The Facebook Project

The Missing Box: The Racial Politics Behind The Facebook Interface

Authored by Jeff Ginger
Keywords and tags: Facebook, racism, discrimination, identity, digital divide, interface, colorblind, human-computer interaction
Revision 3 | Last updated 08.2008 | PDF

Update: A recent response to this paper is now available at http://facebookprojectqa.blogspot.com/2011/01/filling-in-missing-box.html

Synopsis: There is no race, ethnicity or nationality category available the default Facebook profile.  This, combined with the fact that Facebook is a visually-driven website, suggests that Facebook serves to inadvertently or covertly perpetuate two racist or discriminatory norms: the colorblind mentality and racialized visual classification of others. This paper quickly traverses through a basic theory of identity that dictates that persons incur an on-going dialectic between individual and social identity, both on and offline. It contends that Facebook is an advantageous place to perform and portray one’s social identity and is largely visually-driven, therefore the interface is of paramount importance.  Racial, ethnic, or national identity is of significance to many individuals and the lack of a place to express this is a form of discrimination.  Such a category should be added to Facebook’s set of basic identity options. Guidelines for a successful implementation are provided.

Introduction

The world-renowned Social Networking Service (SNS) Facebook.com is often a daily internet stop for undergraduate college students across the nation and increasingly people of all kinds in the US and worldwide.  The site consistently ranks as one of the most visited websites on the entire internet and commands well over 90 million users, all who have returned to visit at least once in the past 30 days (Facebook Statistics, August 2008).  Among college student populations it has created a digital mirror of face-to-face university communities because nearly everyone (over 85% at most colleges) is on it and participates daily (Ginger 2008a).  Other usage and growth rates are downright staggering and the website shows no sign of slowing down or diminishing in influence.  The enormous participation in the Facebook ecology and its importance in college life is of great importance to educators, policy makers, researchers and really anyone with a stake in the college student population.

The internet has become a staple of the average student’s daily media intake and in many cases enacts their sole or primary source of information, be it of social, news, or value driven nature (boyd^* 2007b).  Compounding issues of content composition and its critical role in informing students is the variation in the different systems of interface used to access this knowledge.  As any system or interface designer knows all too well, sometimes they have unknown or unforeseen effects.  Often times the designers or dominators of a system fail to realize or adequately assess the needs of all of its users, leaving some at a disadvantage or with unmet needs.  Facebook is a remarkable interface and arena of information, but despite its unprecedented opportunity for user control it has some significant limitations.  This paper captures an example of an insidious implication found amongst the profile interface of the system: a race/ethnicity/nationality box is missing from the array of social analytic categories available for defining ones identity.  This interface decision encourages two racist or discriminatory norms in society: denial of self identification and an ignorant colorblind regard for identity.  This paper quickly traverses through an explanation of identity and interface on the web building up to the problem of the missing race/ethnicity box and then offers comprehensive solution and various implications are presented in response.

Identity remains one of the most controversial realms of academic study.  Countless scholars have pondered the composition, construction and meaning of identity for as long as history has been remembered.  Regardless of specific definitions of the perplexing abstraction, which can even only be spoken about because it is given dubitable, emergent form by dynamic, contingent recognition (Waters 2004 referencing Dominguez 1986), identity remains at the core of our understandings of self and existence as human beings.

By addressing the perspectives presented by Adam Smith, Mead, Freud, and Lacan one arrives at two primary notions of the self: an internal perception and an external social identity (boyd 2002 summarizes this concept, which was also assumed by Goffman).  We as human (sentient) beings all have a comprehension of self (self-concept), which is often understood through self-evaluations that involve consistent attributes (e.g. “I am enthusiastic”).  In other words, one’s internal identity consists of physical, psychological, philosophical and moral aspects of self (boyd 2002).  This self-concept is a prerequisite (but distinguished from) self-consciousness (or awareness), which is a comprehensive sense of self that is dependent upon context.  No aspect of one’s persona is self-evident, however.  They are demonstrated relative to other actors who serve as the basis of reflexive measurement.  That is to say we can’t really know who we are without comparing and relating ourselves to others.  Our internal identity is assembled from history, experience, and interaction, which in turn gives rise to social identity as people group and share experiences in various ways.  Both the complex actor conveying a representation and the context in which it is extant form the social ipseity (identity).  The incessant and necessary interplay between the two worlds, which is retroactive, perpetual, and heterogeneous, is a fascinating dialectic.  This duality can be likened to many other classic debates, such as situationism verses traits and motivations¹ or to structure verses agency. ²

Regardless, this paper does not seek to provide an essentialist or all-defining answer to the question of identity.  Instead it assumes the basis of social identities, or the shared histories, experiences, and understandings between individuals—in particular those of racial, ethnic, or national identities.  It goes without saying that these aspects of identity are crucially important to many persons throughout the world, much less the college student population on Facebook.

As the internet became mainstream in the developed world during the 1990’s researchers investigated the various possibilities and implications it would hold for the social conception of identity.  Two of the most famous were Sherry Turkle and Donna Haraway.  Turkle investigated the potential for fragmented, multiple postmodern selves and surmised that “computer-mediated communication can serve as a place for the construction of identity.” (Turkle 1995:14).  Haraway explored discursive feminist concepts such as affinity politics and oppositional consciousness through the idea of the cyborg and sought to reject boundaries such as those between man and machine (Haraway 2004). In many ways identity as we know it in an everyday sense could not immediately port to the web, much like it could not be directly replicated in written form or over the telephone.  By the end of the decade some researchers were addressing how the internet had encouraged the vision of identity as a sort of commodity to be valued, verified, and transferred (Abelson and Lessig 1998, Camp 2004). Others sought to dispel the negative associations that had begun to accrue regarding the online interactions that mediate identity and personality.  One such pair was McKenna and Bargh (2000) who were among the first to assert (in response to popular opinion and fears) that the internet could have no single, simple effect upon all people, defining who they are in ways such as inducing loneliness or causing introverts.  They instead explained that people use the web for all kinds of reasons and motivations and that it was not, like most technologies, inherently good or bad in terms of the kinds of interactive social effects it could have on individuals.  Their article, Plan 9 From Cyberspace: The Implications of the Internet for Personality and Social Psychology, calls upon the works of dozens of authors to spot several major digital architectural differences that could alter the conditions in which identity works. 

One of the primary points discussed by McKenna and Bargh is the lack of a physical self online.  In person our identity is constructed, in part, by instrumental physical characteristics and interactions involving non-verbal cues (somatic, tonal, accessory-driven). In many places online, from chat rooms to websites, this is turned upside down, so much to the point that people who meet online are more likely to like one another than if they had met in person.  McKenna and Bargh (1999, cited in 2000) found that people who met first online walked away with a conception of the person they had just met that more closely resembled that person’s own identified image.  With services like Skype and social networking becoming more popular (not to mention those such as Second Life and 3D games) this disembodiment, as boyd (2002) refers to it, is less and less prevalent.  It would seem that many people wish to extend their physical-selves online as much as their intellectual personas.

This analysis offers two important notions.  First, if we’re more likely to make good impressions on the internet with the management of our identities in places like Facebook, then it’s especially important that we be able to represent all of our appropriate personal and social identities, if we so choose.  In fact, Facebook fits boyd’s proposition for a self-awareness enabling tool; that is an interface or system that provides data that “goes above and beyond the magnitude of information that people have offline” and would also be “useful for users to be aware of what is out there about them.” (boyd 2002: 58).  It does so by (implicitly) letting users know which applications their profile spans (the applications available on their profile is manageable, as well as the way their profile appears within them on other pages), by providing high –level data about issues of interaction pertinent to the user (alerts and the newsfeed), and by giving both raw data (say, total number of friends) and more useful impressions (a listing of all friends in a given group or a random pick of six thumbnail images of friends).  Facebook is both compelling and functional, so much so that users invest a great deal of time and energy into it.

Second, much of Facebook’s success has been induced by its visual interface.  Besides being user friendly and aesthetically eloquent, the interconnectivity and interaction between profiles and users is considerably tied to pictorial displays.  Users are more likely to check out profiles of others with pictures they find attractive or interesting (Ginger 2008a), often engage with picture galleries on a profile when possible, and easily jump from one profile to another through profile picture thumbnails or by clicking on tags of users in pictures or videos.  Indeed, the first thing viewed about a given profile in the return in a search queue is the picture.  With its heavy reliance on pictorial representation the creation of a Facebook profile also helps to fight the feeling of disembodiment that afflicts digital travelers as they embark on their journeys through the web.  Fragments of real world spaces can be indirectly mapped into the digital space through the use of pictures, audio, and video.

So Facebook is an advantageous place to perform and portray one’s identity and it is largely visually driven, but one of the most immediate quandaries of the transition from face-to-face to digital existence is the essential but connotation-plagued and contrived issue of metaphor.  Sociotechnical systems are fundamentally tied to the use of metaphor to make themselves accessible.  Interfaces are distributed throughout and indivisible from their systems, monitor and control a reductive-oriented, indexical map of separate elements of multiple (potentially infinite) states, and act as an associational structure that permit agents to manipulate, alter, create, destroy, and replicate processes and objects to which they are independent (Fuller 2003).  The digital representation of identity, at root, must be tied to metaphor in its manifestation and interpretation.  People cannot intrinsically understand the 1’s and 0’s that make up the operations within a computer; interfaces are designed to make meaning and symbols out of the data to convey information. This introduces all manner of limitations and potential avenues for reinterpretation (or misinterpretation) of identity.  Systems are erected to verify or authenticate ‘users’ (who are also emulations of code) that pay special attention to attributes and authorizations to confirm or shape identity (Camp 2004, Nakamura 2002, Lessig 2006).  Naturally this adaption is ridden with deficiencies—the spatial properties of the physical world do not often translate properly in cyberspace, save for virtual worlds like Second Life or World of WarCraft, and even those realities have substantial discrepancies when compared to reality.

The internet is far from the egalitarian utopia once pitched during its conception.  Many individuals do not have physical access and others do not have the skills to operate web technologies (DiMaggio and Hargittai 2001).  Still others do not have experiential access; perceived barriers to access (or usefulness or ease of use) play just as much of a role in preventing people from getting online as actual barriers (Porter and Donthu 2006).  As a result, group identities belonging to marginalized or disadvantaged populations could be setback or hindered in the world of the digital.  Even once people are established online studies demonstrate that gendered, sexual, classed, raced, and age-based identities and corresponding conflicts continue to be salient factors in determining the character of online relationships (Kendall 1998). 

Naturally Facebook, like the rest of the net, is a contested zone.  It can be a place of discursive activity that might come in positive forms, like the expression of African American identity through AAVE that was sighted on Black Planet (Banks 2006), or in negative ones, like racist or sexist groups whose latent function is to keep majority groups in power (Ginger 2007, 2008a).  Participation in Facebook may be an example of the latest form of true experiential access to cyberspace for youth today (DiMaggio and Hargittai 2001) and those left out could easily be at a disadvantage in a college atmosphere (Ellison et al. 2006).  The interface of Facebook is both visual and selective; some choices are unavoidable while others are unavailable.  Users are able to give a variable amount of feedback in this regard as well as force alternative interpretations through agency manifested in groups, events, and deviant answers to Facebook’s categorical imperatives. The encoded responsiveness of the system, however, does not bestow equal consideration to all types of users.  Though anyone can build an application or start a group or event, not everyone has the motivation or education required for such efforts.  The primary creators of content are likely those who are active and those who are savvy (Ginger 2008b).  These institutionalized aspects allow for a certain brand of identity to be propagated on Facebook. 
Facebook allows you to explicitly specify your sex, birth date, and political and religious views in the basic information section and sexuality³ and relationship status in the relationships section.  Furthermore by studying the assemblage of current locale, high school or home town, education and occupation researchers can actually deduce socio-economic class with a moderate level of accuracy.  The only primary social analytic category missing in the fundamental and uniform set of profile identity characteristics on Facebook is race/ethnicity/nationality.  In lieu of no technical category for race on Facebook users can make estimations of another’s ethnicity or race based on their picture.

What happens, then, when you combine these two constraints, the missing box and visually imperative interface, together?  If the underlined function of White Nationalism (also known as the racial project) is a way to keep the white majority in the US comfortable and in power (Walters 2003), then Facebook is a slave to this ideal. Members of the white majority can remain comfortable by not talking about race and maintaining their privilege without having to consciously think about it.  On some level this objective makes sense; most people have some measure of shared suffrage and aptitude to care about others and so to think about race might force them to reconcile the disparities.  We see all manner of defensive mechanisms to combat this eventuality: denial of unequal opportunity to avoid the development of a consciousness of privilege, claiming race is only perpetuated by obsessing over it, and pledging personal colorblind outlooks in a society not founded on them.  From the perspective of Digital Divide studies unequal systems of power relationships are “sustained by shaping people’s worldviews, controlling resources, and constraining opportunities” (Goodman 2000) and Facebook’s deprivation of a racial-ethnic or nationality category is just that.  The corresponding view of the world sanctioned by the absence of opportunity paints a sort of normalcy the gives way to a sense of superiority and parallel dominant ideology (Goodman 2000, Bonilla-Silva 2003).  Just because white people don’t find race important enough to include in their basic notion or entourage of identity doesn’t mean this value is one that can or should be adopted by everyone.  The implication of Facebook, therefore, is corollary to White Nationalism.  The one-sided racial identification magnifies this problem as well.  Preventing individuals from asserting their racial, ethnic or national identity prevents them from making its importance known to the majority.  Dominant members just make subjective categorizations of others and have the power to decide if they are a racial minority or not.  People feel disempowered when they cannot assert their own identity and are treated inaccurately or unfairly depending upon the race assigned to them.  The issues of general unjust treatment of different races suspended, people aren’t even afforded control over their simple racial label.

By choosing to make a race, ethnicity, or nationality category unavailable and knowing that Facebook is a visually-driven SNS, Facebook serves to inadvertently or covertly perpetuate two racist norms: the colorblind mentality and racialized visual classification of others.

Though this is not a white cloak, flag burning, immigrant-hating form of racism, most modern day forms of institutionalized or covert racism are not.  Most progressive race scholars agree that indifference and apathy in regards to race or ethnicity is another tool and form of white prejudice and is engineered and perpetuated on a societal level (Forman and Lewis 2006).  Likely the Facebook designers are not even contemplating the role of race or ethnic identity as part of their value set when fabricating the features and interfaces of the SNS, but this unintentional continuation of ambivalence is harmful.  Pretending that historically established and contemporarily sustained inequalities do not exist, or simply ignoring them in hopes that they will go away, is a preposterous and appallingly inadequate solution to social disproportions.  What’s more is that we know identity to be co-producing, our racial and ethnic identities are some combination of who we feel and think we are, and who others feel and think we are.  Facebook only allows for the racialised, initially visual, assignment of identity.  Though Facebook communication with another person could easily penetrate racial barriers and fashion new alliances, participants are often seldom ever able to reach this point because their internalized race and racist norms that prevent them from making the connection in the first place. The effect is an echo of offline-world hierarchies where reciprocation and initiation to counteract racial fragmentation and desolation are definitively stacked against the norm. 

Research, as it turns out, has verified this, as Mayer and Puller discovered in 2007 in their study of group formation and network composition on Facebook. “Race is strongly related to social ties, even after controlling for a variety of measures of socioeconomic background, ability, and college activities… changes in the school environment that affect the likelihood that two students interact have only a limited potential to reduce the racial segmentation of the social network.” (Mayer and Puller 2007).  Currently Facebook does make much of a difference in bridging between people of different racial and ethnic backgrounds.  Though Mayer and Puller investigate the possible differences affirmative action policies make they don’t ponder the possibility of making the interface of the medium more accurately representative.

Simply adding a category for race, ethnicity, or nationality on Facebook is not a solution.  It is a compelling first step, but truly we must change our discourse and understanding of race and ethnicity on a societal level for this change to be truly effective.  The change must simultaneously consist of individual and institutional agency – educators, media, and social movements ought to be embodied by individuals who wish to reflect positive social change.  Adding a race or ethnicity category to Facebook can accompany their mobilization and impact.  Regardless, even this first step should be taken with caution.

Race isn’t the only category missing from Facebook that might be disputed.  Sociologists generally recognize disability, be it physical, mental, or mental health related to be a significant manner in which ways individuals might identify.  Such a category would be difficult to place on Facebook as there is no good language for being ‘differently-abled’ or ‘abled’ and would likely be misinterpreted.  Student users already make a joke of Facebook, and while categories like race or religion are serious social considerations and are less likely to be abused, a box for ‘ability’ would merit responses ranging from super hero powers to sexual jokes.

Race is very different in America then it is in many other countries.  Ethnicity certainly exists globally but race may not be a category that even makes sense in places like Korea, where the population is close to 98% homogeneous.  Identifying as ‘white’ might not carry the same connotation in the US as it would in some Latin-American countries, as people who are ‘white’ there might be identified as Hispanic in the US.  Facebook, however, has already solved this problem with their built in network clustering.  They could choose to add the category for relevant populations (again, dangerous, but possible) or better yet just rest assured in the fact that the populations using the service will reference it in perspective of their own network and at worst it’s just another box to skip, just like the way the religion might be interpreted in China.  Thus there is a need for a category that covers race, ethnicity, and nationality.  The addition of nationality to the repertoire also gives a voice to white Americans who might not otherwise have a way they could identify.

The honest reality is that they’re all related anyway.  The politics of racial and ethnic differences are ridden with volatile disputes and strong emotional ties to identity.  Numerous studies of census information collection and quality assurance have illuminated the dilemmas in counting race (Rodriguez 2000).  The contended differences between race and ethnicity, their intersectionality, and dynamic, emergent nature all complicate the objective as higher level questions of simultaneously confuse it by asking who is being counted, to what end, and if numbers even truly reflect diversity (Prewitt 2004).  Menus that insist on a limited range of choices deny ways of being raced not considered by system architects (Nakamura 2002).  Facebook has a clear advantage in this game, as it does not need a strict categorized count or to encourage a certain countable degree of racial diversity, so response uniformity is not a necessity.

Given these aggravations, this paper presents a proposal.  Facebook should add a racial-ethnic-nationality category which should include the following parameters:

1. The category must be added to the basic identity set – Racial, ethnic and national identification is arguably as important as gender, age, political/religious views, or any of the other major ways of being or thinking and as such needs to receive equal representation within the options presented by the interface.  It should be offered front and center, up top.
2. The label for the text box should be “Race / Ethnicity / Nationality” – Since there are so many disputes between the differences and relationships between these concepts it would be best to leave it up to the respondent to navigate the murky waters for themselves.  Avoiding treacherous definitions also enables white Caucasian participants to better know how to identity (though potentially a downside) as they can exercise their ethnic options, or the ability to pick and choose whatever ethnic history from their past they like (Waters 1990).  What’s more is that populations that don’t have a conceptualization of race can still use ethnicity or nationality if it makes more sense.
3. The category should NOT be required – Just like most other elements of identity are not required on a Facebook profile this one should follow suit.  Not everyone will want to reveal this aspect of their persona and respect for privacy is crucial.  If the system designers really wanted to give race or ethnicity a connotation of importance they could make it required, but the opinion of this paper is that simple availability is sufficient.  This becomes especially helpful when regarding populations that don’t have a lot of racial variation and such a category would become akin to marking ‘human’ or something equally irrelevant or superfluous.
4. The feature should be open response – More or less along the same rationale as the first proposal, since there are a near indefinite ways to racially or ethnically identify, just like religion, the spot ought to be left open to interpretation.  As noted above, Facebook doesn’t need to do demographic counts and people could still like to one another’s profiles through common key words in the category.  The box should be like religious or political views and open for users to indicate whatever they like (with a limitation on number of characters, of course).

By creating a social analytic category that complies with these specifications Facebook can successfully offset the colorblind norms previously encouraged by inactivity on the issue.

• Abelson, Hal and Lawrence Lessig. 1998. “Digital Identity in Cyberspace.” White paper submitted for 6.805/Law of Cyberspace: Social Protocols, December 10.
• Banks, Adam. 2006. Race, Rhetoric, and Technology: Searching for Higher Ground. Mahwah, CT: Lawrence Erlbaum Associates.
• Bonilla-Silva, Eduardo. 2003. “’New Racism,’ Color-Blind Racism, and the Future of Whiteness in America.” Pg. 271-85 in White Out: The Continuing Significance of Racism. Oxford: Routledge.
• boyd, danah. 2002. “Faceted ID/Entity: Managing Representation in a Digital World.” Masters Thesis on Danah.org, September 2002.  (http://smg.media.mit.edu/people/danah/thesis/danahThesis.pdf)
• boyd, danah. 2007a. “Information Access in a Networked World.” Talk Presented to Pearson Publishing. Palo Alto, California, November 2. (http://www.zephoria.org/thoughts/archives/2007/12/03/information_acc.html)
• boyd, danah. 2007b. “Social Network Sites: Public, Private, or What?” Knowledge Tree, May 13. (http://kt.flexiblelearning.net.au/tkt2007/?page_id=28)
• Camp, L. Jean. 2004. “Digital Identity.” IEEE Technology and Society Magazine. An earlier version of this paper was presented at ISTAS’03, Amsterdam, The Netherlands. Fall 2004: 34-41.
• Ellison, Nicole, Charles Steinfield, and Cliff Lampe. 2006. “Spatially Bounded Online Social Networks and Social Capital: The Role of Facebook.”  Annual Conference of the International Communication Association. Dresden, Germany: June 19-23.
• DiMaggio, Paul and Eszter Hargittai. 2001. “From the ‘Digital Divide’ to ‘Digital Inequality’: Studying Internet Use as Penetration Increases.” Princeton: Center for Arts and Cultural Policy Studies, Woodrow Wilson School, Princeton University.(http://www.princeton.edu/~artspol/workpap15.html)
• Forman, Tyrone A. and Amanda E. Lewis. 2006. “Racial Apathy and Hurricane Katrina.” W.E.B. Du Bois Institute for African and American Research 3:175-202.
• Fuller, Matthew. 2003. “The Impossibility of Interface.” Behind the Blip. Brooklyn, NY: Autonomedia: 99-120.
  Ginger, Jeff. 2007. “Social Capital and the Chief.” TheFacebookProject.com. December 2007. (http://www.thefacebookproject.com/research/jeff/publications/socialcapitalchief.html)
• Ginger, Jeff.  2008a. “Performance and Construction of Identity.” TheFacebookProject.com. July 2008. (http://www.thefacebookproject.com/research/jeff/publications/masters.html)
• Ginger, Jeff. 2008b. “Gender Roles and Group Discourse.” TheFacebookProject.com. May 2008. (http://www.thefacebookproject.com/research/jeff/publications/genderroles.html)
• Goodman, Diane J. 2000. “About Privileged Groups.” Pg. 13-36 in Promoting Diversity and Social Justice. Thousand Oaks, CA: Sage Publications.
• Haraway, Donna. 2004. The Haraway Reader. New York, NY: Routledge.
• Kendall, L. (1998). “Meaning and Identity in “Cyberspace”: The Performance of Gender, Class, and Race Online.” Symbolic Interaction 21:129-53.
• Kenneth Prewitt. 2004. “The Census Counts, the Census Classifies.” Not Just Black and White: Historical and Contemporary Perspectives on Immigration, Race, and Ethnicity in the United States. Ed. Nancy Foner and George M. Fredrickson. New York: Russell Sage Foundation Publications, Pg. 145-164.
• Lessig, Lawrence. 2006. Code version 2.0. New York, NY: Basic Books.
• Mayer, Adalbert and Steven L. Puller. 2007. “The Old Boy (and Girl) Network: Social Network Formation on University Campuses.” Journal of Public Economics, doi: 10.1016/j.jpubeco.2007.09.001.
• McKenna, Katelyn Y.A. and John A. Bargh. 2000. “Plan 9 From Cyberspace: The Implications of the Internet for Personality and Social Psychology.” Personality and Social Psychology Review 4:57-75.
• Nakamura, Lisa. 2002. Cybertypes: Race, Ethnicity, and Identity on the Internet. New York, NY: Routledge.
• Porter, Constance Elise and Naveen Donthu. 2006. “Using the technology acceptance model to explain how attitudes determine Internet usage: The role of perceived barriers and demographics.” Journal of Business Research 59:999-1007.
• Rodriguez, Clara E. 2000. Changing Race: Latinos, the Census, and the History of Ethnicity in the United States. New York, NY: New York University Press.
• Roediger, David R. 2003. “Still White.” Colored White: Transcending the Racial Past.  Berkeley: University of California Press, Pg. 3–26.
• Turkle, Sherry. 1995. Life on the Screen: Identity in the Age of the Internet. New York, NY: Touchstone.
  Walters, Ronald. 2003. White Nationalism, Black Interests: Conservative Public Policy and the Black Community. Detroit, : Wane State University Press.
• Waters, Mary C. 1999. Black Identities: West Indian Immigrant Dreams and American Realities. New York, NY: Russell Sage Foundation.
• Waters, Mary C. 1990. Ethnic Options: Choosing Identities in America. Berkeley, CA: University of California Press.