they, for the first time, are unable to participate in some workplace discussions and because the Spanish speakers' conduct destroys the atmosphere of intimacy (in which their experience was culturally dominant). Alternatively, they may react negatively simply because they believe that Spanish is stigmatized and that these conversations destroy the higher-status "American" or English speaking culture in the workplace. Similarly, white employees may react negatively to blacks wearing cultural hairstyles or headwraps to work because these appearance choices reject the existing aesthetic baseline of the workplace, making white workers suddenly feel marginalized. The black workers' acts disrupt the pleasant fiction that all workers share the same aesthetic values (and may be particularly disturbing to white workers if they feel a more Anglo aesthetic is being disfavored). Additionally, white workers may worry that the minority workers' appearance will detrimentally affect public perception of the company by making it appear culturally infected by African American aesthetics, a development that threatens to lower their personal status as employees of the company.
Group position theory's potential to help clarify race/ethnicity performance disputes becomes increasingly clear once we control for a phenomenon that Samuel Gaertner and John Dovidio call "aversive racism." 206 Gaertner and Dovidio explain that "changing norms" and "legislative interventions," such as the Civil Rights Act, have made discrimination "not simply immoral but also illegal" 207 and, as a consequence, "overt expressions of prejudice have declined significantly over the past 35 years." 208 Racism, however, has not been abolished; rather, it has morphed into a different form. This modern form of discrimination, which they call "aversive racism," differs from "old-fashioned" racism, which was blatant. 209 Aversive racism is a more "subtle, often unintentional form of bias" displayed by whites who have "strong egalitarian values and who believe that they are nonprejudiced." 210 [*1191] Whites who suffer from aversive racism will deny any personal prejudice but still harbor underlying or unconscious negative feelings or beliefs about minority groups. Gaertner and Dovidio explain that "because aversive racists consciously endorse egalitarian values ..., they will not discriminate directly and openly in ways that can be attributed to racism. However, because of their negative feelings, they will discriminate, often unintentionally, when their behavior can be justified on the basis of some factor other than race ... ." 211
Gaertner and Dovidio suggest that aversive racists need to disguise prejudiced behavior as seemingly objective neutral complaints because they suffer from cognitive dissonance. They explicitly sympathize with the victims of prior racial injustice and in principle are liberal 212 but, almost unavoidably, have closeted negative feelings about certain minorities that they reveal in limited circumstances. 213 These negative unconscious feelings about minorities stem from instructions provided by parents and peers that minority groups are relatively low-status, 214 and these lessons prove much more difficult to unlearn than explicitly racist beliefs. Dovidio explains, "With experience or socialization, people change their attitudes. However, the original attitude is not replaced, but rather it is stored in memory and becomes implicit, whereas the newer attitude is conscious and explicit." 215 As a consequence, aversive racists experience unconscious negative feelings about minorities and feel discomfort, uneasiness, fear, and even disgust in their presence. 216
[*1192] Dovidio and Gaertner's work suggests that employers and employees that are agitated by "group positioning" concerns rarely will explicitly acknowledge that they are racially or ethnically biased. However, because of their general discomfort with minorities and their unconscious negative feelings towards them, they will feel much more comfortable articulating this discomfort as disgust over race/ethnicity performance behavior. Indeed, when an employer can prohibit race/ethnicity-associated behavior on seemingly neutral grounds, she can satisfy her conscious desire to support racial equality and simultaneously satisfy a subconscious desire to maintain the dominance of her racial group in the workplace, as well as decrease the number of minority applicants and workers. 217
Taken together, these studies provide a roadmap for understanding workplace race/ethnicity performance disputes. They suggest that most Americans largely have internalized antiracist messages, and, as such, they will avoid at all costs being labeled a prejudiced person. 218 However, despite their professed antiracism, most Americans remain acutely aware of racial/ethnic group status issues and feel threatened by cultural changes that disfavor their group. 219 At present, common sense tells them that they will be branded as bigots if they discriminate against other groups based on race/ethnicity-associated morphology or because of racial/ethnic status; as such, they will work actively to avoid having their behavior characterized in this manner. Americans will, however, find ways to express their antipathy towards outgroups in "neutral" terms: This justification allows the individual to affirm her egalitarian antiracist identity, while at the same time satisfying her subconscious desire to express antipathy towards outgroup members or strengthen her own group's position in the racial and ethnic status hierarchy. 220
[*1193] Jack Balkin lays out a similar template that, while not explicitly indebted to group position theory, understands workplace discrimination disputes as struggles over status hierarchy. 221 Balkin's primary purpose is to defend Title VII's hostile environment doctrine against attacks from free speech absolutists who contend that this doctrine unreasonably inhibits individual speech liberties. 222 Balkin argues that hostile environment rules should not be understood as an attempt to muzzle employees under general civility codes but, rather, are an attempt to prevent coworkers from imposing status hierarchies by exacting dignitary and psychological injuries on workers from low-status groups. 223
Balkin's concerns about status hierarchy prove extremely useful in understanding why Title VII should prohibit conflicts in the workplace that create a hostile environment based on race/ethnicity performance. As he explains, since "material benefits and social status are so deeply interconnected in the workplace, status-based harms that significantly alter people's working conditions for the worse constitute employment discrimination under Title VII." 224 Stated more simply, he points out that the subtle messages of inferiority sent about minorities in the workplace have real material repercussions for minority workers. For example, rules that are hostile to race/ethnicity performance may deleteriously affect an employee's psychology and her willingness to attempt advancement. These rules also may silently confirm any negative perceptions other employees may have about [*1194] the targeted group, increasing the potential for other kinds of discrimination.
Given these realities, one questions why the law permits employers to inflict status-based injuries through grooming codes and other rules without violating Title VII's protections. Indeed, if an employer can be required to institute hostile environment rules that prohibit employees from generally communicating their view about the low status of particular races or ethnic groups, why should the employer then be permitted to communicate the same message of inferiority under cover of a grooming or disciplinary code?
C. Understanding Modern Title VII Cases
The scholarship discussed above provides a number of insights that are immediately helpful in understanding the challenges that courts will face in resolving future discrimination cases, particularly as they affect the individual psychology of the complaining worker, and the psychological calculus of the alleged discriminators. As a baseline proposition, the aversive racism model suggests that courts in the future should expect to see fewer cases about race/ethnicity-associated morphology, and more about "discrimination by proxy," which involve rules that on their face or by application prohibit behavioral or aesthetic attributes that are associated with particular races or ethnic groups. The way courts sort through these cases should be informed by the psychology of both the target and the discriminator if Title VII is to achieve its goal of making discrimination too costly for employers and employees to indulge in this behavior.
First, the aversive racism model teaches that the new frontier in employment antidiscrimination law is "discrimination by proxy," cases involving a neutral policy that either specifically identifies a cultural practice (or statistically correlated practice) associated with a particular racial/ethnic group for prohibition, or a neutral policy that is interpreted to prohibit racial-or ethnic-specific behavior. Again, the aversive racism model posits that almost no employer will institute a policy saying that she will not hire blacks. The employer may, however, achieve the same goal by instituting a policy prohibiting all-braided hairstyles. This "neutral" rule could trigger a lawsuit by a worker who experiences the rule as an assault on the relative status of her racial group, or one who feels compelled to defend the group as an attempt to perform a black identity. 225 Under the current regime, [*1195] the employer wins under both a disparate impact and disparate treatment model, as the employer's rule simply does not concern any immutable aspect of racial identity. 226
Prejudice studies suggest that we must interrogate these allegedly "neutral" grooming codes and recognize that they are often used to harass low-status ethnic groups and races. 227 Skillfully drawn, these policies appear colorblind but, when viewed against the backdrop of the default behaviors of the high status or dominant group in the workplace, it becomes clear that they subsidize practices that are favored by one race or ethnic group, and thereby improve their relative standing in the workplace, but are hostile to low-status races and ethnic groups, and potentially build an atmosphere for other kinds of discrimination. 228 Minority workers often experience these policies with built-in preferences for certain groups as creating an atmosphere of intimidation: They communicate the employer's continuing belief in certain racial/ethnic status hierarchies and send a clear message of the minority employee's relative inferiority. 229
To address squarely the inequities caused by these so-called "neutral" grooming codes, judges conducting Title VII inquiries must interrogate aggressively the motives behind policies that disproportionately burden or effectively screen out minority workers because of voluntary race/ethnicity-associated behavior. 230 For example, in a series of cornrow discrimination cases that involved American Airlines, the company contended that braided hairstyles simply did not comport with their "professional" aesthetic. 231 If the [*1196] company had been asked to justify this position, it would have been forced to articulate its actual justifications and confront any racially biased reasoning behind this policy. I suggest that if employers were more often required to expand upon their "objective" reasons for prohibiting race/ethnicity performance, they would reveal their reliance on historical stereotypes about minority groups and their policies could be rejected on that basis. One thing, however, is clear. As long as race/ethnicity performance is exempted from Title VII's definition of protected status, "discrimination by proxy" will continue. These seemingly "neutral" policies will be justified as employer prerogative under common law rules, a doctrinal field simply ill-equipped to address the equal protection concerns raised in these circumstances.
The second area of court emphasis that is suggested by the aforementioned prejudice studies is the need to interrogate what is now treated as "unconscious" racism. Stated simply, there will be some who believe Title VII is overreaching when it seeks to address unconscious racist and ethnically biased impulses, despite the teachings of aversive racism. Proponents of this view would suggest that since the employer has not attempted consciously to deny a person benefits and privileges because of race or ethnicity, she should not be held liable if her decisions inadvertently cause the same result. They would argue that Title VII has done all it was supposed to do: It has altered employers' conscious intent and made it too costly for them to engage in explicit discrimination. Now that employers have adopted policies that are non-discriminatory on their face, the law simply should not attempt to reach further and deconstruct unexamined preferences that may have discriminatory effects. Critics also may cite the practical difficulties of shaping legal doctrine to address subconscious or unconscious behavior. 232 As a basic matter, they question whether courts reliably can discern when an ostensibly neutral reason is actually a cover for prejudice. 233
Constructing a legal inquiry to examine these unconscious attitudes will prove challenging; however, ignoring this problem would exact too high a cost. Aversive racism is a powerful force in personnel decisions. Dovidio points to data showing that when black and white [*1197] candidates have equivalent credentials but are perhaps only marginally qualified for a position, white personnel directors tend to prefer white candidates. 234 He concludes that in these marginal cases, whites are willing to give other whites "the benefit of the doubt" that they can perform adequately. 235 Additionally, other sociological studies have shown that employers tend to interview and hire those candidates that they perceive as similar to themselves, unfairly burdening minority candidates who display unfamiliar race/ethnicity-associated behaviors. 236 When these subtle preferences are rendered invisible, one is left without satisfactory answers regarding minorities' failure to advance in certain workplaces. Having recognized that prejudice has not entirely abated, but has simply morphed into another form, how can we defend the decision not to proscribe this conduct? It seems wholly unconscionable to ignore this bias, particularly when it is known to have material effects on workers' opportunities for advancement.
Furthermore, the claim that courts cannot interrupt unconscious discriminatory impulses proves untrue. In describing the advances of prejudice studies, Dovidio highlights that prejudiced persons display two kinds of behavior, each of which is measurable and studiable. The first kind concerns "explicit attitudes," which are deliberative or actively chosen. Explicit attitudes are those for which people have the opportunity to weigh the costs and benefits of various courses of action. 237 Thus far, antidiscrimination law has devoted itself solely to this problem. Second, there are "implicit attitudes" that are more difficult to monitor and control because people do not view them as representative of their true attitude or outlook and, therefore, have no investment in trying to control them. 238 Note, however, that this second group of implicit or "unconscious" attitudes is not intrinsically different from conscious thoughts; they are not thoughts we cannot control; they are simply thoughts that we do not try to control. 239
Our history suggests that the law has proven quite effective in bringing unconscious impulses into the realm of conscious thought [*1198] through legal sanction. 240 Indeed, much of the complaint about sexual harassment/hostile environment codes as being muzzles or civility codes is based on resentment from employees who dislike having to think more carefully about what they say and are nostalgic for the time when their discriminatory comments were treated as innocent, innocuous behavior. Yet requiring workers to recognize the effects of casual comments that create hostile environments unquestionably has had a transformative effect on women's and minorities' workplace experiences. If the law were to make the more subtle kinds of aversive racism the basis for Title VII sanctions, these attitudes and behaviors would also be drawn into the realm of conscious thought. Barbara Flagg similarly has argued in favor of a standard that would challenge employers to recognize that certain "unconscious" behavior has discriminatory effects and hold them accountable for their decisions. 241 What is required is that the employer must be given the motivation to consider the repercussions of seemingly "innocent" decisions that burden racial or ethnic minorities engaged in group-specific behavior. Several scholars have already developed promising tools for probing objective sentiments to discern their discriminatory underpinnings. 242
Given these insights, the challenge for courts is to weigh the equities in each race/ethnicity performance case and determine whether the employer's hostility to certain performances is motivated by discriminatory impulses - because of concerns about maintaining the cultural dominance of one group or because she automatically stigmatizes the race/ethnicity-associated practice. 243 In examining the employee's claim, the court first must be concerned about whether the [*1199] race/ethnicity performance concerns a trait that, because of dignitary concerns, we would not ask the individual to change. Under this model, the burden is on the employee to show the strength of the association between a practice and her race or ethnic group. Once the burden is established, an employer must do more than simply offer a "neutral" value or justification for its decision; instead, it must specifically explain how the race/ethnicity performance thwarts the realization of that value, rather than simply offering its unsupported opinion. In Part III, I review a number of cases to show that, with a careful eye towards these issues, courts can better resolve race/ethnicity performance cases.
These aforementioned factors will continue to figure in the discussion in Section III, as we address the current doctrinal model for addressing racial and ethnic discrimination based on voluntary behavior. In discussing these cases, the Article proposes that a narrative which is more sensitive to these factors inevitably produces a more equitable resolution of race/ethnicity performance discrimination cases, and one which more directly deals with the equality and justice issues at stake in these conflicts.
III The Doctrinal Definition of Discrimination: Understanding the Status/Conduct Distinction
Having explored multiple ways in which individuals are subject to discrimination, the next question is: How does Title VII currently [*1200] attempt to intervene in this problem? 244 The statute provides no simple answer, for it does not define discrimination but instead broadly prohibits employers from engaging in "unlawful employment practices," 245 namely, the hiring and firing of workers based on race and national origin, as well as instituting procedures that tend to "limit, segregate, or classify" workers on these impermissible bases. 246 Also, although it generally refers to race and national origin, Title VII does not provide a detailed definition of either term. 247 In the absence of clear statutory definitions that would limit the scope of Title VII's protections, courts have filled the void with judicial definitions based on Fourteenth Amendment equal protection jurisprudence.
The courts' reliance on Fourteenth Amendment cases is to be expected, as Title VII was passed in order to help actualize some of the Fourteenth Amendment's equal protection guarantees. Most of the courts' analogies between the Fourteenth Amendment context and Title VII cases have proven relatively uncontroversial. For example, the Supreme Court's Fourteenth Amendment analysis suggests that one of the reasons that races and ethnic groups are offered antidiscrimination protection is because they possess visible, identifiable characteristics that function as irrational bases for stigma; 248 consequently, courts interpreting Title VII have construed the statute as intended primarily to address employment discrimination that is triggered by race/ethnicity-associated morphology. 249 More controversially, however, courts interpreting the Fourteenth Amendment also have held that groups which are defined primarily by practice instead of morphology are not entitled to antidiscrimination protection. 250 Extrapolating from this premise, courts interpreting Title VII similarly have ruled that any part of a morphology-based identity that is practice-based [*1201] is not entitled to antidiscrimination protection. 251 This more controversial assumption has prevented plaintiffs from prevailing on race and national origin discrimination claims when the discrimination is triggered by voluntary practices associated with protected classes. 252 In short, it has resulted in an involuntary/voluntary or "status/conduct" distinction in Title VII cases.
This Part additionally shows how the seminal cases on race/ethnicity performance are based on under-theorized analogies comparing race and national origin discrimination to gender-based discrimination. In making these comparisons, courts have adopted doctrinal tools and rhetorical constructions that starkly distort the interests raised in discrimination cases concerning race/ethnicity performance. Section A outlines the involuntary/voluntary or natural/artifice framework and shows how several cases decided under this framework would have been resolved better under a race/ethnicity performance analysis. Section B shows that the voluntary/involuntary distinction created in the seminal race/ethnicity performance discrimination cases is based on a framework called the "sex-plus analysis," a doctrine that was narrowly cabined in the face of the threat that the doctrine could disrupt traditional heterosexually-oriented gender distinctions. In contrast, national origin and race discrimination cases about race/ethnicity performance present an entirely different set of thorny concerns, namely, the proper scope of assimilation pressures and the social stratification effects of performance-based discrimination. This Part argues that because the sex-plus framework is not structured to address these concerns, it is wholly unsuited for understanding the issues raised in race/ethnicity performance discrimination cases. Special emphasis is placed on the problems caused by the immutability construct and the "preference" language borrowed from gender performance cases, as these rhetorical tools have caused courts to give short shrift to the social stratification concerns raised by discrimination triggered by race/ethnicity performance.
Section C of this Part outlines the consequences of the courts' misconceived application of the sex-plus framework in race/ethnicity performance cases. The discussion demonstrates that the propositions that inform sex-plus analysis do not hold true in race/ethnicity performance cases. Specifically, courts do not believe there is a general consensus about the social importance of preserving race or ethnicity or the manner in which they should be maintained (as opposed to gender); nor do they generally agree that employers can be trusted to [*1202] assist in the maintenance of these identity categories. Whether courts' assumptions about the law's role in the maintenance of gender are correct is another issue; however, sex-plus analysis is based on the proposition that courts and employers can be entrusted to identify and preserve gender categories in their current form. Indeed, Part III suggests that because the sex-plus framework sanctions and endorses the social maintenance of a particular consensus about a narrow band of gender difference, this framework should not have served as the template for understanding the much broader unresolved question about the social value of race/ethnicity performance.
A. The Importance of a Shift in Paradigm: Applying the Race/Ethnicity Performance Model
A shift in paradigm can have seismic effects on the understanding and explanation of legal problems. Vicki Schultz's work on the narratives employed in sex discrimination cases amply demonstrates this view. 253 Schultz has shown how the current focus on sexual harassment cases targeting conduct motivated by sexual desire tends to render invisible the experiences of women who face non-sexualized gender discrimination. The consequence is that courts and employers often try to squeeze these non-sexual claims into a desire framework. 254 In her other work, Shultz has demonstrated how courts' common sense arguments about women's interest in particular fields has tended to justify sex segregation in employment, with courts relying on the "lack of interest" trope rather than do the analytic work necessary to address these discrimination cases. 255 Collectively, this strand of her work demonstrates the importance of being vigilant in observing how the use of language, tropes, and certain constructs can limit our thinking in ways that cause us to overlook or misrepresent factors that play a central role in causing and maintaining discrimination. 256
The courts' reliance on a flawed paradigm in the race/ethnicity performance cases has had similar deleterious effects. Specifically, courts construing Title VII have treated ethnic and racial identity as having two dimensions: a "status" component, which refers to those characteristics that can be traced directly to the stigmatized morphological [*1203] features that define the group, and characteristics referred to as "conduct," that is, practices, aesthetics, and other traits that occur as a secondary consequence of an "identity" developing around the stigmatized feature. 257 They have concluded that Title VII protects only against "status"-based discrimination and is not concerned with discrimination triggered by "conduct," associated with a protected class. 258 As the status/conduct distinction has been applied in various Title VII cases, it has taken on a variety of rhetorical constructions. Specifically, courts are conducting an inquiry into the status/conduct divide when they attempt to distinguish between the "involuntary" attributes of a group and those which are "voluntary," 259 or, alternatively, when they distinguish between the "immutable" and "mutable" characteristics of a protected class identity. 260
The status/conduct distinction is problematic on a number of levels, the most obvious being that the line between voluntary and involuntary attributes is neither bright nor clear. Indeed, many of the so-called voluntary or "conduct-based" aspects of identity are extremely difficult to unlearn because routine practice has caused physical changes in a person's body (e.g., accents caused by the shape of one's palate) or because cognitive barriers develop over time. 261 Additionally, the status/conduct divide fails to consider the personal dignity concerns that inform race/ethnicity performance behavior. Plaintiffs rightly might question whether it is fair or appropriate to ask a person to abandon race-associated or ethnically-marked conduct when it does not interfere with the person's ability to do her job. They would argue that these features of identity should not be summarily disregarded because, when a morphologically-marked, stigmatized group develops a positive conduct-based component of its identity, the conduct often serves a special psychological purpose and therefore should enjoy special protection. 262 Specifically, this conduct may provide a kind of dignitary armor that allows a person to tolerate subtle discriminatory slights that the law does not address or provide [*1204] an independent basis for self-esteem that combats the negative stereotypes about the group in larger society.
Legal scholars also are critical of the status/conduct distinction, largely because of its conceptual instability. They explain that the status/conduct divide tends to be interpreted and rationalized differently, depending on the identity category at issue. 263 As applied to the Title VII context, it suggests that courts will articulate different rationales to explain why they exclude conduct or "voluntary" behavior from Title VII's protection for different protected classes. 264 Equally troubling, judges may borrow rhetorical constructions from one category of discrimination cases and inaccurately conceptualize the interests at issue in another category of discrimination cases, 265 often with disastrous effects for the parties involved.
The paradigmatic case on race performance is Rogers v. American Airlines, Inc., 266 where the natural/artifice distinction was first articulated. In Rogers, an African American woman brought a disparate impact claim challenging her employer's policy of prohibiting women from wearing all-braided hairstyles. 267 Rogers argued that the policy constituted discrimination because it violated black women's dignitary interest in wearing the cultural hairstyle. 268 Recognizing that her claim of race discrimination was not premised on race- [*1205] associated morphology, Rogers sought to establish that braids were a kind of race performance by explaining that black women in America perceived the hairstyle to be a means of expressing African American identity. 269 To support this view, Rogers noted that Cicely Tyson, a popular African American actress recently had worn braids to the Academy Awards as a sign of black pride and that this act had inspired her, and that the hairstyle historically had been worn by women in Africa. 270 The district court flatly rejected Rogers's claim, explaining that Title VII, like other antidiscrimination statutes, was designed only to protect against discrimination based on the immutable, biological characteristics that are constitutive of blackness. 271 The court explained that Title VII and section 1981 might prohibit discrimination based on the "Afro/bush" hairstyle because this hairstyle is a biological or immutable feature of blackness. 272 However, it [*1206] explained that an "all-braided hairstyle" was not a protected racial trait because it was an "'easily changed characteristic'" and "even if socioculturally associated with a particular race or nationality, [it was] not an impermissible basis for distinctions in the application of employment practices by an employer." 273 Additionally, the court scoffed at Rogers's evidence proffered to show that braids were a part of African American identity. It noted that a popular white actress (Bo Derek) recently had worn braids in the movie "10," and therefore the practice was not particularly constitutive of African American identity. 274 Consequently, American Airlines's no-braids policy did not trigger Title VII concerns. 275
The insights provided in the previous two Parts of this discussion explain the injustice of this view and lay the groundwork for a race/ethnicity performance analysis, which would produce a fundamentally different result. As a preliminary matter, I note that the disparate impact analysis theoretically should not have required such a strong showing about the immutable race-associated nature of Rogers's braids, as disparate impact analysis is designed to address policies that are neutral on their face, but tend for some reason to disproportionately compromise the interests of a protected group. However, assuming, arguendo, that the burden to establish a connection between race and a practice is relatively high in this analysis, the race/ethnicity performance analysis would allow that connection to be made without reference to immutability, and precisely because it is based on cultural practice. Indeed, the court would have avoided the woeful error it made in evaluating the evidence Rogers had offered to demonstrate the symbolic importance of the all-braided hairstyle. The court concluded that a single white actress's decision to wear a hairstyle in a film rebutted the more substantial historical evidence Rogers offered showing the long association between Africans and African Americans and braids. 276 While the transmission of cultural [*1207] or race/ethnicity-associated practices is to be celebrated, the race/ethnicity performance framework teaches that courts should be wary of defendants' attempt to use a lone example of an outgroup member who engages in the race/ethnicity-associated practice as a means to invalidate a plaintiff's claims. If the court had applied a race/ethnicity performance analysis, it would have been required to treat Rogers's proffer of evidence in support of the race-associated nature of the practice more seriously and proceed to the second phase of the inquiry: an investigation of the employer's justifications and motivations.
The race/ethnicity performance framework would also have warned the Rogers court to be wary of American Airlines's seemingly "neutral" justification for its policy and to avoid allowing general palliatives about professionalism to improperly short circuit the court's analysis. In the Rogers case, American Airlines never was asked to define what they meant when they required employees to wear "conservative" or "business-like" hairstyles, or to explain why it believed that if her braids were visible Rogers could not convey a business-like image. 277 As a consequence, the company was able to use the value of professionalism 278 (an undefined, highly subjective value) as a cover for a policy which expressed the company's hostility towards blacks. Even if we assume that some hairstyles might interfere with job performance, American Airlines should have been required to explain why visible braids simply could not be viewed as professional. Under a race/ethnicity performance framework, unless American Airlines provided some clear, well-defined, and supported reasons for why the plaintiff's hairstyle interfered with the projection of its image, the plaintiff should have prevailed on her claim.
McGlothin v. Jackson Municipal Separate School District 279 presents another opportunity to explore the insights that the race/ethnicity performance framework provides in these cases. McGlothin was an African American teacher's aide who brought a claim of religious discrimination under Title VII, alleging that her employer, a municipal school district, had subjected her to disparate treatment when it terminated her for wearing African head wraps and dreaded hairstyles as required by her Rastafarian and Hebrew-Israelite [*1208] beliefs. 280 The school alleged that her appearance violated its dress code and that she had never communicated the religious basis for her preferences, instead citing her reasons as being related to her normal "practice and heritage." 281 Although McGlothin submitted documentary evidence establishing that she had informed the district of the reasons she wore dreaded hairstyles, the court credited the testimony of her supervisor that McGlothin had never represented these activities as religious but rather as associated with race. The court therefore dismissed her claim, indicating that it did not raise Title VII concerns. 282
Once the technical legal justifications for the court's decision are set aside, the case shows the hallmark traits of a race performance scenario. The facts of the case indicated that McGlothin repeatedly was warned that her unkempt hair set a bad example for her young charges and responded accordingly by changing her natural hairstyle and intermittently wearing headwraps. The evidence provided in the case and the testimony offered would have strongly supported a claim of race performance discrimination. However, recognizing that these kinds of protections were not available, McGlothin sought to recharacterize her claims as religious discrimination.
When McGlothin's actions are viewed under the race/ethnicity performance framework, several points become clear. First, the notice problems that plagued her case evaporate, for she provided clear evidence that she gave her employer specific notice that the hairstyles she had chosen were part of an expression of her heritage and nothing was presented to refute this claim. 283 Second, the race/ethnicity performance framework focuses our attention on the context in which the dispute occurred and the event that caused McGlothin to increase her race performance behavior. A review of the facts shows that the precipitating event that encouraged her to adopt a "natural" hairstyle was the school's Black History Month celebration and its announcement that it was adopting a diversity initiative. 284 [*1209] McGlothin also began wearing headwraps to school more consistently and dreadlocking her hair after the district adopted a general program celebrating diversity. 285 Viewed in context, her behavior seems more logical; she felt safer in expressing ethnically marked aesthetic preferences because she believed that expressions of cultural or racial pride would be tolerated in line with the school's new policy.
Third, the race/ethnicity performance model suggests that McGlothin's behaviors must be viewed in the aggregate, rather than as single practices. Courts must understand that disparate aesthetic and behavioral choices may be part of a comprehensive effort to enact a racial/ethnic identity. Indeed, the facts of McGlothin's case suggest that she did not understand why her race-associated practices were offensive in light of the district's multiculturalism policy. Therefore, she vacillated between headwraps and dreaded hairstyles in an attempt to find some means of expressing her racial identity that did not offend her employer. Despite her efforts, she was chided for being unkempt and inappropriately dressed, and she was informed that her appearance violated the dress code. 286 Her attempts at accommodation and compromise were recast as willful recalcitrance.
Finally, the race/ethnicity performance framework suggests that the school district should have been required to explain why it perceived McGlothin's appearance to be in violation of the dress code, particularly in light of its newly adopted multiculturalism policy. It could not have prevailed based only on the assertion that dreadlocks are dirty, or that it perceived McGlothin's headwraps and cultural hairstyles to be unkempt, in part because these race-neutral justifications are disturbingly resonant with stereotypes about blacks, and it had no apparent basis for declaring McGlothin's hairstyles unsanitary or disruptive. Under the race/ethnicity performance framework, McGlothin presented sufficient evidence to establish a prima facie case of discrimination, and she should have been allowed to proceed to trial on her claims.
The framework proves equally helpful in addressing national origin or ethnic discrimination claims. For example, in Jurado v. Eleven-Fifty Corp., 287 a bilingual Latino disc jockey brought a Title VII disparate treatment and disparate impact claim against his radio station employer, seeking wrongful termination damages based on his employer's decision to fire him for his refusal to abide by an English-only rule when hosting his radio program. 288 When Jurado began his [*1210] tenure at the radio station, he hosted an English-only radio program and never expressed an interest in speaking Spanish on the air until his employer instructed him to do so. His employer hoped that by providing dual language programming, the radio station would attract more listeners. 289 Ultimately, the increase in listeners failed to materialize, and a consultant concluded that listeners were confused by the Spanish interludes and might lose interest in the radio station because of them. 290 Jurado, therefore, was ordered to resume using an English-only format. 291 When he refused to abide by the English-only rule, he was terminated. 292
On review of Jurado's claims, the district court granted summary judgment to his employer. 293 On appeal, the Ninth Circuit affirmed the lower court's decision, explaining that since Jurado was bilingual, his language choice was a voluntary, mutable characteristic, and therefore it did not concern Title VII. 294 The court concluded that Jurado's rights under Title VII were not violated when the employer ordered him to stop speaking Spanish on the air. 295
In Jurado, the race/ethnicity performance framework proves especially helpful, as the insights it provides again help explain the plaintiff's seemingly irrational behavior. The ethnic performance framework suggests that Jurado refused to stop speaking Spanish on the air because he perceived his employer's programming decision to be a status-based assault on the standing of his racial/ethnic group. Specifically, the model shows that Jurado likely believed that the radio station's abandonment of Spanish programming signaled either that his employer devalued Latinos or that he was pandering to racism. 296 As Jurado explained, the switch back to the English-only format felt like an attack on his identity: He argued that "it would have taken [his] character away." 297
Despite Jurado's hurt feelings, a court applying the race/ethnicity performance framework likely would conclude that the employer should have prevailed. When viewed in toto, the radio station's [*1211] actions do not suggest that its request for an English-only program was based on discriminatory intent. Rather, it was the radio station's program director's idea to add Spanish to Jurado's program, and the desire was motivated by an interest in making a special appeal to Spanish-speaking listeners. When the special appeal failed, the program director, logically, should have been permitted to require Jurado to return to an English-only format. 298 Also, unlike the scenario described in previous cases, in which employers attempted to justify discriminatory policies based on speculation about customer preferences or vague notions of "professionalism," in this case the programming director's actions were based on a study produced by a consultant who analyzed the relevant market demographics and came to a reasonable, well-supported decision.
The plaintiff's actions also must be considered. While it seems problematic to force ethnically marked workers to use these indicia to market products, this is not the issue presented here. Rather, in this case, Jurado voluntarily consented to allow his employer to use his Spanish-speaking capability in his radio program and, having agreed to use this performative behavior as a commodity, he could not cry foul when his employer decided that this commodity was not as valuable as it initially seemed. Jurado knowingly consented to allowing part of his identity to be used in a marketing strategy. Similarly, he should have known that his right to engage in this kind of ethnic display could be summarily terminated. In short, because the radio station's treatment of Jurado was not based on hostility to Latino culture, his claim of national origin discrimination was properly denied.
Although Jurado still loses on his claim under the race/ethnicity performance framework, the resulting decision produces an analysis that is much more responsive to the needs of the parties involved. The race/ethnicity performance analysis directly addresses Jurado's concerns about individual dignity and group status, as well as his employer's concerns about his marketing discretion. In contrast, in our current Title VII analysis, the decision turns on the irrelevant fact that Jurado is bilingual. The resolution of this claim under the race/ethnicity performance framework is not only more satisfying, but more logically defensible.
The Jurado case also provides an opportunity to consider some of the basic limiting propositions for the race/ethnicity performance framework, propositions which may comfort employers. The first limiting [*1212] proposition is that an employer can defeat a race/ethnicity performance discrimination claim when she provides a valid, objective, documented explanation for why the behavior compromises her ability to market her product or impedes the employee's job performance. The second limiting factor is based on individual dignity concerns. Stated simply, the employer is free to prohibit any race/ethnicity performance when the behavior tramples upon civil rights of other employees.
A clear example of this second limitation is presented in circumstances when race/ethnicity performance claims conflict with Title VII's goal of gender equality. For example, one can imagine that Clarence Thomas, under a Title VII regime that protected race performance behavior, might attempt to explain his alleged sexual harassment of Anita Hill as a moment of race performance, namely "down home courting." 299 His employer's sanction of the harassing behavior, under this view, would provide grounds for a Title VII race discrimination claim. However, under a race/ethnicity performance regime, Thomas loses. His employer would have been well within his rights to tell Thomas that such behavior violated Title VII because of its effect on Hill, regardless of its independent cultural standing.
Some may argue that the above cases paint an overly optimistic view of courts' potential to resolve cases using the race/ethnicity performance framework, and rather than being branded racist, courts automatically will assume a practice is race/ethnicity-associated for fear of being accused of insensitivity. However, this admittedly politically loaded landscape is easier to navigate than it seems. As shown by many of the cases previously cited, plaintiffs typically come forward with clear, specific references for their race/ethnicity-associated beliefs. For example, in McGlothin, the plaintiff provided Bible references and cited the tenets of Rastafarianism as a basis for her race performance behavior. 300 Similarly, the plaintiff in Rogers cited [*1213] examples from contemporary African American culture and the cultural practices of blacks living in Africa. 301 By requiring some evidentiary basis for the assertion that a performance is racially or ethnically marked, the inquiry becomes more analogous to religious discrimination cases under Title VII, which merely require that the employee identify the source of her belief and communicate the importance of the practice and belief to her employer. The employee must then demonstrate that she was terminated because she violated a policy that conflicted with or prohibited the practice in question, and was not provided with a reasonable accommodation. 302 It seems fair and reasonable to expect that those raising race/ethnicity performance claims would do the same.
B. The Danger Posed By Garcia v. Gloor: Extrapolating from Sex-Plus Analysis in Race and National Origin Cases
Given the limitations of the involuntary/voluntary framework in analyzing race/ethnicity performance claims, the question is: How did we become wedded to this paradigm? The Fifth Circuit was the first appellate court to apply the status/conduct distinction to a national origin claim in Garcia v. Gloor. 303 In Gloor, the Fifth Circuit affirmed a judgment in favor of an employer on a Mexican American employee's disparate treatment and disparate impact national origin discrimination claims. 304 The plaintiff alleged that speaking Spanish, although voluntary, was an essential part of Mexican American identity and, therefore, his employer's English-only rule constituted national origin discrimination. The Gloor court, in affirming the district court's decision, sanctioned the use of gender performance analogies to understand ethnic performance.
Specifically, the Gloor court began its analysis by noting that Title VII offers no definition of national origin and, therefore, no insight into the connection between voluntary behavior and national origin identity. 305 Additionally, it recognized that there was nothing [*1214] in the legislative history of the statute that defined national origin. 306 It then compared the case to Willingham v. Macon Telegraph Publishing Co., 307 a recent en banc decision that concerned a challenge to an employer's ability to promulgate gender-specific grooming rules. The Willingham court held that, unless a Title VII claim concerned a fundamental right or an immutable characteristic, it did not state a claim. 308 Applying this framework, the Gloor court ruled that since the plaintiff did not raise a claim about an immutable feature associated with national origin or a legally established fundamental right, his claim must fail. 309 Importantly, the opinion never explicitly stated that it was applying a "national origin plus" analysis to analyze the plaintiff's claim. Structurally, however, it mirrors the sex-plus analysis offered in Willingham.
By relying uncritically on the Willingham decision, the court caused race/ethnicity performance analysis to be shaped by a doctrinal field that was preoccupied with a much different question; namely, to what degree was Title VII intended to transform gender categories? Before Willingham, sex-plus analysis simply provided that when an employer instituted a rule that used sex plus another neutral characteristic to discriminate against a subclass of women, this could constitute gender discrimination, particularly when it appeared to reinforce stereotypes about women. 310 Initially, the employer rules challenged under the sex-plus analysis concerned obvious stereotypes about women, such as rules preventing women of childbearing age from working in certain employment, or rules preventing women with children from qualifying for certain jobs. 311 However, litigants soon began using these claims to challenge grooming codes intended to maintain gender differences (i.e., rules that discriminated against effeminate men or masculine women). These litigants interpreted the standard in a manner that allowed them to bring a Title VII sex-plus claim based on any neutral employment rule that distinguished between subclasses of men or women and enforced a sex stereotype, [*1215] even an appearance-based one. 312 In these appearance cases, the plaintiff would attempt to show the invalid nature of the prohibition being applied to her group by showing the employer permitted the very same practice for women that it had prohibited for men. For example, a plaintiff might challenge a rule indicating that no men with earrings could be hired, by showing that the employer hired women with earrings. The Willingham plaintiff raised a claim based on this broader version of the sex-plus analysis. He argued that the telephone company he had applied to for employment had discriminated against him based on sex because it had declined to hire him under a sex-specific grooming code rule that prohibited the hiring of men with long hair. 313 He alleged that the rule discriminated based on sex plus a neutral characteristic (hair length) and enforced a stereotype about men. He noted that women were allowed to have long hair, and therefore the employer had no legitimate basis for prohibiting the practice for members of his gender.
From a legal realist perspective, the Willingham decision must be understood as a maneuver to limit the legal and social repercussions of the "sex-plus" analysis to ensure that it did not become a tool that required employers to participate in the fundamental dismantling of the aesthetics of gender. Although Willingham's claim comported with the general guidelines for a sex-plus challenge, the Fifth Circuit was unprepared to tell employers that they could not enforce grooming codes that preserved certain gender differences. 314 Therefore, the court's project in the Willingham case was to interpret the sex-plus analysis in a manner that afforded employers discretion to regulate the aesthetics associated with gender, but also ensured that the analysis still could be used for other sex-stereotyping challenges. In order to achieve this goal, the Willingham court held that unless a Title VII claim concerned a fundamental right or immutable characteristic, it must fail. 315 It recognized that some of the most odious discriminatory employment rules concerned stereotypes about procreation, or the fundamental right of childbirth; therefore, it concluded that as long as the sex-plus analysis prohibited discrimination based on a fundamental right or an immutable feature, it still would be able [*1216] to address the problem of gender discrimination. 316 In analogizing the Willingham and Gloor plaintiffs' claims, the Gloor court failed to consider the practical purposes served by the Willingham decision.
Indeed, as the rest of this Part will explore, when the Gloor court sought guidance from the Willingham decision, it created a host of new problems because it unquestioningly borrowed the constructs and rhetoric of Willingham without considering whether they fully captured the political issues and values that inform race and national origin discrimination cases. First, the Gloor court failed to recognize that the Willingham sex-plus analysis was structured by social anxieties unique to the issue of gender performance. Second, the Gloor court adopted the Willingham court's understanding of the immutability construct. Instead of using immutability to identify groups in need of protection, it used the construct to identify traits held by protected classes that deserved protection. As a consequence, the court began using the immutability construct to shave off portions of protected class identities from statutory protection. The third error was that the Gloor court incorporated rhetoric from the Willingham decision equating performative behavior with the expression of "preferences," not recognizing that these comments in the gender cases were in response to specific autonomy and freedom claims raised with regard to grooming codes and their impact on gender diversity. Each of these errors is discussed in more detail in the sections that follow.
1. Sex-Plus Analysis and the Dissolution of Gender
The court of appeals in the Willingham case began its analysis with two foundational premises that, when laid bare, raise serious questions. Its first assumption was that there existed a social consensus about the need to preserve heterosexually informed gender categories. 317 The court's second assumption was that employers could be trusted to assist in maintaining heterosexually informed gender identity categories. 318 The district court, in its earlier review of the case, makes these concerns explicit and clear. The court explained that it feared a world in which "employers would be powerless to prevent extremes in dress and behavior totally unacceptable according to [*1217] prevailing standards and customs recognized by society." 319 Moreover, the court warned that:
if it [were] mandated that men must be allowed to wear shoulder length hair ... because the employer allows women to wear hair that length, then it must logically follow that men, if they choose, could not be prevented by the employer from wearing dresses to work if the employer permitted women to wear dresses... . It would not be at all illogical to include lipstick, eyeshadow, earrings, and other items of typical female attire among the items which an employer would be powerless to restrict to female attire and bedeckment. It would be patently ridiculous to presume that Congress ever intended such result ... . 320
The assumptions made above about the propriety of maintaining gender categories and employers' role in this process seem, on their face, immediately subject to dispute - that is, as soon as one considers the needs and interests of gay Americans or anyone with strong autonomy views about the performance of gender. Indeed, while the district court dismisses the idea of men in earrings and makeup as "ridiculous," many men (gay and straight) already may find pleasure in such adornments. Rather than presenting a ridiculous and tangential concern, a victory for the plaintiff in Willingham would have had substantial repercussions, giving individuals broad license to disrupt gender categories. I submit that the assumptions the court made about the maintenance of gender were only logical because the court rendered certain communities invisible and disregarded their interests. Also, I submit that these assumptions only seemed plausible because the subject matter being regulated was the aesthetic content of gender categories, and the court could not contemplate how regulation of this aspect of social life could have serious material repercussions or have any impact on financial realities.
Having laid bare the anxieties and concerns that surround gender performance cases, it is clear they have little or nothing to do with the anxieties triggered by the plaintiff in Gloor or, for that matter, any kind of race/ethnicity performance. Rather, the Gloor complaint was based on the concern that rules prohibiting ethnic performance in themselves inflicted a status-based harm on the plaintiff's minority group and would have serious social stratification repercussions. The concerns in Willingham are about exclusion as well, but are, in that case, pitched in the language of individual autonomy, self-determination, and freedom of expression.
[*1218] Also, in a critical miscalculation, the Gloor court failed to realize that persons raising race/ethnicity performance discrimination claims find that the social attitudes about difference and employer discretion are exactly the opposite of those that hold for gender performance cases. Specifically, even if there is some consensus about the value and preservation of heterosexual gender categories, there is no such consensus about the value of racial or ethnic categories or whether they should be preserved. Also, assuming we do believe that race and ethnicity are worthy of preservation, there certainly is no consensus that employers should play a key role in regulating and maintaining these identity categories.
Indeed, one of the critical errors in the Gloor court's analysis is the assumption made about employers' attitude towards difference. Employers in gender cases are assumed to be enforcing rules that encourage and maintain a given set of established gender identity categories or differences and to prevent a blurring of those categories. In contrast, the rules employers enforce in race/ethnicity performance cases typically are designed to quash expressions of ethnic or racial difference in favor of maintaining an "unmarked" baseline culture of the workplace, which is typically Anglo or European. Stated more simply, one typically does not encounter employer rules that explicitly subsidize a given form of ethnic identity or rules designed to maintain differences between ethnic groups. One does, however, find workplaces that subsidize a certain kind of gender performance (e.g., rules that require women to wear skirts). Ironically, the discretionary authority that employers are granted in gender cases under the "plus" analysis to maintain a limited specific form of difference for each gender actually allows employers in the race/national origin cases to demand that racial and ethnic difference be wholly eradicated. 321 Of course, one could characterize employer rules about cultural difference in an alternative manner, and argue that they actually subsidize a certain kind of ethnic identity. For example, if an employer can create a policy that discourages blacks from wearing dreadlocks, she is arguably subsidizing some other presentation of black identity. However, [*1219] under the current regime, since she can do this for every visible voluntary identifier of a subgroup, the regime effectively allows her to prohibit cultural difference entirely.
Because the Gloor court failed to acknowledge the Willingham court's investment in heterosexually informed gender categories and the unique anxiety caused by the prospect of disrupting traditional gender aesthetic standards, it applied Willingham's analytics to a race/ethnicity performance case with no modifications. In doing so, it created a world in which employers are granted broad authority to manipulate the performance of national origin and race, as long as these regulations are drafted in colorblind terms.
With these insights, the resolution of the Gloor case makes perfect sense. Applying the "plus" framework, the Gloor court ultimately grants the employer authority to regulate the performance of Latino identity because the regulation appears as a neutrally formulated English-only rule. While this was not the Gloor court's stated intention in the case, once it borrowed the sex-plus framework to analyze ethnic performance issues, this result was a foregone conclusion. 322
2. The Immutability Requirement
The second mistake in Gloor is that the court adopts the proposition that only "immutable" features of protected class identity are entitled to Title VII protection. Again, citing Willingham as the basis for its decision, the Gloor court explains:
Save for religion, the discriminations on which [Title VII] focuses its laser of prohibition are those that are either beyond the victim's power to alter, or that impose a burden on an employee on one of the prohibited bases. No one can change his place of birth (national origin), the place of birth of his forebears (national origin), his race or fundamental sexual characteristics. 323
Although neither Gloor nor Willingham cites a case in support of this proposition, the rhetoric and language used seems suspiciously similar to that employed in Frontiero v. Richardson, 324 an equal protection case that introduced "immutability" as a partial justification for extending special protection to women, first under the Fifth Amendment of the Due Process Clause, and later as its logic was extended and interpreted under the Fourteenth Amendment. Other scholars have commented on how the immutability construct has been given an unnecessary and unintended prominence in equal protection analysis and has become a substantial barrier to new groups finding protection under these amendments. 325 In the Title VII context, the immutability construct has a different, but equally restrictive and aggressive, role. It works to limit claims within protected classes, screening out certain aspects of protected class identity from statutory protection.
Indeed, the immutability construct's screening role is apparent in Gloor, as the court uses the construct to establish a distinction between the plaintiff's language abilities and his ethnic status, thereby denying his national origin discrimination claim. 326 Ironically, however, even as the Gloor court introduces the immutability construct into the discussion of ethnic performance, it is unprepared to accept its full repercussions. The court recognizes that once it severs language from ethnic identity, monolingual Spanish speakers who suffer discrimination on this basis are left without adequate discrimination protections. 327 To avoid this problem, the court distorts the immutability construct to address these concerns. The Gloor court explains:
To a person who speaks only one tongue or to a person who has difficulty using another language than the one spoken in his home, language might well be an immutable characteristic like skin color, sex or place of birth. However, the language of a person who is multi-lingual elects to speak at a particular time is by definition a matter of choice. 328
As this excerpt shows, the Gloor court refuses to apply faithfully the Frontiero court's definition of "immutability," as describing those characteristics that are "determined solely by the accident of birth." 329 Instead, it treats language capacity as both mutable and immutable. For the bilingual plaintiff, it argues that language is mutable, because she can switch back and forth between languages and therefore cannot be injured by an English-only rule. For the monolingual speaker, however, it argues that language is immutable and therefore can be the basis for a Title VII claim. The Gloor court, however, should not be able to have it both ways. Language capacity either is immutable or not. In the alternative, the court should have confronted more directly the fact that the immutability construct was inappropriate for addressing the issues raised by the case.
The court's doublespeak about immutability is directly tied to the issue that should have dominated the court's discussion of the Gloor plaintiff's claim: the potential socially stratifying effects of rejecting race/ethnicity performance discrimination claims. The Gloor court was aware that if it announced a standard that did not protect monolingual workers from language-based discrimination, its decision likely would have severe stratifying effects; consequently, its goal was to preserve Title VII claims for these workers. 330 If the court had announced that language capacity, uniformly and in all cases, should be treated as a mutable characteristic, then employers could refuse to hire large numbers of monolingual, Spanish-speaking immigrants under an English-only rule, even if English language proficiency was not a requirement for the job at issue. This ruling would have catastrophic effects for many of the most marginalized and vulnerable workers - workers from ethnic enclaves where Spanish is the only language spoken or newly arrived immigrant workers. Because the Gloor court felt compelled to resolve this issue using an immutability analysis, it could not give voice to these social justice concerns in its discussion.
3. Preference Rhetoric
The last concern raised by Gloor stems from its adoption of a rhetorical formulation used in the Willingham decision - the distinction between rights and "preferences." In Gloor, once the court concluded that the plaintiff's claim did not concern a fundamental right or an immutable feature, it decided that his language claim was simply a [*1222] "preference," a "de minimis" issue beyond the scope of Title VII. 331 Commenting on Garcia's "language preference," the Gloor court explains:
That this [English-only] rule prevents some employees, like Mr. Garcia, from exercising a preference to converse in Spanish does not convert it into discrimination based on national origin. Reduced to its simplest, the claim is "others like to speak English on the job and do so without penalty. Speaking Spanish is very important to me and is inherent in my ancestral national origin. Therefore, I should be permitted to speak it and the denial to me of that preference so important to my self-identity is statutorily forbidden." 332
This discussion about language preference mirrors the discussion about appearance preferences in Willingham. 333
Certainly, one could argue that this "preference" rhetoric was appropriate in Willingham, as the plaintiff's gender performance or "hair length" claim primarily rested on concerns about the freedom to transgress gender norms as a matter of individual expression or autonomy. 334 However, the Gloor plaintiff was making a different argument. He argued that Spanish was already a historically established constitutive aspect of his ethnic identity, the regulation of which had potentially social-stratifying repercussions. 335 He raised the concern that the prohibition of this highly common feature of Latino identity might function as a proxy for hostility against the group itself. 336 By dismissing his interest in speaking Spanish as a preference, the court rhetorically individuates his claim and avoids any [*1223] detailed treatment of the broad effects of the English-only rule on Mexican American workers as a group.
C. Second Order Cases: The Repercussions of Gloor
Although not all of the race/ethnicity performance cases explicitly acknowledge their debt to the Gloor court in charting the unknown waters in race/ethnicity performance analysis, one can see the problems inherent in the Gloor court decision reflected in numerous cases issued after the decision. In those that explicitly acknowledge these connections, the repercussions of its faulty logic are particularly clear.
1. Garcia v. Spun Steak Co.: The Rhetoric of Choice
Indeed, because Gloor was the first case to apply the status/conduct distinction to race and ethnic identity, it has played a seminal role in cases on this issue. For example, in Garcia v. Spun Steak Co., 337 the Ninth Circuit deployed the Gloor court's "preference rhetoric" to defeat another Mexican American plaintiff's national origin challenge to an employer's application of an English-only rule. The Spun Steak court, however, elaborated on the Gloor analysis, worrying that the plaintiff's claim about language rights and "preferences" was the beginning of a wave of minority plaintiffs' claims seeking "special rights" in the workplace. The Spun Steak court explained:
It cannot be gainsaid that an individual's primary language can be an important link to his ethnic culture and identity. Title VII, however, does not protect the ability of workers to express their cultural heritage at the workplace. Title VII is concerned only with disparities in the treatment of workers; it does not confer substantive privileges. It is axiomatic that an employee must often sacrifice individual self-expression during working hours. Just as a private employer is not required to allow other types of self-expression, there is nothing in Title VII which requires an employer to allow employees to express their cultural identity. 338
In making this argument, the Spun Steak court refused to consider the cultural context in which the English-only rule operated. The plaintiff in Spun Steak specifically argued that he was seeking the vindication of a neutral right already afforded other workers - the [*1224] right to speak in the language with which he was most comfortable. His goal was to address the interests of a class of Mexican American workers who were unfairly burdened by the English-only rule because they were forced to communicate in a language with which they were less familiar. 339 The court, however, reversed the district court's decision for the Spun Steak plaintiff. In doing so, it rejected the idea that rules drafted in "neutral" terms might provide certain workers (in this case, English-speaking workers) with substantive privileges and special rights. Indeed, in this case, the English-only rule was neutral only in the formal sense that it imposed the same requirements on all employees. However, the Spun Steak plaintiff could have used the court's own logic to argue that the English-only rule was an attempt to create a class of special rights for English-speaking Americans by prohibiting any worker in their presence from speaking a foreign language. 340
Importantly, the preference rhetoric that the court pulled from the gender performance cases was a result of the kinds of arguments raised by gender performance litigants. Indeed, analysis of the gender performance cases shows that they frequently have featured arguments about autonomy and freedom of expression. 341 In these cases, plaintiffs allege that by regulating the practice of gender with sex-specific workplace rules, employers deny employees the opportunity to [*1225] experiment with their gender identities. In order to explain why these expressive concerns are an insufficient basis for a Title VII claim, the court describes them as "preferences," 342 a construction which both individuates these claims and suggests that they have limited symbolic import. The problem with using the preference rhetoric in race/ethnicity performance cases, however, is that it makes these claims seem like individual squabbles between employer and employee instead of symbolic cultural contests. 343 Indeed, the preference rhetoric, by focusing attention on the individual, masks the fact that these rules have broad repercussions for entire classes of workers, resulting in decreased opportunities for those who find it hard to abandon these behaviors and exacting a high dignitary cost on those who are compelled to give them up.
2. The Rogers Decision: The Rhetoric of Immutability
In contrast to the national origin cases, where the focus was on preferences, when the Willingham decision was applied to race performance claims, courts tended to focus on its discussion of immutability. Rogers v. American Airlines, Inc., 344 a case earlier discussed in this Part, provides the clearest example. In Rogers, the court transforms the immutability construct into an analytical device called the "natural/artifice" distinction. 345 The court rejected the plaintiff's discrimination claim, in which she alleged that her employer discriminated against her because of her all-braided hair style. 346 In the course of rendering its decision, the court noted that Rogers's braids must be distinguished from some involuntary, immutable race-associated trait, such as the "Afro." 347 Given that Rogers could have chosen to style her hair in a manner that did not offend her employer but refused to do so, her race discrimination claim failed. 348
Similar to the Gloor case, the distinction the Rogers court attempted to establish between the immutable and voluntary features of an identity broke down even as it was articulated. The court failed to account for the fact that there is a broad range of morphological [*1226] diversity within African American communities and that hair texture is not consistent across the group. Some persons who bear morphology that causes them to be categorized as black are born with Afros; some are not. The Afro is simply not a constitutive morphological marker of blackness. Additionally, the court failed to recognize that the Afro, at that time, was what I have referred to here as an "active race performance" feature. Many people voluntarily had attempted to create and cultivate the hairstyle for social and political reasons. Indeed, in the only case prior to Rogers that mentioned the specter of Afro discrimination, the plaintiff had voluntarily taken on the hairstyle after having worn her hair in another style for years. 349 The Rogers court, however, ignored this problem and endorsed the fiction that there are natural and artificial features of black identity. Other courts adopted a similar analysis and used variants of the natural/artifice distinction to reject minority employees' Title VII claims challenging the employer rules preventing them from wearing braids, 351 and headwraps. 352
3. Echoes of Frontiero in the Race/Ethnicity Performance Cases
Rogers is helpful for this analysis because it is the most explicit about the fact that its discussion of immutability is derived from Frontiero v. Richardson. 353 Although the court never refers to Frontiero by name, the Rogers decision borrows from its language and logic, using Frontiero's explanation of why certain groups are offered special protections to identify those traits persons in protected classes [*1227] have that are subject to protection. As this Section shows, with a better understanding of Frontiero, it becomes clear that the courts' application of the immutability requirement in Title VII jurisprudence rests on a fundamental error about the role of the concept in equal protection jurisprudence.
Although it was decided under the Fifth Amendment, Frontiero has proven to be a seminal Fourteenth Amendment equal protection case, as its antidiscrimination logic was applied to regulate the states. In the decision, the Court explains why women, like blacks and other protected classes, are entitled to the Constitution's equal protection guarantees. 354 Specifically, in Frontiero, the Court reviewed a servicewoman's challenge to a United States Army rule that provided that wives of servicemen were presumptively dependents and automatically provided with medical benefits, but husbands of servicewomen were not dependents unless it was established that they relied on their wives' salaries for more than fifty percent of their support. 355 The servicewoman plaintiff alleged that this benefits rule arbitrarily deprived her of property in violation of the Fifth Amendment's Due Process Clause because it discriminated based on sex. 356 Because a government agency typically need only show a rational relationship between its classification system and its purpose in order to withstand a Fifth Amendment challenge, the servicewoman was required to demonstrate that women were a special "protected class," and therefore regulations based on sex should be subject to more stringent scrutiny. 357
In the course of its discussion, the Frontiero Court offered the definition of "immutability" that informs Title VII cases. The Frontiero Court explained that protected classes share one commonality: They are wedded by "immutable characteristics determined solely by the accident of birth." 358 The Frontiero Court explained that the reason equal protection law shields these groups from societal discrimination is because of the moral view in America that one's opportunities should not be constrained by features that bear no [*1228] relationship to ability and are beyond one's power to control. 359 Because sex, like race, is immutable, the Court concluded that classifications based upon sex deserved heightened scrutiny, 361
The Rogers court's reliance on this Frontiero-inspired logic was understandable because of Title VII's relationship to the Fourteenth Amendment, and Frontiero has been a guiding reference for Fourteenth Amendment jurisprudence. Again, although the court does not explicitly cite the case, it assumes that protected traits are only those that one possesses by accident of birth - and cannot change - making its focus immutability. However, by relying exclusively on this feature, the Rogers court obscures the fact that protected classes are defined by more than their immutability. Indeed, in addition to immutability, courts inquire into the political power of the concerned group; 363 whether the immutable trait is visible; 364 and whether the immutable trait validly may be treated as relevant in the assessment of inherent qualities. 365 Indeed, equal protection cases that turn on these other considerations [*1229] reveal that the immutability requirement, standing alone, is insufficient to establish that a group is entitled to special protection. 366
Consequently, when the Frontiero decision is placed in its proper context, it is clear that the immutability construct should not serve as a barrier to race/ethnicity performance discrimination claims. Read broadly, the decision merely identifies one of several criteria necessary to demonstrate the right to special antidiscrimination protections. Read more narrowly, it identifies the minimum qualifying characteristics a group must possess in order to merit special protection. However, nothing in the Frontiero decision establishes that the immutability factor was intended to be used against already recognized protected classes as a means to sift out which features of their identity should be shielded from discrimination. Stated alternatively, nothing in the decision suggests that we should take one aspect of protected class status and assume that protected classes have no other dimension apart from that discussed in the articulation of their qualifying characteristics. 367
The above discussion provides a detailed analysis of the repercussions that the gender performance constructs borrowed from Willingham had for the race/ethnicity performance cases. However, I emphasize that by understanding the basic principles that informed the Willingham case, the Court could have avoided these errors, and determined that this approach was not the proper framework for understanding race and national origin performance claims. 368 Stated simply, Willingham is based on the proposition that we have a fundamental commitment to the maintenance of gender categories, and, [*1230] moreover, that we are comfortable with employers exercising their discretion to manipulate these categories. The Gloor court, therefore, had an obligation to ask: Do we have the same understanding about race and ethnicity? Are we similarly committed to the maintenance of these social categories? If so, are we comfortable with or confident in the idea that employers can assist us in this goal? The Gloor court failed to ask these questions before it applied the sex-plus analysis to Garcia's language claim. If it had, it would have realized that the sex-plus framework assumes a different social orientation towards difference, and it should not have served as the template for understanding the social value of race/ethnicity performance.
Part IV addresses the most likely concerns about race/ethnicity performance protections. Section A addresses arguments that these protections are contrary to the rules of statutory construction. Section B turns to concerns that the protections will compromise employers' ability to market their products or discipline their employees, as well as unnecessarily increase their potential for Title VII liability. Lastly, Section C considers the political question of whether these protections are contrary to America's goal of creating a unified, cohesive citizenry, one that transcends the dangers of identity politics.
A. Legal Concerns
Legal concerns about race/ethnicity performance protections form two groups. The first claims that the analysis goes too far and invites judges to transform an issue by judicial fiat that should be left to the legislature. The second group suggests that these protections do too much of the same, by further instantiating racial and ethnic categories, a dynamic which cannot advance us in the project to end discrimination. Each group of concerns is explored in the Section that follows.
1. Statutory Construction Claims
Critics may argue that whatever improvements the race/ethnicity performance regime might bring to our current antidiscrimination regime, the model simply cannot be used because there is no authorization for it in Title VII's language or legislative history. 369 According [*1231] to this view, when courts create doctrine to apply statutes, they are simply giving effect to congressional mandates; they are transforming abstract concepts into tests that can be applied to assess the array of particular problems within the statute's ambit. 370 In the case of Title VII and its provisions on race and national origin, the absence of any mention of voluntary race performance or ethnic-associated behavior bars courts from creating doctrine under Title VII to address these concerns.
When framed in this manner, the statutory construction argument seems impossible to overcome. Without question, Congress left no indication that it considered discrimination against race/ethnicity-associated behavior to be a legitimate source of governmental concern. 371 However, the import of Congress's silence seems less clear when one recognizes that it also nowhere indicated that courts should be concerned exclusively with discrimination based on morphologically associated racial and ethnic characteristics. Since Congress has not explicitly defined race, national origin, or the concept of discrimination in terms that distinguish between morphological markers of racial or ethnic status and voluntary and behavioral features, any prior judicial pronouncements on these issues are only rationalizations of "common sense" views about race and ethnicity during the period in which they were made. 372 Consistent with this view, Part III showed that the doctrine which establishes that Title VII is concerned only with morphological race/ethnicity-associated traits is actually a product of a misguided period of judicial activism in the 1980s, when courts were preoccupied by concerns about creating special protections for groups. 373 Therefore, when critics argue in favor of maintaining [*1232] the "original" definitions under Title VII for race, ethnicity, and discrimination, they are actually making an argument in favor of preserving the judicial interpretation of race or ethnicity during a particular historical moment. We cannot, however, flatly conclude that earlier judicial definitions capture these concepts adequately or disregard the understandings of the current period.
Indeed, Reva Siegel argues that we must maintain a critical perspective, one that recognizes the historical situatedness of our anti-discrimination efforts and their potential limitations when confronting new problems. She argues that antidiscrimination law must respond to changes in the nature of discrimination if we are to achieve our goal of equality. 374 Siegel explains:
The body of equal protection law that sanctioned segregation was produced as the legal system endeavored to disestablish slavery; the body of equal protection law we inherit today was produced as the legal system endeavored to disestablish segregation. Are we confident that the body of equal protection law we inherit today is "true" equal protection, or might it stand in relation to segregation as Plessy and its progeny stood in relation to slavery? 375
Siegel's observations are not intended to diminish the critical role that statutes like Title VII have played in disrupting workplace discrimination. Rather, her goal is to cultivate both a recognition of the historical context of any definition of race, ethnicity, and discrimination, and the willingness to question these definitions. Once one recognizes the distinctions in the way discrimination is defined over time, it seems appropriate - even necessary - to ask whether our current definition actually captures the phenomenon that is the subject of our current concerns, or if it was simply developed with an eye towards disrupting the most common, prevalent manifestation of discriminatory attitudes when the definition was crafted. The same must be asked of Title VII doctrine. The concept of discrimination that is addressed by today's Title VII was developed in a period when the most pressing issue was that people of color and ethnic minorities simply could not gain access to the workforce at all. Now, forty years later, having made partial inroads, Title VII should be concerned with [*1233] assuring that these workers have equal dignitary rights and inquiring into other ways hostile employers might seek to discourage or thwart minority participation.
Armed with these insights, the statutory constructionists' reification of the current doctrine on discrimination seems less defensible, yet the consequences are enormous. When definitions of discrimination change due to shifts in American culture, concepts of race and ethnicity change as well. 376 Relatedly, as our understandings of assimilation and integration change, race and ethnicity definitions change as well. These cultural shifts are neither clean nor easy, and they develop over time. 377 However, as litigants have discovered, these shifts in cultural understanding typically outpace shifts in the law's understanding of these issues. 378 Part II of this Article provided some insight into how in the late nineteenth and early twentieth centuries, a variety of institutional actors, including biologists, social scientists, and judges, were called upon to justify cultural understandings codified by the legislature and proved quite useful in this role. 379 I suggest, however, that the problem is that non-judicial institutions are more responsive to changes than the judiciary, which on the whole tends to continue to build additional justifications on existing frameworks unless the legislature intervenes. The consequence is that the law quite often reflects an antidiscrimination orientation that is outdated and at odds with contemporary antidiscrimination logic.
Consider, for example, that even as the 1964 Civil Rights Act was being first introduced, several identity politics movements were gaining force that emphasized the importance of maintaining and celebrating difference. 380 However, as I read them, the Civil Rights statutes and the doctrine developed under them were still wedded to a concept of "assimilation" that was fundamentally at odds with the identity politics movements that had begun to mobilize. The understanding of discrimination held by judges, as illustrated in Part III, was that the statute was merely intended to be applied to address discrimination based on physical, immutable difference. Title VII then was interpreted and applied based on a "melting pot" orientation which encouraged assimilation, and was hostile to cultural difference. Under this model, race/ethnicity-associated behavior was an obstacle to be overcome. The intended message was that although racial and ethnic [*1234] difference was real, the law was blind to these differences because they bore no relationship to any meaningful measure of worth.
The Frontiero Court gave full expression to this view when it explained that "race and national origin [are] ... characteristics determined solely by the accident of birth" and "the imposition of special disabilities upon the members of a particular [group] because of ... [race or] sex would seem to violate 'the basic concept of our system that legal burdens should bear some relationship to individual responsibility ... .'" 381 Additionally, the Court explained that race-and sex-associated morphological traits "bear no relation to ability to perform or contribute to society" and antidiscrimination protections are designed to protect against those traits "having the effect of invidiously relegating the entire class ... to inferior legal status without regard to the actual capabilities of its individual members." 382
Importantly, the melting pot model described above still treats race and sex as fundamental biological differences, assigned "solely by the accident of birth." 383 Therefore, the antidiscrimination laws as interpreted under this model provided that employers could discourage or penalize persons when racial or ethnic difference appeared to be a voluntary feature and was not biological. In fact, these workplace sanctions were treated as a necessary part of assimilating non-whites into American culture. Minorities were, in essence, told that in exchange for antidiscrimination protections based on racial or ethnic status, they were required to shed voluntary or cultural markers of racial and ethnic identity and adopt the cultural demands required for public life, specifically those mandated by employers. Workers who refused to accept these terms were required to bear the cost of their recalcitrance, namely, social marginalization. 384
By the mid 1990s, however, Americans had decisively shifted from the melting pot paradigm to the "salad bowl" or mosaic paradigm. 385 Scholars and politicians began to suggest that it was wrong to ask racially and ethnically marked persons to shed their distinctiveness [*1235] both because of dignitary concerns 386 and because this difference could be harnessed to make America a smarter and stronger economic force in the world community. 387 The mosaic paradigm was the logical outgrowth of a number of diversity initiatives that began after the Civil Rights movement and grew in prominence over time. 388 The new model's benefits for ethnic minorities were clear. The model required that cultural differences were not automatically treated as regressive and opened the possibility for accommodation of these differences. 389 The doctrine under Title VII, however, failed to change in conjunction with this shift, which I believe resulted in the dynamic observed in many of the cases discussed in Part III. Specifically, courts cracked down on employees who were acting based on an expectation that Title VII would be responsive to the values of diversity and multiculturalism and protect them from discrimination based on voluntary difference. 390 Importantly, even as this Article encourages judges to consider this diversity or multicultural model in their interpretations of antidiscrimination laws, I am also wary of the diversity model's limits, as it opens the door to a variety of new antidiscrimination problems that must be negotiated carefully. Several of these diversity-related problems are identified below.
First, the diversity paradigm tends to assume that behavioral-or practice-based differences associated with subgroups are "cultural" and have some substantive content. Some of these differences, however, are not actively mobilized as conscious symbolic gestures, but are "accidental" characteristics that develop as a consequence of segregation. Title VII at present does not indicate whether these "accidental" traits should be afforded the same kind of antidiscrimination protection as we would provide for cultural difference, or provide any guidance on whether they can and should be discouraged.
Second, the diversity paradigm turns race-and ethnic-associated difference into a commodity; it argues that these differences, once harnessed, will strengthen the community. Now Americans are told [*1236] that diversity is not a "problem" to be managed, but rather is a resource for communities and employers. The model, however, fails to consider that employers might demand that workers of color engage in certain kinds of race/ethnicity performance that workers find objectionable, while simultaneously refusing to compensate the employee for engaging in this performance, resulting in a kind of exploitation. Under the current Title VII regime, employers could theoretically encourage, and even require, an employee to engage in race/ethnicity performance at work, safe in the knowledge that the law's refusal to "protect" these aspects of identity gives them the discretion to crack down on these same behaviors as "disruptive" when they fail to provide an economic payoff. 391 The diversity paradigm on its face provides no basis for resolving these disputes.
Third, this paradigm may tend to increase cultural status contests in the workplace as each group, told that its culture and views are of equal worth, demands equal representation, resources, and time. Indeed, some of the cases discussed in this Article stem from the fact that employees felt slighted under diversity initiatives. 392 The diversity model, at present, is ill-equipped to handle these disputes.
Fourth, the diversity model does not consider how to resolve conflicts when the cultural imperatives of two groups are inconsistent, and one group's perspective must be regulated or limited in order to preserve the peace, promote efficiency, or protect the rights of an individual worker. 393 This area, again, promises to be an important antidiscrimination frontier.
In spite of its problems, the diversity model still holds substantial advantages over the assimilation model. Under the assimilation model, one cultural perspective dominates all others, and employers have enormous latitude to force employees to behave in ways that are consistent with this hegemonic cultural perspective. My hope is that the journey we have taken through the timeline of Americans' understandings about assimilation and discrimination suggests that any judicial decision based on the assimilation paradigm will not only seem disrespectful of subgroup difference, but, also, will fail to capture the [*1237] concerns that should inform disputes about the management of diversity. If Title VII is interpreted using a diversity model, the primary question will be: How do we regulate the chorus of Lilliputian voices clamoring for recognition and respect? The race/ethnicity performance framework articulated here tries to provide some preliminary answers to that question.
The short response to the statutory constructionists is that the legitimacy of constructs which exempt voluntary aspects of race and national origin identities from protection are already questionable, as these constructs are not based on the plain language of Title VII or its legislative history. What is more, these judicial constructs no longer comport with our "common sense" understanding of the stakes in cases concerning voluntary race-and ethnic-associated behavior, further undermining their legitimacy. If statutory constructionists want the judiciary's interpretation of race and national origin to be based on clear legislative mandate, they can and should bring this question to the attention of Congress. They cannot, however, claim that the current definitions of race and ethnicity used in Title VII cases enjoy any special legitimacy, or that there is any statutory barrier to prevent the reformulation of antidiscrimination doctrine.
2. Instantiating Racial Identity
Another critical view is that race/ethnicity performance protections actually may prove worse than the current regime because they invite courts to build a behavioral portrait of each racial and ethnic group that inevitably will be based on, and therefore reify, stereotypical images of these groups. 394 When logically extended, this criticism suggests that courts could end up aggressively shaping individuals' discrimination cases to fit a range of standardized scenarios and, consequently, would leave the litigant in the equally disadvantageous position of being required to comply with some previously recognized static racial or ethnic portrait in order to garner protection. These concerns suggest that we should hold firm in our longstanding commitment to a colorblind America and the eradication of racial and ethnic categories. Under this view, the race/ethnicity performance framework takes us in the wrong direction because it further instantiates potentially regressive stereotypical racial and ethnic identities. 395
[*1238] The first step in addressing this argument is to recognize that even current antidiscrimination law is not committed to the eradication of racial and ethnic categories. Rather, antidiscrimination statutes can operate only by maintaining these categories, by creating "a vocabulary of [racial and ethnic] identities and sometimes even channeling claims (and thus claimants) into recognized identity categories with conventional scripts for behavior." 396 Robert Post espouses this view, explaining that antidiscrimination law is designed to intervene selectively in the process of racial ascription. He explains that its goal is to limit the stigma associated with race groups by preventing this stigma from having an effect on an individual's political, social, and economic opportunities. 397 Also, as I explained in Part I, as a practical matter, it is impossible for antidiscrimination law actually to dismantle race or ethnicity because these constructs are believed to be a valid, reliable basis for gleaning information. Individuals will continue to rely on these constructs in private life, regardless of the nature of public sphere protections. 398
The second concern - that judges may create stereotypical, regressive portraits of minority communities - deserves more serious consideration. American courts litigating the issue of racial identity in the eighteenth and nineteenth centuries were all too eager to associate regressive behavior with minority groups in racial determination cases concerning slave codes and citizenship statutes. 399 Not surprisingly, courts ruled that low-status groups possessed the negative traits from which whites hoped to disassociate themselves. 400
Leti Volpp's work suggests that this predisposition continues at present, even more insidiously, as part of the professed project to respect and preserve ethnic difference. 401 Volpp's project is to analyze the destructive effect that the "cultural defense" used in criminal cases has for the project of equality generally. She argues that courts are more likely to recognize a cultural defense when it comports with negative stereotypical representations of an ethnic group. 402 These stereotype-based [*1239] cultural defenses often seem reasonable to courts because they resonate with the "common sense" view that minority persons come from primitive, backward, or regressive cultures. 403 In contrast, when illegal acts are committed by whites, the behavior is characterized as aberrant, individualized behavior. 404 She explains that by recognizing minorities and minority culture as regressive, courts can "cast certain individuals outside the boundaries of our social body." 405
Cultural protections of this order have two costs for minority Americans. First, these "protections" further instantiate the idea that minority culture is regressive and emphasize one strand of the ethnic or racial subgroup's experience at the expense of others. 406 Second, it allows the West to avoid interrogating its own troubling cultural practices that bear resemblance to those in immigrant cultures, including various manifestations of sexism and violence against women. 407 Volpp's analysis, however, is not complete until we consider that this cultural relativism is more likely to play a role when the victim is of the same ethnicity as her assailant. In this situation, her injury is likely to be represented as a minority community concern. 408 If cultural relativism is allowed to play a role in these circumstances, it affords people of color less protection under Title VII than it does their white or outgroup counterparts.
Applying these insights to the Title VII context, the concern is that courts only will recognize employees' race/ethnicity performance when the voluntary behavior at issue comports with stereotypical negative representations of minority communities. These insights also suggest that the need to engage in race/ethnicity performance is more likely to be recognized when the other workers burdened by the plaintiff's identity performance are from the plaintiff's minority community. Disturbingly, this analysis explains the credibility afforded Orlando Patterson's "down-home courting" argument offered to explain Clarence Thomas's sexual harassment of Anita Hill. 409 Patterson argued that Americans were wrong to use "American" values to judge Thomas's culturally specific flirtatious behavior with another member of his cultural community. 410 If Hill had taken [*1240] offense at this behavior, it was unfortunate, but she was being too sensitive. Thomas should not be punished for addressing her in this way, as, based on their shared cultural background, Thomas's actions toward her were reasonable and to be expected. 411 As Volpp explains, the corollary to this argument is the proposition that members of an ethnic or racial group that allegedly gave rise to a cultural practice must be offered less legal protection than their white counterparts from so called "cultural" behavior. 412
While these concerns about judicial stereotyping are valid, I think they are often overstated. We have progressed in our understanding of diversity and continue to make progress. The courts' tendency to correlate negative behaviors with minority groups, while not entirely eliminated, is less prevalent and can be further regulated if they abide by the tenets discussed in Part I. Additionally, the race/ethnicity performance framework provides that behaviors which are truly disruptive to an employer's business or cause dissent between employees are prohibited. 413 An employee should not be able to secure race/ethnicity performance protections for any behavior that violates the civil rights of other employees, and the employer can and should cite the need to protect other employees' rights as the reason this behavior is being prohibited. These factors will encourage courts to deny protection for truly regressive behavior. 414 For example, in the Anita Hill case, Clarence Thomas's fictional supervisor could have prohibited "down-home courting" because it created a hostile environment for Hill as a woman, regardless of her cultural background.
3. Group Identity Subsidies
J<um u>rgen Habermas's critique of group identity subsidies provides opportunity for a thought experiment about certain concerns about race/ethnicity performance protections. 415 Habermas argues that group subsidies may stunt an ethnic group's growth because they immunize the group from having to respond to social and economic [*1241] pressures. 416 The analogous concern about the race/ethnicity performance framework would be that race/ethnicity-associated traits in fact may be dysfunctional and should face social pressures that will require the group to adapt its practices to contemporary circumstances. 417
While this concern about race and ethnic identity subsidies is valid, the race/ethnicity performance framework outlined above is designed to ensure that groups' practices do remain reactive to social conditions. First, the model has built in protections to avoid the creation of a single, frozen, static portrait of a given group. Because it is reactive, and only "protects" a feature upon a showing that the trait was racialized or ethnically marked in the plaintiff's workplace, there will be variation across jurisdictions as to what traits are racially or ethnically coded. For example, an individual in the Northeast might be able to establish that she was discriminated against because she spoke Black English. The same claim could fail in the rural South where the accent might be harder to distinguish from a Southern dialect. Indeed, this requirement of contextual proof ensures that there will not be one unified "script" for any racial/ethnic group.
The race/ethnicity performance framework also requires that some race/ethnicity performance behaviors must be conceded when they unavoidably interfere with a legitimate business purpose, or when they trample on the rights of other employees. These are precisely the kinds of social pressures these practices would otherwise be subject to, but it raises the bar for challenging these practices to ensure that an employer offers some legitimate reason for discouraging employees from engaging in them.
As a separate concern, Anthony Appiah worries that racial scripts created by groups pose limitations for individual freedom; consequently, if groups are able to organize and encode these scripts in law, they may work as strong symbolic referents further spurring conformity. 418 He argues that even when there are multiple scripts for acting out an identity, an individual still experiences these scripts as limitations on her identity. 419 Gesturing towards Gayatri Spivak's [*1242] work on "strategic essentialism," Appiah recognizes that racial categories are an important tool for social justice at present. 420 However, because of his initial concerns, it is likely Appiah would view the race/ethnicity performance framework with some apprehension, as a move to further instantiate a set of racial identities.
To this complaint, I can respond only that I do not believe there to be a clear relationship between the law and individual race/ethnicity performance. Many performative behaviors are never the subject of litigation, yet they remain vital and important parts of racial identity. 421 Others, like braided hairstyles, are litigated, fail to achieve protection, and still remain important features of identity. Moreover, those recognized under the law as being racially or ethnically marked, such as the Afro, 422 do not experience a surge in importance merely because they are recognized by courts. 423 The forces of identity construction are too varied and complicated to be defined completely by the features the law identifies as being worthy of protection.
4. Special Rights Claims
Finally, some will argue that race/ethnicity performance protections afford minority groups "special rights." When persons mobilize this kind of rhetoric, the careful scholar will recognize that this rather fuzzy term conflates two distinct types of claims. 424 The first claim is that the rights offered by an initiative are "special" because they are not extended to all groups and, therefore, constitute unfair preferential treatment. 425 The second argument posits that all antidiscrimination measures are "special rights" and although a national consensus has developed that these protections are available for race, sex, [*1243] religion, and disability, there is no mandate for these protections to be extended beyond their initial formulation. 426
The first special rights claim can be refuted by showing that the benefits of the race/ethnicity performance paradigm inure to the benefit of both minorities and whites. The paradigm is neutral in its formulation: It offers protection from discrimination triggered by ethnic behavior given stigmatic meaning because of racial/ethnic constructs, or similarly non-cultural behavior that, by accident or design, triggers stigmatic racial/ethnic associations. While ethnic minorities of color likely will be the first group to use the paradigm to establish their cases, whites will find it equally useful, particularly in cases involving intra-race ethnic discrimination. That is, the paradigm will prove useful in demonstrating that certain voluntary features associated with white ethnic subcategories are stigmatized as low-status "white" behaviors and provide the basis for discrimination against them. Additionally, those who experiment with racially marked practices will enjoy a safer context for exploration.
Still, some will insist that the race/ethnicity performance paradigm is a special tool designed to assist minority groups, based on their expectation that minority workers will be able to secure exemption from grooming and behavior rules of general application. The model, however, does little to alter facially neutral rules that are applied evenly. Instead, the model focuses on employer rules that prohibit specific race/ethnicity-associated practices, requiring the employer to demonstrate a legitimate business reason for its policies. 427 For example, it targets rules that prohibit braids and dreadlocks or Spanish speaking between employees. These are rules that have no bearing on groups with no interest in these practices. In this sense, the race/ethnicity performance cases help to level the playing field by treating rules that appear to target certain groups as suspect on their face.
In disparate treatment cases, the challenge is that a facially neutral rule is applied more aggressively to a minority employees' behavior. Here, the burden is on the employee to demonstrate that her conduct was comparable to outgroup members' conduct but was [*1244] regarded less favorably because it was racialized. For example, if an employer routinely permits employees to wear ponytails at work, she should not be able to sanction a minority employee for wearing a similar ponytail composed of braids or dreads. As these cases show, the goal is not to provide special rules that benefit ethnic subgroups; instead, it is to require employers to relax their focus on subgroups' differences when these differences are irrelevant to the job at issue, and to apply neutral rules consistently across groups.
B. Preserving Employer Discretion
The second group of arguments against race/ethnicity performance protections stems from concerns that they will interfere with employers' ability to run their businesses. 428 This Section explores and responds to those arguments.
1. Respecting Employer Autonomy and Expressive Concerns
One group of arguments under this banner is based on the view that the law should not interfere unnecessarily with employers' expressive rights and interests. Proponents of this view argue that the employer, as a creator of business opportunity, is entitled to regulate workplace "culture," regardless of whether her preferences actually relate to the business's purpose. Under this view, the politics of culture are a zero-sum game, and while the employer must tolerate morphological difference, she is not required to tolerate cultural performances that disturb her sensibilities. The absolutist view of an employer's prerogative to control workplace culture is fundamentally libertarian in its orientation, and it treats the employer's right to operate a business as a space for her expression of individual freedom, with the employee's interests being subordinate to the employer's.
This libertarian orientation is reflected in common law employment rules, which provide that an employer may fire an employee "at will" 430 The sole limitation is that these preferences may not be based [*1245] on race/ethnicity-associated morphology. 431 However, this libertarian account fails to consider that wage-based work is an integral, unavoidable part of life for most Americans and that it consumes the vast majority of most people's waking hours. Indeed, these same libertarian principles could suggest that it would be fundamentally unfair to mandate that a person who is required to support herself financially by seeking work also should be forced, based on an employer's whim, to surrender features that are integral to her sense of personal dignity and identity.
Additionally, those that espouse a libertarian view of employers' rights rely on a na<um i>ve perception of the availability of employment opportunities for individual workers. They assume the availability of a large, accessible pool of jobs, suggesting that an employee who is particularly invested in race/ethnicity performance behavior is free to shop for an employer who will not prohibit these practices. Therefore, they argue, no individual employer should be required to tolerate a particular kind of race/ethnicity-associated behavior if she believes it is inconsistent with her business. 432 In reality, however, a single employer or group of employers may dominate the employment market in a particular venue, leaving the employee with few choices. Exit simply may not be an option. Additionally, this libertarian argument ignores the fact that employers with strict rules as to appearance and behavior tend to involve higher-status positions with greater economic advantages. 433 Under this analysis many ethnically and racially marked workers must "choose" to remain locked in marginal or low- [*1246] status jobs, or make various compromises with regard to personal dignity.
Unfortunately, the judiciary has failed to problematize the common law or libertarian framework for the employment relationship, instead waxing on at length about the importance of preserving the employer's right to make judgments about how to run her business. 434 Fortunately, the common law's generous grant of employer autonomy is now fundamentally at odds with most Americans' understanding of the employer-employee relationship. Because most Americans' work experience has been during the era of federal and state employment protections for race and sex, as well as protections based on pregnancy, disability, and religion, they operate under the inaccurate perception that the employer-employee relationship provides them with some protection from random adverse treatment by employers. This break between workers' expectations and their actual protections is indicative, I believe, of a larger hegemonic shift in view about the proper rights balance in the employer-employee relationship. 435 Stated more simply, the common man no longer finds it natural, or "common sense," that employers should be permitted unilaterally to impose their will on workers when cultural interests are at stake. Rather, the new social expectation is that when an employer imposes a rule, she will justify her decision on some rational, cost-benefit analysis. 436 Consequently, employers' attempts to rebut race/ethnicity performance discrimination claims typically should identify some rational, economic, or practical reason for their actions, in contrast to the current regime which allows them to argue that no claim lies because the rule only targets voluntary behavior. 437 Of course, an employer's "rational calculations" can have discriminatory consequences as well, an issue discussed further below.
2. Concerns about Employer Functionality
The second group of arguments in support of preserving employer discretion consists of functionality claims. Specifically, employers argue that their discretion to control workplace culture is critical to their ability to: 1) control the marketing of their products; 438 and 2) maintain control over social dynamics in the workplace. With regard to the second factor, they may argue that they need to prohibit certain kinds of race/ethnicity performance in order to ensure worker harmony, and thereby increase workplace cooperation and efficiency. 439
a. A Closer Look at Efficiency and Marketing Claims
Courts tend to treat functionality claims with substantial deference, most likely out of concern that judge-made rules can have unanticipated effects on a company's productivity and profitability, and therefore business administration and personnel decisions are best left to individual employers. As a separate matter, judges often are confounded by these seemingly rational efficiency-and marketing-based reasons for forbidding race/ethnicity performance because they remain trapped in the model proposed by the first wave of prejudice studies, which counseled that prejudice was an irrational, individual problem, an aberration from businessmen's otherwise accurate cost-benefit calculations. 440 Consequently, when an employer presents a rational, facially neutral reason for prohibiting racially/ethnically-marked behavior, judges conclude that the calculation itself was not discriminatory, and the rule cannot serve as the basis for Title VII liability. 441
Many legal scholars, however, tell a different tale about the role of employer racial and ethnic discrimination in efficiency and marketing decisions, debunking the claim that discrimination is an irrational personal aberration and arguing instead that it may be a logical and powerful motivator in employer decisionmaking. 443 Robert Post, 445 have identified myriad ways in which patterns of racism form part of the cost-benefit analysis in employer decisionmaking. The simple truth is that employers historically have adopted marketing and workplace efficiency strategies that take account of and capitalize on historically sedimented and contemporary patterns of social discrimination and racial stratification. Prior to the passage of the 1964 Civil Rights Act, employers refused to hire minorities out of a fear of losing customers 446 and out of concern that these morphologically distinct new hires would disrupt and antagonize their majority white workforce. 447
The remaining question is whether there is any relevant distinction between the aforementioned clearly illegal discriminatory behavior based on morphology and a situation in which employers forbid race/ethnicity-associated behavior because of concerns about a white customer base or to avert workplace conflict. This problem has been aggravated by hegemonic shifts in the ways in which Americans feel comfortable articulating discriminatory attitudes, as employers have learned to avoid explicitly stating their concerns about upsetting a white customer base or their employees and instead tend to articulate these same concerns in "neutral" terms. 448 However, as I demonstrate below, these ostensibly "neutral" reasons are intimately tied to their desire either to harness contemporary discriminatory attitudes or to avoid upsetting consumers who hold them.
b. Deconstructing Claims Regarding Customer Preferences
Employer rules prohibiting workplace race/ethnicity performance are commonly justified based on an employer's need to market her product effectively. 449 While employers rarely explicitly articulate the [*1249] reasons race/ethnicity-associated practices affect marketing strategy, the underlying premise of this claim is that the employer's goal is to deliver products and services that appeal to customers' tastes, and that cultural, racial, or ethnic signifiers are an important part of this process. 450 Race/ethnicity-associated practices common to low-status groups carry meaning and run the risk of becoming associated with the product. Consequently, the employer explains, she must have the power to prevent workers from engaging in this kind of behavior. As explained above, 451 because of shifts in social attitudes about discrimination and the rise of aversive racism, employers are unlikely to articulate these concerns about customer preference in such stark terms. 452 Instead, employers typically offer race-neutral justifications for prohibiting race/ethnicity-associated practices, characterizing these practices as "unprofessional," "dirty," "eyecatching," or some other seemingly uncontroversial but negative description.
In Fagan v. National Cash Register Co., 454 a Title VII sex-plus discrimination case involving grooming rules, the D.C. Circuit provides a cogent discussion of this view. The Fagan court explains:
Perhaps no facet of business life is more important than a company's place in public estimation. That the image created by its employees dealing with the public when on company assignment affects its relations is so well known that we may take judicial notice of an employer's proper desire to achieve favorable acceptance. Good grooming regulations reflect a company's policy in our highly competitive business environment. Reasonable requirements in furtherance of that policy are an aspect of managerial responsibility. 455
Armed with this premise, the Fagan court concluded that Fagan's employer's policy, which prohibited him from wearing long hair, did not violate his rights under Title VII to be protected against sex discrimination because Fagan's appearance hampered the company's ability to market its services. The Fagan analysis, when extended to [*1250] race or national origin discrimination, thus provides that any time an employer identifies a rational reason to argue that a voluntary race/ethnicity-associated behavior compromises her ability to provide a service or sell her product, she has presented sufficient justification to police employees' behavior. 456 The quote could just as easily have been used to justify the Rogers v. American Airlines decision.
Although it appears uncontroversial in the abstract, this rule has disturbing repercussions when applied in cases of race/ethnicity performance. Consider, for example, a hypothetical in which an employer such as Federal Express which, in the wake of September 11th, issued a directive to midwestern field offices not to use persons with accents on any delivery run in a federal building because these couriers were more likely to be stopped by security, which would compromise the company's ability to ensure the timely delivery that is the hallmark of its business. The company imposes this rule knowing that the majority of workers impacted by the decision are Middle Eastern based on the demographics of the region. Consider another directive, which prohibited the hire of any person wearing dreadlocks on similar grounds: Couriers wearing the hairstyle were likely to be stopped on delivery runs, compromising fast delivery. Under the existing Title VII framework, these cases would not trigger antidiscrimination concerns, as both are facially neutral and involve voluntary, changeable behavior. Moreover, both meet the highest standard for employers' business purpose: They are directly linked to the core task that is the nature of the employers' business. Yet despite these features, these rules do have discriminatory implications.
These kinds of employer functionality decisions are worrisome because they conflict with one of the basic tenets of antidiscrimination law: the view that the instrumental value or overall functionality of an employee is entirely independent from her race or ethnicity. 457 However, this is simply untrue. Assuming for a moment that morphology no longer plays any role in an employee's decisionmaking, an employee's value to her employer is determined by the degree and kinds of racial/ethnic markers she bears, independent of her possession of the baseline physical or cognitive abilities necessary to perform her job. These voluntary racial/ethnic markers play a role because the employer is aware of outgroup members' reaction to the employer's racial/ethnic group, and the employer must assess whether or not he is willing to bear the risk that these marks of distinctiveness will compromise [*1251] the employee's ability to effectively perform her job. As the Federal Express hypotheticals show, these calculations need not be tied to bare prejudice; instead, they may concern indisputable realities about an employee's ease or difficulty in moving through society, her ability to trigger immediate liking or aversion, as well as a host of other subconscious aversive reactions. Also, these hypotheticals demonstrate that an employer who is aware of discriminatory trends in its customer pool can translate the realities posed by social prejudice into myriad race-neutral and rational justifications for treating racially and ethnically marked employees unequally.
Given these realities, how should courts presented with marketing or customer preference claims respond in cases concerning discrimination based on race/ethnicity performance? As an initial matter, they should acknowledge that we live in a world that still harbors lingering effects of historically sedimented patterns of discrimination, one that engenders new forms of cultural bias with new national conflicts, and one in which voluntary behavior, as much as physical difference, plays a role in triggering discrimination. 458 Moreover, they should be mindful that employers know that customers, who may be neutral with regard to race/ethnicity-associated morphology, still may be adverse to voluntary markers associated with certain groups and may pay a premium to avoid these behaviors. Lastly, they should recognize that employers who want to appeal to this animus may attempt to institute rules that prohibit distinctiveness associated with a particular low-status group or invoke ostensibly neutral reasons to crack down on a race/ethnicity-associated practice.
Additionally, judges must take cues from the lessons learned in the sex discrimination cases concerning customer preference, which counsel that these claims should be vigorously challenged and interrogated before being accepted. In the 1970s and 1980s, there were a number of suits in which employers were challenged for hiring only one gender, in which employers justified their rules based on customer preference. 459 As in the racial/ethnic performance cases, the concern in these cases was the socially stratifying effects of the employers' [*1252] decision. Indeed, the concern in the gender/customer preference cases was that, in satisfying customer preferences, the court would have to endorse certain gender stereotypes, lock women out of high paying jobs and ensure a certain amount of social stratification. 460 Indeed, in the 1960s and 1970s, courts struck down a number of employer policies forbidding the hiring of women based on customer preferences and the claim that women would not be accepted in certain positions. 461 These customer preference cases show that courts are able to deconstruct "common sense" claims about customer preference and must work hard to recognize those which rest on propositions that have become naturalized. 462 Similarly, when neutral justifications for policies appear to comport with the social stereotypes associated with low-status races and ethnic groups, courts are less likely to recognize their offensive basis. 463 For example, an [*1253] employer's claim that braided hairstyles are prohibited because they are dirty or unprofessional comports with the low-status stereotypes associated with blacks and, consequently, normally might not be subject to the scrutiny it requires.
In summary, there are some circumstances in which an employer can make a valid claim that race/ethnicity-associated behavior may interfere with workplace performance, but they usually will be cases concerning more concrete functionality claims. 464 For example, Title VII is not implicated when a rule requiring the wearing of a safety helmet leads to the termination of an employee whose dreadlocks will not fit under the helmet and the employee refuses to alter her hairstyle. Rather, the concern is when employers frame business concerns as neutral justifications for workplace rules when they really are, in effect, trying to capitalize on prejudice. When courts allow such justifications to stand unchallenged, they unwittingly permit employers to make discrimination part of their marketing strategies.
c. Deconstructing Efficiency Claims
Employers also rely on racial and ethnic status hierarchies as a tool in promoting workplace harmony, hoping to improve worker efficiency. A culturally homogenous workforce typically is more congenial; therefore, employers can forestall dissent among workers by avoiding the hire of persons with strong racial and ethnic markers or by hiring workers of predominately one race or ethnic group. 465 Alternatively, having hired a strongly racially/ethnically marked worker, the employer may adopt a laissez-faire attitude towards harassment when workers from the majority ethnic/racial group target the minority worker; the employer's decision is based on the assessment [*1254] that this dynamic will enforce cultural conformity over time and eventually lead to less conflict. Alternatively, the employer may permit the harassment to continue because the majority employees' hazing of a minority employee fosters a sense of camaraderie between these otherwise valuable employees. After assessing the potential difficulty of finding replacement, nondiscriminatory workers or the costs of training new hires, the employer may make the rational decision to ignore the discrimination triggered by racial or ethnic practice, for it does not violate current Title VII standards. In the employer's view, it is cheaper to replace the target of the harassment than the perpetrators. J.M. Balkin nicely summarizes this view, explaining that "[employers] will even tolerate employee behavior that is racist, sexist, unjust, or anti-social, as long as it promotes workplace cohesion and morale and is not bad for business." 466
The role that racial animus and ethnic bias play in employers' workplace harmony strategies is far easier to discern than it is in their marketing and customer preference claims. Stated simply, the employer not only tolerates, but capitalizes on her employees' belief in racial and ethnic status hierarchies, relying heavily on their feelings of ingroup racial and ethnic solidarity to facilitate workplace cooperation, 467 instead of taking on the harder task of independently cultivating a company identity that encourages cooperation. In the more extreme cases, the employer relies on racial/ethnic harassment to further galvanize ingroup cooperation and a sense of belonging among nonminority workers, while requiring her minority or outgroup workers to bear the costs of this community-building strategy or to find other employment.
Strangely, courts seem unaware of the discriminatory premises that inform the view that race/ethnicity performance must be forbidden in order to promote workplace harmony. This confusion stems from two sources. For one, many of the cases in which this claim is raised concern one group of minority employees complaining about another minority employee's behavior. 468 The court fails to consider that an employer may hire employees predominately from one minority group - a group that claims cultural hegemony in the workplace - and then that group may attempt to stamp out the markers of [*1255] other racial or ethnic groups in an attempt to protect the high status position of their group. Disputes about group position are particularly acute when they occur between racial groups with low social status as compared to whites, for "the more that members of a racial group feel that they are alienated and oppressed, the more likely they are to regard other racial groups as competitive threats to their own group's social position." 469
A second reason for confusion is that many of the cases up for review involve hazing by non-homogenous groups of employees with different racial and ethnic backgrounds who, united by an "American" identity, attempt to quash racial or ethnic difference. That is, a racially diverse group of employees grows concerned that the traits associated with an American identity are under attack by the behavior of newly arrived immigrant groups, and unite across racial lines to quash this difference. The courts' confusion in these cases is unwarranted, for it seems relatively obvious that members of various ethnic or racial groups will and often do cross racial and ethnic lines for the purpose of terrorizing others: White and Latino workers join to mistreat African Americans, or African Americans and whites join in an effort to prevent Spanish from being spoken by Latino employees.
A few examples help make these points clear. In Garcia v. Spun Steak, 470 the Ninth Circuit granted summary judgment to an employer on a Latino employee's discriminatory discharge claim when the employee was discharged for violating an English-only rule instituted in the interest of workplace harmony. The employer presented evidence showing that it had instituted the rule because it had received complaints alleging that Latino employees had made racist remarks in Spanish about two other employees, one African American and one Chinese American. 471 The court concluded that the English-only rule did not overly burden the fluently bilingual Spanish employees; therefore, it held that the employer had a right to enforce the English-only rule. 472 The case illustrates how workers who occupy a low-status position in America's racial/ethnic hierarchy are often the first to complain about race/ethnicity performance and find Title VII a useful tool in this endeavor. That is, these employees may claim that the [*1256] race/ethnicity performance of other low-status groups creates a hostile environment and request that their employers crack down on these displays. The Spun Steak case may be correctly decided; however, the reader is left without a clear sense of what it was about the speaking of Spanish that constituted evidence of hostile environment.
This issue is revisited in Webb v. R&B Holding Co., 473 where an African American plaintiff filed a discriminatory discharge, hostile environment, and retaliatory discharge claim, arguing that she was discriminated against by coworkers who spoke predominately Spanish at work, and that she suffered retaliation when she complained about their behavior. 474 The district court granted the defendant employer summary judgment on the hostile work environment claim because it concluded that the there was no evidence in the record that her coworkers' "speaking of Spanish constituted harassment 'sufficiently severe or pervasive to alter the conditions' of Plaintiff's job." 475 The court next granted summary judgment to the employer on the plaintiff's retaliation claim, concluding that the plaintiff-employee had not engaged in protected activity. 476 The court indicated that in order to establish that she had engaged in protected activity, the plaintiff was required to show that her complaints to the employer specifically characterized the Spanish-speaking as an act of discrimination, rather than, for example, a practice to which she objected for personal reasons. 477 Furthermore, because her employer likely was aware at that time that an EEOC ruling had prohibited English-only rules, the employer reasonably concluded that plaintiff could not be asking it to institute a rule that would make it liable under Title VII. 478
Without a clear framework to resolve this conflict between groups of low-status minority workers, the court is unsure how to address the African American employee's discrimination concerns as balanced against the Spanish-speaking employees' interest in ethnic performance. After reading the decision, one still is unclear about its meaning. Although the court placed weight on the fact that the EEOC had provided some protections for Spanish-speaking in the workplace, it failed to address whether the exercise of this privilege could be used in a pattern of aggression sufficient to intimidate another minority worker.
[*1257] The race/ethnicity performance framework teaches that the most important consideration in the Webb case was the content and context of the Spanish-speaking that was the subject of the complaint. If the employee could establish that the Spanish-speaking contained racial epithets about African Americans, or was being used to exclude her from workplace discussions, this kind of behavior would trigger hostile environment concerns. If the African American employee simply objected to the occasional use of Spanish at work, this would suggest that she was attempting to preserve the status of her group and prevent the perceived encroachment of Latino culture. The court, however, does not explore these points; it instead summarily concludes that the harassment was not severe or pervasive. 479 Additionally, the court's resolution of the retaliation claim avoids the most difficult, but central issue in the dispute: whether Webb reasonably believed that the Spanish-speaking was sufficient to constitute a hostile environment under Title VII. Again, the court avoids looking at the content and context of the workers' use of Spanish and instead summarily concludes that Webb's complaint was "personal." 480 Interestingly, the primary focus of the court's concern appears to be Webb's employer's perception of her complaint, specifically that, given the EEOC's directive indicating that English-only rules in some circumstances were violative of Title VII, her employer reasonably could have believed that her complaint was personal and not a violation of Title VII. 481 Again, by failing to use the race/ethnicity performance framework the court failed to resolve the dispute equitably.
This case presents a snapshot of some of the factional race-and culture-based battles being fought in workplaces. The scenario painted by the Webb case provides a glimpse into the contentious workplace politics that can occur in urban centers with large, competing, and strongly culturally identified immigrant populations. In the future, we are likely to see Title VII cases between and within "racial" groups based on ethnicity, including disputes between Haitians and Dominicans, El Salvadorians and Puerto Ricans, African Americans and Haitians. Also, there likely will be a large number of cases concerning traditional or historically familiar conflicts between whites and other racial groups.
3. Concerns about Employer Liability
Employers already are embroiled deeply in workplace conflicts involving race/ethnicity performance, 482 and, therefore, contrary to critics' projections, this new paradigm will not expand the bases for their liability. Rather, it will provide a more reasoned basis for resolving employees' complaints. 483 Additionally, predictions that race/ethnicity performance protections will transform the workplace into the Tower of Babel are unwarranted. With or without Title VII protections, workers will engage in race/ethnicity performance and already are bringing their own race-and ethnicity-specific styles and practices to bear on the workplace, triggering conflict. 484 At present, employers face these disputes with no sense of their obligations: While some capitalize on racism or ethnic prejudice, others adopt a laissez-faire attitude that leaves their minority employees vulnerable to harassment. Still others make the intrepid attempt to impose disciplinary sanctions against persons who discriminate based on race/ethnicity performance, with the risk that they may be accused of violating Title VII.
Again, the question becomes, what are courts to do with these insights? First, they should treat workplace rules prohibiting race/ethnicity performance on "workplace harmony" grounds as inherently [*1259] suspect, and only validate rules narrowly designed to address circumstances in which the performative behaviors at issue clearly were being used to harass or exclude outgroup employees. However, an employer's mere assertion that difference is a distractor is per se unreasonable. Also, courts should be aware of the racial/ethnic composition of the workforce when viewing these claims to determine whether one ethnic group has seized the cultural center in the workplace and is using this power to intimidate other workers. They also should be aware that the employer may be engaging in laissez-faire discrimination or capitulating to pressure from the group of ethnic employees with the strongest presence in the workplace to prevent any behavior that threatens to disrupt that group's hegemonic control over the workplace's cultural baseline. Additionally, they should consider whether workers' complaints about race/ethnicity performance simply represent thinly veiled antipathy for a new, rising immigrant group.
Rather than aggravate the tendency for these disputes, the race/ethnicity performance model provides courts with a better paradigm to resolve these problems, accounting for the interests of all involved as well as the employer's concerns in adjudicating these disputes. Most importantly, it provides us with a bright line rule for resolving these workplace cross-racial conflicts: An employee's right to engage in race/ethnicity performance ends when she begins to trample upon the interests of other employees. 485
Finally, employers will argue that they should not be required to bear the social cost of eradicating discrimination triggered by voluntary race/ethnicity-associated behaviors and that the proposed new framework increases their obligations just when they are beginning to achieve success in combating morphology-based discrimination. 486 [*1260] However, employers' substantive obligations would not increase if Title VII imposed a prohibition barring employers from marketing their products in ways that capitalize on discrimination or made them liable for allowing the hazing of employees engaging in race/ethnicity performance behavior. Rather, the new prohibition simply would take the additional steps needed to ensure that employers do not use racial/ethnic animus to achieve their business objectives. This is the purpose of Title VII and other workplace discrimination laws: These laws "push employers toward functional justifications for their actions" and ensure that "employers have strong incentives to articulate 'legitimate reasons' for their decisions," or to show that their selection procedures "'are demonstrably a reasonable measure of job performance.'" 487 The race/ethnicity performance framework does much to advance us in this regard.
It is important to remember that, at present, employees have no protection in this area. Karl Klare explains:
It is one thing to say that managerial interests must be weighed in a balance, where the employer produces evidence of a genuine conflict between employee appearance choices and an agency's efficient performance of its mission... . But the cases do not call for such an inquiry. Rather, they effectively allow the employer merely to state its attitudes in order to make out a showing of "managerial interests," and then put the employee to the impossible task of demonstrating that these attitudes are wholly irrational. 488
By increasing the employer's burden to demonstrate a rational basis for her decisions, the race/ethnicity performance framework locates the balance at a level that better comports with our stated goal of workplace equality.
C. Political Issues
The last group of arguments against the race/ethnicity performance framework centers on concerns about the framework's effect on American political life. Scholars are suspicious of any model that will aggravate the separatist and essentialist trends set in motion by the identity politics movements of the past two decades, namely proprietary claims of races or ethnic groups over culture, 490 and the tendency to brandish difference as an insurmountable barrier to coalition politics. 491 Each of these concerns, however, is accounted for in the race/ethnicity performance framework. Moreover, the model promises to help create the cross-cultural respect and understanding necessary for a cohesive citizenry in a multiracial and multi-ethnic country.
1. Race as Distractor
Various scholars have expressed the view that paradigms which encourage individuals to focus on race or ethnicity may do us a disservice, as they encourage people to affirm racial and ethnic identities instead of exploring more productive, substantive bases for coalition that more accurately reflect their shared interests. One view is that race is in some ways a distractor - it encourages people from a variety of different backgrounds to imagine that there are connections between their racial groups when, in fact, the only thing that unites them is a similar morphology that has triggered oppression in the United States. 492 Another view is that efforts to achieve equality that are made based on categories like race and sex are inherently limiting, as they petition for rights based on the parcel of benefits offered the white, male, middle-class subject without ever challenging whether the parcel of rights accorded this position are just and fair in their own right. 493
[*1262] While these concerns are certainly valid with regard to certain aspects of the identity politics movement, neither holds true with regard to the race/ethnicity performance framework. The race/ethnicity performance framework remains attentive to discontinuities within racial groups and does not assume a shared culture. The paradigm also accounts for the fact that the markers of race are as multiple and varied as the ethnic groups captured by racial constructions, and acknowledges that sometimes these markers are merely side effects of segregation. 494 Because of this focus on diversity, the race/ethnicity performance model discourages the simplistic understanding of race that seduces people into thinking that they have affinities beyond the shared experience of being marked as a member of a stigmatized racial group. 495 Indeed, it focuses on the fact that these markers, rather irrationally, have been stigmatized because of racial constructs, focusing race group members' attention on the actual source of their problem.
Also, the model does not fall prey to the concern that race-based movements simply further legitimize the goal of affording each individual the parcel of rights provided to the white, male, bourgeois subject instead of imagining something beyond it. To the extent that race/ethnicity performance behaviors are motivated by or expressive of values that depart from the norms associated with a white male, bourgeois subject, it opens the door for a discussion of competing values and world views. 496
2. Proprietary Claims over Culture
Appiah gives us insight into the concern that the race/ethnicity performance paradigm might encourage further proprietary race-based claims over culture. 497 Specifically, he argues that identity politics often gives rise to attempts by individuals to claim exclusive credit for the creations of their race. 498 I argue that this phenomenon must be understood within the logic of group position theory, as yet another method of establishing the relative status of their race as compared with other races. Identity politics, it is argued, often causes some individuals, based solely on racial identity, to claim credit for the cultural, political, and social products of a wide array of civilizations' contributions over time merely because the icons most associated with these cultural products display morphology or claim membership in a particular racial group. 499 Whites therefore, for example, might try to claim the accomplishments of every civilization constructed as "white," from the Ancient Greeks, to America's founding fathers, to country music. Similarly, African Americans might claim the cultural legacy of the Ancient Egyptians, the birth of American jazz and gospel music, as well as the creation of rap.
The irony is that often the individuals who make race-based claims about culture have little or no relationship to the cultural and political products they claim as part of their so-called racial heritage. 500 Moreover, the effect of this kind of discourse is to make certain areas of culture "off limits" to outgroups, discouraging efforts at cross-cultural understanding and discrediting the efforts of persons who do feel affinities with the cultural products associated with other races. 501 The repercussions of this critique are already in clear view, as persons who attempt to cross the "race-culture" divide often face challenges as to their "authenticity."
The race/ethnicity performance regime, however, actually undermines proprietary claims over culture. Certainly there will be simplistic interpretations of the model which suggest that because it recognizes the link between racial stigma and certain practices, it cordons off certain cultural practices for a particular race or ethnic group. Again, this error is based on a facile and oversimplified reading. Because the model establishes that racial/ethnic signification operates [*1264] through practices, it rejects the fundamental premise that in order to trigger social codes for a particular racial/ethnic status, one must exhibit the physical markers associated with that racial group. 502 Because the model posits that persons of ambiguous racial/ethnic morphology - or a person morphologically marked as a member of a high-status racial/ethnic group - still may engage in behavior that signifies a different, historically subordinated racial/ethnic status, it compromises any one racial/ethnic group's ability to claim these practices as their sole province. Therefore, under this model, persons morphologically marked as black are not the only persons who potentially can raise race performance claims concerning black signification. Rather, the regime makes room for the possibility that, in some circumstances, a morphologically "white person," who wears braids or dreads and plays reggae music at her workstation, may be discriminated against because she has signified blackness in a way that upsets the cultural tenor of the workplace. These kinds of cases stand in opposition to essentialist claims by a group attempting to bind its cultural practices unto itself and, moreover, protect those who in good faith do engage in cross-cultural exchange out of a desire for affiliation. 503
3. Race/Ethnicity Performance as a Vehicle to Cultivate a Cosmopolitan Citizenry
In his well-regarded book, Postethnic America, David Hollinger echoes numerous scholars who argue that, in order to achieve a cohesive, cosmopolitan, multiethnic citizenry, we must move beyond race. 504 These scholars contend that existing racial constructs power a dysfunctional identity politics: They attempt to forge solidarity among ethnic groups that have no actual cultural connections and encourage them to continue to identify based on a shared experience of exploitation. 505 This dynamic, Hollinger argues, is the biggest threat to our effort to achieve a cohesive American citizenry. A survey of the divisive race-based political debates of the late 1980s and 1990s makes this [*1265] view attractive, even common sense. But what if Hollinger and his admirers are wrong? What if the society they prefer could be achieved with existing racial constructs, corrected by the existing antidiscrimination paradigm? If this were true, the unity they seek through a "postethnic" citizenry might be closer than they imagine.
The "postethnic" community Hollinger prefers, in place of the current racially dominated American culture, would organize people into groups with permeable boundaries, which allow potential group members to join and defect at will. 506 Individuals would be taught to see their ethnic orientations not as genetically predetermined "identities," but as a long process of social formation, the ultimate goal of which is to allow them voluntarily to choose their communities of affiliation. 507 For the postethnic community to be successful, the individual's affiliation choices must be respected, even if they abandon the ethnic community into which they were born. 508 Once racial and ethnic groups lose their ability to invoke genetics as a basis for proprietary claim over their members, he argues, we will see an end to the divisive identity politics debates that fracture our society. 509 Instead, individuals will be more willing to recognize that, despite these ethnic differences, they are united by a unique American democratic culture, one that shapes their identity just as much as ethnic affiliation.
Hollinger's "postethnic" America, however, fails to account for the role stigma plays in individual choice and social relations. That is, one of the unspoken premises of his work is that, for the postethnic America to be realized, we must create a context in which ethnic groups' practices can compete fairly in the "affiliation" contest. Stated alternatively, in order for a person's affiliation choices to be truly free, she must come to form her opinions about outgroups' ethnic practices without being affected by irrational stigmas that trigger her to devalue certain groups and celebrate others. Ironically, [*1266] Hollinger imagines that the most difficult part of creating the postethnic America is to make groups realize that their boundaries are permeable or to teach individuals that they have choices about their ethnic or racial affiliations. 510 However, individuals already are acting based on these premises on a number of fronts, contesting the racial labels imposed by their communities, by the government, and even at work. 511 The biggest project in a country dominated by racial status hierarchies is not to encourage permeable boundaries between groups, but to eliminate the stigma that racial constructions impose on certain ethnic communities. 512
The question is: How do we eliminate the powerful stigma imposed upon ethnic groups because of their association with historically disfavored race groups? I believe that if courts adopt the race/ethnicity performance paradigm, they could use our existing antidiscrimination regime to interrupt the processes by which race is used to stigmatize ethnic groups in the workplace. This would, in effect, create the atmosphere of free choice necessary for the development of a cosmopolitan citizenry, willing to make its affiliation choices purely on individual desire.
Indeed, much of the race/ethnicity performance framework is consistent with Hollinger's theory of "affiliation" or identity formation. The race/ethnicity performance model posits that individuals experiment with racially or ethnically marked practices from which they ultimately select particular behaviors and aesthetics to express "affiliation" with particular groups. 513 Moreover, each racial construct covers more than one ethnic community, and individuals who sample within these categories are more likely to respect or even adopt practices from several of these ethnic identities in the creation of their own "racial identity." However, as explained above, the race/ethnicity performance model recognizes that, in many cases, stigma forestalls the transmission of racialized ethnic practices beyond a particular racial group. It also recognizes that reaction to stigma makes disfavored racial and ethnic groups more willing to attach significance to racial constructions; to make proprietal claims over all positive aspects associated with these racial constructions; and to disrespect [*1267] the practices of outgroup ethnic communities associated with other "races."
By focusing attention on the varied nature of race/ethnicity performance, the model requires a recognition of the multiple ethnic communities associated with a particular racial construct and the role race plays in making these practices the basis for social sanction. Once actors are prevented from sanctioning individuals on this basis, these practices can proliferate undeterred and face a far better prospect of competing for respect and admiration on their own terms. 514 Also, the paradigm allows courts to protect members of outgroups who are willing to experiment with and affiliate themselves with ethnic communities and practices that are associated with disfavored race groups. In circumstances where they are subject to sanction because of this "race" performance, an employer can be held accountable for this kind of sanction. 515
It is tempting to be dismissive of race/ethnicity performance cases; however, the stakes at issue in these disputes should not be underestimated. As this Article shows, in addition to the individual expression concerns, race/ethnicity performance disputes involve the same racial and status hierarchies that inform morphological discrimination. [*1268] Allowing workers to engage in these practices at work can have transformative effects beyond the dignitary benefit that accrues to individual workers. Exposure to different cultural practices in an atmosphere of mutual respect is necessary for the development of a cosmopolitan citizenry. When employees are forced to cooperate with strongly ethnically marked outgroup members, they learn that these cultural differences need not be a barrier to friendships or mutual cooperation. 516 Moreover, they are encouraged and given the opportunity to ask questions about cultural differences, a necessary precondition to becoming a cosmopolitan citizen. Karl Klare goes further, suggesting that "nonconforming appearance choices can be highly subversive of the status quo." 517 He notes that performance practices "sometimes disrupt and shake-up settled understandings and roles, and they may dramatically suggest ... the need for new discursive possibilities and altered power relationships." 518 For example, the stark differences in aesthetic values between Italian American and Caribbean American employees might emphasize the separate nature of their worlds and lead employees to question why such dramatic differences exist. Individuals will have more opportunity to compare their ethnic communities' disparate ways of coping with similar social pressures. These productive interchanges initially will lead to more productive conversations about race and ethnicity and ultimately will require an examination of the ethnic orientation that actually informs these coping mechanisms. 519
I recognize that this atmosphere of mutually respectful questioning about ethnic practice is not possible in many workplaces at present. Indeed, the aversive racist, when presented with unfamiliar race performance behavior, feels "anxiety and uneasiness." 520 Also, group position theory teaches that groups will be tempted to "crack down" on outgroup members' cultural performances out of fear and [*1269] concern that their race group will lose standing in the workplace. 521 What is required is a legal regime that interrupts this dynamic, one that prevents scenarios in which persons from a contingently more powerful ethnic or racial group can coerce an outgroup member to suffer sanctions for her behavior. 522
This Article is intended to demonstrate how current interpretations of Title VII do little to advance the cause of equality because the manner, circumstance, and rationalization of discriminatory behavior have changed. This Article has shown that the natural/artifice distinction, and other attempts to divide race into its involuntary and volitional components, fail to provide a principled basis for identifying the aspects of racial and ethnic identity that should be afforded protection from discrimination. Because these doctrinal constructs fail to account for the role that volitional behavior or race/ethnicity performance plays in defining individual identity, courts cannot construct coherent narratives in cases involving race/ethnicity performance, a prerequisite to equitable resolution of these claims. Moreover, because they fail to understand these cases as group status contests involving racial hierarchies, courts cannot equitably sort out reasonable business judgment claims from attempts at cultural domination. The hope is that this Article having marshaled the scholarship in prejudice studies and provided a theory that shows how these insights are born out in race/ethnicity performance cases, courts will recognize the need for a shift in paradigm.
Some scholars may find flaws with the race/ethnicity performance concept as this Article has defined it, arguing that it is not the proper vehicle for a more comprehensive analysis of discrimination. Whether or not these criticisms find their mark, one thing is clear: We cannot [*1270] avoid the contemporary phenomenon of discrimination simply because we fear that the process of identifying and labeling race/ethnicity performance behavior may be too complicated or political to be risked by courts. Our present regime's failure to account for the role that race/ethnicity performance behavior plays in the experience of racial and ethnic identity simply subsidizes "discrimination by proxy": It allows discriminatory employers to wield "neutral" employment rules to perpetrate the same exclusionary and subordinating practices that they enforced before Title VII prohibited morphology-based race and ethnic discrimination. These patterns will continue unless we develop a more comprehensive model of racial and ethnic identity. This Article provides a start - a paradigm for understanding race, ethnicity, and discrimination - that takes into account both how behavior and physical traits become racially and ethnically inflected, and the circumstances in which these coded signatures work to the disadvantage of the person who claims them.
For related research and practice materials, see the following legal topics:
Labor & Employment LawDiscriminationRacial DiscriminationEnforcementLabor & Employment LawDiscriminationRacial DiscriminationCoverage & DefinitionsLabor & Employment LawAffirmative ActionVoluntary Plans
n1. In the early 1980s, courts reviewing Title VII race and national origin discrimination claims began distinguishing between those claims involving morphological race-or ethnic-associated traits and those involving voluntarily chosen race/ethnicity-associated or ethnic-associated traits. Ultimately, they held that Title VII did not provide relief for discrimination claims based on voluntary traits associated with protected class identities. See, e.g., Garcia v. Gloor, 618 F.2d 264, 268-72 (5th Cir. 1980) (denying bilingual Mexican American employee's challenge to employer's English-only rule because language choice was purely voluntary and therefore not included in definition of "national origin" under Title VII); Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 232 (S.D.N.Y. 1981) (denying African American woman's claim challenging company policy prohibiting all-braided hairstyles because hairstyles are voluntary and therefore not included in definition of race under Title VII); see also Carswell v. Peachford Hosp., No. C80-222A, 1981 U.S. Dist. LEXIS 14562, at 6 (N.D. Ga. May 26, 1981) (denying Title VII claim challenging employer policy prohibiting employees from wearing beads in their hair because "the wearing of beads in one's hair is [not] an immutable characteristic, such as national origin, race or sex").
n2. Claims concerning voluntary race/ethnicity-associated behavior may be raised as disparate treatment or disparate impact claims; however, both kinds typically fail. In a disparate treatment case, the plaintiff must show that she suffered discrimination under a facially neutral policy applied in a discriminatory manner (e.g., a rule against "unprofessional hairstyles" that is interpreted to prohibit cornrows, or a rule prohibiting a specific racialized or ethnic practice: "No one may wear African-inspired hairstyles"). The employee simply must show that she was treated differently than similarly situated employees who are not members of her race or ethnic group. See, e.g., Gloor, 618 F.2d at 267-69. In a disparate impact case, the plaintiff must show that a facially neutral rule disproportionately affects minority employees in an adverse manner. The rule suspiciously must exclude members of the protected class at a higher rate than outgroup members (e.g., a rule that provides that "no one may wear all-braided hairstyles"). See Marvin F. Hill, Jr. & James A. Wright, Employee Lifestyle and Off-Duty Conduct Regulation 76 (1993). Even if plaintiffs meet these initial burdens, they still might lose if their employers can offer a legitimate business justification for the rule being challenged. See, e.g., Fragante v. City and County of Honolulu, 888 F.2d 591, 595 (9th Cir. 1989).
n3. No. 1:96- CV-0196-CC, 1996 U.S. Dist. LEXIS 16190 (N.D. Ga. May 28, 1996).
n4. Id. at 4.
n5. An individual may bring a Title VII retaliatory discharge claim against her employer when she suffers an adverse employment action as a consequence of engaging in a protected activity listed in the statute. 42 U.S.C. 2000e-3(a) (2000). Protected activity includes all activities related to the filing of a discrimination complaint as long as the employee has a reasonable, good faith belief that the behavior or policy, which is the basis for her complaint, is a violation of antidiscrimination law. See EEOC v. Crown Zellerbach Corp., 720 F.2d 1008, 1013 (9th Cir. 1983); McBride, 1996 U.S. Dist. LEXIS 16190, at 5-6. In McBride, the court concluded that the plaintiff did not have a reasonable, good faith belief that the "no braided hairstyles" policy violated Title VII because cases since 1978 had established that voluntary aesthetic choices, such as hairstyle, were not part of the definition of racial status protected under Title VII. Id. at 6-7. Therefore, because McBride had not engaged in protected activity as defined by Title VII, her employer legitimately could fire her for challenging their hiring policies. Id. at 7.
n6. McBride, 1996 U.S. Dist. LEXIS 16190, at 7.
n7. Cases concerning race/ethnicity performance discrimination are not limited to the employment context. Students have raised race performance claims challenging school grooming codes as well. See New Rider v. Bd. of Educ., 480 F.2d 693 (10th Cir. 1973) (rejecting First Amendment challenge to school grooming code prohibiting Pawnee students from wearing traditional Native American hairstyles). Additionally, prisoners have raised challenges to grooming codes and disciplinary rules that prohibit or have a disparate impact on religious or race performance behavior. See, e.g., Hines v. S.C. Dep't of Corr., 148 F.3d 353, 356 (4th Cir. 1998) (rejecting Muslim, Rastafarian, and Native American prisoners' challenge to prison disciplinary code requiring short hair and clean-shaven faces); May v. Baldwin, 895 F. Supp. 1398, 1404-05 (D. Or. 1995), aff'd 109 F.3d 557, 564 (9th Cir. 1997) (rejecting black plaintiff's First Amendment challenge to prison regulation prohibiting dreadlocks). These claims typically fail, as courts find that the institution's interest in discipline, order, or encouraging conformity outweighs the individual's interest in expression. Hines, 148 F.3d at 358; see also May, 895 F. Supp. at 1404 (holding that rule prohibiting dreadlocks serves compelling governmental interest in preventing inmates from concealing contraband in their braids).
n8. For a detailed discussion of the relationship between public opinion and legislators' willingness to sponsor civil rights legislation, see Paul Burstein, Discrimination, Jobs, and Politics (2d ed. 1998). Burstein explains that though there was some interest in passing an antidiscrimination statute like Title VII in the early 1900s, members of Congress remained cautious about advancing any proposals into the 1940s, largely because they were unsure whether they would alienate portions of the electorate. Id. at 98-100. Additionally, Burstein explains that the Supreme Court stood as a formidable barrier to the enforcement of these kinds of statutes. Prior to the New Deal, the Court, almost by reflex, struck down laws regulating workplace conditions. As a consequence, legislators were not motivated to address this problem because they realized that their proposals could not survive judicial review. Id. at 16-17.
n9. Title VII provides a narrow list of groups afforded protection, the relevant ones for this discussion being race and national origin, which I refer to as "ethnicity." The statute protects each protected class slightly differently. Importantly, the statute does not define race, nor does it delineate the boundary between natural, involuntary racial/ethnic traits and voluntary, performed racial/ethnic features. See 42 U.S.C. 2000e-2(a) (2000). The legislative history of the statute and its amendments also fail to address this issue. See, e.g., S. Rep. No. 872 (1964), reprinted in 1964 U.S.C.C.A.N. 2355; H.R. Rep. No. 92-238 (1972), reprinted in 1972 U.S.C.C.A.N. 2137 (discussing 1972 amendments); H.R. Rep. No. 102-40(I) (1991), reprinted in 1991 U.S.C.C.A.N. 549 (discussing 1991 amendments). The legislative history of Title VII does offer a limited definition of national origin. However, it provides no guidance on where to draw the status/conduct divide with regard to this identity category either. See Espinoza v. Farah Mfg. Co., 414 U.S. 86 (1973) (discussing plain language and legislative history of 42 U.S.C. 2000e-2(a)(1)). Recognizing this ambiguity, the Equal Employment and Opportunity Commission (EEOC) has issued regulations attempting to bring voluntary behavior, such as language choice and accent, under the protection of the statute. See 29 C.F.R. 1606.7(a) (2003) (explaining that English-only rules can constitute national origin discrimination because they can "create an atmosphere of inferiority, isolation and intimidation based on national origin which [can] result in a discriminatory working environment"). The courts, however, have been somewhat hostile towards strict interpretations of these regulations. See, e.g., Garcia v. Spun Steak Co., 998 F.2d 1480, 1489 (9th Cir. 1993) ("Nothing in the plain language of [Title VII] supports [the] EEOC's English-only rule Guideline.").
n10. The majority view in the scientific community is that there are no truly biologically distinct races, given the small degree of genetic difference between races. See, e.g., R. C. Lewontin, The Apportionment of Human Diversity, in 6 Evolutionary Biology 381, 397 (Theodosius Dobzhansky et al. eds., 1972) (arguing against biological theories of race because there are no stable genetic differences that correlate with racial categories). Compare Ian F. Haney Lopez, The Social Construction of Race: Some Observations on Illusion, Fabrication and Choice, 29 Harv. C.R.-C.L. L. Rev. 1, 11-12 (1994) (summarizing biological studies showing that morphological traits are not stable within race groupings and that many paradigmatic racial traits do not track specific genes), and Kenneth L. Karst, Myths of Identity: Individual and Group Portraits of Race and Sexual Orientation, 43 UCLA L. Rev. 263, 305-06 (1995) (summarizing scientific critiques of racial categories claiming that (1) morphological traits associated with any given race are also found in other races, and (2) there is more genetic variation within races than between them), with Masatoshi Nei & Arun K. Roychoudhury, Genetic Relationship and Evolution of Human Races, in 14 Evolutionary Biology 1, 11, 41 (Max K. Hecht et al. eds., 1982) (concluding that rate of statistical difference between races is small but remains statistically significant and noting that some morphological traits are tied to particular genes). Anthony Appiah's discussion of this debate is cited frequently. He explains that most geneticists and biologists agree that there is a 14.3% rate of genetic variation within racial groups and a 14.8 % variation across racial groups. Kwame Anthony Appiah, In My Father's House: Africa in the Philosophy of Culture 35 (1992). This 0.5% genetic difference provides a rather thin evidentiary basis for the claim that there are biologically distinct races. Additionally, Appiah notes that the vast majority of scientists concede that even if this 0.5% difference is sufficient to establish that there are distinct races, these biological and morphological differences cannot be used as proxies for measuring inherent qualities, such as moral character or intelligence. Id. at 37. The Supreme Court has recognized this view. See, e.g., Saint Francis Coll. v. Al-Khazraji, 481 U.S. 604, 610 n.4 (1987) (enumerating biological and sociological studies acknowledging that race is "for the most part sociopolitical, rather than biological, in nature"). However, thus far, it has remained a peripheral issue, an important truth only granted the status of a footnote in American discrimination jurisprudence.
n11. See infra Part I.A.
n12. This model of "race/ethnicity performance" is based, in part, on Judith Butler's theory of "performativity," which describes gender identity performance. Judith Butler, Bodies that Matter: On the Discursive Limits of "Sex" (1993); see infra Part II.A for a further discussion of Butler's theory and its applicability to race and ethnicity.
Some scholars have indicated that ethnic behaviors or race performance should be understood as part of race and national origin identities without citation to Butler's model. See Karst, supra note 10, at 316-18 (discussing cultural and political aspects of black identity); Juan F. Perea, Ethnicity and Prejudice: Reevaluating "National Origin" Discrimination Under Title VII, 35 Wm. & Mary L. Rev. 805, 857-70 (1994) (arguing that Title VII should be amended to protect ethnic traits); Stephen M. Cutler, Note, A Trait-Based Approach to National Origin Claims Under Title VII, 94 Yale L.J. 1164 (1985) (arguing that courts should adopt analysis that includes ethnic traits when considering national origin discrimination claims); see also Kenji Yoshino, Covering, 111 Yale. L.J. 769, 878-923 (2002) (applying performance theory to analyze gender, race, and sexual orientation identities and investigating how assimilationist demands made of gays to "cover" their identity practices may parallel demands made of African Americans and women to "cover" voluntary aspects of their identities).
n13. See, e.g., Hollins v. Atl. Co., 188 F.3d 652, 661, 663 (6th Cir. 1999) (accepting black female plaintiff's Title VII disparate treatment claim based on her employer's scrutiny of her hairstyles); Fragante v. City and County of Honolulu, 888 F.2d 591, 596-97, 599 (9th Cir. 1989) (rejecting Title VII disparate treatment and national origin claim based on evidence that accent served as basis for denial of position because communication skills were integral to employee's position); Garcia v. Gloor, 618 F.2d 264, 269, 272 (5th Cir. 1980) (rejecting disparate treatment challenge based on employer's imposition of rule prohibiting workers from speaking Spanish at work); Upshaw v. Dallas Heart Group, 961 F. Supp. 997, 1000 (N.D. Tex. 1997) (rejecting disparate treatment claim based, in part, on plaintiff's allegation that she was terminated because her supervisor believed that she "sounded too black").
n14. See Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 232 (S.D.N.Y. 1981).
n15. Common law employment rules provide that a private employer has "the right, in the absence of statute or contract to the contrary, to fire an employee for personal reasons, unrelated to job function, that appeal to the employer, the color of hair, a dislike of men who smoke, or have a tattoo, etc." Gregory B. Reilly, Employees' Personal Appearance, 11 Lab. Law. 261, 262-63 (1995) (quoting Carter v. United States, 407 F.2d 1238, 1244 (D.C. Cir. 1968)); see also Hill & Wright, supra note 2, at 75 (noting that "courts accord significant discretion to management" regarding regulation of employees' appearance). The only limits on this discretion are: (1) Title VII; (2) union rules that prohibit employers from prohibiting certain employee behavior; and (3) state antidiscrimination statutes which may extend to issues beyond the purview of Title VII. See id.
n16. This discussion should not be read as an argument against "lookism," or part of the effort to create protections that would prohibit employers from using subjective appearance determinations in making personnel decisions. See Robert Post, Prejudicial Appearances: The Logic of American Antidiscrimination Law, 88 Cal. L. Rev. 1, 2-5 (2000) (describing local California ordinance prohibiting employers from using "personal appearance" as basis for employment decisions). Certainly, appearance-based decisions are often unfair and irrational, but they are also unavoidable. We have an inherent tendency to make predictions about others based on the social medium of appearance. See id. at 2 (arguing that human beings inevitably rely on physical traits in making judgments about others, as these features serve as symbols that trigger social meaning); see also Judith Butler, "Appearances Aside," 88 Cal. L. Rev. 55, 58-59 (2000) ("Appearance provides the epistemological condition for judging another person's worth or skill, even if that worth or skill is not, as it often is not, reducible to appearance itself."). This Article, however, does suggest that we all should be vigilant about how stigma associated with certain race and national origin groups tends to shape how we regard styles and behavior associated with these groups, independent of these traits' objective value. See infra notes 87-88 and accompanying text.
n17. Indeed, with a sufficient number of grooming and behavioral requirements that prohibit race/ethnicity-associated behaviors, employers can create such a hostile environment that minority employees voluntarily "elect" to leave their employ. For a discussion of the negative psychological effect these policies have, see Martha Chamallas, Structuralist and Cultural Domination Theories Meet Title VII: Some Contemporary Influences, 92 Mich. L. Rev. 2370, 2407-08 (1994). Chamallas explains that workplace rules that prohibit culture-specific behavior create an atmosphere of intimidation for minority employees and send a clear message of inferiority. For an illustration of a claim based on this view, see Garcia v. Spun Steak Co., 998 F.2d 1480, 1487-88 (9th Cir. 1993).
n18. See, e.g., Paulette M. Caldwell, A Hair Piece: Perspectives on the Intersection of Race and Gender, 1991 Duke L.J. 365 (1991). Caldwell's seminal article is the first scholarly discussion of the race-based dignitary and identity issues that shape Title VII grooming code challenges. See also Chamallas, supra note 17, at 2408 (explaining that workplace rules that are hostile to cultural practice exact dignitary harm); John M. Kang, Deconstructing the Ideology of White Aesthetics, 2 Mich. J. Race & L. 283, 316 (1997) (arguing that by validating such rules, courts strip race/ethnicity-associated behaviors, such as hairstyles, of their heritage and cultural meaning).
n19. See infra Part I.A.2.
n20. This view relies heavily on the "group position" model, a model drawn from the field of social psychology. The "group position" model is based on the proposition that "prejudice ... involves most centrally a commitment to a relative status positioning of groups in a racialized social order." Lawrence D. Bobo, Prejudice as Group Position: Microfoundations of a Sociological Approach to Racism and Race Relations, 55 J. Soc. Issues 445, 447 (1999). For an application of the model to workplace culture, see infra Part II.B.
n21. See infra Part I.A.
n22. See generally Milton M. Gordon, Assimilation in American Life: The Role of Race, Religion, and National Origins 115-31 (1964) (discussing melting pot model of assimilation).
n23. See id. at 132-59 (describing beginnings of cultural pluralism theories).
n24. Gordon W. Allport, The Nature of Prejudice (1979).
n25. See generally Butler, supra note 12.
n26. When discussing issues of racist antipathy or ethnic discrimination, I use the term "outgroup" to refer to persons outside of a particular individual's race or ethnic group.
n27. Most of the mistakes made in this analysis stem from a radically oversimplified understanding of the concept of immutability as it is used in equal protection doctrine. For a discussion regarding some of the concerns about the immutability doctrine, see Janet E. Halley, Sexual Orientation and the Politics of Biology: A Critique of the Argument from Immutability, 46 Stan. L. Rev. 503 (1994) (explaining that immutability criterion in Fourteenth Amendment's equal protection inquiry is based on biological construct of race and sex); Kenji Yoshino, Assimilationist Bias in Equal Protection: The Visibility Presumption and the Case of "Don't Ask, Don't Tell," 108 Yale L.J. 485 (1998) (same).
n28. See generally David A. Hollinger, Postethnic America: Beyond Multiculturalism 105-29 (2d ed. 2000) (expressing concerns about how identity politics tend to encourage groups to treat their experiences of oppression as irreducible and encourages individuals to form coalitions based on facile constructions of ethnic and racial identity); see also Martha Minow, Not Only for Myself: Identity, Politics, and the Law 32-46 (1997) (discussing various arguments against identity politics).
n29. Specifically, Gordon Allport explains in his seminal book, The Nature of Prejudice, that "the human mind must think with the aid of categories" or generalizations. "Once formed, categories are the basis for normal prejudgment." Allport, supra note 24, at 20. Once trained to recognize racial categories, an individual typically assigns persons to these categories based on conspicuous, visible cues. Id. at 21.
n30. Id. at 131-36.
n31. For the purposes of this discussion, I recognize four races: blacks, whites, Latinos, and Asians. I recognize, however, that many ethnic groups cannot be assigned consistently to any of the four race categories. For example, Filipinos are recognized variously as Latino or Asian. Samoans also defy easy categorization. My decision to treat Latinos as a race may raise some concerns, particularly since Title VII claims brought by members of this "race" typically frame their claims as national origin claims, each concerning a specific Latin ethnicity. Latinos themselves often resist characterizations that assume connections between Latin ethnicities, explaining that American race categories fail to capture adequately their experience of race, ethnicity, and culture. See Bernardo M. Ferdman & Placida I. Gallegos, Racial Identity Development and Latinos in the United States, in New Perspectives on Racial Identity Development: A Theoretical and Practical Anthology 32, 33 (Charmaine L. Wijeyesinghe & Bailey W. Jackson, III eds., 2001) [hereinafter New Perspectives]. Because various Latino ethnic groups are composed of persons with varying skin tones, members of the same family or ethnic enclave might fall into different racial categories according to American race categories. Michael Omi & Howard Winant, Racial Formation in the United States: From the 1960s to the 1980s, at 61 (1986); see also Mirta Ojito, Best of Friends, Worlds Apart, N.Y. Times, June 5, 2000, at A1. Further complicating matters, studies investigating Latinos' self-identification indicate that an individual may represent himself as a member of different racial groups depending on the circumstances. Ferdman & Gallegos, supra, at 43. While recognizing these problems, I characterize Latinos as a racial group because much of their experience of discrimination in the United States is premised on stigmatic associations broadly attributed to a Latin identity, real or imagined.
n32. Allport, supra note 24, at 132-34.
n33. Omi & Winant, supra note 31, at 62 ("One of the first things we notice about people when we meet them ... is their race."); see also id. at 23-24 (describing how Americans aggregate individuals of various ethnicities into broader racial categories, thinking they "all look alike").
n34. Allport, supra note 24, at 172 ("Many people find ... that the more they know about a group the less likely they are to form monopolistic categories.").
n35. See, e.g., Perkins v. Lake County Dep't of Util., 860 F. Supp. 1262, 1270 (N.D. Ohio 1994) (discussing Native American expert's claim that he could identify Native Americans by their physical features but admitting that others had misinterpreted his features to assume that he was Mexican or Italian).
n36. Allport recounts a stark moment in the 1950s South when a Hindu woman's sophistication about morphological paradigms allowed her to negotiate the tiers of legal rights afforded under Jim Crow. He explains:
A Hindu woman traveling in a southern state was denied a hotel room by a clerk who noticed her dark skin. The woman thereupon took off her headdress and showed that she had straight hair - and obtained accommodations. To the clerk it was color that cued his first behavior. The Hindu lady, with her keener sense of "small differences," forced the clerk to alter his perception, and reclassify her.
Allport, supra note 24, at 134.
n37. See id. at 134 (explaining that we categorize others by race even when we find their ethnicity ambiguous). The more prejudiced a person is, the more inclined she is to notice and accord significance to features or behaviors that are potential markers of racial difference. Id. at 133.
n38. Id. at 173-75.
n39. Marilynn Brewer discusses this disturbing phenomenon, explaining:
Ingroup membership is a form of contingent altruism. By limiting aid to mutually acknowledged [racial] ingroup members, total costs and risks of nonreciprocation can be contained. Thus, ingroups can be defined as bounded communities of mutual trust and obligation that delimit mutual interdependence and cooperation. An important aspect of this mutual trust is that it is depersonalized, extended to any member of the ingroup whether personally related or not.
Marilynn B. Brewer, The Psychology of Prejudice: Ingroup Love or Outgroup Hate?, 55 J. Soc. Issues 429, 433 (1999) (citation omitted).
n40. In the nineteenth century, scientists developed a number of studies to prove that there were essential genetic and biological differences between the races, including studies documenting differences in skin color, hair texture, facial angles, jaw size, cranial capacity, brain shape, and overall brain size. See Haney Lopez, supra note 10, at 14 & nn.53-54; Cheryl I. Harris, Whiteness as Property, 106 Harv. L. Rev. 1707, 1739 n.139 (1993). For a more detailed description of these theories, see generally Thomas F. Gossett, Race: The History of an Idea in America 54-83 (Oxford Univ. Press 1997) (1963); see also Michael Banton, Racial Theories 87-97 (2d ed. 1998). These theories lost sway as sociologists began to challenge the belief that there are any inherent qualitative differences between racial subjects and began to recognize pluralist models of society that recognized the same essential human potential in all races. Id.; see also supra note 10.
n41. See generally Ariella J. Gross, Litigating Whiteness: Trials of Racial Determination in the Nineteenth-Century South, 108 Yale L.J. 109 (1998) (analyzing court definitions of "whiteness" in racial determination cases concerning slave codes in late nineteenth and early twentieth centuries); Harris, supra note 40, at 1739-40.
n42. See generally Ian F. Haney Lopez, White By Law: The Legal Construction of Race (1996) (analyzing court-constructed definitions of race in citizenship cases in late nineteenth and early twentieth centuries).
n43. Daniel C. Littlefield, Rice and Slaves: Ethnicity and the Slave Trade in Colonial South Carolina 8-31 (1981).
n44. Gross, supra note 41, at 111-13 (discussing courts' role in constituting racial identity during antebellum period); id. at 123-76 (discussing trials).
n45. See infra notes 48-66 and accompanying text.
n46. Littlefield, supra note 43, at 8-31 (discussing early Americans' intense interest in ethnicity of slaves based on perception that some were hardier or more suited for certain kinds of labor than others).
n47. Gordon, supra note 22, at 95-97.
n48. See Gross, supra note 41, at 137-41 (explaining that morphology often did not track bloodlines in manner that permitted clear demarcation of races).
n49. Hypodescent rules were used to quantify the amount of black "blood" or ancestry that required an individual to be socially recognized as black. See Harris, supra note 40, at 1740 n.144 (explaining that hypodescent rules were designed to ensure that racially mixed persons were assigned status of socially subordinate parent); see also Neil Gotanda, A Critique of "Our Constitution Is Colorblind," 44 Stan. L. Rev. 1, 33-34 (1991) (discussing use of descent rules in institutionalizing and legitimizing subordinate minority class).
n50. See Gross, supra note 41, at 137-41 (discussing cases in which testimony about physical appearance was countered with testimony about ancestry). Gross analyzes sixty-eight trials in the nineteenth century in which courts were required to litigate the racial status of plaintiffs. She shows that, because mixed race plaintiffs in these cases often displayed morphology associated with "whiteness," courts were forced to receive evidence and engage in an extended inquiry into these persons' racial status. Id. at 122, 137-41. These trials were based on a fundamental paradox. The law posited that race was an obvious, surface observation that could be proven by reference to morphological difference. However, these racial determination trials often turned on the question of "blood," with courts arguing that morphology was misleading and blood lines were the only true determinant of racial status. Harris, supra note 40, at 1739. Review of the racial determination cases shows that courts offered a variety of justifications for their conclusions: scientific studies, morphological paradigms, common sense, and race-associated behavior (of course, the behaviors attributed to groups reflected the stereotypes and racist antipathy of the era). Gross, supra note 41, at 132-76; see also Gotanda, supra note 49, at 26-27 (discussing political importance of hypodescent rules); Harris, supra note 40, at 1740 (noting that legal standards applied in racial determination cases were increasingly "designed to accomplish what mere observation could not: 'That even Blacks who did not look Black were kept in their place'") (quoting Raymond T. Diamond & Robert J. Cottrol, Codifying Caste: Louisiana's Racial Classification Scheme and the Fourteenth Amendment, 29 Loy. L. Rev. 255, 281 (1983)).
n51. For a detailed examination of this issue, see Charles S. Mangum, Jr., The Legal Status of the Negro (1940). Cheryl Harris also provides a useful discussion contrasting the different hypodescent rules applied in different jurisdictions in the nineteenth century. See Harris, supra note 40, at 1738 & nn.136-38. Compare People v. Dean, 14 Mich. 406, 424-25 (1866) (applying rule that persons with less than one-quarter black blood are white for purposes of state voting statutes), with State v. Chavers, 50 N.C. (1 Jones Eq.) 11, 14-15 (1857) (applying rule that persons with less than one-sixteenth black blood are white).
n52. Dean, 14 Mich. at 423.
n53. The Dean court was asked to determine whether a voting statute that limited the franchise to white citizens should be construed to prohibit the racially mixed plaintiff from voting. Although the court openly criticized blood line rules and morphological paradigms of difference, citing their inconstancy as the basis for its complaint, id. at 417, it ultimately turned to a blood line analysis for its conclusion. It ruled that any person with more than three-quarters white ancestry should be treated as white and would be granted the right to vote in the state of Michigan. Id. at 425.
n54. Indeed, the progeny of these unions were destined to become slaves. Contrary to America's traditional reliance on patrilineal descent to determine one's social standing, children born of slave mothers inherited the status of the mother. See Haney Lopez, supra note 10, at 1.
n55. 11 Va. (1 Hen. & M.) 134 (1806).
n56. Id. at 139.
n57. In White By Law, Haney Lopez performs a close reading of appeals of citizen determination trials between 1870 and 1950, specifically exploring the race-based claims petitioners were forced to make to demonstrate their right to citizenship. Haney Lopez, supra note 42. Prior to the Civil War Amendments, a person could petition for citizenship only on the grounds that he was "white." Id. at 39-44. Although the franchise ultimately was extended to blacks, immigrants did not attempt to gain the rights of citizenship by offering proof that they were black, recognizing that there were significant privileges attached to whiteness and that blacks were actually second class citizens. Id. at 52-53. Courts in these cases sometimes refused to base their decisions on the conclusions produced by eugenics, particularly when these scientific studies contradicted social expectations about the membership of particular races. Id. at 7.
n58. Id. at 5-8.
n59. 1 F. Cas. 223, 223 (C.C.D. Cal. 1878).
n61. Id. The court ultimately conceded that neither the plain language of the immigration statutes nor their legislative history established Caucasians as the sole group entitled to be called "white." However, the court argued that it could be inferred from these materials that Congress did not intend for "Mongolians" to be treated as white persons. Id. at 224.
n62. 205 F. 812, 813 (E.D.S.C. 1913).
n65. Id. at 814.
n66. Id. at 816-17. In some cases, courts explicitly would reject the conclusions dictated by eugenics when these conclusions contradicted common sense understandings. See, e.g., United States v. Thind, 261 U.S. 204 (1923) (rejecting Indian man's claim of whiteness based on inclusion of Indians in scientific definition of Caucasian because common sense definition of Caucasian established that Indians were not considered white); Ozawa v. United States, 260 U.S. 178 (1922) (rejecting proposition that white skin was determinative feature of whiteness, accepting instead scientific definition of whiteness, which was persons of Caucasian descent, and denying Japanese man's citizenship claim).
n67. Haney Lopez argues that courts made their determinations about racial status by relying on one or more of the following four sources: (1) common knowledge; (2) science; (3) legislative history; and (4) precedent. See Haney Lopez, supra note 42, at 54, 64-65. See also Gross, supra note 41, at 138-39 (discussing lack of consensus about proper basis for determinations about "whiteness"); Harris, supra note 40, at 1739 (same).
n68. See generally Omi & Winant, supra note 31 (describing how seemingly objectively defined racial categories or constructs change over time in response to social pressures).
n69. See Haney Lopez, supra note 42, at 53 (explaining that racial definitions in play during a particular historical period are "products of their particular historical setting," and that judicial opinions in citizenship cases he discusses "do not record the facile recognition of racial difference, but rather the convoluted processes through which race is socially and legally constructed").
n70. Omi & Winant, supra note 31, at 61.
n72. Howard Winant, Racial Conditions: Politics, Theory, Comparisons 23-24 (1994).
n73. Id. at 24.
n74. Id. at 24-29 (discussing changes in political and cultural meaning of race from mid-to-late twentieth century).
n75. Id. at 29-30, 34.
n76. Id. at 24.
n77. See Harris, supra note 40, at 1736-41 (explaining that "the right to exclude was the central principle ... of whiteness as identity, for mainly whiteness has been characterized, not by an inherent unifying characteristic, but by the exclusion of others deemed to be 'not white'").
n78. See Gotanda, supra note 49, at 40-52 (discussing Supreme Court cases exploring meaning of racial constructs in cases concerning minority rights). In addition to showing how particular racial constructs are the product of the sensibilities of particular time periods, Gotanda also shows how different racial constructs compete for supremacy by contrasting the way race is discussed in the Supreme Court majority decisions as compared with the dissents and concurrences. Id.
n79. See, e.g., Perkins v. Lake County Dep't of Util., 860 F. Supp. 1262 (N.D. Ohio 1994) (inquiring about morphology of alleged Native American plaintiff).
n80. Mashpee Tribe v. Town of Mashpee, 447 F. Supp. 940 (D. Mass. 1978), aff'd sub nom. Mashpee Tribe v. New Seabury Corp., 592 F.2d 575 (1st Cir. 1979). This case is described in detail in Gerald Torres & Kathryn Milun, Stories and Standing: The Legal Meaning of Identity, in After Identity: A Reader in Law and Culture 129 (Dan Danielson & Karen Engle eds., 1995). The Mashpee tribe historically had defined membership in terms of voluntary affiliation as opposed to blood lines and morphology. Consequently, the tribe was relatively matzo in appearance, as it had never discouraged mixed unions. Id. at 132-33. The court, however, instructed the jury to determine whether the Mashpee were a tribe based on a legal inquiry that required (among other things) some finding of racial purity. Mashpee Tribe v. New Seabury Corp., 592 F.2d at 582; Torres & Milun, supra, at 130. Not surprisingly, the Mashpee ultimately failed to meet this test and were denied the right to be called a "tribe." Mashpee Tribe v. Town of Mashpee, 447 F. Supp. at 950; Torres & Milun, supra, at 131.
n81. These challenges typically concern the right to receive benefits through affirmative action programs. See, e.g., Susan Diesenhouse, In Affirmative Action, a Question of Truth in Labeling, N.Y. Times, Dec. 11, 1988, at E26. Diesenhouse discusses a case in which two Boston firemen who were brothers were dismissed after officials concluded that they were not entitled to be hired under the department's affirmative action hiring program because the firemen had stated falsely in their job applications that they were black. The firemen did have some "black" ancestry, as they had an African American grandmother. Id.
n82. See Omi & Winant, supra note 31, at 62.
n83. See infra Part I.B.
n84. In this discussion, I distinguish between voluntary and involuntary markers. Voluntary traits and markers are those behaviors and aesthetic displays that are capable of being eliminated or changed without chemical or surgical intervention. The analysis assumes that some "voluntary" markers may in fact be extremely difficult to change, either because they are second nature or because the body has changed in response to a certain practice in ways that make it difficult to avoid displaying these markers (e.g., changes in oral musculature which produce certain accents). However, these traits are still considered voluntary for the purposes of this analysis.
n85. Claire Dwyer & Peter Jackson, Commodifying Difference: Selling Eastern Fashion, 21 Env't & Plan.: Soc'y & Space 269 (2003) (explaining that marketing of cultural products transforms those products as they are disseminated and creates additional pressure to produce cultural difference).
n86. See Nelson George, Hip Hop America 60-75 (1998) (discussing white teenagers as fans and purchasers of rap and hip hop music).
n87. See generally Robin R. Means Coleman, Elmo Is Black! Black Popular Communication and the Marking and Marketing of Black Identity, 1 Popular Comm. 51 (2003) (discussing marketing of black cultural practices to mainstream culture and resulting reductionism and identity clashes within black communities that often are associated with these exchanges).
n88. For a discussion of group position theory, see infra Part II.B.
n89. Race/ethnicity performance ostensibly is always about "voluntary" displays and behavior. Therefore, most would assume that the motivations for and needs served by this behavior are primarily about identity expression. However, as explained above, the interests involved are more complex and involve a number of displays that are not chosen consciously.
n90. See, e.g., Upshaw v. Dallas Heart Group, 961 F. Supp. 997, 1000 (N.D. Tex. 1997) (discussing employee's allegation that she was fired because supervisor believed that "she sounded too black"); see also Jill Gaulding, Against Common Sense: Why Title VII Should Protect Speakers of Black English, 31 U. Mich. J.L. Reform 637, 645 (1998) (summarizing studies which found that "Standard English speakers are more successful in job interviews and that recruiters are less likely to offer a job to a Black English speaker").
n91. See, e.g., Fragante v. City and County of Honolulu, 888 F.2d 591, 597 (9th Cir. 1989) (rejecting applicant's claim that employer's complaints about his accent constituted national origin discrimination because good communication skills were central requirement for position). Mari J. Matsuda identifies the status valuations implicit in these determinations in Voices of America: Accent, Antidiscrimination Law, and a Jurisprudence for the Last Reconstruction, 100 Yale L.J. 1329 (1991). Specifically, Matsuda explains that when a person encounters a speaker with an accent associated with a low-status group, she is more likely to conclude that the speaker is unintelligible and fail to do the necessary cognitive work that would enable her to understand the speaker. Id. at 1355. Such a conclusion might be based on a "hidden assumption of an Anglo accent at the center." Id. at 1394. Matsuda argues that we should question employers' motives in cases where they deny a minority worker a job based on "accent" because often the alleged difficulty in understanding the speaker is a product of an intolerance that should be understood properly as racial or ethnic discrimination. Id. at 1377-78 (explaining that some people automatically associate certain accents with laziness or untrustworthiness).
n92. Mari Matsuda demonstrates the prevalence of this phenomenon by describing a study examining the diction and pronunciation of store clerks at three department stores catering to different classes. The highest status store had clerks with more standard, proper English pronunciation. The discount store clerks displayed more variation in their pronunciation of certain words, specifically dropping certain consonants from their pronunciation. She explains that the more expensive, higher-status store tended to prefer candidates for employment with "waspier" accents. Matsuda, supra note 91, at 1377.
n93. Gaulding, supra note 90, at 644 (recognizing that some employers do not give employees opportunity to seek help to correct racially coded speech patterns).
n94. The case can be made that passive race performance does not fit traditional notions of voluntary behavior. Indeed, scholars have indicated that some behaviors are so ingrown that either the process of unlearning them proves prohibitively difficult or the burden is so great that it offends our notions of fairness and dignity to ask a person to do so. See Chamallas, supra note 17, at 2407-08 (noting that rules prohibiting certain types of cultural performance are experienced as sending message of inferiority to their targets); Karl E. Klare, Power/Dressing: Regulation of Employee Appearance, 26 New Eng. L. Rev. 1395, 1400 (1992) (noting that employees experience great shame when their appearance choices are rejected by employers).
n95. See McManus v. MCI Communications Corp., 748 A.2d 949, 956-57 (D.C. Cir. 2000) (rejecting black plaintiff's disparate treatment claim against employer alleging that she was terminated for wearing African style attire as well as dreadlocked and braided hairstyles); McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 865-66 (S.D. Miss. 1992) (rejecting disparate treatment claim concerning employer's enforcement of neutral grooming rules to prohibit headwraps and dreadlocks).
n96. Cf. Garcia v. Gloor, 618 F.2d 264, 266-67 (5th Cir. 1980) (rejecting Title VII claim challenging English-only rule that prohibited worker from speaking Spanish to coworkers during working hours).
n97. Cf. Roman v. Cornell Univ., 53 F. Supp. 2d 223, 237 (N.D.N.Y. 1999) (rejecting challenge to English-only rule and noting that employers may prohibit speaking Spanish in order to prevent non-Spanish speakers from "feeling left out" or "feeling that they are being talked about in a language they do not understand").
n98. In these types of cases, an employer typically contends that no discrimination has occurred, pointing to his fair treatment of other minority workers who do not display these "voluntary" racial and ethnic features. This evidence, however, is completely irrelevant to the problem at issue. Kenji Yoshino's concept of "cover" can be used to describe the minority individual's status in the workplace under such a regime. The individual is allowed to keep her job as long as she "mutes the difference between herself and the mainstream." See Yoshino, supra note 27, at 500. The purpose behind the demand to cover is to minimize "how much the visible, [community-specific] trait impedes the flow of 'normal' interaction." Id. at 501. This notion of "cover" suggests that these features are part of a predetermined natural identity. I find it more helpful to emphasize the idea of race/ethnicity performance because it emphasizes that these practices typically are affectations, none of which can make claims to a more authentic form of representation.
n99. These cases promise to be a significant part of courts' Title VII docket in the future as the number of interracial unions increase and the morphologically indistinct children produced by these unions face discrimination because of cultural or behavioral patterns they learn from their parents.
n100. Specifically, courts' analysis of race has not accounted for the fact that race is the byproduct of an interpretive exchange of racial signifiers between the subject and viewer. Yoshino, supra note 27, at 498 ("There is no such thing as a purely biologically visible trait, for visibility is always relational, requiring a performer and an observer."). Whether a trait is visible thus will depend not only on the performer's attempts to conceal the trait, but also on the "'decoding capacity' of her audience." Yoshino, supra note 12, at 822 (quoting Erving Goffman, Stigma: Notes on the Management of Spoiled Identity 48-51 (1963)).
n101. See Kathleen M. Sullivan, The Supreme Court, 1991 Term - Foreword: The Justices of Rules and Standards, 106 Harv. L. Rev. 24, 112 (1992) ("All the Justices seek to distinguish the judicial enterprise from politics; none seeks to 'legislate from the bench.'").
n102. See supra Part I.A.1.
n103. In my research on Title VII race and ethnic discrimination cases, I found only one case in which an employer defended against a plaintiff's claim by arguing that a plaintiff was not entitled to the protections of Title VII, challenging whether the plaintiff in fact could rightfully claim membership in a protected class. See Perkins v. Lake County Dep't of Util., 860 F. Supp. 1262, 1264-77 (N.D. Ohio 1994) (addressing employer's claim that bloodlines and legal records assigning race were determinative of plaintiff's racial status for purpose of Title VII inquiry).
n104. Although there are a wide range of perspectives represented in the Critical Legal Studies (CLS) movement, some of the seminal texts include Robert W. Gordon, Unfreezing Legal Reality: Critical Approaches to Law, 15 Fla. St. U. L. Rev. 195 (1987); Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685 (1976); Mark Tushnet, Critical Legal Studies: An Introduction to its Origins and Underpinnings, 36 J. Legal Educ. 505 (1986). Although not always denominated as CLS, after the late 1980s, a number of scholars continued in this tradition. CLS scholars deconstruct legal decisions to reveal judicial decisionmakers' possible political motivations and the ways in which their decisions are both produced by and reflective of certain types of social knowledge and social institutions. The subsection of CLS work that focused most centrally on race is sometimes categorized as Critical Race Theory. See, e.g., Critical Race Theory: The Key Writings that Formed the Movement (Kimberle Crenshaw et al. eds., 1995).
n105. See generally Critical Race Theory: The Key Writings that Formed the Movement, supra note 104; see also Thomas Ross, Just Stories: How the Law Embodies Racism and Bias (1996).
n106. See Post, supra note 16, at 31-33 (contrasting dominant approach to antidiscrimination law with sociological approach he advances); Katharine T. Bartlett, Only Girls Wear Barrettes: Dress and Appearance Standards, Community Norms, and Workplace Equality, 92 Mich. L. Rev. 2541, 2544-45 (1994) (discussing judges' reliance on community norms as unavoidable in legal analysis).
n107. See, e.g., Bartlett, supra note 106, at 2569.
n108. Post, supra note 16, at 31-33.
n109. Id. at 32.
n110. Bartlett, supra note 106, at 2569.
n112. See, e.g., Jurado v. Eleven-Fifty Corp., 813 F.2d 1406, 1409-10 (9th Cir. 1987) (rejecting Latino plaintiff's claim regarding expressive interest in using Spanish in his radio program); Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 231-33 (S.D.N.Y. 1981) (rejecting African American employee's dignitary and expressive concerns regarding her braided hairstyle).
n113. Post, supra note 16, at 31.
n114. Bartlett, supra note 106, at 2569.
n116. See id. at 2569-70; see also Matsuda, supra note 91, at 1350-55 (arguing that court in Fragante v. City and County of Honolulu, 888 F.2d 591 (9th Cir. 1989), failed to interrogate its own common sense views about need for intelligibility and what constituted intelligibility when analyzing Filipino plaintiff's claim that his employer was engaging in accent discrimination).
n117. Indeed, this Article's comparison of the similar cognitive processes that inform morphology- and performance-based discrimination is intended to provide judges with the tools to challenge their "common sense" views about how discrimination works in society. Hopefully, with an improved understanding of the cognitive bases of discrimination, judges will conclude that the reasons for sanctioning morphology-based discrimination suggest that we should be concerned about performance-triggered discrimination as well.
n118. Indeed, plaintiffs rarely appear before the court without evidence in support of their claim that a voluntary practice has a significant race-associated history or evidence showing that the practice has, in fact, been correlated with racial animus in the particular case.
n119. See generally William E. Cross, Jr. & Peony Fhagen-Smith, Patterns of African American Identity Development: A Life Span Perspective, in New Perspectives, supra note 31, at 243, 243-44 (discussing various models of Nigrescence). For an earlier, more complete discussion of Cross's model, see William E. Cross, Jr., The Thomas and Cross Models of Psychological Nigrescence, 5 J. Black Psychol. 13 (1978).
n120. See generally Bailey W. Jackson III, Black Identity Development: Further Analysis and Elaboration, in New Perspectives, supra note 31, at 8 (offering enhanced and updated Black Identity Development (BID) model).
n121. See generally Butler, supra note 12, at 1-3; Judith Butler, Gender Trouble: Feminism and the Subversion of Identity (1990) (discussing gender as effect of performances).
n122. See Bobo, supra note 20, at 447 (analyzing and elaborating on Blumer's "group position" theory).
n123. Samuel L. Gaertner & John F. Dovidio, The Aversive Form of Racism, in Prejudice, Discrimination, and Racism 61 (John F. Dovidio & Samuel L. Gaertner eds., 1986) [hereinafter Gaertner & Dovidio, Aversive Racism]. Gaertner and Dovidio's work is built on their earlier studies of whites' helping behavior in accident situations as compared between white and non-white victims. See David L. Frey & Samuel L. Gaertner, Helping and the Avoidance of Inappropriate Interracial Behavior: A Strategy That Perpetuates a Nonprejudiced Self-Image, 50 J. Personality & Soc. Psychol. 1083 (1986); Samuel L. Gaertner et al., Race of Victim, Nonresponsive Bystanders, and Helping Behavior, 117 J. Soc. Psychol. 69 (1982); Samuel L. Gaertner & John F. Dovidio, The Subtlety of White Racism, Arousal, and Helping Behavior, 35 J. Personality & Soc. Psychol. 691 (1977).
n124. Importantly, race/ethnicity performance behaviors should not be understood as synonyms for culture, although some of these behaviors are cultural in nature. See Hollinger, supra note 28, at 36-37, 48-49 (warning against simplistic analyses that conflate race with culture). I argue that while many cultural behaviors are strongly racially marked and stigmatized (e.g., all-braided hairstyles), thus triggering what I call race performance discrimination, other race-associated behaviors have no traditionally defined cultural dimension and are more likely the side effects of residential segregation, or a shared cross-ethnic reaction to stigma. For example, some might argue that accents, or dialects, such as Spanglish or Black English, are not "cultural," as they are not the product of conscious design, but instead are a natural development caused by high degrees of racial and ethnic segregation. Neil Gotanda begins discussion on this issue with his concept of "culture-race," which he argues "includes all aspects of culture, community, and consciousness." See Gotanda, supra note 49, at 56. This Article gives more dimension to that view, articulating the myriad ways in which racial and ethnic identities may come to be characterized by acts, aesthetics, and behaviors, whether by conscious design or inadvertent effect.
n125. Since the late 1950s, the majority of racial identity development scholars focused on African American identity development processes. See, e.g., Racial and Ethnic Identity: Psychological Development and Creative Expression (Herbert W. Harris et al. eds., 1995) (presenting series of studies on African American identity development); Cross & Fhagen-Smith, supra note 119 (offering newly updated and "repositioned" Nigrescence model of racial identity development); Jackson, supra note 120 (offering BID model based on African American racial identity issues). In the 1990s, scholars began to focus on other racial groups' identity development processes, at first comparing them with those of African Americans. See, e.g., Maurianne Adams, Core Processes of Racial Identity Development, in New Perspectives, supra note 31, at 209, 211 (identifying "generic processes of identity development across various racial and ethnic groups"); Jennifer Crocker et al., Collective Self-Esteem and Psychological Well-Being Among White, Black, and Asian College Students, 20 Personality & Soc. Psychol. Bull. 503 (1994) (discussing collective self-esteem theory); Jean S. Phinney, The Multigroup Ethnic Identity Measure: A New Scale for Use with Diverse Groups, 7 J. Adolescent Res. 156 (1992) (discussing common identity development processes across racial/ethnic groups); Jean S. Phinney, When We Talk About American Ethnic Groups, What Do We Mean?, 51 Am. Psychologist 918 (1996) (discussing culture, ethnic identity, and minority status as aspects of ethnicity and race).
However, recognizing the limitations of the African American-centered models, scholars are beginning to explore the specific challenges faced by other racial groups in developing racial identity. See, e.g., Ferdman & Gallegos, supra note 31 (discussing Latino identity development); James E. Coverdill, White Ethnic Identification and Racial Attitudes, in Racial Attitudes in the 1990s: Continuity and Change 144-79 (Steven A. Tuch & Jack K. Martin eds., 1997) (discussing rise in whites' interest in ethnic identification interest and theorizing possible relationship to racist impulses). The works of these scholars can be placed within a larger field of identity development studies and are particularly indebted to Erik Erikson's work on ego identity development. See Erik Homburger Erikson, The Problem of Ego Identity, 4 J. Am. Psychoanalytic Ass'n 56 (1956); see also Cross & Fhagen-Smith, supra note 119, at 247-49 (contextualizing Nigrescence model within larger ego identity model); Jackson, supra note 120, at 11-12 (noting that Erikson's work influenced development of BID theory).
n126. Two competing antidiscrimination approaches were present in the early years of the civil rights movement, each associated with the specific groups that employed them. See Kristin Bumiller, The Civil Rights Society: The Social Construction of Victims 46-48 (1988). One approach was premised on working within the legal system, as advanced by the National Association for the Advancement of Colored People. The other targeted racism through grass roots activism, such as boycotts; this approach was promoted in the late 1950s and early 1960s by the more moderate Southern Christian Leadership Conference (SCLC), led by Martin Luther King, as well as the more radical Student Nonviolent Coordinating Committee (SNCC) and the Black Power Movement.
n127. Ferdman & Gallegos, supra note 31, at 33, 38-39 (noting that "much of the thinking on race in the United States stems from the history of Blacks and Whites and their relationship" and explaining that American race categories are irreconcilable with understandings of race, ethnicity, and culture that order various Latino ethnic groups living outside of United States).
n128. See Cross & Fhagen-Smith, supra note 119, at 244 (describing Cross's work creating "Cross Nigrescence Model").
n129. See Jackson, supra note 120, at 8-11 (contextualizing BID and Nigrescence theories). For a more detailed treatment of BID theory, see Bailey Jackson, Black Identity Development, in Urban, Social, and Educational Issues 158 (Leonard H. Golubchick & Barry Persky eds., 2d ed. 1976).
n130. Under the Nigresence model, an individual transitions through five stages: (1) pre-encounter - the "stable identity that will eventually be the object of the metamorphosis"; (2) encounter - the stage in which the individual experiences the event or group of events that challenge the stable identity; (3) immersion-emersion - "the simultaneous struggle to bring to the surface and destroy the moorings of the old identity, while decoding the nature and demands of the new identity" (assuming there is no regression); (4) internalization - "the habituation, stabilization, and finalization of the new sense of self"; and (5) internalization-commitment - the point at which the individual translates this stable racial identity into a basis for cultural transmission and political action. Cross & Fhagen-Smith, supra note 119, at 244.
Under the BID model, an individual goes through a similar five-stage liberatory sequence: (1) a naive stance - "the absence of a social consciousness or identity"; (2) acceptance - adoption of majority white norms and relative worth of black people; (3) resistance - whether active or passive, in which the individual rejects "the prevailing majority culture's definition and valuing of Black people"; (4) redefinition - "the renaming, reaffirming, and reclaiming of one's sense of Blackness, Black culture, and racial identity"; and (5) internalization - "the integration of a redefined racial identity into all aspects of one's self-concept or identity." Jackson, supra note 120, at 15-16, 18-26.
n131. Cross & Fhagen-Smith, supra note 119, at 248, 261; Jackson, supra note 120, at 8-12. Both theorists argue that an individual must progress through all stages in their psychological models in order to become a fully realized, psychologically stable subject that is comfortable with her racial identity.
n132. See Jackson, supra note 120, at 13-14 (discussing influence of BID theory on development of white, Asian, Jewish, and multiracial identity development theories as well as general theory of minority identity development).
n133. See, e.g., Aaron Celious & Daphna Oyserman, Race From the Inside: An Emerging Heterogeneous Race Model, 57 J. Soc. Issues 149, 154-55 (2001) (describing middle-class blacks' ability to employ multiple frameworks for expressing African American identity in different contexts).
n134. See id. at 151 (explaining that black subjects' "race, and racial identity can be symbolic, switched into, or deemphasized depending on context").
n135. See Vetta L. Sanders Thompson, Variables Affecting Racial-Identity Salience Among African Americans, 139 J. Soc. Psychol. 748, 748 (1999) ("Not all possible members identify with the [racial] group, nor do all members identify equally. Members may differ in their willingness to identify with specific issues or aspects of the group, and their differences are theoretically related to the salience of their identities.").
n136. Although Afrocentricity is not explicitly mentioned in Jackson's BID theory, he indicates in other sections of his analysis that he is referring to an Afrocentric perspective when he speaks of the development of a positive black identity. Jackson, supra note 120, at 27.
n137. For example, new immigrants' racial identity development issues are likely to focus equally on the development of an Americanized version of their racial identities and on whiteness as an identity category. See, e.g., Ferdman & Gallegos, supra note 31, at 50-54 (exploring new Latino identity models including those that feature whiteness as aspect of identity).
n138. See Cross & Fhagen-Smith, supra note 119; Jackson, supra note 120.
n139. See generally Butler, supra note 121.
n140. Several scholars have suggested that performance theory could be applied usefully to understand racial identity development. See, e.g., Klare, supra note 94, at 1409 (explaining that "people constitute their identities, including identities of race, gender, class, sexual orientation, and so on, through social practices, including appearance practices"); see also Yoshino, supra note 12, at 879-905 (comparing insights from gay identity performance with insights into understanding race identity performance issues). However, thus far, no one has offered a comprehensive account of how these principles might be applied to further understand the individual's relationship to and use of racial identity categories.
n141. See generally Butler, supra note 121.
n142. See generally Butler, supra note 12.
n143. Butler, supra note 121, at 16-25.
n144. Id. at 6-7.
n145. Id. at 142-45.
n146. Id. at 145. Contrasting her model to the traditional models of subjectivity and identity, Butler explains:
The question of locating "agency" is usually associated with the viability of the "subject," where the "subject" is understood to have some stable existence prior to the cultural field that it negotiates. Or, if the subject is culturally constructed, it is nevertheless vested with an agency, usually figured as the capacity for reflexive mediation, that remains intact regardless of its cultural embeddedness. On such a model, "culture" and "discourse" mire the subject, but do not constitute that subject... . And yet, this kind of reasoning falsely presumes (a) agency can only be established through recourse to a prediscursive "I" ... and (b) that to be constituted by discourse is to be determined by discourse ... .
Id. at 142-43.
n147. Id. at 144-45; Butler, supra note 12, at 7.
n148. Butler, supra note 121, at 143-45.
n149. Cf. id. (discussing proposition as applied to gender identity). I would describe this version of performativity as a "strong constructionist model," a version which denies the biological or physical fact of race and sex categories.
n150. Cf. id. at 144-45 (discussing gender) ("The rules that govern intelligible identity, i.e., that enable and restrict the intelligible assertion of an 'I,' rules that are partially structured along matrices of gender hierarchy ... operate through repetition.").
n151. See generally Lise Funderburg, Black, White, Other: Biracial Americans Talk About Race and Identity (1994) (discussing consequences of choosing - or not choosing - to accept one racial identity); Stephen L.H. Murphy-Shigematsu, The Voices of Amerasians: Ethnicity, Identity, and Empowerment in Interracial Japanese Americans (2000) (same), at www.dissertation.com/library/112080xa.htm (last visited Sept. 6, 2004).
n152. Cf. Butler, supra note 12, at 2-3 (clarifying that performativity does not seek to deny material realities of sex but asks us to question whether what has been naturalized as biological is actually culturally determined).
n153. Charmaine L. Wijeyesinghe, Racial Identity in Multiracial People: An Alternative Paradigm, in New Perspectives, supra note 31, at 129, 140-41.
n154. Butler, supra note 121, at 8 (discussing relationship between morphology and gender: "'The body' it [sic] itself a construction, as are the myriad 'bodies' that constitute the domain of gendered subjects.").
n155. See, e.g., Wijeyesinghe, supra note 153, at 140-41 (quoting multiracial woman, who looked white, as saying, "Because of my physical appearance, I'm not gonna be taken serious. It's hard in a way that I almost want to be able to wear a sign or something letting people know of my background and whatever, so that they can accept me ... .").
n156. See id. For a literary treatment of this problem, see Malcolm X, The Autobiography of Malcolm X (Ballantine Books 1999) (1965).
n157. See Butler, supra note 121, at 8 (discussing proposition in context of gender).
n158. Id. at 144-45.
n159. See id. at 128-41 (discussing "subversive bodily acts" through which identity may be performed "on the surface of the body") (emphasis omitted).
n160. See, e.g., Hollins v. Atl. Co., 188 F.3d 652, 655-57 (6th Cir. 1999) (describing African American employee changing hairstyles when faced with harassment by her employer); McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 854-59 (S.D. Miss. 1992) (same).
n161. Kertzer describes more than one definition of the term "ritual." However, because he emphasizes the need for repetition and the symbolic importance of these acts in interpreting the events of everyday life, his model is consistent with the definition of "active race/ethnicity performance" as used in this discussion. See David I. Kertzer, Ritual, Politics, and Power 8-9 (1988) (describing ritual as "means of channeling emotion, guiding cognition, and organizing social groups").
n162. Id. at 10.
n164. Cf. Butler, supra note 121, at 144-45 (discussing proposition in context of gender).
n166. Race/ethnicity performance behaviors do, to some degree, rely upon social stereotypes to communicate meaning. Indeed, as Karst explains, a person is often forced into "enactment (even in all sincerity) of the narratives of stereotype attached to the various identity categories [in order to signal his identity category] - or, for people who want to avoid these caricatures, a conscious performance against type." Karst, supra note 10, at 288 (emphasis omitted). This is because our world is made out of already established cultural meanings, which we recombine and reinterpret to make new associations. Because these understandings which link behaviors and physical characteristics to groups are not always negative, this analysis refers to such understandings in less loaded terminology: social codes.
n167. Butler suggests, regarding gender, that these idiosyncratic practices are the results of pastiche, combining old behaviors or shaping behaviors to fit contemporary circumstances. Butler, supra note 121, at 145 (explaining that subject's agency, "located within the possibility of a variation on that repetition," is expressed through "complex reconfiguration and redeployment" of available social norms and symbols); see also Klare, supra note 94, at 1408 (explaining that individual innovation on identity performance comes from "acknowledging, rearranging, recombining, and altering the bits and pieces [of behaviors and signs] already available within the cultural context").
n168. Butler, supra note 12, at 6-8.
n169. Id. at 7.
n170. Kertzer, supra note 161, at 8-12 (defining and describing characteristics of rituals).
n171. Id. at 9-11. Kertzer aptly notes that an individual's participation in certain group-based symbolic practices arouses an emotional response through which a complex ritual drama is permitted to take hold. As applied here, Kertzer's work suggests that race performance may help individuals actualize more complicated political ideas about race. For example, an individual may not be familiar with all of the constitutive political and religious tenets associated with the Nation of Islam; however, he can affirm his connection to the group by reading the group's newspaper or wearing a Kufe during his period of experimentation with the identity. Lastly, Kertzer indicates that symbolic behaviors like race performance can build bonds of solidarity among community members. Id. at 61-67.
n172. See Cross & Fhagen-Smith, supra note 119, at 250-51; Wijeyesinghe, supra note 153, at 138-40.
n173. See Cross & Fhagen-Smith, supra note 119, at 248, 250-51 (recognizing that cues for performance of racial identity are drawn primarily from family and community).
n174. See id. at 250-51, 253.
n175. This is consistent with Cross's model of Nigrescence, which indicates that an individual may move from a racially neutral position to a more race-affirming outlook in the face of a traumatic event. Id. at 260 (describing this kind of experience as part of "encounter" phase of his model).
n176. See, e.g., McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 855-57 (S.D. Miss. 1992) (describing plaintiff's changes in hairstyles and head coverings); see also Caldwell, supra note 18, at 383-85 (discussing changing hairstyles as means of expressing cultural identity).
n177. See Kertzer, supra note 161, at 11. Kertzer explains that "creating a symbol or ... identifying oneself with a popular symbol can be a potent means of gaining and keeping power, for the hallmark of power is the construction of reality." Id. at 5.
n178. Id. at 11; see also Kenji Yoshino, Suspect Symbols: The Literary Argument for Heightened Scrutiny for Gays, 96 Colum. L. Rev. 1753, 1770 (1996) (explaining that symbols "subsume plural meanings in order to have plural appeal" and have "ability to draw together communities of adherents precisely because they do not force believers to articulate what it is about the symbol that draws them together"). Kertzer agrees and explains that symbols' ambiguity increases their power; "the complexity and uncertainty of meaning of symbols are sources of their strength." Kertzer, supra note 161, at 11.
n179. See Yoshino, supra note 178, at 1770.
n180. May v. Baldwin, 895 F. Supp. 1398, 1403-05 (D. Or. 1995), aff'd 109 F.3d 557, 559-64 (9th Cir. 1997) (discussing plaintiff's claim that Rastafarian religious beliefs required him to wear dreadlocks); McGlothin, 829 F. Supp. at 861-63 (discussing "Hebrew Israelite" and African American culture justifications).
n181. See Matsuda, supra note 91, at 1333-39 (discussing accent discrimination in Fragante v. City and County of Honolulu, 888 F.2d 591 (9th Cir. 1989)).
n182. See id. at 1338-39 (noting that employer prevailed in Fragante).
n183. Bobo, supra note 20.
n184. See Gaertner & Dovidio, Aversive Racism, supra note 123.
n185. John F. Dovidio, On the Nature of Contemporary Prejudice: The Third Wave, 57 J. Soc. Issues 829, 830 (2001).
n187. Id. at 831.
n188. See id. at 831-32 (noting that second wave focused on "how normal processes associated with socialization and social norms [could] support prejudice and aid in its transmission").
n189. Id. at 835.
n190. Id. at 838-41.
n191. See id. at 835-38 (describing studies on aversive racism in emergency intervention situations and hiring decisions).
n192. See, e.g., id. at 844-45.
n193. Bobo, supra note 20.
n194. Id. at 449 (summarizing Herbert Blumer's theory of group position).
n195. Id. at 454 (explaining that dominant group's concerns run along two axes: "One axis involves the more obvious dimension of domination and oppression, of hierarchical ordering and positioning. [The] second ... involves a dimension of exclusion and inclusion, of socioemotional embrace or recoil").
n197. Id. at 459.
n198. Id. at 449.
n200. There are two schools of thought about the relationship between ingroup pride and outgroup prejudice. The first presumes that ingroup preference can exist without outgroup hostility; the second presumes that the two are reciprocally related. Brewer, supra note 39, at 430 (noting that majority of scholarship presumes ingroup bias and outgroup prejudice are always combined). A group of scholars, however, building on the first premise, has sought to identify the circumstances when these ingroup preferences and solidarities will correlate with outgroup prejudice and hostility. Amelie Mummendey & Hans-Joachim Schreiber, Better or Just Different? Positive Social Identity by Discrimination Against, or by Differentiation from Outgroups, 13 Eur. J. Soc. Psychol. 389, 391 (1983) (investigating which types of comparisons reflect outgroup discrimination in addition to ingroup bias); Amelie Mummendey et al., Categorization Is Not Enough: Intergroup Discrimination in Negative Outcome Allocation, 28 J. Exp. Soc. Psychol. 125, 125 (1992) (investigating whether group members exhibit ingroup favoritism in distributing negative outcomes between ingroup and outgroup); Naomi Struch & Shalom H. Schwartz, Intergroup Aggression: Its Predictors and Distinctness From In-Group Bias, 56 J. Personality & Soc. Psychol. 364, 364 (1989) (investigating relation of ingroup bias to outgroup aggression). Marilynn Brewer argues in favor of the first premise, claiming that ingroup solidarity and outgroup antipathy are not necessarily related. Brewer, supra note 39, at 430. She indicates that much of contemporary prejudice is not motivated by hostility towards outgroups, but rather by the attempt to exhibit favor for and limit preferences to ingroup members. Id. at 433-34. In a context where workers participate in a zero-sum game for opportunities and benefits, particularly when one of the central issues in dispute is the cultural backdrop of the workplace, this ingroup preference can ripen into outgroup hostility. See id. at 435. In all, Brewer discusses five factors that tend to make ingroup preferences dangerous: (1) feelings of ingroup moral superiority; (2) perceived threat of competition from outgroup members for social resources; (3) perceived advancement of a shared threat or goal, which highlights a lack of basis for trust between group members; (4) shared values, which can promote competition on the same or similar scales of differential worth; and (5) formal political contests between groups. Id. at 435-38. Brewer explains that the best way to achieve social stability is to get people to recognize the multiple bases for their identities as opposed to encouraging ingroup loyalty on a single axis. Id. at 439-41.
n201. Bobo, supra note 20, at 449 (discussing premises from Herbert Blumer's work).
n205. Id. at 450.
n206. Gaertner & Dovidio, Aversive Racism, supra note 123, at 61-62.
n207. John F. Dovidio et al., Why Can't We Just Get Along? Interpersonal Biases and Interracial Distrust, 8 Cultural Diversity & Ethnic Minority Psychol. 88, 90 (2002); see Gaertner & Dovidio, Aversive Racism, supra note 123, at 66.
n208. Dovidio et al., supra note 207, at 90.
n210. Id. The authors recognize that minorities can display the same kind of aversive racism but explain that they focus on whites in their analysis because this group holds a disproportionately greater amount of political, social, and economic power. Id.
n211. Id.; see Gaertner & Dovidio, Aversive Racism, supra note 123, at 66. ("Even when normative guidelines are clear, aversive racists unwittingly may search for ostensibly nonracial factors that could justify a negative response to blacks.").
n212. Gaertner & Dovidio, Aversive Racism, supra note 123, at 62.
n213. Id.; Dovidio, supra note 185, at 834-35.
n214. See Dovidio et al., supra note 207, at 94 (noting that such attitudes "commonly arise developmentally").
n216. Gaertner & Dovidio, Aversive Racism, supra note 123, at 63. They note that sometimes the negative feelings start out as a fear of discussing race issues with blacks out of concern that they may be regarded as racist, and subsequently, this general anxiety and unease becomes directly associated with blacks (and other minorities). See id. at 64.
One of the reasons people find it difficult to identify and problematize aversive racist thinking is because they have been taught to villainize old-guard, traditional, explicit racism and build their identities around rejecting this kind of extreme behavior. Once their identities are constructed in this manner, they tend not to be attuned to the fact that racist attitudes and behaviors fall on a continuum, and that their rejection of extreme forms does not obviate the need for them to address racist behaviors (including their own) that take a more subtle form. See Barbara J. Flagg, "I Was Blind, But Now I See": White Race Consciousness and the Requirement of Discriminatory Intent, 91 Mich. L. Rev. 953, 981 (1993) (explaining that most whites "view Klan and other overtly white supremacist attitudes as extreme, perhaps pathological, deviations from the norm of white racial thinking, as if those attitudes can be comprehended in complete isolation from the culture in which they are embedded").
n217. See Gaertner & Dovidio, Aversive Racism, supra note 123, at 85 (arguing that "even when norms are clear, whites continue to be more sensitive to ostensibly nonracial factors that could permit them to rationalize a negative response toward blacks").
n218. See id. at 62.
n219. Gaertner and Dovidio explain that the ethnic or racial group with the greatest social power "is motivated to ensure its advantages by initiating" (or by extension defending) policies that favor their group. Id. at 64 (discussing theory of internal colonization).
n220. With this understanding about the cognitive process that underlies contemporary prejudiced behavior, one can see how complaints about race performance are likely to serve the same role that more explicit discriminatory statements served in the past. In cases regarding race/ethnicity performance, no defendant ever says, "We don't want X to engage in this behavior because it is race or ethnically marked." Rather, the defendant's justifications are that the behavior is disruptive, causes safety concerns, degrades morale, sends the message of bad hygiene, or other "neutral" concerns. See, e.g., Matsuda, supra note 91, at 1350 (discussing how arguments about accent discrimination often involve racially coded claims about intelligibility which, properly understood, are a cover for aversive racism); see also Robert H. Kelley, The Washington Civil Rights Initiative: The Need for a Meaningful Dialogue, 34 Gonz. L. Rev. 81, 97 (1998-99) (recognizing that traditional Title VII law does not address problems posed by aversive racism and quoting John Dovidio's hypothesis that aversive racist is likely to re-evaluate criterion necessary for job when faced with hiring black or white applicant in order to favor white applicant).
n221. J.M. Balkin, Free Speech and Hostile Environments, 99 Colum. L. Rev. 2295, 2308 (1999).
n222. Id. at 2306-09.
n223. Balkin explains:
It would be a great mistake to understand hostile environment doctrine simply as a set of rules designed to preserve civility, to protect individual dignity, or to prevent offense. Sexual [or race-based] harassment is prohibited because it is a status-enforcing mechanism - it employs offense, insult, and indignity to maintain the inferior status of women [and minorities]. Prohibitions on the use of this mechanism are designed to dismantle social subordination and to achieve civil equality, both within the workplace and, through their effects on the structure of work, in society as a whole.
Id. at 2308.
n224. Id. at 2308-09.
n225. See supra notes 3-7 and accompanying text (discussing case in which employee challenged temporary employment agency's policy of not referring people who wore braided hairstyles to temporary jobs).
n226. See Gaertner & Dovidio, Aversive Racism, supra note 123, at 85 (arguing that "it is unlikely ... that aversive racism can be alleviated by such direct methods [as social and legal pressures]," in contrast to "old-fashioned racism"). Although disparate impact analysis only requires that one identify a practice that has a disproportionately negative impact on a protected group, when presented with race performance claims, courts have asked for evidence that the practice stems from an immutable characteristic. See Rogers v. Am. Airlines, Inc. 527 F. Supp. 229, 231-32 (S.D.N.Y. 1981).
n227. See also Susan Bisom-Rapp, Discerning Form from Substance: Understanding Employer Litigation Prevention Strategies, 3 Employee Rts. & Emp. Pol'y J. 1, 18-19 (1999) (explaining that employers are counseled by defense law firms to avoid any reference to race and sex when conducting performance evaluations and to frame their comments in neutral terms to protect themselves against litigation).
n228. See generally Barbara J. Flagg, Fashioning A Title VII Remedy for Transparently White Subjective Decisionmaking, 104 Yale L.J. 2009 (1995).
n229. Chamallas, supra note 17, at 2408.
n230. See id. (arguing that jobs which force one to "suppress one's cultural identity to suit the image of the business [are] insulting and demeaning").
n231. Indeed, American Airlines never was required to demonstrate why braids were unprofessional in the cases I reviewed on this issue. See Cooper v. Am. Airlines, Inc., No. 97-1901, 1998 U.S. App. LEXIS 10426 (4th Cir. May 26, 1998) (per curiam); Rogers v. Am. Airlines, Inc., 527 F. Supp. 229 (S.D.N.Y. 1981).
n232. See Amy L. Wax, Discrimination As Accident, 74 Ind. L.J. 1129 (arguing that practical difficulties of controlling unconscious or subconscious discrimination militate against creating protections). But see Flagg, supra note 216, at 957-58 (arguing for importance of capturing phenomenon currently characterized as unconscious racism). See also id. at 991-1005 (describing how heightened scrutiny of criteria for employment decisions could be used to address "unconscious" discrimination).
n233. See Dovidio, supra note 185, at 836-37 (discussing results of study that show that "unconscious" racism has powerful consequences but can be quite subtle).
n234. Id. at 837.
n236. Devon W. Carbado & Mitu Gulati, The Law and Economics of Critical Race Theory, 112 Yale L.J. 1757, 1795-1803 (2003).
n237. Dovidio, supra note 185, at 838.
n238. Id. at 840.
n239. See id. at 838 ("Implicit attitudes and stereotypes ... are evaluations and beliefs that are automatically activated by the mere presence ... of the attitude object. They commonly function in an unconscious fashion.").
n240. See generally Jolynn Childers, Note, Is There a Place for a Reasonable Woman in the Law? A Discussion of Recent Developments in Hostile Environment Sexual Harrassment Law, 42 Duke L.J. 854 (1993) (exploring how hostile environment law has created debates about social interactions and perceptions of social interaction in ways that force individuals to think more critically about previously unexamined comments and behavior).
n241. Flagg, supra note 216, at 991-1005.
n242. See Jody Armour, Stereotypes and Prejudice: Helping Legal Decisionmakers Break the Prejudice Habit, 83 Cal. L. Rev. 733 (1995) (outlining model under which discriminatory responses based on automatic processes can be inhibited and replaced by responses based on controlled processes). Armour also discusses several other studies targeting unconscious racism, including Margo J. Monteith et al., Self-Directed Versus Other-Directed Affect as a Consequence of Prejudice-Related Discrepancies, 64 J. Personality & Soc. Psychol. 198, 200-08 (1993); Margo J. Monteith, Self-Regulation of Prejudiced Responses: Implications for Progress in Prejudice-Reduction Efforts, 65 J. Personality & Soc. Psychol. 469, 471-78 (1993).
n243. Paul Brest notes that outgroup members may fall prey to the "phenomenon of racially selective sympathy and indifference" towards minority persons, which is "the unconscious failure to extend to a minority the same recognition of humanity, and hence the same sympathy and care, given as a matter of course to one's own group." Paul Brest, The Supreme Court 1975 Term - Foreword: In Defense of the Antidiscrimination Principle, 90 Harv. L. Rev. 1, 7-8 (1976). Brest explains:
Although racially selective sympathy and indifference ... is an inevitable consequence of attributing intrinsic value to membership in a racial group, it may also result from a desire to enhance our own power and esteem by enhancing the power and esteem of members of groups to which we belong. And it may also result - often unconsciously - from our tendency to sympathize most readily with those who seem most like ourselves.
Id. at 8. Brest's insights make clear that workplace rules and practices that appear to penalize minorities for voluntary exhibition of racial difference must be viewed as part of a project to maintain racial hierarchies and that they implicate antidiscrimination concerns.
Barbara Flagg also builds on this work, describing how these tendencies affect interactions between white employers and minority personnel in the enforcement of appearance codes. She describes a problem called the "transparency phenomenon: the tendency of whites not to think about whiteness, or about norms, behaviors, experiences, or perspectives that are white-specific." Flagg, supra note 216, at 957. Flagg explains that "transparency operates to require black [or minority] assimilation even when pluralism is the articulated goal; it affords substantial advantages to whites over blacks even when decisionmakers intend to effect substantive racial justice." Id. Whether the behavior is conscious or unconscious is irrelevant for the purposes of our inquiry.
n244. This analysis adopts Robert Post's understanding of the logic of antidiscrimination law, specifically that it is not intended to eradicate racial categories entirely. Rather, Title VII, like other antidiscrimination statutes, only attempts to partially disrupt racial and ethnic ascription. These laws are not designed to make us entirely colorblind, but instead to disrupt the stigmatic associations made with certain races and ethnic groups. Post calls this approach to studying antidiscrimination statutes the "sociological" study of antidiscrimination law. See Post, supra note 16, at 31.
n245. 42 U.S.C. 2000e-2(a) (2000).
n247. See 42 U.S.C. 2000e (2000).
n248. See Yoshino, supra note 27, at 493-500 (discussing connection between immutability requirement and visibility of trait in equal protection jurisprudence).
n249. Garcia v. Gloor, 618 F.2d 264, 269-70 (5th Cir. 1980); Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 232 (S.D.N.Y. 1981); see infra Part III.A.
n250. See Yoshino, supra note 27, at 497-98 (arguing that courts requiring visibility of traits for Fourteenth Amendment protection are referring to physical "corporeal visibility," and not "social visibility").
n251. See infra Part III.A (discussing Rogers); infra Part III.B (discussing Gloor).
n252. See generally Gloor, 618 F.2d 264; Rogers, 527 F. Supp. 229.
n253. Vicki Schultz, Women "Before" the Law: Judicial Stories About Women, Work, and Sex Segregation on the Job, in Feminists Theorize the Political 297 (Judith Butler & Joan W. Scott eds., 1992).
n254. See Vicki Schultz, The Sanitized Workplace, 112 Yale L.J. 2061, 2076-77 (2003); Vicki Schultz, Reconceptualizing Sexual Harassment, 107 Yale L.J. 1683, 1686-87 (1998).
n255. Schultz, supra note 253, at 298-99, 304-05.
n256. Id. at 311-14.
n257. See generally Karen Engle, The Persistence of Neutrality: The Failure of the Religious Accommodation Provision to Redeem Title VII, 76 Tex. L. Rev. 317 (1997) (comparing status/conduct divide in cases of race, sex, national origin, and religious discrimination). The sole exception is disability discrimination, which is defined in a manner that protects both status and conduct associated with being disabled. See 42 U.S.C. 12101 (2000).
n258. See Garcia v. Gloor, 618 F.2d 264, 269-70 (5th Cir. 1980).
n259. See id. at 270-71.
n260. See, e.g., Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 232 (S.D.N.Y. 1981).
n261. See Gaulding, supra note 90, at 685-87 (explaining that "most Black English speakers cannot simply choose to speak Standard English").
n262. See id. at 692.
n263. See Halley, supra note 27, at 509-10 (critiquing immutability); Yoshino, supra note 27, 494-96 (arguing that "courts do not characterize all traits that are hard to change ... as immutable," only those traits "perceived to be defined by nature rather than by culture").
n264. See Yoshino, supra note 27, at 495-96.
n265. While I am sensitive to the fact that there may be problems in drawing analogies between the discrimination suffered by racial and ethnic groups, I believe that the animus involved in these groups of cases is functionally identical. Indeed, many of the claims framed as "race discrimination" claims concern a plaintiff from an ethnic group that has been so marked by racial constructs and their accompanying stigmas that the reference to ethnic identity is treated as irrelevant and improper. See, e.g., Rawlins-Roa v. United Way of Wyandotte County, Inc., 977 F. Supp. 1101, 1106-07 (D. Kan. 1997) (holding that Dominican national was required to show evidence of race discrimination to prevail on Title VII claim); Cuello Suarez v. Puerto Rico Elec. Power Auth., 798 F. Supp. 876, 891 (D.P.R. 1992) (holding same in case raising section 1981 claim). National origin claims, in contrast, are brought by plaintiffs who perceive that their ethnic group has retained its distinct identity, independent of racial constructs and that the discrimination suffered is triggered by that specific identity. In my view, the main distinction between national origin cases and race cases is that the plaintiff in the national origin case can offer evidence which shows a tighter fit between a stereotype and her ethnic identity. These distinctions between race and national origin discrimination prove irrelevant to our analysis, as it generally concerns negative animus triggered by voluntary behavior associated with disfavored groups. See also Perea, supra note 12, at 857 (analogizing between race and national origin discrimination and noting that they are "neither relevant nor detrimental in the performance of a job").
n266. 527 F. Supp. 229, 231-33 (S.D.N.Y. 1981).
n268. Id. at 231.
n269. Id. at 231-32. Rogers is not the only plaintiff who attempted to raise this issue; indeed, although she lost her claim, id. at 234, black women plaintiffs have continued to litigate this issue as late as the 1990s. See, e.g., Cooper v. Am. Airlines, Inc., No. 97-1901, 1998 U.S. App. LEXIS 10426 (4th Cir. May 26, 1998) (per curiam). Cooper challenged American Airlines' pre-1993 grooming code forbidding employees from wearing all-braided hairstyles. American Airlines revised its policy in 1993 to allow all "'braided hairstyles without beads or trim' so long as any loose braids were 'secured to the head or at the nape of the neck.'" Id. The court concluded that Cooper's claim was moot. Id. at 3.
n270. Rogers, 527 F. Supp. at 232.
n271. Id. at 232-33.
n272. Id. at 232; see also Kang, supra note 18, at 312-20 (analyzing Rogers decision). Kang explores the irony of the Rogers court's statement, noting that the insistence that black women's hairstyles be natural in order to be afforded protection ignores the fact that most white women's hairstyles are not natural. Id. at 315. This results in black women having fewer aesthetic choices than whites. Id. Of additional note, courts assume a variety of white hairstyles are natural because the whiteness is unmarked or invisible. Against this backdrop, any hairstyle a black woman adorns that is inconsistent with these styles is deemed deviant. Id. at 312-13.
Indeed, the only circumstance in which courts have recognized the validity of a race-associated practice was in this "Afro" scenario. See Jenkins v. Blue Cross Mut. Hosp. Ins., Inc., 538 F.2d 164, 168 (7th Cir. 1976) (en banc). In Jenkins, an African American employee claimed that she had an uneventful relationship with her supervisor until she appeared at work one day wearing an Afro. In her complaint to the EEOC, she stated, "I have worked for Blue Cross and Blue Shield approx. 3 years during which time I [had] no problem until May 1970 when I got my natural hair style." Id. at 167. Importantly, Jenkins' race-associated morphology had never offended her employer prior to the change in hairstyle. In order for the court to treat Jenkins' EEOC complaint concerning Afro-discrimination as race discrimination, it must have concluded that Afros are a genetic or immutable component of blackness. However, the facts of Jenkins refute this proposition. Jenkins indicated that she had suddenly changed her aesthetic and donned an Afro, clearly indicating that the hairstyle was not a function of biology. Id. In this case, her employer regarded her new aesthetic as a threat to the cultural hegemony of the workplace. The Jenkins case stands as a reminder that the aesthetic choices of minority employees can trigger race discrimination even when the discriminators profess to be resolutely indifferent to race-associated morphology. See supra notes 206-217 and accompanying text (discussing aversive racism). These cases decisively establish that, although there are employers who may have shed their antipathy towards the plain morphological characteristics associated with races, many remain deeply offended by practices associated with low-status racial "others."
n273. Rogers, 527 F. Supp. at 232 (quoting Garcia v. Gloor, 618 F.2d 264, 269 (5th Cir. 1980)).
n274. Id. at 232.
n275. Id. at 233. The court summarily dismissed Rogers's claim as a mere matter of individual choice and personal expression. See id. at 231. ("This type of regulation has at most a negligible effect on employment opportunity. It does not regulate on the basis of any immutable characteristic of the employees involved. It concerns a matter of relatively low importance in terms of the constitutional interests protected by the Fourteenth Amendment ... .").
n276. Id. at 232. See Kang, supra note 18, at 316-17 (describing Judge Sofaer's analysis in Rogers as falling prey to this problem). Kang argues that Judge Sofaer reports on Bo Derek's appropriation of the cornrow style to explain that the style was simply hot and faddish, offering this singular example as proof that the hairstyle was not communicative of black aesthetic pride or any racially associated political message. See id.
n277. Rogers, 527 F. Supp. at 233.
n279. McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853 (S.D. Miss. 1992).
n280. Id. at 854-55. Though she alleged religious discrimination, most of the evidence that McGlothin submitted indicated that she represented her desire to wear headwraps or a "natural" hairstyle as part of her performance of African American identity. Id. at 857-58. Indeed, at the hearing convened after she was terminated, she repeatedly represented the headwraps and dreaded hairstyles as part of her African American culture, or her Hebrew-Israelite culture. Id. at 863.
n281. Id. at 860 (quoting testimony of Dr. Joseph Pete, Assistant Superintendent of the District) (internal quotation marks omitted).
n282. Id. at 865-66.
n283. Id. at 865 (concluding that "the explanations which the District's witnesses credibly maintain she provided ... were reflective of the African culture and Ms. McGlothin's African heritage").
n284. Id. at 856.
n285. Id. at 858.
n286. Id. at 855.
n287. 813 F.2d 1406 (9th Cir. 1987).
n288. Id. at 1408-09.
n289. Id. at 1408.
n290. Id. at 1408, 1410.
n293. Jurado v. Eleven-Fifty Corp., 630 F. Supp. 569 (C.D. Cal. 1985).
n294. Jurado, 813 F.2d at 1411, 1412.
n296. Jurado presented evidence on this point, as the consultant that recommended that the station switch to an English-only format indicated that the radio station was "preoccupied with ethniticity [sic] to a frightening degree," that Jurado's show was "too ethnic," and that the station did not "need the Mexicans or the blacks to win in L.A." Id. at 1410 (internal quotation marks omitted) (alternation in original).
n297. Id. at 1409-10 (internal quotation marks omitted).
n298. The case does, however, raise questions about when it is permissible for an employer to ask an employee to use some aspect of her racial or ethnic identity to market his product or business. This issue will be explored in a subsequent article.
n299. See Kimberle Crenshaw, Whose Story Is It, Anyway? Feminist and Antiracist Appropriations of Anita Hill, in Race-ing Justice, En-gendering Power: Essays on Anita Hill, Clarence Thomas, and the Construction of Social Reality 402, 427-31 (Toni Morrison ed., 1992) (discussing claim that Clarence Thomas's behavior towards Anita Hill was example of "down home courting"). Although no one yet has explored possible cultural defenses to sexual harassment and other Title VII claims, there is a developing scholarship on cultural defenses to criminal behavior. See Leti Volpp, Blaming Culture for Bad Behavior, 12 Yale J.L. & Human. 89, 110-16 (2000) (questioning whether cultural defenses are defensible interpretation of pluralist values); James J. Sing, Note, Culture as Sameness: Toward a Synthetic View of Provocation and Culture in the Criminal Law, 108 Yale L.J. 1845, 1867-69 (1999) (discussing provocation defense as form of cultural defense).
n300. McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 855 (S.D. Miss. 1992).
n301. Rogers v. Am. Airlines, Inc., 527 F. Supp. 229, 231-32 (S.D.N.Y. 1981).
n302. See, e.g., Turpen v. Missouri-Kansas-Texas R.R. Co., 736 F.2d 1022, 1026 (5th Cir. 1984) (describing plaintiff's and employer's burdens in cases alleging religious discrimination); Brener v. Diagnostic Ctr. Hosp., 671 F.2d 141, 144 (5th Cir. 1982) (same).
n303. 618 F.2d 264 (5th Cir. 1980).
n304. Specifically, for the disparate impact claim, Garcia argued that the English-only rule denied native Spanish speakers the opportunity to speak to each other in the language they felt most comfortable speaking - a privilege already granted to primarily English-speaking employees. Id. at 268.
n305. Id. ("Neither the statute nor common understanding equates national origin with the language that one chooses to speak.").
n306. Id. at 268 n.2 (noting that "legislative history concerning the meaning of 'national origin' is 'quite meager'" (quoting Espinoza v. Farah Mfg. Co., 414 U.S. 86, 88 (1973)).
n307. 507 F.2d 1084 (5th Cir. 1975) (en banc).
n308. Id. at 1091-92.
n309. Gloor, 618 F.2d at 269-70.
n310. "Sex-plus"-based decisions are those that differentiate among employees of the same gender on the basis of an additional characteristic. In these cases, employers differentiate among subclasses of men or women. The concept first gained judicial recognition in Phillips v. Martin Marietta Corp., 400 U.S. 542, 544 (1971) (per curiam).
n311. See Int'l Union v. Johnson Controls, Inc. 499 U.S. 187, 191-92 (1991) (challenging rule prohibiting women of childbearing age to work for company in lead processing); Phillips, 400 U.S. at 543 (challenging policy of refusing to hire women with preschool-age children).
n312. See, e.g., Barker v. Taft Broadcasting Co., 549 F.2d 400, 401 (6th Cir. 1977) (holding that grooming code, which limited manner in which men's hair could be cut, did not violate Title VII's prohibition on sex discrimination in employment); Longo v. Carlisle DeCoppet & Co., 537 F.2d 685, 685 (2d Cir. 1976) (same).
n313. Willingham, 507 F.2d at 1087-88.
n314. Id. at 1090.
n315. Id. at 1091-92.
n316. Id. at 1091.
n317. See id. at 1087 (noting that employer "believed that the entire business community it served - and depended upon for business success - associated long hair on men with the counter-culture types who [had] gained extensive unfavorable national and local exposure").
n318. Id. at 1091 (noting that hiring policy based on hair length is within employer's discretion as to "how to run his business").
n319. Willingham v. Macon Tel. Publ'g Co., 352 F. Supp. 1018, 1020 (M.D. Ga. 1972).
n321. These concerns, however, do not abate when we consider a circumstance in which an employer wants to institute a rule that seeks to preserve ethnic difference. In these circumstances, most Americans still would be uncomfortable with employers manipulating race and national origin identities or creating special codes for these groups that distinguish them from other races or ethnicities. Indeed, it shocks the conscience to think that an employer could require an employee to "blacken" up and take on more African American affectations in order to sell her product, and we would not be comforted by the fact that she put the same pressure on white employees to display white ethnic traits in service of her goals. When viewed from this vantage point, it seems a terrible idea to give employers discretion to control the development of race and ethnic identities.
n322. The Gloor court explains:
As [we] said in Willingham, "Equal employment opportunity may be secured only when employers are barred from discriminating against employees on the basis of immutable characteristics, such as race and national origin... . But a hiring policy that distinguishes on some other ground, such as grooming codes or length of hair, is related more closely to the employer's choice of how to run his business than to equality of employment opportunity."
Garcia v. Gloor, 618 F.2d 264, 269 (5th Cir. 1980) (quoting Willingham, 507 F.2d at 1091) (alteration in original).
n323. Id. at 269 (emphasis added) (citation omitted). Initially, the court appears to be making a distinction between immutable characteristics (those things beyond one's power to change), and the "prohibited bases" explicitly listed in Title VII as a basis for protection. However, in the next sentence, it reveals that it believes the two are exactly the same, listing the "prohibited bases" or identity variables named in the statute as examples of things that are beyond an individual's power to change. Id.
n324. 411 U.S. 677, 686 (1973) (plurality opinion).
n325. See Halley, supra note 27, at 507-10 (noting theoretical and practical problems with immutability construct, including its poor "fit with the political realities of gay, lesbian, bisexual and queer life"); Yoshino, supra note 27, at 487, 490-91 (citing critiques of immutability doctrine and arguing that it encourages groups to assimilate by hiding or changing immutable characteristics).
n326. Gloor, 618 F.2d at 268-69.
n327. Id. at 270.
n329. Frontiero, 411 U.S. at 686.
n330. Gloor, 618 F.2d at 270 ("In some circumstances, the ability to speak or the speaking of a language other than English might be equated with national origin ... .").
n331. Id. at 271.
n332. Id. (quoting Mr. Garcia).
n333. Willingham v. Macon Tel. Publ'g Co., 507 F.2d 1084, 1091 (noting that employee may "choose to subordinate" appearance "preference").
n334. See id. at 1089. Summarizing its view of the Willingham claim, the Fifth Circuit explained, "Nothing that we say should be construed as disparagement of what many feel to be a highly laudable goal - maximizing individual freedom by eliminating sexual stereotypes. We hold simply that such an objective may not be read into the Civil Rights Act of 1964 without further Congressional action." Id. at 1092 (emphasis added).
n335. Gloor, 618 F.2d at 267, 270.
n336. Though it refused to offer plaintiffs protection from discrimination on this basis when they easily could conform with reasonably crafted English-only rules, the Gloor court did recognize that language was an important part of ethnic identity. The court explained, "We do not denigrate the importance of a person's language of preference or other aspects of his national, ethnic or racial self-identification." Id. at 270. Moreover, the court seemed to recognize that its ruling missed a critical point - that language and other national origin-associated features may serve as a proxy for discriminating against status, warning that "differences in language and other cultural attributes may not be used as a fulcrum for discrimination." Id. It likely hoped that its twist on the immutability paradigm would prevent this proxy type of discrimination from occurring. However, having concluded that the English-only rule that Garcia's employer had instituted was in fact motivated by legitimate business reasons, the court determined that this was not a problem in the instant case.
n337. 998 F.2d 1480 (9th Cir. 1993).
n338. Id. at 1487 (citations omitted). The plaintiff in Gloor also raised freedom of expression and autonomy arguments; however, they played a far smaller role in his case than in gender performance cases. Gloor, 618 F.2d at 270.
n339. Spun Steak, 998 F.2d at 1487.
n340. Again, recognizing the concern that workers who speak only Spanish may be penalized unfairly under its analysis, the Spun Steak court offered the same caveats as the Gloor court, recognizing that, for monolingual speakers, speaking Spanish is not a preference. Consistent with Gloor, it argued that language is immutable for those who are able to speak only one language. Id. at 1488.
Ironically, prior to Spun Steak, a series of Ninth Circuit national origin "language" discrimination cases had cited approvingly to the EEOC's regulations on national origin discrimination, which indicated that language was part of national origin identity and rules regulating its use were presumptively violative of Title VII. See, e.g., Gutierrez v. Mun. Court, 838 F.2d 1031, 1039-40 (9th Cir. 1988), reh'g en banc denied, 861 F.2d 1187 (1988), vacated as moot by 490 U.S. 1016 (1989) (rejecting employer's English-only rule as national origin discrimination based on EEOC guideline, which recognized that any language prohibitions should be presumed to have disparate impact in violation of Title VII unless justifiable by legitimate business necessity); cf. Jurado v. Eleven-Fifty Corp., 813 F.2d 1406, 1411 (9th Cir. 1987) (recognizing validity of EEOC rule but concluding that no discrimination occurred because employee was fluently bilingual and English-only rule was justified by business necessity). In Spun Steak, however, the court reversed course and definitively rejected the EEOC regulations, ruling that language regulations do not presumptively offend Title VII and that language is not an immutable part of national origin identity. Spun Steak, 998 F.2d at 1489-90 (rejecting EEOC rule creating presumption in favor of employee and refusing to adopt per se rule that English-only policies negatively impact workplace).
n341. Bartlett, supra note 106, at 2558. Some equality arguments are made as well; however, because they offer even fewer parallels to the race and national origin discrimination context, they are not explored here.
n342. Spun Steak, 998 F.2d at 1487 (quoting Garcia v. Gloor, 618 F.2d 264, 270 (5th Cir. 1980)).
n343. To the extent that the gender cases are about symbolic exchanges, they do share parallels with race/ethnicity performance cases, in that the behavior is taken up in order to communicate social belonging to a particular group. These behaviors also often have social and political meaning. See supra Part II.A.
n344. 527 F. Supp. 229 (S.D.N.Y. 1981).
n345. Id. at 232; see supra notes 266-278 for additional discussion of Rogers.
n346. Rogers, 527 F. Supp. at 233-34.
n347. Id.; see supra note 272 & accompanying text.
n348. Rogers, 527 F. Supp. at 232.
n349. Jenkins v. Blue Cross Mut. Hosp. Ins., Inc., 538 F.2d 164, 168 (7th Cir. 1976) (en banc) (ruling that plaintiff stated claim under Title VII for race discrimination when her EEOC charge concerned discriminatory treatment after she began wearing Afro). For a more detailed discussion of Jenkins, see supra note 272.
n350. McPherson v. Shoney's Colonial, Inc., No. 95-0069- C, 1996 U.S. Dist. LEXIS 17627, at 9-12 (W.D. Va. Nov. 21, 1996) (denying Title VII constructive discharge claim concerning plaintiff's desire to wear all-braided hairstyle); see also Carswell v. Peachford Hosp., No. C80-222A, 1981 U.S. Dist. LEXIS 14562, at 5 (N.D. Ga. May 26, 1981) (finding that "plaintiff's discharge was not based on an immutable characteristic," but on "her refusal to remove beads from her [braided] hair").
n351. Eatman v. United Parcel Serv., 194 F. Supp. 2d 256, 262 (S.D.N.Y. 2002) (rejecting discrimination claims concerning employer rule requiring covering of dreadlocks); Hines v. Hillside Children's Ctr., 73 F. Supp. 2d 308, 317 (W.D.N.Y. 1999) (rejecting discrimination claims concerning employer's claim that dreadlocks were unprofessional); Miller v. CCC Info. Sys., Inc., No. 95 C 6612, 1996 WL 480370, at 3 (N.D. Ill. Aug. 22, 1996) (rejecting discrimination claims concerning allegations based on discriminatory statements about dreadlocks).
n352. McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853 (S.D. Miss. 1992) (rejecting disparate treatment claim concerning employer's reaction to employee's desire to wear "natural" hairstyles and headwraps).
n353. Frontiero v. Richardson, 411 U.S. 677 (1973).
n354. Id. at 682-88.
n355. Id. at 678-79.
n356. Id. at 679-80.
n357. Yoshino discusses this issue as it relates to a Fourteenth Amendment analysis. The three criteria used to determine whether a group is a protected class under the Fourteenth Amendment are: 1) the group's history of disempowerment; 2) their current political powerlessness; and 3) the possession of an immutable physical characteristic that defines them as a group. See Yoshino, supra note 27, at 496-98.
n358. Frontiero, 411 U.S. at 686. Several scholars have noted that the discussion of immutability in Frontiero should not be understood as a claim that a group need only demonstrate that it is organized around an immutable characteristic in order to be recognized as a protected class under the Fourteenth Amendment. See Halley, supra note 27, at 507-08; Yoshino, supra note 27, at 504.
n359. Frontiero, 411 U.S. at 686.
n360. Id. at 688.
n361. Craig v. Boren, 429 U.S. 190, 197 (1976) (articulating standard of review for sex-based classifications later denominated "intermediate scrutiny").
n362. City of Cleburne v. Cleburne Living Ctr., Inc., 473 U.S. 432, 443-45 (1985) (noting that mentally retarded cannot be considered politically powerless, and thus in need of judicial protection, because legislators have been responsive to their needs).
n363. Yoshino, supra note 27, at 496.
n364. Id. (explaining that immutability and visibility are required in order to be conferred suspect class status). Specifically, Yoshino points out that courts tend to conflate immutability with corporeal visibility, presuming that most corporeal, visible traits are an irrational basis for forming conclusions about a person's capabilities. Id. at 495-96.
n365. Halley, supra note 27, at 507-510 (explaining that Court uses immutability-plus test, with Court making subjective determination into whether immutable characteristics are relevant bases for discrimination). Indeed, the Supreme Court has refused to recognize the immutable characteristic of mental retardation as the basis for recognizing the developmentally disabled as a protected class deserving of heightened scrutiny under equal protection law. Id. at 510 (discussing Cleburne). The Court recognized that the developmentally disabled are burdened based on a characteristic that is merely an accident of birth but argued that this characteristic is sufficiently related to their abilities so as to provide a valid basis for treating them differently than other individuals. Cleburne, 473 U.S. at 442-48; see Halley, supra note 27, at 510. Janet Halley therefore concludes that the Court extends groups protected class status based on an immutability-plus test. The plus factor works as an inquiry to determine whether the immutable trait actually is related to the opportunity to which the group is being prohibited access. For example, intelligence and physical disability are considered to be valid bases for discrimination when these immutable differences are related directly to a person's ability to perform certain kinds of work. Id. at 507-08.
n366. See supra notes 364-365. Both Halley and Yoshino have indicated that the Supreme Court seems to recognize the immutability criterion's conceptual limitations, as it has been applied inconsistently in the Fourteenth Amendment context. Halley, supra note 27, at 507-08 (noting Frontiero court's admission that "many immutable characteristics ... form the basis of discriminatory decisions that are widely regarded as unproblematic"); Yoshino, supra note 27, at 495 ("If a trait is perceived to be defined by nature rather than by culture, then the courts will be more likely to call it immutable."). This suggests that the Supreme Court is aware that immutability, standing alone, is not determinative of what affords a group the right to seek relief under equal protection.
n367. Indeed, this proposition is counterintuitive given the range of protections racial groups are afforded in other equal protection-based circumstances. For example, courts make efforts to protect ethnic communities' wishes in framing public debate through redistricting. See Shaw v. Hunt, 517 U.S. 899, 918 (1996) (explaining that districts drawn on basis of race to increase voting power of minority group must demonstrate compelling state interest for justification).
n368. Indeed, this perhaps explains why fundamental rights and immutable characteristics are the basis for protection. The immutability requirement has a tight fit with biological sex-based characteristics - the social identity the court wants to preserve. However, there is no biological underpinning for the race cases, and consequently, the immutability requirement ensures only that discrimination based on physicality is punished.
n369. Title VII does not define race or identify where the status/conduct distinction should be drawn with regard to voluntary race-associated behavior. See supra note 9. Other sections of Title VII and similar workplace antidiscrimination statutes do outline where the status/conduct distinction is to be drawn. Specifically, Title VII defines religious discrimination and discrimination "on the basis of sex" by defining these concepts to cover actual status as well as some practices associated with each particular status. See 42 U.S.C. 2000e(j) (2000) ("The term 'religion' includes all aspects of religious observance and practice ... ."); 42 U.S.C. 2000e(k) (2000) ("The terms 'because of sex' or 'on the basis of sex' include, but are not limited to, because of or on the basis of pregnancy, childbirth, or related medical conditions ... .").
n370. The prevailing understanding of the federal courts' role is that when federal courts interpret statutes, their primary goal is to give effect to congressional intent.
n371. Review of the 1964 House and Senate reports shows that three issues dominated congressional debates on the statute immediately prior to its passage: (1) concerns about the "eleventh hour" passage of the bill; (2) the related concern that earlier, more conservative versions of the bill had sub silentio been discarded in favor of the version set before the full House for a vote; and (3) concerns about the bill's potential to violate principles of federalism. See generally Charles Whalen & Barbara Whalen, The Longest Debate: A Legislative History of the 1964 Civil Rights Act 102, 106, 121 (1985).
n372. Winant, supra note 72, at 24.
n373. The charge of judicial activism could be leveled easily at the Fifth and Ninth Circuit courts that, independent of textual support, summarily concluded that the statutory definition of race for Title VII referred solely to morphological, biological features, and severed race into its involuntary and voluntary elements. See supra Part III.B (discussing Garcia v. Gloor, 618 F.2d 264 (5th Cir. 1980)) and notes 287-298 and accompanying text (discussing Jurado v. Eleven-Fifty Corp., 630 F. Supp. 569 (C.D. Cal. 1985)).
n374. Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of Status-Enforcing State Action, 49 Stan. L. Rev. 1111, 1140-46 (1997) (explaining that antidiscrimination laws of all kinds must contend with forces that work to limit their disruptive effects, even as they transform society).
n375. Id. at 1114.
n376. Omi & Winant, supra note 31, at 62-68.
n378. See supra notes 348-350.
n379. See supra notes 44-66 and accompanying text.
n380. See Bumiller, supra note 126, at 46-48 (discussing rise of black activist groups during civil rights movement).
n381. Frontiero v. Richardson, 411 U.S. 677, 686 (1973) (quoting Weber v. Aetna Casualty & Surety Co., 406 U.S. 164, 175 (1972)).
n382. Id. at 686-87.
n383. Id. at 686.
n384. See Peter D. Salins, Assimilation, American Style: Universalist Ideals, Capitalism, a Plethora of Associations, and a Love of Progress are the Secret to Interethnic Identity, 28 Reason 20, 20 (1997) (describing "up or out" model of assimilation in which immigrants must conform to majority culture or suffer marginalization).
n385. John Rhee, Theories of Citizenship and Their Role in the Bilingual Education Debate, 33 Colum. J.L. Soc. Probs. 33, 37 (1999-2000).
n386. See Stephen Steinberg, The Ethnic Myth: Race, Ethnicity, and Class in America 3-4 (1981); see also Salins, supra note 384, at 20-22 (discussing criticisms of melting pot model of assimilation). The melting pot theory also has foundered because some ethnic groups proved "unmeltable." See generally Nathan Glazer & Daniel P. Moynihan, Beyond the Melting Pot: The Negroes, Puerto Ricans, Jews, Italians, and Irish of New York City (2d ed. 1970); Michael Novak, The Rise of the Unmeltable Ethnics: Politics and Culture in the Seventies (1972).
n387. See generally Steven A. Ramirez, Diversity and the Boardroom, 6 Stan. J. L. Bus. & Fin. 85 (2000).
n388. See Rhee, supra note 385, at 37-46.
n389. See id.
n390. See supra Part III.A-B.
n391. See Jurado v. Eleven-Fifty Corp., 813 F.2d 1406, 1408 (9th Cir. 1987) (noting that employer radio station requested that employee disc jockey speak Spanish on his radio program in order to attract Hispanic listeners, then told him to stop speaking Spanish due to concern about ratings).
n392. See, e.g., McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 856 (S.D. Miss. 1992) (noting that employee felt that her attire should be acceptable under school's new "multi-cultural educational directive").
n393. See, e.g., Webb v. R&B Holding Co., 992 F. Supp. 1382, 1388-89 (S.D. Fla. 1998) (alleging derogatory use of Spanish word "negra" and speaking of Spanish in workplace created hostile environment).
n394. See K. Anthony Appiah, Identity, Authenticity, Survival: Multicultural Societies and Social Reproduction, in Multiculturalism: Examining the Politics of Recognition 149, 162-63 (Amy Gutmann ed., 1994).
n395. See id.
n396. Karst, supra note 10, at 295.
n397. Post, supra note 16, at 20 (explaining that law is "seeking to alter the particular meanings of ... [gender-and race-based] conventions as they are displayed in specific contexts").
n398. Id. at 32 ("We can expect judicial opinions to reach conclusions accepting social practices in implicit and indirect ways.").
n399. See supra Part I.A; see also Gross, supra note 41, at 113 (noting that race determination cases implied that "'negro blood' ... made a person act in certain ways").
n400. See supra Part I.A.
n401. Volpp, supra note 299, at 89-91.
n402. Id. at 94-96 (explaining that American cultural common sense assumes third-world culture is static and monolithic and emphasizes events which represent these cultures as primitive and regressive).
n403. Id. at 97-98.
n404. Id. at 96.
n405. Id. at 90.
n406. Id. at 95.
n407. See id. ("Many of the behaviors attributed to people of color are similarly ascribed to working-class whites, e.g., early marriage and excessive promiscuity.").
n409. But see Crenshaw, supra note 299, at 421-29 (criticizing Patterson's argument).
n410. Id. at 421-22.
n412. Volpp, supra note 299 (discussing cultural defense in context of early marriage and forced marriage cases involving minorities and lack of such defense in cases involving whites).
n413. For the process courts should follow in recognizing an employer's defense in this circumstance, see infra Part IV.B.
n414. As explained above, Title VII's gender and religious discrimination protections can shield employers from liability in some circumstances when they need to police race/ethnicity performance that infringes on the equality rights of other workers.
n415. See J<um u>rgen Habermas, Struggles for Recognition in the Democratic Constitutional State, in Multiculturalism: Examining the Politics of Recognition, supra note 394, at 107, 130-31.
n416. See id. at 132 ("Cultures survive only if they draw the strength to transform themselves from criticism and secession.").
n417. See id.
n418. Appiah, supra note 394, at 159-63. This is distinguished from a classic assimilation model which demands that cultural differences must be surrendered.
n419. See id. at 163; David B. Wilkins, Introduction: The Context of Race to K. Anthony Appiah & Amy Gutmann, Color Conscious: The Political Morality of Race 3, 7 (1996). Relatedly, Wendy Brown also questions whether we should want to continue with these scripts based on racial (or ethnic) identity if they previously have been the basis for oppression because, whatever positive political movements have formed based on these identities, they still were originally crafted based on an experience or condition of subordination. Wendy Brown, Wounded Attachments: Late Modern Oppositional Political Formations, in The Identity in Question 199, 220-22 (John Rajchman ed., 1995).
n420. Appiah, supra note 394, at 160-63 (discussing role of group identity in black struggle to be treated with respect and dignity).
n421. For example, certain hairstyles or clothing styles may be popular with an ethnic group in a particular period, yet they do not become the subject of litigation.
n422. Jenkins v. Blue Cross Mut. Hosp. Ins., 538 F.2d 164 (7th Cir. 1976) (en banc).
n423. See Martha Minow, Not Only for Myself: Identity, Politics, and Law, 75 Or. L. Rev. 647, 654 (1996) (explaining that articulation of trait as "essential" to particular group identity tends to lead to debate within group, rather than unification).
n424. Samuel A. Marcosson, The 'Special Rights' Canard in the Debate over Lesbian and Gay Civil Rights, 9 Notre Dame J.L. Ethics & Pub. Pol'y 137, 140 (1995). Marcosson's project is to show how these arguments are deployed in debates challenging gay antidiscrimination statutes. They are equally applicable in the context of race/ethnicity performance.
n425. Id. at 140-44.
n426. Id. at 140, 144-45.
n427. The need for more aggressive inquiry into the business reasons asserted by an employer is explored in detail in Part IV.B. In the Gloor case, for example, the employer justified his English-only rule by explaining that he had received objections from English-speaking customers about the Spanish, and that he needed to cultivate workers' English skills so that they could better communicate with customers and understand trade literature (written in English). Garcia v. Gloor, 618 F.2d 264, 267 (5th Cir. 1980). Also, he argued, the English-only rule allowed for better supervisory control by supervisors who spoke only English. Id.
n428. See generally Richard A. Epstein, Forbidden Grounds: The Case Against Employment Discrimination Laws 24-28 (1992) (explaining that freedom of contract is basic social norm). Economists who hold this view also argue that because discrimination is irrational, employers who continue to engage in such behavior and ignore talent for invalid reasons will find that they are disadvantaged by the practice. See id. at 41-42.
n429. An employer may dismiss an "at will" employee whenever the whim strikes her, subject to a few judicially recognized policy-based exceptions that do not implicate cultural expression concerns. See Hill & Wright, supra note 2, at 9-18.
n430. Reilly, supra note 15, at 262-63.
n431. Several scholars have commented on the limited protection from religious discrimination that Title VII offers, as the statutory test requires only that the employer show that she will suffer "undue hardship" if she adopts the proposed accommodation. Typically, courts only require a de minimis showing to establish that burden. See, e.g., Sidney A. Rosenzweig, Comment, Restoring Religious Freedom to the Workplace: Title VII, RFRA and Religious Accommodation, 144 U. Pa. L. Rev. 2513, 2518-20 (1996).
n432. This view is adopted in Willingham with regard to gender performance. Rejecting the employee's complaint that sex-specific grooming rules constituted an unfair burden to employment, the Fifth Circuit explained, "If the employee objects to the grooming code he has the right to reject it by looking elsewhere for employment, or alternatively he may choose to subordinate his preference by accepting the code along with the job." Willingham v. Macon Tel. Publ'g Co., 507 F.2d 1084, 1091 (1975).
n433. Ruth P. Rubinstein, Dress Codes: Meanings and Messages in American Culture 83-102 (2d ed. 2001) (discussing cultural meanings sent by executive attire). Ironically, the shift to casual dress in the workplace triggered even more anxiety in professional settings about maintaining hierarchies and projecting a professional look. Publishers are responding to the need to observe the now unwritten rules about projecting the appropriate authoritative image and "cope" with Casual Friday. See generally Susan Bixler, The New Professional Image: From Business Casual to the Ultimate Power Look (1997); Mark Weber & The Van Heusen Creative Design Group, Dressing Casually for Success ... for Men (1997).
n434. See, e.g., Willingham, 507 F.2d at 1091 (rejecting grooming code challenge as improper attempt to "interfere with the manner in which an employer exercises his judgment as to the way to operate a business").
n435. In a series of case studies, Kristin Bumiller documents that this is particularly true of minority workers who often are extremely wary of bringing suit because of the social repercussions that would flow from actively identifying and opposing oppressive practices in the workplace. See Bumiller, supra note 126, at 26-29, 52-54.
n436. See Post, supra note 16, at 13-14 (explaining that "functional rationality ... is ... broadly regarded by American antidiscrimination law as a justification for employer decisions"). This expectation may develop because employers, wary of running afoul of employment discrimination law, may prophylatically justify policies which appear to have racial effects by way of some rational, business-related explanation. See Bisom-Rapp, supra note 227, at 14-31 (explaining that employer discretion has not decreased under antidiscrimination law but that lawyers have found ways to exercise that discretion in manner that comports with and may even rely on antidiscrimination doctrine).
n437. See Post, supra note 16, at 13 (explaining that logic of antidiscrimination law pushes employers to find functional justifications for their decisions).
n438. Klare, supra note 94, at 1427 (recognizing that arbitrators regard employers' justifications based on "company image" as valid managerial interest).
n439. Balkin, supra note 221, at 2319.
n440. See Dovidio, supra note 185, at 830-31; supra notes 185-186 and accompanying text.
n441. See infra notes 453-456 and accompanying text.
n442. Brest, supra note 243, at 7. Brest explains:
If all race-dependent decisions were irrational, there would be no need for an antidiscrimination principle, for it would suffice to apply the widely held moral, constitutional, and practical principle that forbids treating persons irrationally. The antidiscrimination principle fills a special need because - as even a glance at history indicates - race-dependent decisions that are rational and purport to be based solely on legitimate considerations are likely in fact to rest on assumptions of the differential worth of racial groups or on the related phenomenon of racially selective sympathy and indifference.
n443. Cass R. Sunstein, Three Civil Rights Fallacies, 79 Cal. L. Rev. 751, 753-54 (1991) (arguing that employer may be motivated by economics instead of irrational prejudices in making discriminatory decisions).
n444. Post, supra note 16, at 20 (identifying customer preference as employer justification for discrimination).
n445. Balkin, supra note 221, at 2319 (arguing that employer may allow discriminatory behavior by "team player" if it promotes workplace cohesion and does not injure business).
n446. Cf. Post, supra note 16, at 20 (discussing sexist customer preferences and their relationship to employer justifications for discriminatory hiring decisions).
n447. Balkin, supra note 221, at 2319.
n448. Indeed, in many cases, their own personal need to maintain a non-prejudiced self-image requires that they characterize their requests in race-neutral terms. See supra note 211 and accompanying text.
n449. Klare, supra note 94, at 1427.
n450. Bartlett, supra note 106, at 2573-76.
n451. See supra Part I.B.
n452. See id. But see Garcia v. Gloor, 618 F.2d 264, 267 (5th Cir. 1980) (justifying English-only policy on grounds of customer preference).
n453. Indeed, in McGlothin, an African American woman plaintiff who taught elementary school children was informed by her supervisor that she was concerned that the employee's all natural hairstyles were teaching the children bad hygiene and setting a bad example for them. McGlothin v. Jackson Mun. Separate Sch. Dist., 829 F. Supp. 853, 857 (S.D. Miss. 1992). For a more detailed discussion of this case, see supra notes 279-286 and accompanying text.
n454. 481 F.2d 1115 (D.C. Cir. 1973) (rejecting male plaintiff's challenge to his employer's rule prohibiting long hair).
n455. Id. at 1124-25.
n456. Klare, supra note 94, at 1138-42, 1433-34 (offering economic arguments for why employer's business necessity justifications for maintaining control over employee appearance should fail).
n457. Post, supra note 16, at 12-16.
n458. As Cass Sunstein explains, "[A] social or legal system that has produced preferences, and has done so by limiting opportunities unjustly, can hardly justify itself by reference to existing preferences." Cass R. Sunstein, The Anticaste Principle, 92 Mich. L. Rev. 2410, 2420 (1994).
n459. See, e.g., Wilson v. S.W. Airlines, 517 F. Supp. 292, 302 (N.D. Tex. 1981) (rejecting claim that airline's love-themed image required female workers, on grounds that image concerns were too tangential to service provided); Diaz v. Pan Am. World Airways, Inc., 442 F.2d 385, 387-89 (5th Cir. 1971) (rejecting claim that airline passengers' preference for female employees was sufficient to justify exclusion of male workers from flight attendant position).
Additionally, an employer's belief that a practice will result in loss of customers, in the absence of any proof, cannot be used to establish sex as a bona fide occupational qualification (BFOQ). Moreover, sex does not become a BFOQ simply because an employer wants to market female sexuality. Guardian Capital Corp. v. N.Y. State Div. of Human Rights, 46 A.D.2d 832, 833 (1974) (rejecting employer's claim that he needed to fire waiters to hire sexually attractive waitresses and boost sales).
n460. Several of these cases concerned wait staff positions in high status restaurants which only employed male waiters. Univ. Parking, Inc. v. Hotel & Rest. Employees & Bartenders Int'l Union, 71-2 ARB. P 8622 (1971) (Peck, Arb.) (rejecting claim that waitresses must be fired to hire waiters as part of classier image); see also EEOC v. Joe's Stone Crab, Inc., 15 F. Supp. 2d 1364 (S.D. Fla. 1998) (holding restaurant chain responsible for policy of only hiring male waiters).
n461. See supra note 459 and accompanying text.
n462. Post, supra note 16, at 27-28.
n463. Courts have recognized that in some narrow circumstances when an independent and objective justification explains the preference, gender is sufficiently related to job function to justify the exclusion of one sex. To test for these circumstances, courts have interpreted the legal exception in a manner that is designed to ensure that customer preference claims are narrowly tailored to meet specific, clearly delineated and reasonable needs. See, e.g., Dothard v. Rawlinson, 433 U.S. 321, 333-40 (1977) (interpreting 42 U.S.C. 2000e-2(e) narrowly, but allowing exception in case of prison guard height and weight requirements which excluded some women).
Some would argue that race never may be marketed in a manner that does not offend Title VII's goal of eradicating a belief in essentialized racial difference. However, while it is not well known, the legislative history of Title VII indicates that there is some limited legal ground for making a case for BFOQ based on "national origin" or ethnicity in certain circumstances. See 110 Cong. Rec. 2548-49 (1964), reprinted in EEOC, Legislative History of Titles VII and XI of Civil Rights Act of 1964, at 3179-81 (1968) (noting that Title VII should not be interpreted to prevent employer from imposing BFOQ requiring Italian national in enterprise making pizzas, or someone of French national origin for waiter position in French restaurant). But see Ray v. Univ. of Arkansas, 868 F. Supp. 1104, 1126 (E.D. Ark. 1994) (arguing that plain language of BFOQ exception prohibits race-based BFOQ). Given current marketing trends, it seems clear that race-based marketing might force broader consideration of this doctrinal issue. If the BFOQ distinction in Title VII ultimately is interpreted to cover race-based claims, employees would fare far better under this test than under the more amorphous customer preference inquiry currently conducted. These questions will be discussed in further detail in my next article on this issue, regarding the permissive marketing of race. Stated simply, antidiscrimination scholars simply cannot have it both ways: Either racial difference has no place in an employer's economic calculus about the value of workers, or it does have value and its absence or presence can be controlled, shaped, and marketed.
n464. In some instances, the practice may be forbidden, but this should not occur before an inquiry as to whether accommodation is possible and necessary. Cf. Matsuda, supra note 91, at 1353-57 (arguing for analysis in accent discrimination cases that separate out practices that are necessary and related to job performance from practices that simply give advantage to whites).
n465. Sunstein, supra note 443, at 754. Indeed, Balkin and Sunstein also indicate that an employer may retain workers who engage in discriminatory behavior simply because, in every other respect, they are good employees, and they are reluctant to lose valued workers. Balkin, supra note 221, at 2319; Sunstein, supra note 443, at 754; see, e.g., Saucedo v. Bros. Well Serv., 464 F. Supp. 919, 920 (S.D. Tex. 1979) ("The court believes that Mr. Nohavistza [the employer] tolerated the [discriminatory] supervisory conduct described herein because good tool pushers are undoubtedly hard to find.").
n466. Balkin, supra note 221, at 2319.
n468. See, e.g., Webb v. R&B Holding Co., 992 F. Supp. 1382, 1388 (S.D. Fla. 1998) (denying African American woman's hostile environment claim based on Spanish-speaking co-worker's use of word "negra"). Plaintiff also had complained previously about her Latino co-workers' use of Spanish in the workplace, which led to a reprimand of those co-workers by a supervisor, but no permanent halt to their Spanish speaking. Id. at 1384.
n469. Bobo, supra note 20, at 460 ("In a multiracial context, one may find some of the highest levels of perceived threat from other groups among members of the most disadvantaged racial minority group, not among dominant racial group members.").
n470. 998 F.2d 1480 (9th Cir. 1993). For a more detailed discussion of this case, see supra Part III.C.1.
n471. Spun Steak, 998 F.2d at 1483.
n472. Id. at 1487-88, 1490.
n473. 992 F. Supp. 1382 (S.D. Fla. 1998).
n474. Id. at 1384-85.
n475. Id. at 1389 (quoting Meritor Sav. Bank, FSB v. Vinson, 477 U.S. 57, 67 (1986)).
n478. Id. at 1389-90.
n479. Id. at 1389. The court then granted summary judgment to the defendant employer because there was no evidence that the employer was aware that the plaintiff was being harassed or that it endorsed the practices she found troubling. Id.
n481. Id. at 1389-90.
n482. See, e.g., Upshaw v. Dallas Heart Group, 961 F. Supp. 997 (N.D. Tex. 1997). In Upshaw, plaintiff offered two comments as direct evidence of her employer's discriminatory intent: her employer's directive that she "'get that nigger rap music off' the telephone hold" and a comment her coworker told another coworker that she overheard in which the supervisor who terminated the plaintiff said that she was being fired because "she sounded too black" and that she intended to pad plaintiff's personnel file with unfavorable evaluations in order to establish a nondiscriminatory basis for firing her. Id. at 1000. The court concluded both that the various party admissions described above were inadmissible on evidentiary grounds and, even if admissible, were insufficient to establish the requisite racial animus. Id.; see also Jeffery v. Dallas County Med. Exam'r, 37 F. Supp. 2d 525 (N.D. Tex. 1999) (granting summary judgment to employer on employee's disparate treatment and hostile environment claim when evidence presented alleged that supervisor made comments about his wearing gold jewelry and listening to "gangsta rap" and had inappropriately targeted him for criticism about his allegedly substandard performance).
n483. Balkin explains that hostile environment rules and other antidiscrimination laws should not be assumed to reduce employer control over the workplace. Rather, in some ways these laws give them increased control, as they use federal antidiscrimination statutes as reasons to regulate employee behavior for often tangential concerns and actually are addressing issues of workplace efficiency and productivity. See Balkin, supra note 221, at 2318-20.
n484. Indeed, the expression of prejudice itself may be a form of "race performance" that triggers hostile environment claims. For example, in Swartzentruber v. Gunite Corp., 99 F. Supp. 2d 976 (N.D. Ind. 2000), the court granted summary judgment to an employer on a white plaintiff's Title VII religious discrimination claim. The employee objected to his employer's requirement that he cover a tattoo on his arm indicating his affiliation with the Klu Klux Klan after several black employees complained about it to the employer. Id. at 978, 981.
n485. Courts already have been called upon to recognize these conflicts as religious discrimination claims. See id. (granting summary judgment to employer on white employee's Title VII religious discrimination claim to challenge employer's rule requiring him to cover KKK tattoo at work). The employer in this case requested that the employee cover his tattoo because several black employees had complained about the tattoo and he worried that it would create a hostile environment for those employees. Id. at 978. Importantly, the tattoo more easily can be understood as germane to the white employee's performance of whiteness in the workplace, a performance which had begun to intrude on black employees' workplace rights.
n486. But see Balkin, supra note 221, at 2304-05. Balkin explains that employers are uniquely positioned to intervene in these workplace cultural conflicts because they have more information about the collective actions of coworkers which, individually, might not trigger Title VII concerns but cumulatively create problems. Id. at 2304. Additionally, employers are better positioned to anticipate the types of conflicts that may occur. Id. Lastly, he explains that employees themselves may have little incentive to prevent hostile environments due to collective action problems. Id. at 2305.
n487. Post, supra note 16, at 13.
n488. Klare, supra note 94, at 1404-05 (discussing management prerogative).
n489. See K. Anthony Appiah, Race, Culture, Identity: Misunderstood Connections, in Appiah & Gutmann, supra note 419, at 30, 90.
n490. Minow, supra note 423, at 671-72 (explaining that identity politics fails to teach people how to "talk across differences and divides" in manner that would allow them to recognize similarities in their experiences).
n491. See id. at 647-50 (discussing identity politics). Minow discusses this impulse in the context of identity groups demanding representatives who "look like them," based on the assumption that a person of the same sex or race has the same experience and political perspective and that they cannot rely on a member of another group to adequately represent their experiences. Id. at 650-53.
n492. Appiah, supra note 394, at 160-61. Appiah worries that concepts like race substitute "gross differences" of morphology with "subtle differences" not correlated with such features and are standing in for the more accurate ways of grouping people, by shared culture. See, e.g., Anthony Appiah, The Uncompleted Argument: Du Bois and the Illusion of Race, in "Race," Writing, and Difference 21, 36 (Henry Louis Gates, Jr. ed., 1986). Wendy Brown addresses another dimension of the same problem, explaining that certain oppositional identities founded on the need for "recognition," the hallmark of identity politics, hold questionable value in a liberatory project as they are focused on recrimination, rancor, and disdain freedom rather than hold it as an ideal. Wendy Brown, States of Injury: Power and Freedom in Late Modernity 52, 55 (1995).
n493. Brown, supra note 492, at 60-61. Brown explains that many identities rely on a "white masculine middle-class ideal" as proof of their exclusion but, in doing so, fail to realize that their positioning of this identity as center renders invisible certain social forces (i.e., capitalism) and insulates them from critique. Therefore, reform efforts to give excluded groups the same rights as this middle-class subject already are limited in their revolutionary potential and appeal even as they are conceptualized. Id. at 61.
n494. See supra text accompanying note 94.
n495. Id.; see also Minow, supra note 423, at 663-64 (arguing that common history of oppression may be primary source of group identification).
n496. Its transformative effect might be felt even in second-order practices that do not immediately trigger thoughts about rights and values. For example, because the race/ethnicity performance paradigm recognizes that individuals may need accommodations for cultural and other race/ethnicity-associated differences, it holds the possibility of changing both the requirements that employers impose on employees and customers' expectations with regard to services. Under the current regime, a rule that prohibited workers with strong accents from customer service roles might make business sense, as we are encouraged to treat these workers as invisible, interchangeable, and reduce them to their instrumental value. However, under a race/ethnicity performance regime, where these workers were employed in such positions, the caller would be forced to do more interpretive work in the exchange, the conversation would be more personal as this was negotiated, and the attitude and expectation one brings to the call would be forced to change. Some will argue that ignoring a person's accent in these circumstances or providing special training amounts to unfair special treatment; however, these accommodations may be necessary simply to level the playing field for all workers. See Amy Gutmann, Responding to Racial Injustice, in Appiah & Gutmann, supra note 419, at 106, 109 (explaining that in some instances, it is necessary to treat members of various groups differently in order to treat them fairly).
n497. See Appiah, supra note 489, at 90 (discussing race-based claims specifically).
n498. Id. at 90-91.
n500. See id. at 90 (arguing that racial group members do not necessarily have interest in or know anything about particular cultural practices associated with that group).
n501. See id. (noting that idea of exclusive claims to culture "deprives white people of jazz and black people of Shakespeare").
n502. See supra Part I.A.2 (discussing race/ethnicity performance-based ascription).
n503. Additionally, the model avoids the "authenticity problems" that may prevent a worker from raising a race performance claim. That is, a white worker raised in an urban ghetto may speak "Black English" and offer evidence that she was discriminated against on that basis. Rather than doubting the legitimacy of her experience, the worker could bring a Title VII disparate treatment claim, challenging her employer's adverse treatment of her based on this racialized display.
n504. Hollinger, supra note 28, at 107.
n505. Minow, supra note 423, at 666-67 (worrying that "identity politics may freeze people in pain and also fuel their dependence on their own victim status as a source of meaning"); see also Brown, supra note 492, at 220 (explaining that some identity groups cannot let go of past suffering "without giving up [their] identity as such").
n506. Hollinger, supra note 28, at 116 ("Postethnicity prefers voluntary to prescribed affiliations, appreciates multiple identities, pushes for communities of wide scope, recognizes the constructed character of ethno-racial groups, and accepts the formation of new groups as a part of the normal life of a democratic society... . Individuals should be allowed to affiliate or disaffiliate with their own communities of descent to an extent that they choose, while affiliating with whatever nondescent communities are available and appealing to them.").
n508. Id. at 118 ("Not every descent-community will retain its members; some of these communities can be expected, over time, to decrease their role in the lives of individuals and of the larger society.").
n509. Id. at 107-08 (explaining that postethnic perspective remains attentive to those elements from different cultures that reflect shared ethos); see also id. at 126 (explaining that postethnic perspective resists will to descend, using genetics as basis for proprietary claims about culture).
n510. Id. at 117-19.
n511. See Wijeyesinghe, supra note 153, at 140-42 (discussing multiracial people's options for claiming various racial identities).
n512. Hollinger argues that contemporary models of identity are too genetic or "fixed," arguing that "the concept of identity is more psychological than social, and ... [fixed models] can hide the extent to which the achievement of identity is a social process by which a person becomes affiliated with one or more acculturating cohorts." Hollinger, supra note 28, at 6.
n513. See supra note 176 and accompanying text.
n514. Kenneth L. Karst, The Supreme Court 1976 Term - Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 Harv. L. Rev. 1, 9 (1977) ("The principle of equal citizenship, nourished to fruition, offers all of us the opportunity to belong to a national community while continuing to identify ourselves with smaller groupings of diverse and even conflicting values. Equal citizenship, then, is one institutional response to the tension between autonomy and community.").
n515. Certainly, Hollinger and his colleagues still may resist the creation of race performance protections out of concern that they might subsidize race/ethnicity-associated differences and prompt litigation based on these differences. They may argue that race performance runs the risk of further escalating an already deleterious trend in subordinated minority communities: the tendency to privilege ethnic specificity over their right to participate in, shape, and feel ownership of American culture. This concern takes on additional import in light of the high rates of residential segregation in the United States, which serve to isolate minority communities and increase their distinctiveness. See Douglas S. Massey & Nancy A. Denton, American Apartheid: Segregation and the Making of the Underclass 160-62 (1993); Douglas S. Massey & Nancy A. Denton, Trends in the Residential Segregation of Blacks, Hispanics, and Asians: 1970-1980, 52 Am. Soc. Rev. 802, 823 (1987) (noting blacks trapped in segregated racial enclaves have very narrow horizons and do not often venture beyond their community). Also, in my view, the decreasing interest of the middle class in funding, cultivating, and making use of urban public space, as evidenced by the growth of suburbs and gated communities, aggravates matters, as it further decreases opportunities for cross-cultural contact. Consequently, the workplace takes on additional importance as it remains one of the few places for required cross-cultural interaction. In my view, however, it would be far better to furnish a vehicle for preventing ethnic communities from being further stigmatized by racial constructs and that draws attention to the subtle ways in which this occurs. The focus must be on the atmosphere in the workplace years after the litigation, when different ethnically marked workers are forced into coalition together.
n516. Cynthia L. Estlund, The Workplace in a Racially Diverse Society: Preliminary Thoughts on the Role of Labor and Employment Law, 1 U. Pa. J. Lab. & Emp. L. 49, 60-61 & nn.51-54 (1997-98) (explaining that whites who work in workplaces with high black representation are twelve times more likely to have black friend). Estlund also cites studies suggesting that working in a mixed community seems to be even more effective than living in a desegregated community if the goal is increasing the potential for productive interracial interaction. Id. at n.52.
n517. Klare, supra note 94, at 1411.
n519. Ferdman & Gallegos, supra note 31, at 44 (discussing anti-black prejudice as existing within Latino communities and occasionally causing intra-Latino splits and discrimination within group based on color).
n520. Dovidio et al., supra note 207, at 90; Gaertner & Dovidio, Aversive Racism, supra note 123, at 63-64; see supra Part II.B (discussing aversive racism).
n521. Cf. J.M. Balkin, The Constitution of Status, 106 Yale L.J. 2313, 2336-37 (1997) (arguing that opposition to gay rights is caused by fear of resulting decline in heterosexual status); see supra Part II.B (discussing group position theory).
n522. At this juncture, our discussion about the cultivation of the cosmopolitan citizen nicely dovetails with legal scholarship on the issue of white privilege. Barbara Flagg argues that once whites are deprived of a context in which their race identities exist unproblematically, they will be forced to develop their race identities on a more equal plane with people of color. Flagg, supra note 216, at 957. She explains that, in such a workplace, whites will be called upon to "develop a positive white racial identity, one neither founded on the implicit acceptance of white racial domination nor productive of distributive effects that systematically advantage whites." Id.; cf. Butler, supra note 16, at 59-60 (recognizing that, at present, whiteness is unmarked social sign that structures our existence, whereas blackness is social sign that is marked and imbued with meaning). While these scholars rightly argue that whites should be encouraged to take critical appraisal of how race constructs inform their identities, ethnic minorities would benefit from doing so as well.