Personality and Social Psychology Review€¦ · RACIALPHENOTYPICALITYBIAS EmpiricalAccounts: Evaluations andBeliefs In general, people have exhibited a preference for lighter skin

http://psr.sagepub.com

Personality and Social Psychology Review

DOI: 10.1207/s15327957pspr0804_4 2004; 8; 383 Pers Soc Psychol Rev

Keith B. Maddox Perspectives on Racial Phenotypicality Bias

http://psr.sagepub.com/cgi/content/abstract/8/4/383 The online version of this article can be found at:

Published by:

http://www.sagepublications.com

On behalf of:

Society for Personality and Social Psychology, Inc.

can be found at:Personality and Social Psychology Review Additional services and information for

http://psr.sagepub.com/cgi/alerts Email Alerts:

http://psr.sagepub.com/subscriptions Subscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://psr.sagepub.com/cgi/content/refs/8/4/383 Citations

at UNIV WASHINGTON LIBRARIES on July 20, 2009 http://psr.sagepub.comDownloaded from

http://www.spsp.org/

http://psr.sagepub.com/cgi/alerts

http://psr.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://psr.sagepub.com/cgi/content/refs/8/4/383


Personality aind Social Psychology Review2004, Vol. 8, No. 4, 383-401

Copyright 0 2004 byLawrence Erlbaum Associates, Inc.

Perspectives on Racial Phenotypicality Bias

Keith B. MaddoxDepartment of Psychology

Tufts University

This article reviewls research examining racial phenotypicality bias within-categorystereotypes, prejudice, and discrimination based on race- related phenotypic charac-teristics ofthe face. A literature review of research evamining skin tone bias, drawinglargely from work examining perceptions ofBlacks in the United States, reveals thatindividuals with features typical of members of their racial category are perceivedand treated more negatively by social perceivers. Furthermore, this treatment hasbroad implicationsfor social status and health. Despite this evidence, the tendency to

attend to and use within-race variation in phenotvpic appearance has been over-

looked in social psychological models ofimpressionformation. However, several the-oretical frameworks have recently been proposed to explain the role of pheno-type-based expectancies in social representation and judgment. Drawing on thestrengths of each perspective, a rudimentary model of racial phenotypicality bias isproposed. This analysis suggests that future examinations guided by the current

framework (or similar others) can complement existing evidence toward a greater un-

derstanding of the role ofphenotypic variation in social perception.

Many of us are familiar with vivid examples of ra-cial bias-negative treatment exhibited toward variousindividuals as a function of racial category member-ship. Generally, this bias reflects a negative dispositiontoward non-White individuals. At its core, racial biasstems from the idea that White Eurocentric phenotypiccharacteristics (e.g., lighter skin and eye color, longerand straighter hair, narrower nose, and thinner lips) arepreferable to features toward the other end of the con-tinuum (e.g., darker skin color, kinkier hair, broadernose, fuller lips). As a consequence, White and non-White members of many societies are exposed to thisideal and adhere to it in their evaluations of themselvesand others under many circumstances. A great deal ofresearch in social psychology has explored the ante-cedents and consequences of race-based stereotyping,prejudice, and discrimination (Brigham, 1971; Fiske,1998; Hamilton, 198 1). This research has demon-strated the important role of racial categorization in so-cial perception. The phenotypic facial characteristicswe possess determine the racial category membershipsthat are ascribed to us. These memberships have impli-cations for interpersonal thoughts, feelings, and behav-ior (Fiske, 1998; Fiske & Taylor, 1991; Hamilton,1981).

I thank Irene Blair, Kendrick Brown, Buju Dasgupta, EliotSmith, Linda Tropp, and two anonymous reviewers for their helpfulcomments on previous versions of this article.

Requests for reprints should be sent to Keith B. Maddox, Depart-ment of Psychology, Tufts University. Medford, MA 02155. E-mail:keith.maddox C)tufts.edu.

As the criterion for racial group membership, per-ceived phenotypic variation across individuals plays anextremely important in social perception. But our un-derstanding of this role may be incomplete. A numberof investigations have examined the possibility thatwithin-race variation in phenotypic appearance mayhave consequences for social perception beyond the"mere" determination of racial category membership.These investigations suggest that interpersonal out-comes are also based on the extent to which an individ-ual's physical facial characteristics resemble what isbelieved to be typical of a racial category. Racial groupmembers whose appearance most closely resemblesour representation of the "typical" category memberare more likely to be viewed through the lens of the cat-egory stereotypes and evaluations. This phenomenoncould be termed racial phenotypicality bias.

Using the literature as a barometer, most social psy-chologists seem unfamiliar with incidents of pheno-typicality bias distinguishing members of the same ra-cial category. In the past, the idea that within-categoryvariation in physical features could have a meaningfulinfluence on race perception had been either ignoredor rejected in the development of social psychologi-cal theory governing person perception. However, aglance beyond the disciplinary boundary reveals an ex-isting and expanding body of historical, anthropologi-cal, sociological, and medical evidence indicating thatwithin-race phenotypic variation makes a significantcontribution to social perception. The lack of attentionfrom social psychological theory is somewhat surpris-

383 at UNIV WASHINGTON LIBRARIES on July 20, 2009 http://psr.sagepub.comDownloaded from


MADD)OX

ing given that this approach has provided a number ofuseful insights on racial stereotyping, prejudice, anddiscrimination. Similar ideas have been explored withrespect to gender, suggesting that men described ashaving more "masculine" appearance and women withmore "feminine" appearance more closely resembletheir respective category stereotypes (Deaux & Lewis,1984). However, traditional social psychological per-spectives fail to account for the consequences that vari-ation in racial phenotypic appearance has for the inter-personal and societal outcomes of others.

The following review of research exploring racialphenotypicality bias suggests that the ways that socialpsychologists have thought about racial stereotyping inthe past can be elaborated by considering racial pheno-typicality. The review begins with a consideration ofhistorical and empirical evidence providing perspec-tive on the scope of the phenomenon. This evidence de-scribes a long history of discrimination and conflict,the existence of damaging, negative stereotypes, andmeasurable disparity in societal status among individu-als as a function of phenotypic appearance within ra-cial categories. The similarity to existing disparitiesbetween racial groups is striking, suggesting the use-fulness of a social psychological perspective. This isfollowed by discussion of several theoretical perspec-tives that have been offered to explain various aspectsof racial phenotypicality bias. These approaches offersuggestions for a rudimentary integrative frameworkthat can account for the existing data regarding the roleof racial phenotypicality in social perception. Finally,directions for future research that may be useful towardexploring the nature and impact of racial pheno-typicality bias are discussed. Although multiple racialand ethnic groups are considered, the review is skewedheavily toward research regarding Black Americansbecause this group has been a major focus of attention.Importantly, the evidence reviewed here covers onlythe social perceiver's viewpoint. Several investigationsthat have considered the target's perspective are out-side the scope of the current goals and therefore are notcovered (for a review, see Brown, Ward, Lightbourn, &Jackson, 1999).

A Brief History of RacialPhenotypicality Bias Research

Phenomena falling under the rubric of racialphenotypicality bias have been labeled Afrocentricbias (Blair, Judd, Sadler, & Jenkins, 2002), thebleaching syndrome (Hall, 1994, 1995), colorism(Okazawa-Rey, Robinson, & Ward, 1987; Russell,Wilson, & Hall, 1992), perceptual prejudice (Liv-ingston & Brewer, 2002), phenotyping (Codina &Montalvo, 1994), skin color bias (Hall, 1998), skintone bias (Brown et al., 1999; Maddox & Gray,

2002), and subgroup I)rejudice (Uhlmann, Dasgupta,Elgueta, Greenwald, & Swanson, 2002). All of theseterms reflect differential attitudes, beliefs, and treat-ment of individuals based on variation in phenotyp-ic characteristics of the face traditionally associatedwith particular racial categories. Some of these termsemphasize skin color or tone among those features,reflecting an assumption by the researchers that skintone (as opposed to variation in other features of theface) is the most salient or important dimension onwhich distinctions among group members may bemade. Accordingly, skin tone is often the only phe-notypic variable measured. Other terms do not reflectassumptions concerning the primacy of skin tone. Re-searchers adopting these terms tend to use globalmeasures of phenotypic appearance. In the followingreview, terms reflect the specificity of the measuresthat were used.

Historical and Anecdotal Accounts

Historical and anecdotal accounts of skin tone biasamong Black Americans abound (Drake & Cayton,1945; Frazier, 1957; Parrish, 1944; Russell et al.,1992).1 In general, these episodes suggest that the in-fluence of skin tone was widespread, extending tomany areas of life experience and achievement. Earlyhistorical evidence suggests that both Blacks andWhites in the United States exhibited bias based onskin tone as early as the slavery era. White Eurocen-tric facial features in Blacks were seen as evidence ofWhite ancestry, leading to inferences of racial superi-ority (Russell et al., 1992). After the abolition of slav-ery, lighter skin provided better social, educational,and economic opportunities (Neal & Wilson, 1989;Russell et al., 1992). During this time, employmentand educational opportunities were finally becomingavailable for all Blacks, but the intragroup manifesta-tion of skin tone bias became more evident. All-Black schools and social organizations used a varietyof methods to weed out undesirable applicants as afunction of skin tone. Typically, these methodssought to exclude darker Blacks from positions ofhigher status, thus maintaining the social, educa-tional, and economic distance between Blacks withlight and dark skin (Hall, 1992; Maddox & Gray,2002; Russell et al., 1992).

IBecause most of the empirical research has focused on BlackAmericans, I chose to reflect that same locus in the following reviewby first considering the historical context in which skin tone biases to-ward Blacks developed. Although a few studies that have examinedmembers of other racial categories are considered, their racial histo-ries in the United States are not elaborated here. I suggest that thesehistories share a similar theme that of racial discrimination in theUnited States. However, one should keep in mind that a considerationof the unique histories of these groups may be important toward devel-oping a comprehensive understanding of the phenomenon.



RACIAL PHENOTYPICALITY BIAS

Empirical Accounts: Evaluationsand Beliefs

In general, people have exhibited a preference forlighter skin and Eurocentric facial features in theirmates, friends, acquaintances, and, at times, eventhemselves. Lighter skin tone is preferred to darkerskin tone in general (Livingston, 2001; Livingston &Brewer, 2002; Porter, 1991; Seeman, 1946) and inclose relationships (Averhart & Bigler, 1997; Bond &Cash, 1992; Hill, 1944; Robinson & Ward, 1995; Ross,1997). Ross investigated Black college students' atti-tudes toward dating and mate selection. Men weremore likely to prefer lighter skinned mates in datingand marriage. Bond and Cash found that 70% of theirBlack college student participants indicated that theythought that men preferred lighter skin tone in theirmates.

People also tend to associate more positive person-ality traits to those with light skin, and negative char-acteristics to those with dark skin (Anderson &Cromwell, 1977; Bayton & Muldrow, 1968; Blair etal., 2002; Maddox & Gray, 2002, Study 2; Marks,1943; Sciara, 1971, 1983). Anderson and Cromwellsurveyed Black teenagers about their beliefs concern-ing positive and negative traits that correspond to in-dividuals having very light, light-brown, dark-brown,and black skin. Results showed that light-brown skinwas associated with positive characteristics (e.g., theprettiest skin, the smartest girl, the children fatherlikes best). Black skin was associated with more neg-ative characteristics (e.g., the person one would notlike to marry, the color one would prefer not to be,the dumbest person).

Averhart and Bigler (1997) examined Black chil-dren's memory for stereotype-related informationabout light- and dark-skinned Blacks. These partici-pants were read stories that included light- and dark-skinned Black characters paired with stereotypic andcounterstereotypic traits and occupations. They foundthat children were more likely to remember stories inwhich characters were portrayed in stereotype consis-tent ways (e.g., light-skinned Blacks with positive traitsand high-status occupations; dark-skinned Blacks withnegative traits and low-status occupations). Similar ev-idence was found in White participants' stereotypingof faces as a function of the degree to which faces wereperceived to reflect Afrocentric physical features(Blair et al., 2002). Further evidence comes from thefact that we may be explicitly aware of cultural stereo-types distinguishing Blacks as a function of skin tone.Maddox and Gray (2002, Study 2) found that bothBlack and White participants more closely associateddarker skinned Blacks with the negative cultural ste-reotype associated with the representation of Blacks.

Other investigations, although not focused directlyon skin tone, show that phenotypic variation also con-

tributes to implicit evaluations. Highly prototypic Blackfaces were associated with more negative implicit eval-uations than less prototypic faces (Livingston & Brew-er, 2002). Furthermore, highly prototypic Black facesprimed implicitly led to more negative judgments ofa race-ambiguous target than less prototypical Blackfaces (Livingston, 2001). Uhlmann et al. (2002) foundimplicit preference for less typical Hispanic faces overmore typical Hispanic faces among both light-skinnedand dark-skinned Hispanic participants. Furthermore,these beliefs may be exaggerated as a function of thecultural context in which they are measured. Uhlmannet al. found that implicit preferences for less pheno-typical faces was stronger among Chilean Hispanicscompared to Hispanics in the United States. However,individuals of different racial categories residing in thesame culture may tend to show similar patterns (Mad-dox & Gray, 2002, in press).

Stereotyping and prejudice based on phenotypic ap-pearance are not always revealed empirically (Atkin-son et al., 1996; Secord, 1959; Secord, Bevan, & Katz,1956). Atkinson et al. found no differences in Blackand White clinicians' diagnoses, trait ratings, or feel-ings toward a hypothetical light- or dark-skinned cli-ent. These findings suggest the existence of factorslimiting the expression of phenotypicality biases.

Empirical Accounts: Life Experiences

Sociological and anthropological research corrobo-rates historical accounts, showing that lighter skinnedBlacks enjoy higher status in society than darkerskinned Blacks (Edwards, 1959; Frazier, 1957; Free-man, Ross, Armor, & Pettigrew, 1966; Hill, 2000;Hughes & Hertel, 1990; Hunter, 1998; but see Kreiger,Sidney, & Coakley, 1998; Ransford, 1970; Seltzer &Smith, 1991). These findings have led some authors tosuggest that the socioeconomic status gap betweenlight- and dark-skinned Blacks in the United States isas large as the gap between Whites and Blacks (Hughes& Hertel, 1990; Hunter, 1998; Keith & Herring, 1991).In a study typical of this approach, Keith and Herringexamined data from the 1979-1980 National Survey ofBlack Americans and found that Blacks with lighterskin had higher educational attainment, occupationalstatus, and income levels than Blacks with darker skin.Furthermore, these relationships remain when control-ling for many other factors that would likely create oraugment status differences such as parents' socioeco-nomic status, gender, region of residence, urbanicity,age, marital status, and inheritance of wealth. Similarpatterns of social and economic stratification are evi-dent among Mexican Americans (Arce, Murguia, &Frisbie, 1987; Relethford, Stern, Gaskill, & Hazuda,1983; Telles & Murguia, 1990).

These findings suggest a continuing influence of ra-cial phenotypicality bias on interpersonal outcomes



MADDOX

that mirrors the impact of racial bias. Compared toWhites, Black Americans experience greater racialdiscrimination and are disadvantaged on many indica-tors of morbidity, mortality, and access to health care(Dressler, 1993). Similarly, darker skinned Blacks aremore likely than lighter skinned Blacks to report hav-ing experienced discrimination (Edwards, 1973; Keith& Herring, 1991; Klonoff & Landrine, 2000). Klonoff(2000) found that darker skinned Blacks reported moreracial discrimination in the past year and over their life-times. Darker skinned Blacks were also more likely toperceive the racism they experienced as stressful.Studies suggest that darker skin tone is associated withhigher diastolic and systolic blood pressure, perhapsdue to stress resulting from a higher incidence of dis-crimination (Gleiberman, Harburg, Frone, Russell, &Cooper, 1995; Harburg, Gleibermann, Roeper, Schork,& Schull, 1978). These findings are consistent with abiopsychosocial model that conceptualizes perceivedracism as a significant environmental stressor forBlack Americans resulting in a variety of negativephysical and psychological health outcomes (Clark,Anderson, Clark, & Williams, 1999). Similarly, dark-skinned Chicano men born in the United States showhigher levels of depression compared to their lightercounterparts (Codina & Montalvo, 1994).

Summary

The evidence reviewed here suggests the broadscope and impact of racial phenotypicality bias. With-in-race variation in skin tone and other aspects ofphenotypic appearance have a significant impact onour beliefs, feelings, evaluations, and treatment towardothers with consequences for their interpersonal, eco-nomic, and physical health outcomes. As phenotyp-icality moves away from a White Eurocentric norm,negative evaluations and outcomes increase. Whenconsidering the scope of the phenomenon, severalother tendencies are notable. First, the preference forlighter skin exists when evaluating both African andHispanic Americans. Second, this bias is revealed inthe perceptions of adults and children. Third, pheno-typicality biases are not limited to ingroup members'evaluations of each other-racial outgroup membersare also susceptible. Fourth, phenotypicality biasseems to be reflected in both implicit and explicit eval-uations of others. Fifth, perceived typicality of featuresis strongly related to the degree to which category ste-reotypes and prejudices are applied to a social target.Finally, there are occasions when racial categorizationoverrides more fine-grained distinctions based onphenotypicality. From an examination of the literature,these findings elude the grasp of traditional perspec-tives in social psychology. The next section considersthe role of phenotypicality in traditional models of so-cial representation and judgment.

The (Limited) Roleof Phenotypic Variation

in Social Psychological Theory

Evidence suggests that racial phenotype-basedperceptions influence the social, physical, and psycho-logical outcomes of others. This suggests that a frame-work of social judgment involving racial categoriza-tion, stereotyping, and prejudice derived from socialpsychological theory may be useful in understandingracial phenotypicality bias. However, models of cate-gory representation and impression formation haveplaced no particular emphasis on the role of with-in-race variation in phenotypic appearance. Still, thisfact does not preclude these models from providingsome perspective on the role of phenotype in socialperception.

Traditional Models of SocialRepresentation and Judgment

There are multiple social psychological models thatcan describe the implications of a particular mentalrepresentation for judgments about social targets.Many of these (e.g., Bodenhausen & Macrae, 1998;Kunda & Thagard, 1996) have their bases in two ex-tremely influential models of impression formation.The dual-process (Brewer, 1988) and continuum mod-els (Fiske & Neuberg, 1990), of impression formationhave been compared, contrasted, and extensivelytested in the literature (Brewer & Feinstein, 1999;Fiske, Lin, & Neuberg, 1999). Although theoreticallydistinct, both provide a great deal of explanatory poweracross the range of studies examining person percep-tion. The similarities among the two models, particu-larly with respect to category-based processing, pro-vide the focus for this discussion. Each model beginswith a stage of initial identification of the social tar-get's attributes that act as cues to salient category di-mensions such as age, sex, and race. If importance toself or motivation to process is low, subsequent catego-rization (or recategorization) occurs. Through a com-parison of the target attributes and a relevant mentalrepresentation, perceivers attempt to establish somedegree of fit between the target and the representation,with a bias toward determining subcategory member-ship. Once satisfactory fit is established, this represen-tation will guide the processing of information aboutand behavior toward the target. If fit is sufficientlypoor, perceivers will individuate the target, essentiallydeveloping a unique representation.

Traditionally, discussions of the role of skin tone instereotyping, prejudice, and discrimination have as-sumed that skin tone is simply one of many physical at-tributes that contributes to the determination of racialcategory membership. Subsequently, racial category-based stereotyping potentially determines the infer-




ences and judgments made about an individual. Thisidea was explicit in the writings of some theorists.Take, for example, the model of stereotype change pro-posed by Rothbart and John (1985):

By extension, individual Blacks may differ as to howprototypical they are as examples of Black People.What would be the criteria for determining the good-ness of fit of an individual to a social category such asBlacks? We suspect prototypicality for that group isnot strongly related to variations in "defining" attrib-utes, such as skin color, but rather to stereotypic attrib-utes already associated with the category. (pp. 89-90)

For these theorists, typicality of Blacks is deter-mined by personality traits and behavior rather thanvariation in physical attributes such as skin tone. Fiskeand Taylor (1991) expressed a similar belief in theircomprehensive and influential integration of the socialcognitive literature. "Once a person is categorized asBlack or White, male or female, young or old, the ste-reotypic content of the schema is likely to apply re-gardless of how much or how little the person lookslike the typical category member" (p. 121). A similarsentiment was expressed by Secord (1958), who statedthat "as long as an individual is categorized as a Negro,his place on a continuum of Negro-White featuresmakes little difference in impressions formed" (p. 303).Thus, skin color variation was considered to be of greatimportance to racial categorization. But its contribu-tion to representation and judgment beyond categori-zation was not deemed meaningful.

Seminal Investigations

The neglect of the role of within-race variation re-flected in social psychological research and theory isbased on early experimental research that failed to finda relation between phenotypic features and stereotyp-ing. Secord et al. (1956) were interested in whetherWhite participants associated stereotype-related per-sonality traits to dark- and light-skinned Blacks to dif-ferent degrees. Participants were asked to rate photo-graphs of Black and White men on personality andphysiognomic trait scales. Secord et al. predicted thatcategorization by race, not variation in features, woulddetermine stereotyping. As predicted, they found thatthe presence of Eurocentric phenotypic features inphotographs did not reduce racial stereotyping. Appar-ently, White participants did not make distinctionsbased on phenotypic features and lightness or darknessof skin tone: Once a target was categorized as Black, hewas stereotyped according to that categorization.

Secord (1959) conducted a follow-up experiment toaddress the possibility that the influence of phenotypicvariation in Blacks may have been masked by the pres-ence of Whites in the rating task. In this study, half of

the participants were assigned to make ratings of theBlack photographs alone. The other half were assignedto make ratings of those photographs, as well as photo-graphs of Whites interspersed among the Blacks. Incomparing these two groups, there were no differencesin stereotype use based on the skin tone of the Blacktargets. The absence of the White target contrast cate-gory did not facilitate greater within-category differen-tiation of Black targets, supporting the previous evi-dence that skin tone variation was not meaningful toWhite observers.

The conclusions of these investigations reflect theperceived role of phenotypic variation in influential so-cial psychological theories of category representationand impression formation. Although this may havebeen the dominant perspective, it was not necessarilyubiquitous. Zebrowitz (I1996) was not specifically con-cerned with racial phenotypic variation in discussingher ecological theory of social perception and its rela-tion to stereotyping. Yet, she explicitly supported theidea that "physical qualities can serve as a basis for so-cial impressions in the absence of explicit categoriza-tion processes, and that variations in physical qualitiesshould lead to within-category variations in social im-pressions" (p. 109, Footnote 1).

Racial Phenotypicality Revisited

The Secord investigations are unique and importantin their focus on phenotypic variation and stereotypingof Blacks. However, their conclusions stand in contrastto the majority of published empirical research sincethat time documenting various consequences of racialphenotypicality bias. The conclusions drawn fromthese seminal investigations have been critiqued else-where and are not repeated here (Blair et al., 2002;Maddox, 1998; Zebrowitz, 1996). One example mayilluminate a general concern. In the Secord (1959)study, there was evidence that the Black targets withthe most Eurocentric appearance were often catego-rized as White, not Black. After completing the ratingtask, participants were asked to provide their opinionof racial group membership. The photographs of thethree Black targets with the most extreme Europeanfeatures were categorized as White by 33 of 39 partici-pants. Follow-up analyses showed that the participantswho thought the person in the photograph was Whitestereotyped them less than the four participants whothought the person was Black. The article did notclearly state how the ratings of these incorrectly cate-gorized photos of Blacks compared with ratings ofWhite photos, although their similarity is implied.However, it does state that the four participants whocategorized those photos as Blacks stereotyped themto the same degree as the most unambiguously Blackphotograph. Thus, the study as a whole does not ad-dress the situation where Blacks of varying skin



MADDOX

tones are spontaneously (and "correctly") categorizedas Black.

Consider an alternative interpretation: When partic-ipants were allowed to work with their spontaneouscategorizations, many "incorrectly" perceived the pho-tographs to be White and stereotyped them accord-ingly.2 If so, the study does not address the perceptionof photographs of light-skinned Blacks that are sponta-neously categorized as Black. More important, thesecriticisms c-all into question the empirical underpin-nings of the theoretical discussion outlining a limitedrole of skin tone in social psychological judgmentprocesses.

Alternative Perspectiveson Racial Phenotypicality Bias

The evidence reviewed here suggests that racial cat-egorization does not necessarily imply a terminal stagein processing race-related phenotypic features for so-

cial perception and judgment. In light of this growingbody of cross-disciplinary evidence, several perspec-

tives have been proposed, or could be applied, to rec-

oncile the divergence of social psychological theoryand social reality. These approaches, focusing on theareas of category representation, impression forma-tion, and social judgment, have explored how variationin phenotypic appearance may determine interpersonaltreatment and outcomes. This section explores eachperspective, describes evidence concerning points oftheoretical distinctiveness, and discusses strengths andweaknesses of each perspective in accounting for theavailable data.

Skin Tone Bias

Theory

Maddox and Gray (2002) proposed that discrimina-tion based on skin tone and phenotypic features can beexplained with existing models of category representa-tion (e.g., Brewer & Feinstein, 1999; Fiske et al.,1999). This perspective was not elaborated in that arti-cle, so it is described in some detail here.

The primacy of skin tone. This perspective em-phasizes the assumption that skin tone is of primary sa-

lience and importance among the many phenotypicfeatures used to determine racial category member-ship. The important role of skin tone is supported bythe language choices in the Black and Hispanic com-

munities. Black Americans who display the tendencyto use phenotype to differentiate within a racial cate-gory are sometimes referred to as "color struck," "colorconscious," or said to have a "color complex" (Oka-zawa-Rey et al., 1987; Russell et al., 1992). Theseterms suggest that variation in skin tone is primarilyimportant. Blacks often use terms such as "light-skinned" and "dark-skinned" to describe one another(Maddox & Gray, 2002; Russell et al., 1992). This isalso true of Hispanics, who use the terms "blancos"and "morenos" to distinguish between lighter anddarker category members (Uhlmann et al., 2002).These terms emphasize skin color, but also capture per-ceived correlations between skin color and other fea-tures. The general category "race" and specific grouplabels like "Black" and "White" act as proxies not onlyfor color, but also for dimensions of the nose, hair, andlips. Similarly, labels such as "light-skinned," "blan-co," "dark-skinned," and "moreno" bring to mind indi-viduals who differ not only in skin tone but also inother physical features.

Considering phenotypicality. The model beginswith a stage of initial identification of the target's at-tributes (e.g., phenotypic features, clothing, behavior)that act as cues to salient category dimensions such as

age, sex, and race. Perceivers compare these attributesto a salient mental representation (e.g., race) with thegoal of determining goodness of fit, with a bias towarddetermining subcategory membership. Once satisfac-tory fit is established, this representation will deter-mine the processing of information and behavior. Tar-gets whose physical characteristics are more typical ofthe category will be more likely to be viewed throughits conceptual lens. Less typical targets are also viewedthrough that lens, but with less conceptual clarity. Inother words, the same category stereotypes and preju-dices will be applied, but to a lesser extent. If fit be-tween the racial category and the target is sufficientlypoor, perceivers will individuate the target, essentiallydeveloping a unique representation. Cognitive psy-

chologists have long recognized the important role ofphysical typicality in object category representation(e.g., Medin, 1989). Understandably, social psycholo-gists have focused on the role of behavioral typicalityin past formulations (e.g., Rothbart & John, 1985).This model suggests an integration of the two empha-ses racial phenotypicality is also an important con-

tributor to the process of social representation andjudgment. The degree of category fit increases as a per-

son's phenotypic appearance becomes more typical ofthe racial category representation. Assuming that dark-er skin tone is more typical of the representation ofBlacks, the darker a Black person's skin tone, the morelikely he will be viewed through the lens of the culturalstereotype (i.e., characteristics associated with the rep-resentation of Blacks). Because the cultural stereotype

388

2Curiously, when these same photographs were labeled as Black,the labels did not facilitate use of the stereotype associated withBlacks. This finding contradicts similar research examining the in-fluence of ethnic labels (e.g., Raz.ran, 1950).




of Black Americans is predominantly negative, Blacksshould also be perceived more negatively as skin tonedarkens.

Phenotype-based subcategories. In addition tosuggesting a linear relation between goodness of cate-gory fit and stereotyping, the skin tone bias model alsosuggests the existence of salient, phenotype-based sub-categories. In traditional models, perceivers use targetattributes to establish fit with meaningful, salient sub-types ofthe superordinate category based on social rolesand occupational status (Devine & Baker, 1991) orstereotypicality of their behavior (Rothbart & John,1985). Subtyping, the clustering of individuals bywhether they confirm or disconfirm the stereotype, hasbeen linked to the perception of typicality of groupmembers (Maurer, Park, & Rothbart, 1995; Park, Wol-sko, & Judd, 2001). In this model, subtyping based onphenotype may similarly occur through recategoriza-tion processes (cf. Fiske & Neuberg, 1990). Less pheno-typical targets are more likely to be subcategorized. As aresult, the category representation of Blacks can be dif-ferentiated by various subcategories of group membersbased on their phenotypic appearance.

Essentialist theories of racial category member-ship. Beliefs about skin tone and physical featuresare proposed to originate from our naive biologicaltheories about racial category membership. Our cur-rent system of racial classification was developedlargely based on perceived variation in skin color andother physical dimensions of the face and body (Gould,1994). However, genetic and anthropological researchsuggests that any understanding of human racial cate-gories as a biological certainty is flawed. More accu-rately, racial categories arise as a function of socialconstruction processes (American AnthropologicalAssociation, 1998; American Association of PhysicalAnthropologists, 1996). In part because of our unchal-lenged education concerning the biological bases ofrace, this social construction of race has a very stronglay theoretical basis in biological thinking. Coupledwith any knowledge of distinct cultural practices, peo-ple are disposed to believe that external physical andbehavioral differences between groups are correlated,each reflecting stable genetic differences (Eberhardt,Dasgupta, & Banaszynski, 2003; Haslam, Rothschild,& Ernst, 2002; Hirschfeld, 1996; Rothbart & Tay-lor, 1992). This idea is consistent with theories andevidence in the cognitive literature suggesting that im-plicit causal theories are an important component ofmany of our category representations (Keil, 1989;Medin, 1989; Medin & Ortony, 1989). Implicit causaltheories represent perceiver's beliefs about causal rela-tions among a target's attributes. From an essentialistor implicit causal theory perspective, people are likelyto assume that increasing darkness of skin is a reflec-

tion of a person's genetic makeup and/or culturalheritage. In the United States, legal definitions to deter-mine Black racial category membership hinges onknowledge of a person's Black racial ancestry. His-torically, this was known in the South as the "one droprule," meaning that a single drop of Black blood wassufficient to make someone Black (Davis, 1991). Aperson with dark skin tone might be assumed to havetwo Black parents, whereas light skin tone may suggestone Black and one White parent. These beliefs couldinfluence both biological and social notions of race. Abiological construction of race would lead a perceiverto assume that someone with darker skin tone is more'ipurely" Black. A social construction of race may leadanother perceiver to infer that being raised by mixedrace parents might dilute a "purely" Black cultural up-bringing. In other words, if one drop of Black bloodmakes you Black, two drops makes you Blacker, bothgenetically and culturally.

Evidence

Existence of phenotype-based subcategories.The category-based model also suggests that degree ofphysical typicality is associated with the formation ofsubcategories. Maddox and Gray (2002, Study 1) ex-amined the antecedents of skin tone-based stereotyp-ing using the category confusion paradigm (Taylor,Fiske, Etcoff, & Ruderman, 1978). The results of thememory task revealed that participants made morewithin-skin tone than between-skin tone errors, sug-gesting that they attended to, encoded, and used theskin tone of the discussants in making their statementassignments. Furthermore, contextual factors may op-erate to make skin tone more or less salient. Guided bythe model of social category salience proposed byBlanz (1999), which considers both person and contex-tual factors, Maddox and Gray (in press) used the cate-gory confusion paradigm to investigate contextualinfluences on the salience of skin tone-based subcate-gories. Two experiments demonstrated that a manipu-lation of issue relevance during a group discussionamong light- and dark-skinned Blacks enhanced thecategory salience of skin tone, revealed in the patternof within- and between-category errors. Discussiontopics relevant to the racial politics of skin tone (racerelations in the United States) led to increases in the sa-lience of skin tone compared to neutral topics (poten-tial leisure activities or environmental issues). Thesefindings should be interpreted with caution. Evidencefor within- and between-category confusions does notprovide direct evidence of categorization (or subcat-egorization) because participants were never asked tocategorize the photographs. Nonetheless, there are rea-sons to believe that these data have implications for therole of skin tone in the mental representation of BlackAmericans. The confusion measure used here has been



MADDOX

found to be a reliable indirect measure of subtyping. Asimilar pattern of confusion errors occurs when partici-pants are given explicit instructions to sort individualsusing a subtyping process (Park et al., 2001).

Casual role of skin tone. Simply asking partici-pants to consider Blacks described as light-skinnedand dark-skinned is sufficient to elicit differences inperceived cultural stereotypes (Maddox & Gray, 2002,Study 2). As for the causal role of skin tone in socialperception, encoding and use of skin tone (Maddox &Gray, 2002, Study 1; in press) occurred in the contextof efforts to manipulate skin tone while holding con-stant other facial characteristics that may naturallycovary with skin tone. In those studies, the skin tone ofthe discussants was manipulated digitally across multi-ple replications. However, no direct evidence for thecausal role of skin tone has been offered in studies ex-amining impression formation and judgment of indi-viduals belonging to the same racial category. At-kinson et al. (1996) used a digital manipulation of skintone, yet found no differences in Black and White cli-nicians' diagnoses, trait ratings, or feelings toward ahypothetical light- or dark-skinned client. Dixon andMaddox (in press) also used a digital manipulation ofskin tone and race among a varied-race sample of par-ticipants. There were no differences in impressionjudgments as a function of skin tone. They did find thata brief exposure to a photo of a dark-skinned Blackperpetrator during a crime story led to more emotionaldiscomfort about the story than exposure to a Whiteperpetrator associated with the same story. No differ-ences surfaced when comparing dark Black and lightBlack perpetrators, or when comparing light Black andWhite perpetrators. Dixon and Maddox suggested thatcompared to the light-skinned Black perpetrator, thedark-skinned perpetrator achieved a certain thresholdto activate category stereotypes associating Blacks andcriminality. Bolstering this idea was the finding thatthe aforementioned results distinguishing dark Blackand White targets occurred only among heavy televi-sion news viewers who are more likely to be exposed toBlacks as criminals (Dixon & Linz, 2000). So, al-though there is evidence that supports the causal role ofskin tone in category representation studies and prim-ing of emotional concern, there is no existing supportfor its causal role in impression judgments. Future evi-dence of this nature is crucial to determining just howimportant skin tone is for racial phenotypicality bias.

Implicit causal theories. Only one study has at-tempted to delineate the importance of various pheno-typic features for racial categorization. When asked,participants emphasize skin color among other aspectsof phenotypic appearance in determining racial groupmembership (Brown, Dane, & Durham, 1998). Othershave attempted to measure individual differences among

those who link phenotypic appearance and racial cate-gory membership to stable genetic or cultural causes.Livingston and colleagues have found more extremedifferences in automatic evaluation (Livingston & Brew-er, 2002) and susceptibility to priming effects (Liv-ingston, 2001) for low- versus high-prototypic Blackfaces among individuals who scored high on a scalemeasuring reliance on perceptual cues in impressionformation. Eight of the nine items on this scale reflectphenotypic qualities (facial and vocal) and half ofthose items refer to race or ethnicity. Thus, this scaleseems to include some consideration of the race theo-ries of the respondents suggesting the importance oflay theories of race in social perception. However, otherresearch has failed to find a strong relation betweenessentialist thinking and racial prejudice (Haslam et al.,2002). It may be the case that the consideration of racialphenotypicality is important in linking essentialist be-liefs with racial stereotyping and prejudice.

Afrocentric Bias

TheoryBlair et al. (2002) also proposed that traditional

models of social perception fail to recognize the influ-ence of within-category phenotypic variation. Thismodel suggests that the presence of Afrocentric fea-tures contribute to social perception above and beyondthat of category-based processing. In this view, a pro-cess of reverse generalization characterizes stereotyp-ing and prejudice directed toward individuals as afunction of race-related phenotypic features. Racialphenotypicality bias represents a direct linkage be-tween Afrocentric appearance and cultural stereotypesabout Blacks. This association is created and fortifiedthrough the repeated pairing of dark skin tone, broadnoses, full lips, and short, tightly curled hair with thestereotypes about the group members who possessthose features. As a result, those individuals who ex-hibit Afrocentric features, regardless of group mem-bership, may come to be seen through the lens of theBlack American stereotype.

Evidence

Unique contributions of category- and feature-based processes. Blair et al. (2002) examined therole of Afrocentric features and perceived racial typi-cality in stereotyping of Blacks and Whites. In the firststudy, White participants were asked to rate photo-graphs of Black and White men for the presence ofAfrocentric facial features (dark skin, broad nose, fulllips, and coarse hair). Across three additional studies,separate groups of participants rated subsets of thesephotographs for their degree of fit with descriptive sce-narios that varied in their valence and stereotypicalityof traits distinguishing Blacks and Whites. In addition,




the presence of Afrocentric features contributed tostereotyping and prejudice even when the effect of cat-egory accessibility was controlled. This finding sug-gests that both category-based process and feature-based processes are useful in explaining judgmentsbased on racial phenotype.

Direct linkage between Afrocentric features andcultural stereotypes. The results of Blair et al.(2002) consistently revealed that positive and negativestereotyping of photographs of both Blacks and Whiteswas positively correlated with the presence of Afro-centric features. Stereotyping of Whites on stereo-typically Black personality traits must reflect a directlinkage between those features and the traits of thegroup most closely associated with them. However, theprocess linking features to traits seems to be asymmet-rical. The descriptions that were used in this investiga-tion were constructed from previous investigations thatcontrasted stereotypic conceptions of Blacks relativeto Whites. Thus, descriptions that were counterstere-otypic of Blacks were stereotypic of Whites and viceversa. Blair et al. reported that ratings of Eurocentricfeatures in White faces did not predict stereotyping ofWhites in Study 3. It is not clear why the presence ofEurocentric features in Black and White faces does notpredict stereotyping, but the presence of Afrocentricfeatures does.

Perceptual Prejudice

Theory

Livingston and Brewer (2002) proposed a theorythat governs automatic responses to faces of Blacks.Similar to the Afrocentric bias perspective, this theorysuggests that the influence of phenotypic variation isnot mediated through conceptual knowledge aboutBlacks. Instead, the influence occurs through cue-based, affective responses to the physical features ofthe face. This perspective asserts that faces and the fea-tures they possess do not necessarily activate the socialcategories to which they may belong. According to thedual-process model, the identification stage reflectssimple, automatic recognition of perceptual features.Therefore, tasks that measure automatic responses tofaces are necessarily cue-based (evaluative), but notcategory-based (semantic). Automatic evaluations offaces that are highly typical of the category should bemore negative than evaluations of faces that are lesstypical.

Livingston and Brewer (2002) proposed two pro-cesses to explain the linkage between highprototypicality and negative evaluations. In the first,they suggest that initially strong prejudices against theracial category as a whole may become dissociatedfrom category-based beliefs and attitudes over time.

Early emotional conditioning of negative affect withBlacks will be less intense as cues to racial categorymembership become weaker (i.e., less typical). Thismechanism suggests that affective information associ-ated with the larger racial category becomes associatedwith less prototypical members of the racial category.Livingston and Brewer also posited a familiarity-basedmechanism for the origins of perceptual prejudice. Un-familiar stimuli may elicit strong anxiety and discom-fort that might be expected to decrease with additionalexposure. They suggest that Whites might be more fa-miliar with low-prototypic Blacks than high-proto-typic Blacks for several reasons. Whites living in rela-tively racially segregated communities may get morefrequent exposure to positive images of low-prototypicthan high-prototypic Blacks through media and enter-tainment sources. Furthermore, given that skin tonehas been demonstrated as a limiting factor in social andeconomic mobility, Whites might be more likely to en-counter low-prototypic (i.e., light-skinned) Blacksthan high-prototypical Blacks in occupational and edu-cational settings. Any of these processes would perpet-uate an existing tendency to favor low-prototypic overhigh-prototypic Blacks.

Evidence

Automatic evaluation of low- and high-proto-typic faces. In a series of experiments, Livingstonand Brewer (2002) presented participants with a se-quential priming paradigm that measured affective re-actions to Black faces that were high and low in racialprototypicality. In two experiments, automatic evalua-tive responses indicated negative evaluations of Blackfaces that were high in racial prototypicality comparedto those that were low in racial prototypicality (Experi-ments 1 & 5). Similar evidence using Hispanic faces inan implicit association task was discussed earlier (Uhl-mann et al., 2002).

Failure to activate semantic knowledge. Evi-dence from another experiment (Livingston & Brewer,2002, Experiment 3) suggests that skin tone and phe-notypic facial characteristics elicit preconscious cue-based processing prior to or instead of category-basedprocessing. Under task goals designed to encourage cat-egory activation, participants completed a conceptualjudgment task involving stereotype-relevant words in-stead of the evaluative task used in other experiments.Responses to these words were not affected by a manip-ulation ofracial prototypicality-a finding that suggestssemantic category knowledge was not recruited. Al-though plausible, this point may be subject to some de-bate. Livingston and Brewer present convincing evi-dence that automatic categorization and stereotypingdoes not necessarily occur automatically. But this doesnot mean that it cannot occur under the right conditions.



MADDOX

The results ofExperiment 3 also failed to show evidenceofautomatic stereotyping based on race. This representsa failure to replicate a number of investigations that havedemonstrated that automatic responses to facial stimulican in fact reflect category-based processing throughmanipulations of processing goals (e.g., Macrae,Bodenhausen, Milne, Thorn, & Castelli, 1997; Wit-tenbrink, Judd, & Park, 1997, 2001). This raises ques-tions about the interpretation of the results.

Direct linkage between prototypical features andgroup-based evaluations. There is no direct evi-dence that suggests the existence of this linkage. In the-ory, it is very similar to the feature-trait conditioningprocess proposed by Blair et al. (2002) where bothevaluative and conceptual knowledge become directlyassociated with physical features typical of Blacksover time. However, the process described here sug-gests that only evaluative knowledge becomes asso-ciated with prototypical features. It is unclear howthis process would dissociate affective from semanticknowledge as Livingston and Brewer (2002) suggest.

Increased familiarity with low-prototypic Blacks.Again, no direct evidence exists. Although this mecha-nism seems plausible, some elements seem counter-intuitive. For example, although low-prototypic Blackshave been regarded more positively than high-proto-typic Blacks, it seems unlikely that these individualswould be more familiar as a function of mere exposure.Whites are readily exposed to both negative images ofhigh-prototypic Blacks and positive images of low-prototypic Blacks in the media. Whereas educationaland occupational settings may also provide dispro-portionate (and greater) exposure to low-prototypicBlacks, other settings provide disproportionate expo-sure to high-prototypic Blacks. In addition, low-proto-typic Blacks likely make up a smaller proportion of theBlack population than those with darker skin tone.Using data collected in large sample studies, one mightinfer that low-prototypic (specifically light-skinned)Blacks comprise between 14-21% of the Black Ameri-can population (Hill, 2000; Hunter, 1998; Keith & Her-ring, 1991; Ransford, 1970). Thus, statistically, it isless likely that anyone will encounter a Black personwith lighter skin tone. Furthermore, this mechanismwould not be able to explain Black perceivers' prefer-ences for lighter skin tone. Even if Whites may havegreater familiarity with light-skinned Blacks, there isno reason to believe that dark-skinned Blacks wouldalso have greater familiarity with light-skinned Blacks.

Physical Attractiveness

Theory

Others have explored the hypothesis that differen-tial outcomes among Black Americans are based on

perceptions of physical attractiveness (Breland, 1998;Wade, 1996). Compared to unattractive people, attrac-tive people are generally associated with positive per-sonality characterizations (Dion, Berscheid, & Wal-ster, 1972) with some negative caveats (Dermer &Thiel, 1975). In other words, what is beautiful is good,but self-centered. From this perspective, lighter skintone and Eurocentric features imply attractiveness,leading to perceptions of higher status, financial suc-cess, vanity, lack of sympathy, and arrogance. This ex-planation suggests that the inferences made aboutlight-skinned persons may reflect attractiveness ste-reotyping rather than racial stereotyping.

Evidence

Correlation between skin tone and physical at-tractiveness. Anecdotally, techniques to makephysical features less Afrocentric and more Eurocen-tric such as hair straightening, skin bleaching, coloredcontact lenses, and plastic surgery are often employedby Blacks and others to augment physical attractive-ness. Consistent with anecdotal accounts, recentlypublished data suggested that there is a correlation be-tween perceived attractiveness and skin tone. Usingdata from the 1979-80 NSBA, Hill (2002) examinedratings of physical attractiveness and skin tone madeby the study's Black interviewers. Results indicatedthat lighter skin tone was related to higher ratings ofphysical attractiveness for men and women. This rela-tion was strong and monotonic for women; ratings ofattractiveness steadily increased from "very dark" to"very light." The relation for men was weaker, but fol-lowed the same general pattern. In addition, researchhas shown that Black and White men, as well as menacross various cultures, find women with a more Euro-centric appearance to be more attractive than womenwith a more Afrocentric appearance (Cunningham,Roberts, Barbee, Druen, & Wu, 1995; Hamilton &Trolier, 1986; Martin, 1964). However, empirical evi-dence for the perceived relation between lighter skintone and greater physical attractiveness is not uniform.Blair et al. (2002) found the reverse relation in Whitemale and female participants' ratings of photographsof men, possibly indicating divergence in the conse-quences of phenotypic variation for men and women(Hunter, 1998; Maddox, 1998).

Stereotypes and prejudice as a function of physi-cal attractiveness. Assuming that lighter skin toneis perceived as more physically attractive, perceived at-tractiveness may explain the evidence favoring light-versus dark-skinned Blacks. In one study, ratings ofphysical attractiveness were associated with family in-

come levels that were three or more times greater thanthe poverty level (Hill, 2002). Similar to attractive tar-gets, positive characteristics have also been attributed



RACIAL PHENOlTYPICALITY BIAS

to light-skinned Blacks (e.g., Anderson & Cromnwell.1977). The negative characteristics of attractive targetsare consistent with historical evidence that suggeststhat light-skinned Blacks have enjoyed higher statusand, adopting a bourgeois attitude, have acted to main-tain a status differential through discriminatory prac-tices (e.g., Russell et al., 1992). Despite this evidence,physical attractiveness does not predict racial stereo-typing in judgments of faces. Blair et al. (2002) foundthat ratings of physical attractiveness predicted generalevaluation of faces, but not racial stereotyping. Onlycategory accessibility and Afrocentric appearance pre-dicted evaluation and stereotyping. Automatic evalua-tion effects distinguishing high- and low-prototypicfaces reported by Livingston and Brewer (2002) wereobtained when controlling for perceived attractivenessof faces, suggesting unique contributions of pheno-typicality and attractiveness. Finally, it's not clear thatdifferences in physical attractiveness can explain themagnitude of disparity in social and economic out-comes. Hill (2002) found that, compared to skin tone,perceived physical attractiveness was only weakly as-sociated with educational attainment among Blacks.

The Ecological Perspective

Theory

Zebrowitz and her colleagues (McArthur & Baron,1983; Zebrowitz, 1996) explored an ecological per-spective to social perception. From this perspective, in-formation such as physical appearance, movements,and vocal characteristics provide useful and accurateinformation about a person's underlying qualities anddisposition. These qualities provide us with informa-tion about a person's behavioral affordances charac-teristics that communicate the boundary conditionsgoverning how stimulus and perceiver may interact.Group stereotypes may be explained by the fact that theinformation that these stimulus qualities convey can beovergeneralized to other individuals who share thosequalities. With respect to skin tone and other aspects ofphenotypic appearance, this perspective suggests thatvariation in these features conveys meaningful infor-mation that can be overgeneralized to other individualsas a function of their phenotypic appearance.

Evidence

Zebrowitz (1996) effectively argued that variationin physical appearance, specifically the degree of ba-byish appearance, can account for the content of a vari-ety of group stereotypes. Although no explicit efforthas been made to examine this theory with respect tostereotyping as a function of racial phenotype, it isplausible. One crucial tenet of the ecological theory isthat, at their core, these perceptions are also accurate.This remains an open question due to the fact that accu-

racy is rarely addressed in empirical research (Lee,Jussim, & McCauley. 1995) and difficult to measure(Judd & Park, 1993). Thus the plausibility of the eco-logical perspective rests on future research examiningthe accuracy of information provided by race-relatedphenotypic features. But it has certainly been demon-strated that even inaccurate perceptions can have self-fulfilling effects on behavior (e.g., Chen & Bargh,1997; Word, Zanna, & Cooper, 1974). Determining ac-curacy may be difficult when considering the (some-times) self-fulfilling nature of social perception.

Metaphorical Associations

Theory

In contrast to the bottom-uip process described bythe ecological perspective, other approaches empha-size how existing conceptual knowledge from otherdomains may determine how we interpret the informa-tion provided by facial features in social perception.These approaches have not been specifically examinedwith respect to racial phenotypicality bias. However,they suggest that our positive associations with lightercolors and negative associations with darker colorsmap onto our beliefs about individuals with lighter ordarker skin tones. Gergen (1967) highlighted cross-cultural evidence suggesting that white and light areusually associated with things good, whereas black anddark are associated with things bad. Zebrowitz (1996)suggested that cultural associations with various colorsbecome transferred metaphorically to our beliefs aboutracial categories (for a similar argument, see Gergen,1967; Secord, 1958). This metaphorical overgener-alization perspective is similar to one proposed by Wil-liams, Boswell, & Best (1975), who provided insightconcerning the origin of our light and dark color asso-ciations. They suggested that these associations stemfrom the diurnal nature of human beings and our expe-riences during early childhood. Children's early learn-ing experiences teach them that daytime is associatedwith clarity and safety, whereas nighttime is associatedwith obscurity and danger (see also Gergen, 1967;Goldberg, 1973).

Evidence

Color bias and person perception. Anecdotally,the tendency to make discriminations based on skintone emerges among people of many different ethnici-ties in the United States, as well as in a number ofcountries around the world including Japan (Wagat-suma, 1967), Northern Africa (Brown, 1967), SouthAfrica (Legum, 1967), and India (Beteille, 1967). Wil-liams and his colleagues have conducted extensivecross-cultural research that suggests negative connota-tions with the color black and positive connotationswith the color white among children in the United



MADDOX

States, Europe, and Asia (Best, Field, & Williams,1976; Iwawaki, Sonoo, Williams, & Best, 1978; Wil-liams, Boswell, & Best, 1975; Williams, Morland, &Underwood, 1970). Empirical research by Williamsand others also suggests that adults and children in theUnited States evaluate human figures with black skinor darker black skin tones more negatively than indi-viduals with white skin or lighter black skin tones (e.g.,Filler & Williams, 1971; Williams, Williams, & Beck,1973). Similar evidence comes from doll studies exam-ining children's preferences for lighter or darker skin.These studies usually, but not always, equate skin colorwith race. In a study typical of this genre, Clark andClark (1958) found that Black and White children weremore likely to show preferences for White dolls and re-ject Black dolls.3 These approaches can explain eval-uative distinctions as a function of skin tone. However,they cannot alone explain differences in the degree towhich stereotypes are applied.

Rapprochement

General Conclusions

The perspectives discussed here vary in their abilityto account for phenomena associated with racial pheno-typicality bias. Although none stands out as a clearfront-runner, the existing evidence does not yet warrantelimination ofany ofthe various perspectives from con-sideration. Each perspective attempts to explain differ-ent aspects ofracial phenotypicality bias. Each perspec-tive reveals strengths and weaknesses in accounting forthe available data. This section considers these strengthsand weaknesses, drawing several conclusions from theexisting evidence. It next considers a rudimentary theo-retical model of racial phenotypicality bias and offerssome suggestions for research that may help to illumi-nate aspects of this model.

Racial Phenotype-BasedStereotyping and Prejudice Exists

Individuals with racial phenotypes that are increas-ingly typical of a particular racial category are morelikely to be viewed through the lens of category-spe-cific beliefs and evaluations. Furthermore, sensitivityto variation in racial phenotype is not limited toperceivers who are members of that racial category.These points converge with the skin tone bias, Afro-centric bias, and ecological perspectives. The physicalattractiveness perspective diverges from this point. It

3This research paradigm has been criticized for methodologicalartifacts potentially limiting the interpretation of the results (Banks,1976). More contemporary research using this paradigm has foundthat this tendency still holds, although it is subject to mitigating in-fluence, such as portraying Blacks in positive roles (Powell-Hopson& Hopson, 1988).

assumes that impressions are based on stereotypesabout physical attractiveness rather than racial cate-gory membership. Perceptions of physical attractive-ness may influence, but cannot wholly explain, racialphenotypic bias. The metaphorical association per-spective suggests that evaluations, not beliefs, mayvary as a function of phenotype. Thus, given evidenceof racial phenotype-based stereotyping, the metaphor-ical perspective can also be ruled out as a sole contributorto this process. This is also problematic for the percep-tual prejudice perspective, which discounts semanticassociations with phenotypic appearance. However,this perspective does not make predictions about con-trolled judgments. It was introduced to explain rela-tively automatic responses to faces.

Automatic Evaluation Differsas a Function of Racial Phenotype

The perceptual prejudice perspective alone pre-dicts that variation in phenotypic appearance can in-fluence automatic evaluative responses to categorymembers, but not semantic responses. This is some-what inconsistent with the skin tone bias andAfrocentric bias perspectives, which would predictthat variation in phenotype would elicit bothevaluative and semantic knowledge associated withthe racial category the target's features resemble.However, it may be possible that evaluations governrelatively automatic judgments whereas beliefs influ-ence relatively controlled judgments.

Racial Phenotypicality Contributesto Racial Subtyping

The skin tone bias perspective alone suggests thatracial phenotypicality can be a cue to the formationand use of phenotype-based subcategories (e.g., light-skinned and dark-skinned Blacks) and that the salienceof phenotypic-based subcategories can be influencedby the perceptual context. This perspective also sug-gests that individual phenotypic features, (e.g., skintone) are more important than others are in determin-ing racial typicality. This idea does not conflict withany of the perspectives.

Racial Phenotypicality AffectsCategory-Based and Feature-Based Processes

The skin tone bias perspective suggests that pheno-typicality influences category-based judgments. TheAfrocentric bias and ecological perspectives each sug-gest that racial phenotypicality can also influenceresponses to individuals (category members and non-category members alike) outside of traditional cate-gory-based routes through direct feature-trait associa-tions. The two perspectives differ in the mechanism of

394




that influence. The Afrocentric bias perspective sug-gests that the association with a target group's physicalfeatures and their personality traits is conditioned overtime through activation of the group stereotype, where-as the ecological perspective suggests that this connec-tion an inherent property of the physical features them-selves and contribute to the stereotype. Importantly,these perspectives are not inconsistent with one an-other; both processes may occur.

Conceptual Knowledge InfluencesRacial Phenotypicality Bias

Each of the perspectives attempts to explain possi-ble origins of racial phenotypicality bias. The skin tonebias perspective emphasizes the importance of essen-tialist beliefs about racial category membership. TheAfrocentric bias and perceptual prejudice perspectivessuggest conditioned associations between physicalfeatures and personality traits mediated by racial cate-gory beliefs and/or evaluations. The physical attrac-tiveness perspective suggests that cultural standards ofbeauty govern judgments based on racial phenotype.The metaphorical perspective suggests that associa-tions with colors or early childhood experiences withday and night are applied to people. The ecologicalperspective describes how the behavioral affordancesof physical features are overgeneralized to other cate-gory members. Although the veracity of each approachmay be questioned in future research, the common ele-ment among them is the idea that conceptual knowl-edge concerning physical characteristics either im-plicitly or explicitly influences representations andjudgments about members of a racial category.

A Rudimentary Model ofRacial Phenotypicality Bias

Considering the available evidence, the followingsection explores a rudimentary model of racial pheno-

typicality bias (see Figure 1). The model is an interpre-tation of suggested revisions to traditional models ofperson perception (Blair et al., 2002; Maddox & Gray,2002; Zebrowitz, 1996). Processing begins at the leftof Figure 1 with a cursory consideration of a social tar-get's physical appearance. This represents an initialidentification of the target's attributes that act as cuesto salient category dimensions such as age, sex, andrace. At this stage, the nature of feature processingdiverges. The evidence reviewed earlier suggests thatphenotypic features may have an influence on socialjudgment through two routes of information processingthat operate simultaneously and largely independently.

The Category-Based Route

The top path depicts a category-based route as pro-posed by Maddox and Gray (2002) with general attrib-utes that are consistent with traditional approaches tosocial representation and judgment (Brewer, 1988;Fiske & Neuberg, 1990). Through this route, process-ing of the target's phenotypic features results in racialcategorization, with the possibility that the individualwill be placed into a relevant subcategory (e.g.,light-skinned or dark-skinned) as a function of racialphenotypicality. Racial phenotypicality may be deter-mined through a balanced, global assessment of allrace-relevant features (Blair et al., 2002; Livingston &Brewer, 2002), or more heavily influenced by a singlefeature or subset of features (Maddox & Gray, 2002).Only salient subcategory representations are viablecandidates in the process of determining fit. Subcate-gory use is more or less likely depending on personcharacteristics or contextual cues present in the judg-ment context (Maddox & Gray, 2003)

The Feature-Based Route

The bottom path depicts a feature-based route thatinfluences social perception apart from the traditional

CATEGORY-BASED ROUTEInitial Categorization, Subtyping, Individuation

Target CONCEPTUAL KNOWLEDGEAttributes Implicit Categorization Theories, Reliance on

(Phenotypicality) Perceptual Cues, Essentialist Beliefs

FEATURE-BASED ROUTEDirect Category-Feature Associations

(Beliefs and/or Evaluations)

Figure 1. A rudimentary model of racial phenotypicality bias.

395

--iI~~~~~~~~~~~~~~~



MADDOX

range of category based processing. This route occursthrough the product of direct associations betweenphenotypic features and stereotypic traits (Blair et al.,2002) or race-based evaluations (Livingston & Brewer,2002) formed over time through repeated exposure tocategory members. Alternatively, the features may di-rectly convey social information that may be over-generalized to other individuals with similar features(Zebrowitz, 1996). An important aspect of this route isthat the information that features convey will be ap-plied regardless of the target's racial category member-ship (Blair et al., 2002; Secord, 1958; Zebrowitz,1996). In addition, phenotype may continue to influ-ence target judgments through the feature-based route(Blair et al., 2002) in situations when racial categoriza-tion overrides within-race variation through the cate-gory-based route.

The Role of Conceptual Knowledge

Several varieties of conceptual knowledge mayguide the processing of target attributes through boththe category-based route and the feature-based route.Possibilities include metaphorical associations withvarious colors (Secord, 1958), early childhood experi-ences with light and dark (Williams et al., 1975), es-sentialist beliefs (Haslam et al., 2002), implicit causaltheories (Medin & Ortony, 1989), cultural standards ofphysical attractiveness (Breland, 1998; Wade, 1996),and beliefs about the relation between physical fea-tures and personality (Livingston, 2001). Each of theseand others may contribute to category-based and fea-ture-based processing. Alternatively, certain types oftheories may influence specific routes. For example,essentialist beliefs or implicit causal theories thatguide judgments of racial category membership mayinfluence the category-based route. On the other hand,implicit knowledge concerning the importance ofphenotypic cues in judging personality may influencethe feature-based route.

Summary

The model was constructed in an attempt to incorpo-rate the strengths of various theories and the observedoutcomes reviewed earlier. The model is meant to be in-tegrative. It elaborates on traditional models of socialrepresentation andjudgment through a consideration ofracial phenotypicality. It also includes elements to ac-count for data suggesting phenotypic features have aninfluence on judgments that is distinct from traditionalcategory-based processing. Finally, it considers howvarious types of conceptual knowledge may influencecategory-based and feature-based processing. Althoughsacrificing great specificity and detail, the model mayserve as a guide for future investigations.

Future Directions

Pondering Dual Processes

Based on the available evidence suggesting uniqueinfluences of category- and feature-based processes,the current model assumes that these represent distinctroutes of social information processing. This may notnecessarily be the case. There may be other conceptu-alizations of how features and categories may operateto produce judgments. Several extensive discussionsconcerning the nature of dual processes in social per-ception can be found elsewhere, and are not considerhere in detail (Chaiken & Trope, 1999). The questionof dual processes is a difficult one because the cate-gory- and feature-based processes lead to similar out-comes in many situations. To examine whether theyqualify as a dual-process distinction, future researchersmight create and examine situations where the two pro-cesses may lead to divergent outcomes. For example,in situations where racial category membership mustcompete with another more salient nonracial dimen-sion (e.g., gender), feature-based judgments may con-tinue to reveal stereotypic judgments.

Considering Interactionswith Nonphenotypic Characteristics

To date, the bulk of social psychological researchexamining racial category representation has focusedon the presence or absence of stereotype-relevant be-haviors. This review of research indicates that per-ceived racial phenotypicality of the face also contrib-utes importantly to the representation, perception, andtreatment of category members. Future work may bedone to expand the list of characteristics that may alsodetermine racial typicality and integrate them into ourmodels of person perception. Among these are non-facial phenotypic features, style of dress, and style ofspeech. It seems obvious that all of these features likelycontribute to a determination of racial typicality andcategory membership, but few are studied empirically.Future investigations might investigate the interactionbetween these physical characteristics and behavioraltypicality in producing judgment. For example, sub-typing or individuation through the category-basedroute may be exaggerated when a target's physical typ-icality does not match behavioral typicality with re-spect to his racial category (e.g., a dark-skinned Blackstudent acing a difficult biochemistry exam).

Implications for Stereotype Change

One implication of the subtyping hypothesis in-volves stereotype change. Subtyping refers to the men-tal clustering of atypical individuals who disconfirmthe group stereotype, thus preserving the stereotypefrom change (e.g., Maurer et al., 1995; Rothbart &John, 1985). Through the category-based route, racial



RACIAL PHENOTYPICAL ITY BIAS

phenotypicality is one characteristic that individualsuse in a controlled determination of category fit.Therefore. all things being equal, individuals who areless phenotypical should have less impact on stereo-type change, thus preserving any pre-existing beliefsabout the group as a whole.

Reconsidering the Black Male Normin Stereotyping Research

To date, the literature examining skin tone bias andracial bias has focused heavily on perceptions of Blackmen. More empirical studies that include women, His-panics, Asians, and Whites as targets and perceiverswill be crucial to establishing generalizable models ofracial phenotypicality bias. As an example, very fewstudies attempting to isolate skin tone-based percep-tions and outcomes of men from those of women (e.g.,Hunter, 1998; Keith & Herring, 199 1; Maddox & Gray,2002, Study 2; Thompson & Keith, 2001). Each ofthese investigations indicates that skin tone variationplays a slightly different role as a function of gender.Here is one instance where perceptions of physical at-tractiveness could play a meaningful role. Human be-ings place a great deal of value on physical attractive-ness, particularly in women. Therefore, the interactionof skin tone bias and perceptions of physical attractive-ness may be more important in the outcomes of womenthan men (Russell et al., 1992). The results of the studyby Hill (2002) mentioned earlier suggest that skin toneplays a larger role in the perception of physical attrac-tiveness for women than men. This is also consistentwith findings from Maddox and Gray (2002), whereparticipants listed physical attractiveness as a distin-guishing characteristic among light and dark women,but not men.

Intragroup Intergroup Dynamics

One of the assumptions in this analysis is that thediscrimination from racial ingroup and outgroup mem-bers has led to negative material, physical health, andpsychological effects for dark-skinned Blacks. It ispossible that greater racial discrimination from a Whiteactor toward a dark-skinned Black target might havedifferent psychological implications for a target, com-pared to similar discrimination from a Black actor.Going a step further, conceptualizing light and darkskin tone as an intragroup category marker allows oneto ask how this influence might be different dependingon whether the Black actor has skin tone similar to ordifferent from that of the Black target. In other words,what is the contribution of skin tone bias that comesfrom an actor who belongs to the target's racial in-group, but varies with respect to their membership inthe target's skin tone ingroup'?

This perspective comes partly as an experimentalstrategy to facilitate the study of skin tone effects. My

own work has investigated skin tone variation as a cat-egorical distinction among light- and dark-skinnedBlacks. But there is one difficulty with this approach.Clearly, skin tone varies on a continuum from light todark; to dichotomize this continuum may be consideredartificial. However, a social construction perspective onracial group membership reveals that race is a similarlycontinuous dimension that human beings parse into arti-ficial categories. With respect to skin tone, this perspec-tive is supported by the anecdotal accounts that have his-torically characterized skin tone bias as a dichotomousconflict that mirrors racial bias. In the film School Daze,director Spike Lee illustrates this dichotomy in a scenein which the members of two Black sororities tormenteach other over the virtues and pitfalls of "good and badhair." Although one sorority is dominated bydark-skinned women with "bad" Afrocentric fqatures,the other consists of light-skinned women with "good"Eurocentric features (they refer to each other using thederogatory terms "Jigaboos" and "Wannabes," respec-tively). The "Wannabes" derogate the "Jigaboos" fortheir perceived lack of physical attractiveness. The"Jigaboos" retaliate by criticizing the "Wannabes" fortheir perceived lack of Black consciousness (they"wannabe" White and do not "wannabe" Black).

Intragroup membership may moderate the relationbetween phenotype and evaluation. Blacks with darkskin tone likely do not uniformly display patternsof skin tone outgroup favoritism. This intragroupperspective on skin tone bias is reflected in empiricalresearch suggesting that light skin tone is not uni-formly perceived as a positive characteristic. Andersonand Cromwell (1977) found a general preference forlighter skin tone over dark skin tone, but extremelylight skin tone was also associated with negative char-acteristics. Contextual factors may influence whetherlight or dark skin tone is preferred (Celious &Oyserman, 2001). In a context that emphasizes mate-rial wealth, light skin tone may provide advantages(e.g., Keith & Herring, 1991), but dark skin tone maybe perceived positively in a context that emphasizes ra-cial identity (e.g., Brown et al., 1999). Thus, althoughcategorizing skin tone variation along a continuummay seem artificial, it may accurately reflect the cate-gorization processes that characterize human thoughtand interaction. Although the category-based perspec-tive in the study of racial bias may have masked impor-tant variation within-race, there is no doubt that acategory-based perspective has been useful in illumi-nating our understanding of racial bias. The study ofskin tone bias would benefit in a similar fashion.

Conclusion

Despite some of the conclusions drawn from earlyexperimental investigations, several social psychologi-



MADDOX

cal models of category representation and impressionformation make clear predictions about the role of skintone in social perception. These models suggest thatpeople consider racial phenotypicality in the construc-tion and use of their mental representations of racialgroups. Perceived racial phenotypicality mediates af-fective, cognitive, and behavioral responses to individ-uals. Specifically, those with more typical phenotypicappearance are more likely to be negatively stereo-typed, evaluated, and discriminated against than thosewith less typical appearance. This analysis suggeststhat phenotype-based stereotypes, prejudice, and dis-crimination have broad implications for the treatmentand outcomes of Black Americans and members ofother racial and ethnic groups. Evidence that Blacks,Hispanics, and Whites use race-related phenotypic fea-tures in judgments about themselves and others rein-forces the idea that race-related prejudice has the po-tential to affect us all regardless of racial categorymembership. Acceptance of this reality may have im-plications for the way our society talks about race, andhow we eventually act to deal with it.

Interest in racial phenotypicality bias is slowly be-ginning to re-emerge among social psychological inves-tigators. Future investigations will continue to providedescriptions of the outcomes of phenotypic bias and anunderstanding of the processes through which skin toneand/or other phenotypic features play a role in the per-ceptions and outcomes of Blacks and members of otherracial and ethnic groups. This article focuses on a sam-ple of issues that might be explored when examiningskin tone and phenotypic appearance from theperceiver's perspective. Butwe should not forget the im-plications for the targets of discrimination. Additionalevidence suggests that skin tone plays a significant rolein the development of racial identity (e.g., Brown et al.,1999) that, in turn, may influence susceptibility tothreats to self-esteem and self-efficacy (e.g., Thompson& Keith, 2001). Ultimately, both perceiver-based andtarget-based perspectives are necessary if we are to de-velop our understanding of the complex role of racialphenotypicality in social relations.

The future holds daunting challenges for students ofperson perception. With increasing levels of racial inte-gration and the growing number of interracial childrenin the United States, the diversity in skin tones amongU.S. residents will continue to grow. Over time, racialcategory distinctions will continue to blur in our popu-lation, perhaps magnifying the role of phenotypic vari-ation in interpersonal perception (e.g., Hall, 1999).Even today, a great deal of discussion and research ex-plores the experience and identity development of mul-tiracial individuals (Cunningham, 1997; Root, 1996)but much less on how these individuals are perceivedand treated by others (but see Blascovich, Wyer, Swart,& Kibler, 1997; Eberhardt et al., 2003). Because of ourfocus on racial categories, racially ambiguous social

targets are typically ignored in social perception inves-tigations. Theory and research in social psychologymust remain flexible to the changing characteristics ofthe population it seeks to describe. These investiga-tions of racial phenotypicality represent initial steps to-ward addressing the complexities involved in describ-ing the relationships among the ever-growing numbersof multiracial and multiethnic individuals in this coun-try. Through increased discussion and further researchwe will continue to learn of the consequences of ourpreoccupation with physical types.

References

American Anthropological Association. (1998). American Anthro-pological Association statement on "Race." American An-thropological Association. Retrieved July 18, 2002, fromhttp://www.aaanet.org/stmts/racepp.htm

American Association of Physical Anthropologists. (1996). AAPAstatement on biological aspects of race. American Journal ofPhysical Anthropology, 101, 569-570.

Anderson, C., & Cromwell, R. L. (1977). "Black is beautiful" andthe color preferences of Afro-American youth. Journal of Ne-gro Education, 46, 76-88.

Arce, C. H., Murguia, E., & Frisbie, W. P. (1987). Phenotype and lifechances among Chicanos. Hispanic Journal ofBehavioral Sci-ences, 9, 19-32.

Atkinson, D. R., Brown, M. T., Parham, T. A., Matthews, L. G.,Landrum-Brown, J., & Kim, A. U. (1996). African Americanclient skin tone and clinical judgments of African Americanand European American psychologists. Professional Psychol-ogy: Research and Practice, 27, 500-505.

Averhart, C. J., & Bigler, R. S. (1997). Shades of meaning: Skin tone,racial attitudes, and constructive memory in African Americanchildren. Journal of Experimental Child Psychology, 67,363-388.

Banks, W. C. (1976). White preference in Blacks: A paradigm insearch of a phenomenon. Psychological Bulletin, 83,1179-1186.

Bayton, J. A., & Muldrow, T. W. (1968). Interacting variables in theperception of racial personality traits. Journal ofExperimentalResearch in Personality, 3, 39-44.

Best, D. L., Field, J. T., & Williams, J. E. (1976). Color bias in a sam-ple of young German children. Psychological Reports, 38,1145-1146.

Beteille, A. (1967). Race and descent as social categories in India.Daedalus, 96, 444-463.

Blair, I. V., Judd, C. M., Sadler, M. S., & Jenkins, C. (2002). The roleof Afrocentric features in person perception: Judging by fea-tures and categories. Journal of Personality & Social Psychol-ogy, 83, 5-25.

Blanz, M. (1999). Accessibility and fit as determinants of the sa-lience of social categorizations. European Journal of SocialPsychology, 29, 43-74.

Blascovich, J., Wyer, N. A., Swart, L. A., & Kibler, J. L. (1997). Rac-ism and racial categorization. Journal ofPersonality and SocialPsychology, 72, 1364- 1372.

Bodenhausen, G. V., & Macrae, C. N. (1998). Stereotype activationand inhibition. In R. S. Wyer, Jr. (Ed.), Advances in social cog-nition (Vol. 11, pp. 1-52). Mahwah, NJ: Lawrence Erlbaum As-sociates, Inc.

Bond, S., & Cash, T. F. (1992). Black beauty: Skin color and bodyimages among African-American college women. Journal ofApplied Social Psychology, 22, 874-888.




Breland. A. M. (1998). A model for differential perceptions ofcompetence based on skin tone among African Americans.Journal of Multicultural Coun seling & Development, 26,294-311.

Brewer. M. B. (1988). A dual-process model of impression forma-tion. In T. K. Srull & R. S. Wyer (Eds.), Advances in social cog-nition: A dual process model of impression formation (pp.1-36). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Brewer, M. B., & Feinstein, A. S. H. (1999). Dual processes in thecognitive representation of persons and social categories. In S.Chaiken & Y. Trope (Eds.). Dual-process theories in social pst-chology (pp. 255-270). New York: Guilford.

Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin,76, 15-38.

Brown, K. T., Ward, G. K., Lightbourn, T., & Jackson, J. S. (1999).Skin tone and racial identity among African Americans: A the-oretical and research framework. In R. L. Jones (Ed.), Advancesin African American psychology: Theory, paradigms, and re-search (pp. 191-215). Hampton, VA: Cobb & Henry.

Brown, L. C. (1967). Color in northern Africa. Daedalus, 96,464-482.

Brown, T. D., Dane, F. C., & Durham, M. D. (1998). Perception ofrace and ethnicity. Journal of Social Behavior and PersonalitY,13, 295-306.

Celious, A., & Oyserman, D. (2001). Race from the inside: Anemerging heterogeneous race model. Journal of Social Issues,57, 149-165.

Chaiken, S., & Trope, Y. (Eds.). (1999). Dual-process theories in so-cial psychology. New York: Guilford.

Chen, M., & Bargh, J. A. (1997). Nonconscious behavioral confir-mation processes: The self-fulfilling consequences of auto-matic stereotype activation. Journal of Experimental SocialPsvchology' 33, 541-560.

Clark, K. B., & Clark, M. P. (1958). Racial identification and prefer-ence in Negro children. In E. E. Maccoby, T. M. Newcomb, &E. L. Hartley (Eds.), Readings in social psychology (pp.603-611). New York: Holt.

Clark, R., Anderson, N. B., Clark, V. R., & Williams, D. R. (1999).Racism as a stressor for African Americans: A biopsychosocialmodel. American Psychologist, 54, 805-815.

Codina, G. E., & Montalvo, F. F. (1994). Chicano phenotype anddepression. Hispanic Journal of Behavioral Sciences, 16,296-306.

Cunningham, J. L. (1997). Colored existence: Racial identity forma-tion in light-skin Blacks. Smith College Studies in Social Work,67, 375-400.

Cunningham, M. R., Roberts, A. R., Barbee, A. P., Druen, P. B., &Wu, C.-H. (1995). "Their ideas of beauty are, on the whole, thesame as ours": Consistency and variability in the cross-culturalperception of female physical attractiveness. Journal ofPerson-ality & Social Psychology, 68, 261-279.

Davis, F. J. (1991). Who is Black?: One nation's definition. Univer-sity Park: Pennsylvania State University Press.

Deaux, K., & Lewis. L. L. (1984). Structure of gender stereotypes:Interrelationships among components and gender label. Jour-nal of Personality & Social Psyvchology, 46, 991-1004.

Dermer. M., & Thiel, D. L. (1975). When beauty may fail. Journal ofPersonality & Social Psxchology, 31, 1168-1176.

Devine, P. G., & Baker, S. M. (1991). Measurement of racial stereo-type subtyping. Personality and Social Psychology Bulletin, 1 7,44-50.

Dion, K., Berseheid, E., & Walster, E. (1972). What is beautiful isgood. Journal ofPer~sonalitv & Soc-ial Psychology. 24, 285-290.

Dixon, T. L., & Linz, D. (2000). Overrepresentation and under-representation of African Americans and Latinos as lawbreakerson television news. Journal of Communication, 50, 131 -154.

Dixon, T. L., & Maddox, K. B. (in press). Skin tone, crime news, andsocial reality judgments: Priming the schema of the dark and

dangerous Black criminal. Jouriial o('Applied Social Pscho/l-09Y

Drake. S. C., & Cayton, H. R. (1945). Black metropolis. New York:Harcourt Brace Jovanovich.

Dressler, W. W. (1993). Health in the African American community:Accounting for health inequalities. Medical AnthropologYQuarterly, 7, 325-345.

Eberhardt, J. L., Dasgupta, N., & Banaszynski, T. L. (2003). Be-lieving is seeing: The effects of racial labels and implicit beliefson face perception. Personality & Social Psychology Bulletin,29, 360-370.

Edwards, G. F. (1959). The Negro professional class. Glencoe, IL:Free Press.

Edwards, 0. (1973). Skin color as a variable in racial attitudes ofBlack urbanites. Journal of Black Studies, 3. 473-483.

Filler, J. W., & Williams, J. E. (1971). Conditioning the connotativemeanings of color names to human figures. Perceptual & MotorSkills, 32, 755-763.

Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. InD. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook ofsocial psychology (4th ed., Vol. 2 pp. 357-4 11. Boston:McGraw-Hill.

Fiske, S. T., Lin, M., & Neuberg, S. L. (1999). The continuummodel: Ten years later. In S. Chaiken & Y. Trope (Eds.),Dual-process theories in socialpsychology (pp. 231 -254). NewYork: Guilford.

Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impressionformation, from category-based to individuating processes: In-fluences of information and motivation on attention and inter-pretation. In M. P. Zanna (Ed.), Advances in experimental so-cial psychology (Vol. 23, pp. 1 -74). San Diego, CA: Academic.

Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). NewYork: McGraw-Hill.

Frazier, E. F. (1957). The Negro in the United States. New York:Macmillan.

Freeman, H. E., Ross, J. M., Armor, D., & Pettigrew, T. F. (1966).Color gradation and attitudes among middle-income families.American Sociological Review, 31, 265-274.

Gergen, K. J. (1967). The significance of skin color in human rela-tions. Daedalus, 96, 390-406.

Gleiberman, L., Harburg, E., Frone, M. R., Russell, M., & Cooper,M. L. (1995). Skin colour, measures of socioeconomic status,and blood pressure among Blacks in Erie County, NY. Annals ofHuman Biology, 22, 69-73.

Goldberg, F. J. (1973). The question on skin color and its relation toJapan. Psychologia: An International Journal ofPsychology inthe Orient, 16, 132-146.

Gould, S. J. (1994, November). The geometer of race. Discover, 15,64-69.

Hall, R. (1995). The bleaching syndrome: African-Americans' re-sponse to cultural domination vis-i-vis skin color. Journal ofBlack Studies, 26, 172-184.

Hall, R. E. (1992). Bias among African-Americans regarding skincolor: Implications for social work practice. Research on SocialWork Practice, 2, 479-486.

Hall, R. E. (1994). The "bleaching syndrome": Implications of lightskin for Hispanic American assimilation. Hispanic lournal ofBehavioral S ience.s, 16, 307-314.

Hall, R. E. (11998). Skin color bias: A new perspective on an old so-cial problem. Journal of'Psychologs, 132, 238-240.

Hall. R. E. (1999). The myth of third world solidarity: Hypergamsn-yby skin colour as a vehicle of riacism visais-ls African Amel-cans. 1/k Psvchologia: An International Journal, 7, 1-23.

Hairtilton, D. L. (198 1). Stereotyping and intergroup behavior: Somethoughts on the cognitive approach. In D. L. Hamilton (Ed.).Cognitive p)roces.es5 iii stcreotxpiiig aiid ioitergruo/) lbehl(io0(pp. 333-354). Hillsdale, NJ: Lawrence Erlbauum Associates,Inc.



MADDOX

Hamilton, D. L., & Trolier, 1. K. (1986). Stereotypes and stereotyp-ing: An overview of the cognitive approach. In J. F. Dovidio &S. L. Gaertner (Eds.), Prejudice, discrimination, and racisni(pp. 127-163). Orlando, FL: Academic.

Harburg, E., Gleibermann, L., Roeper, P., Schork, M. A., & Schull.W. J. (1978). Skin color, ethnicity, and blood pressure I: DetroitBlacks. American Journal of Public Health, 68, 1177-1183.

Haslam, N., Rothschild, L., & Ernst, D. (2002). Are essentialist be-liefs associated with prejudice'? British Journal of Social PsY-chology, 41, 87-100.

Hill, M. (1944). Social status and physical appearance among Negroadolescents. Social Forces, 23, 443-448.

Hill, M. E. (2000). Color differences in the socioeconomic status ofAfrican American men: Results of a longitudinal study. SocialForces, 78, 1437-1460.

Hill, M. E. (2002). Skin color and the perception of attractivenessamong African Americans: Does gender make a difference'? So-cial Psychology Quarterly, 65, 77-91.

Hirschfeld, L. A. (1996). Race in the making: Cognition, culture,and the child's construction ofhuman kinds. Cambridge, MA:MIT Press.

Hughes, M., & Hertel, B. R. (1990). The significance of color re-mains: A study of life chances, mate and ethnic consciousnessamong Black Americans. Social Forces, 68, 1105-1120.

Hunter, M. (1998). Colorstruck: Skin color stratification in the livesof African American women. Sociological Inquiry, 68,517-535.

Iwawaki, S., Sonoo, K., Williams, J. E., & Best, D. L. (1978). Colorbias among young Japanese children. Journal of Cross-Cul-tural Psychology, 9, 61-73.

Judd, C. M., & Park, B. (1993). Definition and assessment of accu-racy in social stereotypes. Psychological Review, 100, 109-128.

Keil, F. (1989). Concepts, kinds, and cognitive development. Cam-bridge, MA: MIT Press.

Keith, V. M., & Herring, C. (1991). Skin tone and stratification in theBlackcommunity.American Journal ofSociology, 97,760-778.

Klonoff, E. A., & Landrine, H. (2000). Is skin color a marker for ra-cial discrimination? Explaining the skin color-hypertension re-lationship. Journal of Behavioral Medicine, 23, 329-338.

Kreiger, N., Sidney, S., & Coakley, E. (1998). Racial discriminationand skin color in the CARDIA study: Implications for publichealth research. American Journal of Public Health, 88,1308-1313.

Kunda, Z., & Thagard, P. ( 1996). Forming impressions from stereo-types, traits, and behaviors: A parallel-constraint-satisfactiontheory. Psychological Review, 103, 284-308.

Lee, Y. T., Jussim, L. J., & McCauley, C. R. (Eds.). (1995). Stereo-type accuracy: Toward appreciating group differences. Wash-ington, DC: American Psychological Association.

Legum, C. (1967). Color and power in the South African situation.Daedalus, 96, 483-495.

Livingston, R. W. (2001). What you see is what you get: Systematicvariability in perceptual-based social judgment. Personality &Social Psychology Bulletin, 27, 1086-1096.

Livingston, R. W., & Brewer, M. B. (2002). What are we really prim-ing'? Cue-based versus category-based processing of facialstimuli. Journal of Personality & Social Psychology, 82, 5-18.

Macrae, C. N., Bodenhausen, G. V., Milne, A. B., Thorn, T. M. J., &Castelli, L. (1997). On the activation of social stereotypes: Themoderating role of processing objectives. Journal of Evperi-mental Social Psychology, 33, 471-489.

Maddox, K. B. (1998). Cognitive representations of light- anddark-.skinned Blacks: Structure, content, and use of the AfricanAmerican stereotype. Unpublished doctoral dissertation, Uni-versity of California, Santa Barbara.

Maddox, K. B., & Gray, S. A. (2002). Cognitive representations ofBlack Americans: Reexploring the role of skin tone. Personatl-itY & Social Psychology Bulletin, 28, 250-259.

Maddox, K. B., & Gray, S. A. (in press). Manipulating subcategorysalience: Exploring the link between skin tone and social per-ception of Blacks. European Journal of Social Psvychology.

Marks, E. S. (1943). Skin colorjudgments of Negro college students.Journal of Abnormal and Social Psychology, 38, 370-376.

Martin, J. G. (1964). Racial ethnocentrism and judgment of beauty.Journal ofSocial Psychology, 63, 59-63.

Maurer, K. L., Park, B., & Rothbart, M. (1995). Subtyping versus

subgrouping processes in stereotype representation. Journal ofPersonality & Social Psychology, 69, 812-824.

McArthur, L. Z., & Baron, R. M. (1983). Toward an ecological the-ory of social perception. Psychological Review 90, 215-238.

Medin, D. L. (1989). Concepts and conceptual structure. AmericanPsyvchliologist. 44, 1469-148 1.

Medin, D. L., & Ortony, A. (1989). Psychological essentialism. InS. Vosniadou & A. Ortony (Eds.), Similarity and analogicalieasoning (pp. 179-195). New York: Cambridge UniversityPress.

Neal, A. M., & Wilson, M. L. (1989). The role of skin color and fea-tures in the Black community: Implications for Black womenand therapy. Clinical Ps X6chology Reviesw: 9, 323-333.

Okazawa-Rey, M., Robinson, T., & Ward, J. V. (1987). Black womenand the politics of skin color and hair. Women and Therapy,6(1 -2), 89-102.

Park, B., Wolsko, C., & Judd, C. M. (2001). Measurement of sub-typing in stereotype change. Journal of Experimental SocialPsychology, 37, 325-332.

Parrish, C. (1944). The significance ofcolor in the Negro communityUnpublished doctoral dissertation, University of Chicago.

Porter, C. P. (1991). Social reasons for skin tone preferences of Blackschool-age children. American Journal of Orthopsychiatry, 61,149-154.

Powell-Hopson, D., & Hopson, D. S. (1988). Implications ofdoll color preferences among Black preschool children andWhite preschool children. Journal of Black Psychology, 14,57-63.

Ransford, E. H. (1970). Skin color, life chances, and anti-White atti-tudes. Social Problems, 18, 164-179.

Razran, G. (1950). Ethnic dislikes and stereotypes: A laboratorystudy. Journal ofAbnormal and Social Psychology, 45, 7-27.

Relethford, J. H., Stern, M. P., Gaskill, S. P., & Hazuda, H. P. (1983).Social class, admixture, and skin color variation among MexicanAmericans and Anglo Americans living in San Antonio, Texas.American Journal ofPhysical Anthropology, 62, 97-102.

Robinson, T. L., & Ward, J. V. (1995). African American adolescentsand skin color. Journal ofBlack Psychology, 21, 256-274.

Root, M. P. P. (Ed.). (1996). The multiracial experience: Racial bor-ders as the new frontier. Thousand Oaks, CA: Sage.

Ross, L. E. (1997). Mate selection preferences among African Ameri-can college students. Journal of Black Studies, 27, 554-569.

Rothbart, M., & John, 0. P. (1985). Social categorization and behav-ioral episodes: A cognitive analysis of effects of intergroup con-tact. Journal of Social Issues, 41, 81 -104.

Rothbart, M., & Taylor, M. (1992). Category labels and social real-ity: Do we view social categories as natural kinds'? In G. R.Semin & K. Fiedler (Eds.), Language, interaction and socialcognition. (pp. 11-36). Thousand Oaks, CA: Sage.

Russell, K. Y., Wilson, M., & Hall, R. E. (1992). The color complex:The politics of skin color among Aftican Americans. New York:Harcourt Brace Jovanovich.

Sciara. F. J. (1971). Perceptions of Negro boys regarding color andoccupational status. Child Study Journal, 1, 203-211.

Sciara, F. J. ( 1983). Skin color and college student prejudice. CollegeStudent Journal, 17, 390-394.

Secord, P. F. (I1958). Facial features and inference processes in inter-personal perception. In R. Tagiuri & L. Petrullo (Eds.), Personperception and interpersonal behavior (pp. 300-315). Stan-ford, CA: Stanford University Press.



RACIAL PHENOTYPICAL ITY BIAS

Secord. P. F. ( 1959). Stereotyping and lavorableness in the percep-tion of Negro faces. JIournal of Abnormal and Social Psychol-ogy, 59, 309-314.

Secord, P. F., Bevan, W., & Katz, B. (1956). The Negro stereotypeand perceptual accentuation. Journal o Abnormnal and SocialPsYchologv, 53. 78-83.

Seeman, M. (1946). Skin color values in three all-Negro schoolclasses. Amnerican Sociological Revien: 11, 315-321.

Seltzer, R., & Smith, R. C. (1991). Color differences in the Afro-american community and the differences they make. Journal (1Black Studies, 21, 279-286.

Taylor, S. E., Fiske, S. T.. Etcoff. N. L., & Ruderman, A. J. (1978).Categorical and contextual bases of person memory and stere-otyping. Journal of Personality and Social Psychology, 36,778-793.

Telles, E. E., & Murguia, E. (1990). Phenotype discrimination andincome differences among Mexican Americans. Social ScienceQuarterly, 71. 682-696.

Thompson. M. S., & Keith, V. M. (2001). The blacker the berry:Gender, skin tone, self-esteem and self-efficacy. Gender & So-(ietv, /5, 336-357.

Uhlmann, E., Dasgupta, N., Elgueta, A., Greenwald, A. G., & Swan-son, J. (2002). Subgroup prejudice based on skin color amongHispanics in the United States and Latin America. Social Cog-nition, 20, 198-226.

Wade, T. J. (I1996). The relationships between skin color and self-perceived global, physical, and sexual attractiveness, and self-

esteem for African Americans. Journal of' Black Psychology,22, 358-373.

Wagatsuma, H. (1967). The social perception of skin color in Japan.Daedalus, 92, 407-443.

Williams, J. E., Boswell, D. A.. & Best, D. L. (1975). Evaluative re-sponses of preschool children to the colors white and black.Child Development, 46, 501-508.

Williams, J. E., Morland, J. K., & Underwood, W. L. (1970). Conno-tations of color names in the United States, Europe, and Asia.Journal of Social Psychology, 82, 3-14.

Williams, K. H., Williams, J. E., & Beck, R. C. (1973). Assessingchildren's racial attitudes via a signal detection model. Percep-tual & Motor Skills, 36, 587-598.

Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for racialprejudice at the implicit level and its relationship with question-naire measures. Journal of Personality & Social Psychology,72, 262-274.

Wittenbrink. B., Judd, C. M., & Park, B. (2001). Spontaneous preju-dice in context: Variability in automatically activated attitudes.Journal of Personality & Social Psychology, 81, 815-827.

Word. C. O., Zanna, M. P., & Cooper, J. (I1974). The nonverbal med-iation of self-fulfilling prophecies in interracial interaction.Journal ofExperimental Social Psychology, 10, 109-120.

Zebrowitz, L. (1996). Physical appearance as a basis of stereotyping.In C. N. Macrae, C. Stangor, & M. Hewstone (Eds.), Stereo-types and stereotyping (pp. 79-120). New York: Guilford.



Personality and Social Psychology Review€¦ · RACIALPHENOTYPICALITYBIAS EmpiricalAccounts: Evaluations andBeliefs In general, people have exhibited a preference for lighter skin

Documents