Algorithms of Oppression
HOW SEARCH ENGINES REINFORCE RACISM
RITA ODEY
“Black girl, who do they say you are?”
In 2010, author Safiya Umoja Noble googled the words ‘Black girls’ to find something engaging to show to her young nieces and the search engine told her what it thought she and her nieces were: Overtly sexualized beings present only in pornographic materials. This experience and similar others led Noble to create this work under review: Algorithms of Oppression: How Search Engines Reinforce Racism. Although the book delves into algorithmic systems, its overarching message is one I perceive to be identity politics. As a Black woman, my identity is ever under scrutiny. If I wear my hear in an afro, I must be a social justice warrior. If I straighten it, I must be whitewashed. Who says what I am? Who files me into boxes based on their classification? According to Noble, Google does.
This book is especially important because for far too long, Black women’s identities have been defined by people other than Black women. Tags like ‘mad, loud, mean, pretty for and close enough’ have been created and Black women are fixed into them depending on the situation. This book brings to surface the concept of racial and sexist classification of marginalized groups and the harmful effect of these classifications. Noble sojourns into the often-overlooked intersection of Black and woman to explore the algorithms on search engines like Google and how marginalized groups bear the brunt of racism and sexism encoded on these platforms. She researches into the identities and positioning attached to especially women of color on search engines. Search engines like Google act as a window into the internet and Noble argues that what is being seen or not seen about Black women among others is problematic. She approaches the subject as I would, from a Black Feminist Perspective. In occupying both spaces, she is able to present a complete analysis of the phenomenon.
The Assumption of Search Engine Neutrality
Google positions itself as a neutral and infallible informations system. A staple in our everyday lives, its influence on human culture is so extensive, it is its own verb. Noble explains that internet users see Google as a credible source of information. In 2012, 68% of search engine users believed that search engines were fair and unbiased. However, Noble warns that the assumption of the neutrality or the fairness of Google is destructive simply because it is not. Google is an advertising platform and plays to the advertisers dollars and the user’s clicks. This inherent capacity renders it far from neutral and credible since it can be swayed by the highest bidder.
Search engines oversimply complex phenomena. Stereotypes with long painful racial history are presented simply as page 1 of most relevant items for commercial gain. The pornification of Black bodies on Google is not as simple as Black women in porn but goes back to the sexualization and objectification of Black bodies during slavery and beyond. When search engines use algorithms to prioritize information for profit, it compromises our ability to objectively engage with complex ideas. Classifications and stereotypes are created by the dominant group to the detriment of the marginalized and when these find their way into our knowledge systems, they are legitimized. What then does this hold for the identity of Black girls when porn is at the top of the list associated with them?
Noble does not start off denying the presence of Black female bodies in pornography. She instead goes for the algorithmic system that sets racial and sexist stereotypes about marginalized bodies as the identifier of those bodies for profit. The appearance of sexualized Black girls on searches for Black girls affords a society that already believes in the omniscient power of Google to believe this as the key space Black women occupy. This research presents case upon case of Google’s algorithms identifying Blackness incorrectly and presenting racist and sexist results. From tagging European style hair as professional and Black hair as unprofessional, to the Black woman whose photo was categorized by the algorithm under Gorilla, the lapses are endless. Once brought to the attention of the public, the issues are quickly rectified. The fact that these algorithmic lapses can be corrected leaves the author asking why they exist in the first place. How many of such stereotypes are encoded in the technology that we use? I certainly have never heard of Google misidentifying White people. Who is on the back end and who is not? How often have you searched for anything only to find that you must input the word Black to find something that relates to you? Who categorizes results into dominant and others?
The Make Up of Silicon Valley
An interesting note that the book makes is when Noble talks about the creation of Black Girls Code, a laudable effort that creates opportunities for Black girls to receive technological education. However, Noble discusses this as a move that suggests that presently, there are no Black women with the degrees or experiences to be hired into Silicon Valley and they must be trained from the next generation. This is telling of the marginalization of Black women and women of color in tech spaces. The racial and sexist connotation of the Valley’s preference for historically White colleges is not ignored. In 2015, Google hired only one student from Howard university, a historically Black college. According to a 2020 article in The Economist, Black people make up 3% of top five tech firms in America. Black women account for even less at 3% of the entire technology industry work force. Noble suggests that the racial makeup of the Silicon Valley is mostly White and some Asian males because Blacks are stereotyped as “non-technical” and women even more so. Therefore, we see that problematic algorithms and systems are created by people who see no one but themselves and ultimately code their prejudices into the system. These people whose default for good is White and bad, others. This default is ultimately Google’s default. Marginalized groups are not involved in the conversation around technology design and the impact on different groups of people. Noble points directly at the education of people who work in the industry. Proficient as they are in their field, political and cultural discussions around technology are not held. They are not taught about cultural histories of the users of their products.
Noble goes further to discuss the history of these stereotypes and classifications. She argues that they predate the internet and are present in traditional media like film and the Library Indexing System. As these bodies of knowledge have taken up residence on the internet, so have the stereotypes they hold. She suggests that search engines like Google are not the infallible body of knowledge that one might think. They inform only because they have been informed. She explains that they are informed largely by dominant groups with preconceived notions of the marginalized groups and these in turn inform both the dominant and marginalized shaping identity politics on both sides. Search engines have far too much power over who is and how they are. Today, a lack of presence online supposes a lack of personhood. If a business does not show up on Yelp, it is synonymous with non-existence and if a location is not on Google maps, then does it even exist? We are assumed to be who it says we are, and we cannot be forgotten or deleted because the internet never forgets.
Overall, Noble ties the elements of the book together seamlessly, not minding the fact that it was written over years and technology and the internet have evolved speedily. The key concepts that the book underscores like racism and identity and commodification of marginalized groups unfortunately do not change as quickly. This means that the relevance of this book holds, regardless of the year. The unspoken theme of identity is one that is carried brilliantly throughout the book when others like algorithms fall off. As she has done, Noble suggests that we carry on these conversations about how identity is presented by algorithms and for whose benefit.
“So, Black girl, who are you?”
Bio:
My name is Rita Odey, and I am a freelance culture and lifestyle writer. I also work as an advertising and marketing professional. I developed an overactive imagination from reading, and I began writing at the age of 11. Currently, my writing explores the intersection between Blackness and my personhood. Being born and raised in Nigeria, my Blackness went without saying and remained unexplored. When I moved to Vancouver for a master’s program, I was confronted with the “otherness” that compels Black identity to find expression. This internal struggle is what shapes my thought. Writing affords me the right to be heard; this right that was fought for by Black women before me. When I’m not working, I’m reading everything from topics on Christianity to dairy-free recipes.