Algorithms of Oppression

Home > Nonfiction > Algorithms of Oppression > Page 8
Algorithms of Oppression Page 8

by Safiya Umoja Noble


  Challenging Cybertopias

  All of this leads to more discussion about ideologies that serve to stabilize and normalize the notion of commercial search, including the still-popular and ever-persistent dominant narratives about the neutrality and objectivity of the Internet itself—beyond Google and beyond utopian visions of computer software and hardware. The early cybertarian John Perry Barlow’s infamous “A Declaration of the Independence of Cyberspace” argued in part, “We are creating a world that all may enter without privilege or prejudice accorded by race, economic power, military force, or station of birth. We are creating a world where anyone, anywhere may express his or her beliefs, no matter how singular, without fear of being coerced into silence or conformity.”93 Yet the web is not only an intangible space; it is also a physical space made of brick, mortar, metal trailers, electronics containing magnetic and optical media, and fiber infrastructure. It is wholly material in all of its qualities, and our experiences with it are as real as any other aspect of life. Access to it is predicated on telecommunications companies, broadband providers, and Internet service providers (ISPs). Its users live on Earth in myriad human conditions that make them anything but immune from privilege and prejudice, and human participation in the web is mediated by a host of social, political, and economic access points—both locally in the United States and globally.94

  Since Barlow’s declaration, many scholars have challenged the utopian ideals associated with the rise of the Internet and its ability to free us, such as those espoused by Barlow, linking them to neoliberal notions of individualism, personal freedom, and individual control. These linkages are important markers of the shift from public- or state-sponsored institutions, including information institutions, as the arbiters of social freedoms to the idea that free markets, corporations, and individualized pursuits should serve as the locus of social organization. These ideas are historically rooted in notions of the universal human being, unmarked by difference, that serve as the framework for a specific tradition of thinking about individual pursuits of equality. Nancy Leys Stepan of Cornell University aptly describes an enduring feature of the past 270 years of liberal individualism, reinvoked by Enlightenment thinkers during the rising period of modern capitalism:

  Starting in the seventeenth century, and culminating in the writings of the new social contract philosophers of the eighteenth century, a new concept of the political individual was formulated—an abstract and innovative concept, an apparent oxymoron—the imagined universal individual who was the bearer of equal political rights. The genius of this concept, which opened the door to the modern polis, was that it defined at least theoretically, an individual being who could be imagined so stripped of individual substantiation and specification (his unique self), that he could stand for every man. Unmarked by the myriad specificities (e.g., of wealth, rank, education, age, sex) that make each person unique, one could imagine an abstract, non-specific individual who expressed a common psyche and political humanity.95

  Of course, these notions have been consistently challenged, yet they still serve as the basis for beliefs in an ideal of an unmarked humanity—nonracialized, nongendered, and without class distinction—as the final goal of human transcendence. This teleology of the abstracted individual is challenged by the inevitability of such markers and the ways that the individual particularities they signal afford differential realities and struggles, as well as privileges and possibilities. Those who become “marked” by race, gender, or sexuality as other are deviations from the universal human—they are often lauded for “transcending” their markers—while others attempt to “not see color” in a failing quest for colorblindness. The pretext of universal humanity is never challenged, and the default and idealized human condition is unencumbered by racial and gender distinction. This subtext is an important part of the narrative that somehow personal liberties can be realized through technology because of its ability to supposedly strip us of our specifics and make us equal. We know, of course, that nothing could be further from the truth. Just ask the women of #Gamergate96 and observe the ways that racist, sexist, and homophobic comments and trolling occur every minute of every hour of every day on the web.

  As I have suggested, there are many myths about the Internet, including the notion that what rises to the top of the information pile is strictly what is most popular as indicated by hyperlinking. Were that even true, what is most popular is not necessarily what is most true. It is on this basis that I contend there is work to be done to contextualize and reveal the many ways that Black women are embedded within the most popular commercial search engine—Google Search—and that this embeddedness warrants an exploration into the complexities of whether the content surfaced is a result of popularity, credibility, commerciality, or even a combination thereof. Using the flawed logic of democracy in web rankings, the outcome of the searches I conducted would suggest that both sexism and pornography are the most “popular” values on the Internet when it comes to women, especially women and girls of color. In reality, there is more to result ranking than just how we “vote” with our clicks, and various expressions of sexism and racism are related.

  2

  Searching for Black Girls

  On June 28, 2016, Black feminist and mainstream social media erupted with the announcement that Black Girls Code, an organization dedicated to teaching and mentoring African American girls interested in computer programming, would be moving into Google’s New York offices. The partnership was part of Google’s effort to spend $150 million on diversity programs that could create a pipeline of talent into Silicon Valley and the tech industries. But just two years before, searching on “black girls” surfaced “Black Booty on the Beach” and “Sugary Black Pussy” to the first page of Google results, out of the trillions of web-indexed pages that Google Search crawls. In part, the intervention of teaching computer code to African American girls through projects such as Black Girls Code is designed to ensure fuller participation in the design of software and to remedy persistent exclusion. The logic of new pipeline investments in youth was touted as an opportunity to foster an empowered vision for Black women’s participation in Silicon Valley industries. Discourses of creativity, cultural context, and freedom are fundamental narratives that drive the coding gap, or the new coding divide, of the twenty-first century.

  Part of the ethos of engaging African American women and girls in this initiative is about moving the narrative from African Americans as digitally divided to digitally undivided. In this framing, Black women are the targets of a variety of neoliberal science, technology, and digital innovation programs. Neoliberalism has emerged and served as a framework for developing social and economic policy in the interest of elites, while simultaneously crafting a new worldview: an ideology of individual freedoms that foreground personal creativity, contribution, and participation, as if these engagements are not interconnected to broader labor practices of systemic and structural exclusion. In the case of Google’s history of racist bias in search, no linkages are made between Black Girls Code and remedies to the company’s current employment practices and product designs. Indeed, the notion that lack of participation by African Americans in Silicon Valley is framed as a “pipeline issue” posits the lack of hiring Black people as a matter of people unprepared to participate, despite evidence to the contrary. Google, Facebook, and other technology giants have been called to task for this failed logic. Laura Weidman Powers of CODE2040 stated in an interview by Jessica Guynn at USA Today, “This narrative that nothing can be done today and so we must invest in the youth of tomorrow ignores the talents and achievements of the thousands of people in tech from underrepresented backgrounds and renders them invisible.”1 Blacks and Latinos are underemployed despite the increasing numbers graduating from college with degrees in computer science.

  Filling the pipeline and holding “future” Black women programmers responsible for solving the problems of racist exclusion and misrepresentation in Silicon Val
ley or in biased product development is not the answer. Commercial search prioritizes results predicated on a variety of factors that are anything but objective or value-free. Indeed, there are infinite possibilities for other ways of designing access to knowledge and information, but the lack of attention to the kind of White and Asian male dominance that Guynn reported sidesteps those who are responsible for these companies’ current technology designers and their troublesome products. Few voices of African American women innovators and tech-company leaders in Silicon Valley have emerged to reframe the “diversity problems” that keep African American women at bay. One essay that grabbed the attention of many people, written for Recode by Heather Hiles, the former CEO of an educational technology e-portfolio company, Pathbrite, spoke directly to the limits for Black women in Silicon Valley:

  I’m writing this post from the Austin airport, headed home to Oakland from SXSW. Before pulling out my laptop to compose this, I read a post on Medium that named me as one of three black women known to have raised millions in venture capital. The article began with the startling fact that less than .1 percent of venture capital in the United States is invested in black women founders. I’m not sure what sub-percentage of these are women in tech, but it doesn’t really matter when the overall numbers are so abysmal. The problem isn’t a lack of compelling women of color to invest in; it’s a system in Silicon Valley that isn’t set up to develop, encourage and create pathways for blacks, Latinos or women. Don’t just take my word for it—listen to industry leaders interviewed for a USA Today story on the Valley’s lack of commitment to diversity. Jessica Guynn reports that “venture capitalists tell [Mitch Kapor] all the time that they are ‘color blind’ when funding companies. He’s not sure they are ready to let go of a deeply rooted sense that Silicon Valley is a meritocracy.”2

  Hiles goes on to discuss the exclusionary practices of Silicon Valley, challenging the notion that merit and opportunity go to the smartest people prepared to innovate. Despite her being the only openly gay Black women to raise $12 million in venture capital for her company, she still faces tremendous obstacles that her non-Black counterparts do not. By rendering people of color as nontechnical, the domain of technology “belongs” to Whites and reinforces problematic conceptions of African Americans.3 This is only exacerbated by framing the problems as “pipeline” issues instead of as an issue of racism and sexism, which extends from employment practices to product design. “Black girls need to learn how to code” is an excuse for not addressing the persistent marginalization of Black women in Silicon Valley.

  Who Is Responsible for the Results?

  As a result of the lack of African Americans and people with deeper knowledge of the sordid history of racism and sexism working in Silicon Valley, products are designed with a lack of careful analysis about their potential impact on a diverse array of people. If Google software engineers are not responsible for the design of their algorithms, then who is? These are the details of what a search for “black girls” would yield for many years, despite that the words “porn,” “pornography,” or “sex” were not included in the search box. In the text for the first page of results, for example, the word “pussy,” as a noun, is used four times to describe Black girls. Other words in the lines of text on the first page include “sugary” (two times), “hairy” (one), “sex” (one), “booty/ass” (two), “teen” (one), “big” (one), “porn star” (one), “hot” (one), “hardcore” (one), “action” (one), “galeries [sic]” (one).

  Figure 2.1. First page of search results on keywords “black girls,” September 18, 2011.

  Figure 2.2. First page (partial) of results on “black girls” in a Google search with the first result’s detail and advertising.

  Figure 2.3. First results on the first page of a keyword search for “black girls” in a Google search.

  In the case of the first page of results on “black girls,” I clicked on the link for both the top search result (unpaid) and the first paid result, which is reflected in the right-hand sidebar, where advertisers that are willing and able to spend money through Google AdWords4 have their content appear in relationship to these search queries.5 All advertising in relationship to Black girls for many years has been hypersexualized and pornographic, even if it purports to be just about dating or social in nature. Additionally, some of the results such as the UK rock band Black Girls lack any relationship to Black women and girls. This is an interesting co-optation of identity, and because of the band’s fan following as well as possible search engine optimization strategies, the band is able to find strong placement for its fan site on the front page of the Google search.

  Figure 2.4. Snapchat faced intense media scrutiny in 2016 for its “Bob Marley” and “yellowface” filters that were decried as racist stereotyping.

  Published text on the web can have a plethora of meanings, so in my analysis of all of these results, I have focused on the implicit and explicit messages about Black women and girls in both the texts of results or hits and the paid ads that accompany them. By comparing these to broader social narratives about Black women and girls in dominant U.S. popular culture, we can see the ways in which search engine technology replicates and instantiates these notions. This is no surprise when Black women are not employed in any significant numbers at Google. Not only are African Americans underemployed at Google, Facebook, Snapchat, and other popular technology companies as computer programmers, but jobs that could employ the expertise of people who understand the ramifications of racist and sexist stereotyping and misrepresentation and that require undergraduate and advanced degrees in ethnic, Black / African American, women and gender, American Indian, or Asian American studies are nonexistent.

  One cannot know about the history of media stereotyping or the nuances of structural oppression in any formal, scholarly way through the traditional engineering curriculum of the large research universities from which technology companies hire across the United States. Ethics courses are rare, and the possibility of formally learning about the history of Black women in relation to a series of stereotypes such as the Jezebel, Sapphire, and Mammy does not exist in mainstream engineering programs. I can say that when I teach engineering students at UCLA about the histories of racial stereotyping in the U.S. and how these are encoded in computer programming projects, my students leave the class stunned that no one has ever spoken of these things in their courses. Many are grateful to at least have had ten weeks of discussion about the politics of technology design, which is not nearly enough to prepare them for a lifelong career in information technology. We need people designing technologies for society to have training and an education on the histories of marginalized people, at a minimum, and we need them working alongside people with rigorous training and preparation from the social sciences and humanities. To design technology for people, without a detailed and rigorous study of people and communities, makes for the many kinds of egregious tech designs we see that come at the expense of people of color and women.

  In this effort to try and make sense of how to think through the complexities of race and gender in the U.S., I resist the notion of essentializing the racial and gender binaries; however, I do acknowledge that the discursive existence of these categories, “Black” and “women/girls,” is shaped in part by power relations in the United States that tend to essentialize and reify such categories. Therefore, studying Blackness is, in part, guided by its historical construction against Whiteness as a social order and those who have power given their proximity to it. I make comparisons in this study of Blackness to Whiteness only for the purposes of making more explicit the discursive representations of Black girls’ and women’s identities against an often unnamed and unacknowledged background of a normativity that is structured around White-American-ness. I do believe that the results of my study on identities such as White men, boys, girls, and women deserve their own separate treatment using the extensive body of scholarship in the social construction of Whiteness and a cri
tical Whiteness lens. This study does not deeply discuss those searches in this way. I am not arguing that Black women and girls are the only people maligned in search, although they were represented far worse than others when I began this research. The goal of studying representations of Black girls as a social identity is not to use such research to legitimize essentializing or naturalizing characterizations of people by biological constructions of race or gender; nor does this work suggest that discourses on race and gender in search engines reflect a particular “nature” or “truth” about people.

  It is more interesting to think about the ways in which search engine results perpetuate particular narratives that reflect historically uneven distributions of power in society. Although I focus mainly on the example of Black girls to talk about search bias and stereotyping, Black girls are not the only girls and women marginalized in search. The results retrieved two years into this study, in 2011, representing Asian girls, Asian Indian girls, Latina girls, White girls, and so forth reveal the ways in which girls’ identities are commercialized, sexualized, or made curiosities within the gaze of the search engine. Women and girls do not fare well in Google Search—that is evident. My goal is not to inform about this but to uncover new ways of thinking about search results and the power that such results have on our ways of knowing and relating. I do this by illuminating the case of Black girls, but undoubtedly, much could be written about the specific histories and contexts of these various identities of women and girls of color; and indeed, there is much still to question and advocate for around the commercialization of identity in search.

 

‹ Prev