The inclusion of skin-tone modifiers into the standard emoji set marked a shift from the default white racialization of emoji towards explicit attempts to expand racial representation in the human emoji characters. This study explores the racial logics of emoji as culturally-situated artifacts that rely on linked understandings of race and technology. We conduct an interface analysis of emoji skin-tone modifiers, coupled with user discourse analysis, to explore the design and user interpretations of skin-tone modifiers. Our findings suggest that though the skin-tone modifiers were introduced as an intervention into the lack of racial representation in emoji, they continue to technically center whiteness in the emoji set as an extension of American technoculture.
A new cultural context for emoji
The American technocultural context
Methodological approach: Critical technocultural discourse analysis (CTDA)
Encoding emoji, encoding race
Decoding emoji through the position of whiteness
Decoding emoji as an expression of whiteness
Conclusion: Emoji as American technoculture
The blatant absence of non-white human characters in the early emoji sets available in the United States mobile market drew strong criticism from Black, indigenous, and people of color (BIPOC) emoji users via social media, tech blogs, and opinion pieces in popular online media outlets. This public backlash to what Roxane Gay (2013) termed “the unbearable whiteness of emoji,”  created visibility around emoji representation as a social issue, putting pressure on Apple to respond to these critiques by working closely with the Unicode Consortium to update the encoding standard (Tsjeng, 2014). The result of these efforts was the introduction of the emoji skin-tone modifiers that appeared in Unicode 8.0, subsequently released by Apple in 2015. These modifiers provided a customization option to alter the shade of skin-tone (previously default white in Apple products) for most of the emoji characters that represent human figures and body parts.
Rather than “solve” the problem of lack of diverse racial representation in emoji, the inclusion of skin-tone modifiers only introduced further questions about race and representation in emoji design and use. While Unicode has continued to annually add new emoji characters that offer expanded ways to represent aspects of culture and identity (e.g., the expanded hair texture options released in 2018 under Unicode 11.0), the original skin-tone modifiers have remained one of the primary ways that Unicode has responded to calls for expanded racial representation in their human character sets.
The inclusion of the skin-tone modifiers marked a shift from the default white racialization of emoji towards explicit attempts to expand racial representation in the human character sets. This shift provides a unique opportunity to explore racial logics and meaning of emoji as culturally situated artifacts that rely on linked understandings of race and technology.
We conducted an interface analysis of emoji skin-tone modifiers alongside a discourse analysis of online user conversations about skin-tone modifiers to explore how ideologies of race and technology mutually shape the design, use, and meaning of emoji. We argue that emoji function as extensions of American technoculture where ideologies of whiteness and technological neutrality mutually shaped the design (encoding) of skin-tone modifiers, along with the ways that many white American emoji users interpreted (decoded) skin-tone modifiers. BIPOC emoji users remained critical of the ways that whiteness was continually asserted in the design and use of modifiers, while still taking advantage of opportunities for expanded self-expression and racial identity performance.
A new cultural context for emoji
Emoji, loosely translated from the Japanese 絵 文 字 to mean “picture-character” refers to pictographs and icons used in computer mediated communication (CMC) that represent common objects, faces, people, animals, and weather, as well as feelings, emotions, and activities (Unicode, 2019). Emoji were developed by Shigetaka Kurita in 1999 for the Japanese mobile phone company NTT DoCoMo as a way to add context to text-based conversations in a quick and seamless way. Emoji remained limited to the Japanese mobile phone market until a subset of the characters became incorporated into Unicode 5.2, paving the way for emoji interoperability with international platforms and an expansion to global markets. Berard (2018) provided a comprehensive documentation of rationale and voting processes of the Unicode Technical Committee in 2009 which led to the inclusion of emoji in the Unicode Standard 6 in 2010. She argued that the voting members of the Unicode Consortium at that time were primarily North American technology corporations such as Apple, Google, and Microsoft, reflecting a bloc of corporate interests (Berard, 2018). This was consistent with Lebduska’s (2014) observation that the Unicode standards have historically reflected the relative market position and interests of global industry leaders from North America. This combination of standardization and redeployment, focused particularly on market viability for North American audiences, provided new cultural contexts for emoji design and interpretation.
Apple’s 2011 release of the emoji keyboard (integrated into iOS 5) introduced emoji widely to the United States, with other leading mobile companies (e.g., Google) following soon after. Following Apple’s 2011 release, American users publicly criticized the overwhelmingly white, heteronormative, and gender-normative representation of human characters using social media and popular internet outlets (Kraus, 2013; Cunningham, 2014; Etherington, 2015; Underhill, 2015). Unicode acknowledged that though the technical standards do not specify racial designations for human character emojis, “they are often shown with a light skin tone” following with “precedents set by the original Japanese carrier images” . Unicode observes in their frequently asked questions (FAQs) that emoji have multiple semantic meanings based on their use and cultural context (Unicode, 2019). Similarly, Freedman (2018) described emoji as “unreadable” symbols that require knowledge of a shared cultural context to function through grounded user interpretation. She argued that “the inability to read universal meanings in emoji has given rise to secondary meanings in local contexts” . The representation of homogenous and light-skinned emoji human characters developed out of a culturally-specific Japanese context, likely in ways that encode identity and status in Japanese culture. Yet, American users overwhelmingly “read” these same characters as default white and phenotypically Caucasian, rather than as Japanese (Etherington, 2015). In fact, only two human characters in this set were widely legible by North American audiences as non-white/Caucasian characters, namely the “Man Wearing Turban” (U+1F473) and the “Man with Chinese Cap” (U+1F472). These two characters were widely panned as advancing racial/ethnic stereotypes, and standing in as “token” representations of Otherness in the otherwise Caucasian emoji palette (Cheney-Rice, 2014) . The recirculation of the human emoji characters in the United States, facilitated by the popularity and reach of the iPhone and Android mobile platforms in North American markets, invited users to reinterpret emoji within an American technocultural context.
The American technocultural context
Technoculture refers to the interconnected sets of ideologies that shape technology as both a material and semiotic cultural product. Dinerstein (2006) characterized American technoculture, particularly, as a matrix of six defining qualities: progress, religion, modernity, whiteness, masculinity, and the future. These qualities reinforce each other, producing an illusion of technology as value-neutral, rather than as culturally produced and ideologically aligned. These interconnected ideologies are often mobilized in ways that bolster dominant or elite interests. In the United States, white supremacy, with its attendant legacies of slavery and racism, continues to shape modern structures of wealth, power, and opportunity. White supremacy is bolstered by whiteness, a set of racial ideologies that construct white people and culture as superior, normative, and unmarked (Dyer, 1997). In this way, whiteness operates as an invisible cultural benchmark, status quo, and organizing logic. Sammel (2009) observed that whiteness is “held in place by inherited ideologies and infrastructures” . Computational standards, like Unicode, act as infrastructural substrates that organize ideologies into code form (McPherson, 2012). Code produces authorized, and seemingly neutral, frameworks for communication that become invisible as a kind of “common sense” rubric. Encoding processes normalize, and reproduce, hegemonic ideologies, like whiteness, under the guise of scientific objectivity and rationalization. Matamoros-Fernández (2017) demonstrated that platforms can serve to further entrench and amplify ideologies of whiteness enacting a new form of racism that she termed “platformed racism” . She pointed to the ways that the culture of social media platforms reproduce social inequalities through their design, technical affordances, and modes of governance. In these ways, beliefs about technology and whiteness mutually define each other, obscuring the inherent racialization of technological artifacts and practices.
Brock (2012) argued that American technoculture often “reinforces long-standing American racial practices” , noting that racial discourses shape all aspects of the design, use, and meaning of information and communication technologies. For example, in the American (and more broadly Western) context, white supremacy has historically shaped technological design choices that imagine white people as ideal and universal users. Through this paradigm, people of color become both invisible by absence of representation, and also hyper-visible in their critiques of these systems that are positioned as value-neutral and universal. Nakamura (2002) observed that “default whiteness”  is assumed of unmarked technologies (as well as users) in digital media environments. In these environments, assertions of non-white identity are hyper-visible and often viewed as problematic disruptions (by white people) to the perceived neutrality of the technological apparatus. In other words, technology-as-whiteness reproduces hegemonic power relationships as part of the status quo of racism under white supremacy.
The early emoji sets embodied these racial logics by representing the human characters as uniformly light-skinned, creating a seamless extension of whiteness in the interface for white users. The introduction of skin-tone modifiers constituted a disruption in this technological status quo for white users by 1) by explicitly “racing” a technology that was previously perceived as raceless, or neutral; and, 2) rendering whiteness visible as a racial identity in the interface. This disruption created interesting tension for users — regardless of racial identity — in parsing the particular structuration of racial identity embedded in skin tone modifiers as a new site of racial identity performance and power negotiation. These tensions form the primary sites of inquiry for this project, revealing anxieties, assumptions, and critiques surrounding the use and meaning of skin-tone modifiers when they were first introduced.
Methodological approach: Critical technocultural discourse analysis (CTDA)
We borrow techniques from Brock’s (2016) critical technocultural discourse analysis (CTDA) to approach analyzing the technocultural aspects of emoji skin-tone modifiers. CTDA advocates for an integrated interface analysis with user discourse analysis. We achieve this by analyzing the design (encoding) of emoji skin-tone modifiers using Unicode’s technical literature as rich primary source material. For user perspectives (decoding), we analyzed publicly available online blog posts, popular technology commentary, videos, news articles, editorials, forums, and comments threads that were published after the introduction of skin-tone modifiers (from 2015 through 2016). This broader corpus consists of a selection of around 35 articles, blog posts, videos, podcasts, and opinion pieces with over 600 associated user comments.
The combined approach of interface and discourse analysis situates both the design (artifact) and interpretation (user practice) of emoji skin-tone modifiers as linked processes that rely on mutually formed, and culturally specific understandings of race and technology. The wealth of online user conversations about their interpretations of the skin-tone modifiers allows insight into the technological practices and beliefs of users in their own words. A great strength of CTDA is the explicit application of critical cultural frameworks to expose normative values and assumptions about technology and (particularly underserved) users. Following this approach, our analysis features illustrative examples of some of the main “tensions” that surfaced in early user conversations about skin-tone modifiers, organized around ideas about racial identity, representation, and “appropriate” use of skin-tone modifiers.
Encoding emoji, encoding race
The development and inclusion of skin-tone modifiers directly resulted from the sustained criticism of (particularly Black) American audiences of color. In the wake of the all-white emoji sets that debuted in iOS 5 in 2011, Unicode began to feel public pressure to address diversity in the emoji set. In March 2014 Apple’s then Vice President of Corporate Communication, Katie Cotton, stated to MTVAct that “there needs to be more diversity in the emoji set”  and confirmed that Apple was working with the Unicode Consortium to create a more racially diverse emoji palette (Kelion, 2014). Later in 2014, Unicode added the recommendation to the emoji standards that human characters be displayed with “non-human” colors, ostensibly to curtail the critiques of whiteness in the interface (Unicode Consortium, 2014). This resulted in Apple and Android redesigning their emoji artwork, which previously represented human emoji characters with white skin, to portray human characters in a default smiley-face yellow color instead (Emojipedia, n.d.).
Unicode revealed their expanded, and newly racially diverse, human emoji set (version 8.0) in 2015 (Apple iOS 8.3 and OS X 10.10.3). Among the prominent new additions were emoji characters that expanded depictions of sexuality (same-sex couples and families), as well as new options for users to customize the skin-tone of human emoji characters. Unicode described these additions as an effort to “reflect more human diversity, especially for skin-tone” in the emoji character set . The Unicode skin-tone palette offers a selection of five human skin-tones, along with the “non-realistic” yellow option, that users can select and apply to modify those emoji characters that depict people, hands, faces, or exposed skin.
Figure 1: Emoji skin-tone modifiers.
The Fitzpatrick scale
Unicode based human skin-tones on the Fitzpatrick scale (Figure 1), a dermatological instrument that delineates six different skin types based on a person’s phenotypic characteristics (hair color, eye color, skin tone) and response to ultraviolet light . The Fitzpatrick scale was initially developed, in Fitzpatrick’s own words, to “better classify persons with white skin [emphasis in the original]” , a practice based in exclusionary and racist assumptions. For instance, the scale overrepresents lighter skin-tones both in its original version and in its adaptation in the emoji palette. Given the racist history of scientific classifications of human populations based on skin color and other phenotypical characteristics (e.g., phrenology), the reliance on a “scientific” medical metric for representing racial difference in emoji should be viewed with some skepticism. Stark and Crawford speak to these concerns when they argue that the Unicode Consortium’s selection of the Fitzpatrick scale only further embeds “hierarchies of gendered and racialized authority and inequality”  in the character sets. While the Fitzpatrick scale evades explicit racial categorization, phenotype and skin-typing operate as de facto cultural signifiers for representing race in the character set.
Figure 2: Unicode’s interpretation of the Fitzpatrick scale.
The Unicode consortium members adopted the Fitzpatrick scale as the basis of the skin-tone palette in part because it was viewed by consortium members as a technologically “neutral” solution to the “diversity problem”  in emoji character design. The medical origins of the Fitzpatrick scale as a dermatological instrument (ostensibly) lent a guise of scientific authority to neutralize what was otherwise viewed as a fraught social-political topic. Miltner’s (2015) research explored the decision-making process that resulted in the Unicode 7.0 set, revealing that consortium members discussed concerns about racial representation in emoji as a perceived problem of a “sensitive”  public. In other words, the majority of consortium members saw themselves as responding to a manufactured public relations crisis, rather than remediating platformed racism. Miltner argued that the majority of (predominately white) consortium members insisted on a presumption of neutrality for Unicode, resisting critiques of racism from several members of the consortium that voiced discomfort with skin-tone modifiers. Her research demonstrated that the rhetoric of technological neutrality was mobilized by consortium members as a way to dismiss concerns about the racist implications of emoji design choices, ultimately protecting whiteness as an integral feature of emoji encoding.
The language of technological neutrality is prevalent across the Unicode technical standards literature, particularly in regards to sections dealing with the preferred appearance and representation of human emoji characters. Consider these sections from the emoji technical standards:
General-purpose emoji for people and body parts should also not be given overly specific images: the general recommendation is to be as neutral as possible regarding race, ethnicity, and gender. 
The Unicode Standard does not require a particular appearance for characters that depict people or body parts, such as U+1F474 OLDER MAN or U+270B RAISED HAND. In fact, UTR #51 recommends that such depictions be as neutral or generic as possible with respect to physical appearance, for example using non-realistic colors for skin-tone. 
In these (and other) sections, the developers articulate the desire for “neutral or generic” depictions of human emoji characters. The “neutral” body is defined throughout the technical literature as that which is unmarked by race, ethnicity, or gender. As critical race scholars have long pointed out, proximity to power and privilege shape which identities are perceived as unmarked or “neutral”. For instance, under prevailing racial ideologies in the United States, whiteness is posited as neutral and raceless, rather than as constitutive of racial identity (Bonilla-Silva, 2006; Dyer, 1997). The adherence to the belief of Unicode’s technological neutrality, combined with racial ideologies that posit neutrality as a marker for whiteness, mutually reinforce white racial logics as a foundational infrastructure of skin-tone modifiers.
Prior to the addition of the skin-tone modifiers, default whiteness was assumed in the encoding of emoji human characters, and explicitly interpreted through renderings of the characters (“glyphs”) as light-skinned with features that can be read as phenotypically Caucasian. The inclusion of the skin-tone palette did not change this underlying mechanism. From an encoding standpoint, the skin-tone modifiers operate by overlaying the skin-tone modifier on top of the original emoji base characters (Figure 2). This produces combinations of human characters in a range of six skin-tone shades that otherwise share the same phenotypical features (hair texture, facial features, eye color) as the previously existing base character. The default (white) emoji characters operate as base models that undergird all of the resultant forms of human character representation. This effectually re-centers whiteness in the character set, reinforcing whiteness as the unmarked baseline from which human characters may be otherwise superficially racialized through the application of skin-tone.
Figure 3: Application of emoji skin-tone modifier to human base character.
Despite the directive to render characters as “generic” and “neutral” as possible, there is a peculiar exception to this rule: the “Person with Blonde Hair” emoji. The guidelines for this character states,
No particular hair color is required, however, dark hair is generally regarded as more neutral because people of every skin tone can have black (or very dark brown) hair. One exception is PERSON WITH BLOND HAIR, which needs to have blond hair regardless of skin tone. 
This character is a holdover from the original Japanese character set, and was meant to signify white European phenotyping to stand out from the Japanese-coded characters, shown with dark hair. This character is oddly recontextualized under the skin-tone modifier schema as a unique character, displaying a kind of blonde exceptionalism where blondness (culturally coded as whiteness) is treated an unviable technical rule in Unicode.
The yellow emoji and colorblindness
Unicode recommends using “non-realistic” colors for skin tone as one route to achieving the goals for “neutral or generic” depictions of physical appearance, suggesting a “non-human” yellow/orange color . The yellow emoji operates as a floating signifier that destabilizes the skin-tone modifiers as racial referents. Following colorblind logics, the skin-tone modifier becomes detached from the cultural specificity of race as a socio-cultural category, reframing “race” as “color”. Colorblindness, a dominant racial discourse in the United States since the 1990s, asserts that race is irrelevant in contemporary society as an organizing structure of opportunity (Gallagher, 2003). Under this rubric, merely acknowledging racial difference is itself interpreted as “racist”, precluding interventions into racial inequality. In this way, colorblindness conflates the erasure of racial difference with the end of racial inequality, which simultaneously serves to delegitimize race as a social category while upholding structures of white supremacy (Bonilla-Silva, 2006). The recommendation for “non-realistic colors of skin-tone” similarly works to neutralize the skin-tone modifiers as a direct intervention into racial representation in the interface, depoliticizing skin color as a socio-cultural identity marker, and further denying race as a legitimate signifier in the interface.
Gallagher observed that the “new” colorblind ideology does not ignore race altogether, instead refashioning race from a social category into a commodity “that whites and racial minorities can purchase and share” . The skin-tone modifiers operate very much in this way as they are positioned by Unicode as “feature” of self-expression in the interface. With the addition of skin-tone modifiers, the default human emoji character in many mobile carriers was set to the yellow emoji. This new arrangement meant that users were required to actively select a human skin-tone color for their emoji, otherwise they would de facto be directed to a yellow human character. Through this selection process, “race” is recast as an accessory to enhance or modify the base emoji characters. This allowed Unicode to position skin-tone modifiers as an opportunity for individual user-expressions of “choice”, rather than as structural adherents to an ideological framework rooted in racial categorization and hierarchy. Under this framing, a sanitized, controllable version of non-white racial identity becomes accessible to white users, offering no real challenge of the centrality of white identity in the interface. Rather than a shift away from whiteness, the designation of the “default” yellow emoji can be read as a proxy that retains the cultural signifiers of whiteness in its presumptions of universality, while denying that skin-tone modifiers signify racial social categorization because it remains a “non-realistic” color.
Unicode anticipated questions about racial representation in emoji as they rolled out the skin-tone palette, adding a heading under the emoji frequently asked questions (FAQs), “What about diversity?” . Note, the term “race” never appears in this section, and only occurs once in the technical standards under the previously discussed recommendation to “be as neutral as possible regarding race, ethnicity, and gender” . Instead, the Unicode documentation prefers the term “diversity”, which appears throughout the FAQs and technical standards. Unicode employs a broad definition of the term “diversity” that is largely decoupled from racial identity and representation. Consider the following statement:
Of course, there are many other types of diversity in human appearance besides different skin tones: Different hair styles and color, use of eyeglasses, various kinds of facial hair, different body shapes, different headwear, and so on. 
By this formulation, the skin-tone modifiers are positioned as a variable of “appearance” akin to having different headwear or eyeglasses. This characterization builds on colorblind logics that define “diversity” as a kind of vague catch-all for all the ways a user might customize or accessorize an emoji. Sara Ahmed (2012) noted that the rhetoric of diversity is often mobilized in ways that reproduce rather than challenge social privilege. Unicode’s diversity rhetoric is similarly employed in ways that dodge (or block) opportunities to engage skin-tone modifiers as racial signifiers. This dodge asserts that diversity is both important and has been technically addressed by the expansion of representation options in the encoding, foreclosing further explorations of emoji as already racialized technologies that re/produce power and privilege. Through this framing, race as a system is wholly obscured and subsumed under a celebratory model of “difference” and as a neoliberal opportunity for user expression. This approach mirrors other kinds of white liberal interventions to racism that rely on “additive” models within existing structures (equality), rather than offering an epistemic challenge to (or reorganizing of) the structure itself as status quo (justice).
Decoding emoji through the position of whiteness
Our discourse analysis identified several main tensions that surfaced for users in terms of using and interpreting the newly added emoji skin-tone modifiers. For many white users, the skin-tone palette constituted an ideological disruption by making racial identity explicit in an interface that had been previously experienced as raceless/neutral. This shift created anxiety and produced new questions about emoji use for white users who were not accustomed to confronting their whiteness as both a racial identity, and as an ideological infrastructure that undergirds their technology environment.
Race as “user choice”
The reorganization of race as a visible function of user “choice” and self-expression in the interface, produced new questions about emoji use as an act of identity politics for many white users. These questions revolved around the meaning and appropriate usage of white and non-white skin-tone modifiers. McGill (2016) discussed these same anxieties in his exploratory look at emoji usage on Twitter. He found that the “light-colored emojis are less common than their light-skinned user counterparts” , and that the “medium-light” skin tone (corresponding to Fitzpatrick Type III), is the most popular modifier on Twitter. McGill speculated that white people may experience “squeamishness”  about either unintentionally affirming “white pride” — an ideology linked to white supremacy — or, culturally appropriating emoji that signify Black and brown bodies, opting to hedge their bets by using the middle modifier, corresponding to the “medium-light” skin-tone.
These apprehensions were well-illustrated in a conversation between co-hosts Aminatou Sow and Ann Friedman (2015) on the Call your girlfriend podcast shortly after the skin-tone modifiers were introduced.
Friedman: As a white person, I am as excited as everyone that there are many new racial options for some of the hand and face emoji ... and I want to use them. Obviously, I was someone who was not happy with the default, all of them being white. But, then I felt, especially when I’m texting another white person, is it weird for me to text me brown hands clapping? Is that a weird thing?
Sow: Wow, that’s white people weirdness. Inter-white-person weirdness ... I’m going to say, for one, welcome to our world — where the default was always one thing, and you’re trying to make a new default. Yeah, I guess it’s weird for you. I personally always default to the darkest emoji now ... I live in a world where there was always ever one default.
Sow and Friedman’s responses reflect two different racial and technological realities, grounded in their lived experiences of systemic racism and white privilege, respectively. This example highlights the differential subject positions of white users and BIPOC users in navigating identity and representation in the interface. Sow expresses that the skin-tone modifiers represent a new opportunity for representation and expanded self-expression that was not formerly available, whereas Friedman articulates unease around both explicitly claiming and rejecting white identity via the modifiers. These different framings reinforce how white users have been encouraged to think of technology as an extension of self; an option that has precluded BIPOC users who have been required by dominant/Western technological systems to modify or constrain their racial identity.
Yellow emoji as a racial “opt-out”
Many white users suggested using the yellow emoji as an alternative to selecting one of the “realistic” skin-tone emoji options. For instance, in response to her own question “Is it weird to still use white emoji?,” Kaleigh Rogers (2015) noted that “the obvious solution, of course, is just to stick to the neutral, unnaturally yellow emoji, which is what I’ve been doing” . Commenters in other articles reaffirmed the sentiment that the yellow emoji was perceived by many white users as a raceless option. For example, one commenter suggested about skin-tone modifiers, “use them or don’t — nobody ever thought that yellow, or blue or red emoji actually represent a race.” Or, “What was wrong with the simple yellow emojis? All they need to do is express your emotion, why do they [sic] need to make everything about race and gender?” These comments suggest that some white users perceived the yellow emoji as a “neutral” alternative to human skin-tones, providing a kind of racial identity “opt-out”. McGill noted that, “when white people opt out of racemoji in favor of the ‘default’ yellow, those symbols become even more closely associated with whiteness — and the notion that white is the only raceless color” . These ideas are coherent to ideologies of colorblindness that posit whiteness as a raceless and universal identity.
The default positioning of the yellow emoji technically reinforces the idea of its primacy, reasserting whiteness along with it. For instance, several users noted that the platform “defaults” to the yellow emoji unless the user opts to select another skin-tone option:
It defaults to the yellow one. The yellow one used to be all that was available. Ergo, the yellow one is both more familiar and faster. Therefore that is the one used.
Other users in this comment thread reiterated this point, “The default is yellow, so one doesn’t ‘opt-out’ of white. If you change, you are opting out of yellow.” These comments do not acknowledge that BIPOC users are unable to similarly “opt-out” of being racialized subjects under white supremacy. Thus the yellow emoji does not afford BIPOC users the same opportunity to “evade race” when using the human emoji icons.
Skin-tone modifiers as “reverse racism”
Some white users advanced a kind of “reverse racism” argument against skin-tone modifiers. These comments took the form of calling out the addition of skin-tone modifiers as itself racist for intervening in an otherwise (perceived) raceless technology:
Of course if you take a series of sexless raceless text based emoticons, turn them into graphics and then add race to them you’re injecting race where none was before and jumping right into the ditch behind the obstacle that didn’t exist. They had already solved the race issue by not having any race, they were text, then bright yellow pacman/acid faces, the new emoji disgust me as clearly coming from a fundamentally racist (if to their own way of thinking well meaning) mindset.
These comments mirror broader applications of colorblind ideology that posit social and technological structures as inherently raceless and equitable in their effects, a view that positions even the acknowledgement of race as a racist injunction. Similar to backlash against racial projects like affirmative action that attempt to redress racism, white majority members perceive gains from racial minority groups as a threat and loss of power. These critiques (often willfully) ignore the structural disparities in power that have advantaged white people at the expense of BIPOC.
A few commenters argued that white people were actually underrepresented with skin-tone modifiers:
As a “millennial” who has put some thought ... into which emoji to use, I’ll say my issue is that the “white” emoji woman has blonde hair. but I have brown hair so I don’t want to use her. The one with darker hair has jet black hair and fair skin. She looks blatantly, Asian, which yeah I feel a little uncomfortable using because I’m not Asian. From there it’s an emoji who reads Hispanic to me, (has the lighter brown hair like me but very tan skin) then two shades of dark skin. I’m painting in broad, stereotypical strokes here but I think that the majority of emoji users are women. And every major group is represented, except for white women with brown hair.
This comment, and others like it, conflate the structural issue of racial representation in media with expectations for specific and individuated representation. This framing elides whiteness as a collective racial identity, and as an already dominant ideology structuring emoji design. Instead, white people articulate themselves as individuals, citing a lack of specificity of personalization options within the multiple ways already present to represent white identity through emoji. For instance, several commenters bemoaned the lack of redheaded emoji (now added in the 2018 Unicode updates) as “ginger discrimination” in the updated skin-tone palettes. Though some of these comments seemed to be in jest, the point stands that calls for racial diversity in the interface quickly became critiques for a more generic and depoliticized version of “diversity” that re-centers white feelings and individualized formations of white identity.
Just as in the design process, many users appealed to technological neutrality in their descriptions of selecting which emoji skin tone to use. In response to the McGill article about why white people felt squeamish about using the lightest emoji, one user responded:
One word to entirely upend the premise of the article: CONTRAST. I wouldn’t use a white emoji for the same reason I wouldn’t use a white font, because I’m typing them into a white text box.
Contrast is a particularly interesting argument for why white users opt to not use the lighter-toned emoji selections given the vast research on the ways filmic technologies have historically been developed around ideas about contrast that are optimized for light skin (Dyer, 1997). In filmic technologies, as in these emoji examples, the ideal presentation of whiteness (signified by light skin) serves as the barometer for technological use.
Other users echoed this sentiment in similar comments that point to the technical difficulties of reading the lighter skin-toned emoji arguing that, “Maybe it’s just as simple as ‘you really can’t see the pale color on a white background’ so given choices, nobody chooses the pale colors.” There were many comments that similarly pushed back that this choice of skin-tone selection had anything to do with racial identity, reiterating that it was the technical features that guide this user behavior:
In my case, not using the whitest emoji isn’t about white guilt, or anything racial. It simply doesn’t show up well on a text box. Not a social statement just better visuals.
These comments assert a technocultural understanding of emoji as a race-neutral technology, producing a technical explanation for a racial practice. This argument is presented as a “rational” argument for white people’s avoidance of the lighter toned emoji, that draws authority from its stance as an “objective” technical assessment. These arguments suggest that white users are able to evade complicated questions about emoji skin-tone choices as aspects of racial practice using technical explanations as objective rationale for their self-presentation.
Technological neutrality was also present in comments that proposed technical solutions to the race “problem” in the emoji set. Comments such as, “... randomize the colour [sic] of the emoji so that it comes out with a different skin tone each time.” Others suggested moving to all non-human colors as a way to remove race as a consideration for emoji selection:
Since Black, white, brown, yellow, and red emoji are going to create issues one way or another, simply make all emoji blue or green ... so abstract that they are non-discriminatory, so inclusive that they are infinitely usable.
How about just make everyone mint green?! It’s stupid to try and differentiate every ethnic group in the world. Same for red hair, or glasses, etc. emojis are meant to be GENERIC, not specific.
Colorblind language amplifies ideas about technological neutrality in these statements.
This last comment echoes Unicode’s call for “generic” representations, that remain inoffensive (to white audiences) and presume universal applicability. In fact, users referred back to Unicode’s technical standards multiple times, quoting specifications as a way to lend authority to their arguments for the technological neutrality and racelessness of emoji. Interestingly, many users also referred to Android’s rendering of emoji as a positive example for Apple to emulate, noting that, “The Android emoji works quite well in this respect, as it displays all emoji as green ‘androids’, making race not an issue.” Throughout these examples, skin tone modifiers are positioned as a race “problem” to be “solved” via technical interventions that assume technological neutrality, and ultimately work to restore default whiteness to the human emoji set.
Decoding emoji as an expression of whiteness
While BIPOC commenters expressed excitement about expanded skin-tone options, they also expressed tensions regarding how racial identity was codified and represented in the emoji skin-tone palette. Whereas white commenters tended to employ technological neutrality and colorblindness as ideological frameworks for interpreting emoji, many BIPOC commenters identified emoji as racial projects. Omi and Winant (1986) defined racial projects as “simultaneously an interpretation, representation, or explanation” of racial dynamics that, “connect what race means in a particular discursive practice and the ways in which both social structures and everyday experiences are racially organized, based upon that meaning” . Many BIPOC user comments made these connections by positioning emoji skin-tone modifiers as reflective of, and responding to, broader patterns of race, grounded in the specific racio-historical context of the United States.
Repeating patterns of racism
Only days after Apple introduced the new emoji, Paige Tutt (2015) authored an op-ed critiquing the skin-tone modifiers. She argued that
... there’s nothing specifically “black” about an emoji with browner skin. Deepening the skin color of a previously white emoji doesn’t make the emoji not white. It’s just a bastardized emoji blackface. The blond-haired emoji man and the blue-eyed princess are clearly white, but you can slip them into a darker-colored skin. These new figures are emoji of color; they’re just white emoji wearing masks. 
In this comment, Tutt translated the technical encoding of skin-tone modifiers as a cultural signifier of the racist history of blackface in America. Tutt linked the racist caricatures created through the application of blackface makeup (by predominately non-Black performers), to the technical overlay of dark skin tones over the default white human emoji base. This critique does important technocultural work that enables multi-layered critiques of technology as a site of racial formation.
Similar criticisms were made by users who drew connections between the yellow emoji, orientalist stereotypes, and racial slurs:
All I kept thinking since the icons are released was, “Ok you’re calling me yellow, Apple, really?” As an Asian, my default yellow skin icons bugs me a lot every time it appears.
Though Unicode articulated a vision of the yellow emoji as a “non-human” color option, yellow has deep cultural history in the American context as a racist coding for people of East Asian descent. This was not lost on (particularly) East Asian users who expressed tensions about how the yellow emoji fit into the racial logics of the skin-tone modifier set:
I think it’s interesting that yellow is chosen as the “neutral color” when it is one that is used, and has been used, to code Asian-Americans. So if you’re Asian-American (and not one who feels “brown”), you can be called yellow and have it used against you (see “Yellow Peril” link below), and you can try and reclaim that color (“Black Power, Yellow Peril”), but in the end the color is never really yours, and other people will use it as a so-called “neutral” color.
Like Tutt’s observations about emoji and blackface, this commenter located the yellow emoji in a culturally specific history of racism against Asian people in the United States, specifically referencing the ways that Asian-Americans have reclaimed the phrase “yellow peril” in solidarity with Black radical activists in the 1960s. This critique revealed that the technical placement of the yellow emoji — alongside “human” skin tone options — re-contextualizes yellow from a “neutral” color choice to a fraught signifier of anti-Asian racism.
Limited representations of non-white identities
One tension that surfaced for BIPOC user was the limited options for representing non-white identities in the new emoji set. Though the new modifiers offered a wider range of skin-tone options for self-expression, they did not offer different hair textures or culturally-specific hair styles for BIPOC users. Comments like, “... Where is the kinky hair? This should have been implemented at the outset. What did Apple think that people of color would not purchase their phone? Smdh,” indicate frustration at this obvious absence of representation. Another user joked that “Apple just dropped it’s [sic] new diverse emoji and it looks like it was also co-sponsored by Dark & Lovely relaxer.” This commenter used emoji to signify on a larger cultural conversation about how Black hair types and characteristics are perceived, both within the black community and in white popular culture. Through this rhetorical move, this commenter translated a critique about Black hair relaxers as a technique of “manageability” for Black hair/identity into a critique about the ways emoji skin-tone modifiers are similarly “managing” Black identity. This is a technocultural move that merges understanding of how technologies are implicated in the construction of racial identity. Of course, another user commented on this article, “I’m natural and not offended,” demonstrating that these interpretations remain actively contested and are culturally negotiated within various community groups.
Hypervisibility of BIPOC users
Several BIPOC users voiced concerns about how the skin-tone modifiers made them hypervisible in mediated conversations by making race an explicit variable and part of communication:
As a racial minority I find it offensive for different reasons. I enjoy being colorless or perceived as white in the anonymity of the internet. I feel that people are more open with me about how they really feel about race if they don’t think I have a racial minority chip on my shoulder. It’s very telling when people assume that I’m white, or assume that I’m not of the race I might be speaking of in a discussion about race.
This commenter described the vulnerabilities of participating online as a racial minority, noting that anonymity afforded safer opportunities for participation and interaction. Similarly, Mcgill (2015) cited an interview with engineer Aditya Mukerjee who expressed
Every time I use an emoji, I have to make a choice: Do I use a colored racemoji, and draw attention to my ethnicity (even when it’s not pertinent), or do I use a default emoji, which may misrepresent me altogether?
He notes that this is a unique burden for people of color using emoji that white people do not have to consider. Since whiteness assumed in mediated communication unless otherwise stated, white people are not at risk or made vulnerable by asserting their whiteness either implicitly or explicitly in these media.
Another commenter expressed concern that revealing their race through emoji would invite harm or misuse, depending on the context:
A variety of race representing emoticons to choose from suggests that I should reveal my race in the same way that the Census Bureau or a potential employer might be requesting it. I can’t help but think that the information will be misused, doing more harm than good. I always select “other” if it’s an option and answer that I’m “American” if when asked about my heritage or nationality.
These concerns revisit the persistence of whiteness as a key element of American technoculture.
Conclusion: Emoji as American technoculture
This project revealed that emoji skin-tone modifiers are culturally produced around, and interpreted through, sets of racial logics that extend American technocultural beliefs. Though Unicode introduced skin-tone modifiers as a response to criticism about racial representation from BIPOC audiences, whiteness remained central to the design logics of emoji. Couched in terms of “diversity” and “skin-tone”, Unicode was able to respond to a critical public while sidestepping deeper questions about race and representation in emoji design. The reliance on beliefs about technological neutrality and colorblindness — hallmarks of American technoculture — shaped the resultant encoding mechanisms and recommendations for interpreting emoji glyph design, further entrenching whiteness in the interface.
The user discourse analysis revealed that ideologies of whiteness and technological neutrality were echoed in many white users’ interpretations of the skin-tone modifiers. The beliefs of these users were bolstered by technical encoding, such as being able to “opt out” of engaging with race in the interface by defaulting to the yellow emoji, or dismissing the need for racial representation in an otherwise perceived-neutral interface.
Meanwhile, BIPOC users are not afforded the same opportunities for “opting-out” of engaging hegemonic racial structures in the interface. BIPOC users offered alternative interpretations of the skin-tone modifiers that placed them in conversation with culturally specific histories of race, racism, and representation. These commenters identified the skin-tone modifiers as both a site for expanded representation, while also noting the limitations of these representation options in terms of reliance on stereotypes, flattening of culture, and issues of both invisibility and hypervisibility. This research raises questions about the limits of emoji skin-tone modifiers for providing access to culturally specific and multi-dimensional racial representations outside of the white racial frame.
About the authors
Miriam E. Sweeney is Assistant Professor in the School of Library & Information Studies at the University of Alabama.
Direct comments to: mesweeney1 [at] ua [dot] edu
Kelsea Whaley is a law librarian at Balch & Bingham LLP in Birmingham, Alabama.
E-mail: kelseawhaley [at] gmail [dot] com
1. Gay, 2013, p. 26.
2. Unicode Consortium, 2019.
3. Freedman, 2018, paragraph 3.
4. Some renderings of these character glyphs traded on racist stereotypes (e.g., slanted eyes for the “Man with Chinese Cap”).
5. Sammel, 2009, p. 650.
6. Matamoros-Fernández, 2017, p. 930.
7. Brock, 2012, p. 532.
8. Nakamura, 2002, p. 46.
9. Kelion, 2014, paragraph 11.
10. Unicode Consortium, 2017, section 2.2, paragraph 1.
11. Unicode collapsed Fitzpatrick types 1 and 2 together to make the lightest skin-tone option (U+1F3FB).
12. Fitzpatrick, 1988, p. 869.
13. Stark and Crawford, 2015, p. 7.
14. Miltner, 2015, p. 22.
16. Unicode Consortium, 2017, section 2.0, paragraph 4.
17. Unicode Consortium, 2019, section 9.
18. Unicode Consortium, 2018, section 2.2, paragraph 3.
19. Unicode recommends web safe hex color code #ffcc22 (Unicode Consortium, 2019).
20. Gallagher, 2003, p. 5.
21. Unicode Consortium, 2019, section 9.
22. Unicode Consortium, 2017, section 2.0, paragraph 4.
23. Unicode Consortium, 2018, section 2.2.3, paragraph 2.
24. McGill, 2016, paragraph 5.
25. McGill, 2016, paragraph 7.
26. Rogers, 2015, paragraph 4.
27. McGill, 2016, paragraph 14.
28. Omi and Winant, 1986, p. 56.
29. Tutt, 2015, paragraph 3.
Sara Ahmed, 2012. On being included: Racism and diversity in institutional life. Durham, N.C.: Duke University Press.
Bethany Berard, 2018. “I second that emoji: The standards, structures, and social production of emoji,” First Monday, volume 23, number 9, at https://firstmonday.org/article/view/9381/7565, accessed 29 April 2019.
doi: https://doi.org/10.5210/fm.v23i9.9381, accessed 29 April 2019.
Eduardo Bonilla-Silva, 2006. Racism without racists: Color-blind racism and the persistence of racial inequality in the United States. Second edition. Lanham, Md.: Rowman & Littlefield.
André Brock, 2016. “Critical technocultural discourse analysis,” New Media & Society, volume 20, number 3, pp. 1,012–1,030.
doi: https://doi.org/10.1177/1461444816677532, accessed 29 April 2019.
André Brock, 2012. “From the blackhand side: Twitter as a cultural conversation,” Journal of Broadcasting & Electronic Media, volume 56, number 4, pp. 529–549.
doi: https://doi.org/10.1080/08838151.2012.732147, accessed 29 April 2019.
Zak Cheney-Rice, 2014. “Out of the 250 new emojis you can use on your phone, there’s one glaring omission,” Mic.com (17 June), at https://mic.com/articles/91405/out-of-the-250-new-emojis-you-can-use-on-your-phone-there-s-one-glaring-omission, accessed 29 April 2019.
Andrew Cunningham, 2014. “Unicode 7.0 introduces 2,834 new characters, including 250 emoji,” ArsTechnica (16 June), at https://arstechnica.com/gadgets/2014/06/unicode-7-0-introduces-2834-new-characters-including-250-emoji/, accessed 29 April 2019.
Joel Dinerstein, 2006. “Technology and its discontents: On the verge of the posthuman,” American Quarterly, volume 58, issue 3, pp. 569–595.
doi: https://doi.org/10.1353/aq.2006.0056, accessed 29 April 2019.
Richard Dyer, 1997. White. New York: Routledge.
Darrell Etherington, 2015. “Apple’s latest betas bring more diversity to emoji,” TechCrunch (23 February), at http://techcrunch.com/2015/02/23/apple-emoji-diversity/, accessed 29 April 2019.
Emojipedia, n.d. “Apple iOS 5.0,” at https://emojipedia.org/apple/ios-5.0/, accessed 10 June 2019.
Thomas B. Fitzpatrick, 1988. “The validity and practicality of sun-reactive skin types I through VI,” Archives of Dermatology, volume 124, number 6, pp. 869–871.
doi: https://doi.org/10.1001/archderm.1988.01670060015008, accessed 16 June 2019.
Alisa Freedman, 2018. “Cultural literacy in the empire of emoji signs: Who is crying with joy?” First Monday, volume 23, number 9, at https://firstmonday.org/article/view/9395/7567, accessed 29 April 2019.
doi: https://doi.org/10.5210/fm.v23i9.9395, accessed 29 April 2019.
Ann Friedman and Aminatou Sow, 2015. “Episode 26: Appropriate emoji,” Call your girlfriend [Audio podcast] (22 May), at http://www.callyourgirlfriend.com/episode-26-appropriate-emoji/, accessed 29 April 2019.
Charles A. Gallagher, 2003. “Color-blind privilege: The social and political functions of erasing the color line in post race America,” Race, Gender & Class, volume 10, number 4, pp. 22–37.
Roxane Gay, 2013. “The unbearable whiteness of emoji,” In: Mercedes Kraus (editor). Emoji: A special print edition of Womanzine. New York: Womanzine + Emoji Art and Design Show, at https://issuu.com/lindseyweber5/docs/emoji_by_womanzine, accessed 29 April 2019.
Leo Kelion, 2014. “Apple seeks greater emoji racial diversity,” BBC News (26 March), at https://www.bbc.com/news/technology-26749867, accessed 10 June 2019.
Mercedes Kraus, 2013. “Speaking in picture-letters,” In: Mercedes Kraus (editor). Emoji: A special print edition of Womanzine. New York: Womanzine + Emoji Art and Design Show, at https://issuu.com/lindseyweber5/docs/emoji_by_womanzine, accessed 29 April 2019.
Lisa Lebduska, 2014. “Emoji, emoji, what for art thou?” Harlot: A Revealing Look at the Arts of Persuasion, number 12, at http://harlotofthearts.org/index.php/harlot/article/view/186, accessed 29 April 2019.
Ariadna Matamoros-Fernández, 2017. “Platformed racism: The mediation and circulation of an Australian race-based controversy on Twitter, Facebook and YouTube,” Information, Communication & Society, volume 20, number 6, pp. 930–946.
doi: https://doi.org/10.1080/1369118X.2017.1293130, accessed 29 April 2019.
Tara McPherson, 2012. “Why are the digital humanities so white? or Thinking the histories of race and computation,” In: Matthew K. Gold (editor). Debates in the digital humanities. Minneapolis: University of Minnesota Press.
doi: https://doi.org/10.5749/minnesota/9780816677948.003.0017, accessed 29 April 2019.
Andrew McGill, 2016. “Why white people don’t use white emoji,” Atlantic (9 May), at http://www.theatlantic.com/politics/archive/2016/05/white-people-dont-use-white-emoji/481695/, accessed 29 April 2019.
Kate M. Miltner, 2015. “‘One part politics, one part technology, one part history’: The construction of the emoji set in Unicode 7.0,” paper presented at the National Communication Association 101st annual convention (Las Vegas).
Lisa Nakamura, 2002. Cybertypes: Race, ethnicity, and identity on the Internet. New York: Routledge.
Michael Omi and Howard Winant, 1986. Racial formation in the United States: From the 1960s to the 1980s. New York: Routledge & Kegan Paul.
Kaleigh Rogers, 2015. “Is it weird to still use white emojis?” Motherboard (24 April), at https://motherboard.vice.com/en_us/article/9akm45/is-it-weird-to-still-use-white-emojis, accessed 29 April 2019.
Ali Sammel, 2009. “Turning the focus from ‘Other’ to science education: Exploring the invisibility of whiteness,” Cultural Studies of Science Education, volume 4, number 3, pp. 649–656.
doi: https://doi.org/10.1007/s11422-009-9184-7, accessed 29 April 2019.
Luke Stark and Kate Crawford, 2015. “The conservatism of emoji: Work, affect, and communication,” Social Media + Society.
doi: https://doi.org/10.1177/2056305115604853, accessed 29 April 2019.
Zing Tsjeng, 2014. “Apple pledges to make emojis more ethnically diverse,” Dazed (26 March), at https://www.dazeddigital.com/artsandculture/article/19367/1/apple-pledges-better-ethnic-diversity-for-emoji, accessed 29 April 2019.
Paige Tutt, 2015. “Apple’s new diverse emoji are even more problematic than before,” Washington Post (10 April), at https://www.washingtonpost.com/posteverything/wp/2015/04/10/how-apples-new-multicultural-emojis-are-more-racist-than-before/, accessed 29 April 2019.
Allison Underhill, 2015. “The ‘diversity’ of emojis,” Huffington Post (7 April), at https://www.huffpost.com/entry/the-diversity-of-emojis_b_7038798, accessed 29 April 2019.
Unicode Consortium, 2019. “Emoji and pictographs,” at http://www.unicode.org/faq/emoji_dingbats.html, accessed 29 April 2019.
Unicode Consortium, 2018. “Unicode emoji,” (21 May), at https://www.unicode.org/reports/tr51/tr51-14.html, accessed 29 April 2019.
Unicode Consortium, 2017. “Diversity,” at https://www.unicode.org/reports/tr51/tr51-12.html#Diversity, accessed 29 April 2019.
Unicode Consortium, 2014. “Diversity,” at https://www.unicode.org/reports/tr51/tr51-1-archive.html#Diversity, accessed 10 June 2019.
Received 30 April 2019; revised 10 June 2019; accepted 12 June 2019.
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Technically white: Emoji skin-tone modifiers as American technoculture
by Miriam E. Sweeney and Kelsea Whaley.
First Monday, Volume 24, Number 7 - 1 July 2019