Category: Critical DH

Throughout the years, I have become interested in a variety of different topics. And whenever these interests compels me to begin some form of research, whether it be for a blog or for my own writings, my first step is always to create a new Google Account. Although I recognize that this may be perceived by some as a strange practice, I find that having these separate accounts allows for the information pertaining to my subject of interest to be neatly categorized and to not be lost in the seemingly endless sea of content that I browse through daily.

Some of you may be asking yourselves, “why would one bring up his or her numerous Google Accounts in blog post about Algorithms?” The answer is simple. Because I am constantly creating new Google Accounts pertaining to a single topic, these new accounts are not influenced by previous searches made on the same account. What this means is that I Interact with Google’s search algorithm – tailored to new users – on the daily. Therefore, I can make the safe assumption that the results I receive on these accounts are influenced solely by Google’s search algorithm.

When beginning my research on motorcycle racing for my blog dedicated to the subject, I began by searching the phrase, ‘motorcycle racing’ using Google’s search algorithm. My results did not yield information on individual racers or the most recent racing events but the Cycle World news page. While this website contained brief snippets of information pertaining to cyclists and racing events, its focus was fixed ont he motorcycles being produced by companies such as Kawasaki and Honda. Where were the links to Instagram accounts of cyclists? Where were the photographs of these cyclists on the circuit?

In her essay, Algorithms of Oppression: How Search Engines Reinforce Racism, Safiya Noble states that Google’s search algorithm yields results that favor particular organizations based on their monetary influence. Although I cannot state with impunity that Cycle World has given exorbitant amounts of funds to Google for their place on the top of my search results, their connections with different motorcycle production companies and racing events – both high grossing industries – leads me to believe that they are in the position to reinforce their standing.

However, this problem seems not only to affect racing fans but historians and students as well. When the phrase, ‘Civil Rights Movement’ is searched using Google’s search algorithm, only a text box containing general historical information on the subject appears. To make matters worse, the information provided states that the Civil Rights Movement took place starting in the 1950s and ended in the 1960s. However, this movement began as early as the 1930s during FDR’s New Deal Administration and has continued even to this day.

These examples should remind us that the results we receive using Google’s search algorithm are not always the most reliable – they are not even unbiased. We must rely on ourselves, the users, to find credible, relevant information on our subjects.

Algorithmic criticism is the concept of looking at algorithms with a critical perspective. Scholars studied the idea that algorithms go beyond mathematic and impact race, class, and gender. Scholars adopted intersectional critiques to understand algorithms.

The transition from the old traditional medias and the new digital medias has been adopted by most of us, but it is not necessarily understood. We are framed and driven by algorithms in our everyday life without knowing what it exactly implies. Algorithm in marketing is seen as an objective fact; it is actually an opinion embedded in math. ¹

The public is in need of algorithmic literacy ² In other words, people are unaware of the impact of search engines, digital medias, their mechanisms, and how they are part of it.

Search engines are part of our life and most people believe that it “provides access to credible, accurate, depoliticized and neutral information”³. In fact, a search engine is a “symbiotic process”: it informs, but we also feed it with information which it will measure for our next search.  In other words, we are acting a major part in the evolution of digital media. We both feed from each other.

Unfortunately, search engines use a rank system where paid advertising creates inequalities. Indeed, the negative representation of women through the ranked brings back to life women’s discrimination of lack of status. As a matter of fact, scholars noticed a “direct mapping of old media traditions into new media architecture” ⁴

Search engines algorithms can lead to racism and sexism. Noble mentions that the “pornification of black women” appears as a top search result. How can this be a top-ranked research? Who decides that this was the best information to be top ranked?

We have to understand the way algorithms are built. We first curate our data. Then we should define “success” ⁵ We embed our values into an algorithm.

This leads us to understand that a group of people created algorithm which allowed the “pornification of black women” as being a top search result.

As a matter of fact, if we create this artificial intelligence, this means that we can change it. It is not only math. The latter is the tool to apply the will of a human.

We should remember that many algorithms have been created to satisfy a neoliberal capitalist society with profits in its center. The system leads to misrepresentations and hypersexualisations of black women ⁶. Search engines have commercial goals first, and “labortainment” is a main source of benefits for digital media.

In other words, the latter defines “users who consent to freely give away their labor and personal data for the use of [a private company] and its products, resulting in incredible profit for the company”. ⁷ Users give their agreement to share their private information with digital media for profit use. Most people accept the agreements and policy without reading them.

We must be careful when using search engines, when signing online documents and much more. From commercial use, marketing targets, advertisements purposes, profits, selection, and discrimination, digital media is full of obstacles for us and we need to understand that there is more at stake than what we thought until now.

There is bias in search. It is a fact. What is even more shocking, is that norms reflect racist and stereotypical ideas. Content is screened and assessed using those norms. As a result, racism, sexism, and abuse of humans is normalized in a way that it is considered as perfectly acceptable.

We must learn from algorithm literacy and change the “consciousness embedded in artificial intelligence”, to become better consumers of digital media (Noble, 2018).

The 1930s were also marked by their own challenges. They were years of movements and dynamism, when the Long Civil Rights Era (LCR) was fighting to take the place of an old feudal system.

The New Deal helped to make the LCR era a subject for political debate. Roosevelt himself name women in key positions within the administration. Frances Perkins was the first female cabinet secretary. Mary McLeod Bethune was an African American advisor in the National Youth Administration. The First Lady Eleanor Roosevelt was herself the key advisor of the president and a major voice for economic and racial justice. ⁸

It was difficult for women to attain economic autonomy. While the New Deal programs helped to raise a debate about inequalities in the society, it also kept gendered assumptions. For example, “men serve as breadwinners” and “women serve as mothers, homeworkers, and consumers”.

The emergence of the LCR era was the sign that the nation was turning slightly left. Americans started to favor socialism and movements like the Popular Front brought a left-liberal cooperation with the Socialist Party of America and Communist Party USA (Ritterhouse, 2017).

Before being a party, the Popular Front was also a historical bloc. The Popular Front was an insurgent social movement forged from the labor militancy of the fledging CIO (Congress of Industrial Organizations).

The LCR era was marked by expanding democracy and guarantee the rights of citizenship to a broader segment of the American people, including people of color. ⁹

The Long Civil Rights Era throughout the New Deal and even later, helped to raise debates and fight inequalities. We live in a place today where we take for granted rights which were acquired thanks to people who fought for them at a time in the past. We shall always keep in mind this because we can lose our rights faster than we can obtain them.

There is actually a great parallelism between WPA-era activists and contemporary digital activists. While observing through the critical lens within algorithmic criticism, scholars put in excerpt the idea of the 1930s where we have to fight inequalities in a system where racism, sexism and much more have been normalized.

Scholars are now facing algorithms and the discrimination embedded in mathematics. Contemporary techno and algorithmic critics are working on developing tools to fight against inequalities in algorithms.

Open-source software are one of the ways found to fight against private interest of big companies who seek profits. The algorithms are public, and anyone can have access to it for free and modify it.

While Open-Source software can be a great weapon against inequalities in algorithms, scholars and algorithmic critics are still working hard to point out how we can be manipulated by this era of numbers, controlled by profits, and willingly or unwillingly mapping the mistakes of the past.

Before our readings on Algorithms and the Long Civil Rights Movement, I had never thought that discrimination could cross over into the digitized, artificial intelligence world. Of course, as a white woman, I have been ignorant to the ways this has impacted specifically women of color. However, upon reflection I can now see even more ways this occurs than just the example, given in Safiya Noble’s work, of the google search results of the phrase “black girls” turning up unwanted pornographic searches. I have noticed when I search for makeup products the first results are always aimed at white woman, and even if these companies offer products for darker complexions, they are not nearly as advertised as much or in the same way they are for lighter complexions. That experience especially stood out to me when Noble explained that “search results reflect the values and norms of the…company’s commercial partners and advertisers and often reflect our most demeaning beliefs, because these ideas…are normalized and extremely profitable.”¹⁷ Reading this allowed me to understand better the meaning behind the Long Civil Rights Movement and why the fight for African American civil rights continues in modern America.

The 1930’s were particularly difficult for women and people of color as many of the New Deal programs were directed towards men and left out, or discriminated against, racial minorities. Although this time period provided hard times, particularly for minorities, it did shine a light on many inequalities within the United States. One example of this was shown through the Social Security Act which blatantly discriminated against African Americans. Southerners were “afraid that economic security would allow black southerners to escape the cycle of poverty that kept them tied to the land as cheap, exploitable farm laborers.” When Roosevelt agreed with this, he removed domestic workers and farm laborers, which were primarily black men from the South, from the Act.¹⁸ This relates almost perfectly back to algorithmic criticism. Just like how Southern politicians were profiting from cheap labor in the 1930s, companies are profiting from the objectification of black women, which is only amplified through algorithmic searches.

An algorithm, in essence, is a recipe for an outcome so desired by the person who created the recipe (Cathy O’Neil). Thus, every algorithm, no matter its nature, is subjective to the desires and goals of its creator. Algorithmic criticism has everything to do with morality, and the fight especially within research to make algorithms that will actually draw out a more focused answer to the questions posed, rather than what that most common answer is. Safiya Noble proposes the notion in Algorithms of Oppression, that it is now critical to “think about whether it truly makes sense to outsource all of our knowledge needs to commercial search engines…in lieu of libraries, librarians, teachers, researchers, and other knowledge keepers and resources.” (pg 16). This concept resonates strongly with not just drawing out the more specific knowledge we ask for, rather than the competitive answer that the engine’s algorithm drives to the top, but as well as with the ethics with which this knowledge is derived and shared. The idea of algorithmic criticism is external criticism of the results of what is output by an algorithm, is in its essence, a balance of differing opinions. Given the nature of historical research being one driven by the need to discover the truth (or the closest version of the truth to which we may get), and the biases inflicted upon these truths by the sources and the researchers themselves, it is difficult to truly change the consumption of historic media. It is still important to focus on the ethics and morality which we as historians, employ when we go looking for answers.

Noble’s chapter focusing on the re-imposition of sexism and racism in on the worlds most popular commercial search engine, highlights the longstanding strength of these ideologies, and how their being reinforced by popular majorities, against smaller parties who seek to fight it. The Long Civil Rights Movement, as well as the continuing remnants of the suffragette fights brought a strong, but somehow subtle front to the 1930s- the Civil Rights movement as we knew it to be in the 1960s was finally starting to gain speed in the 1930s, especially with opportunities for work and influence beginning to open more and more with the Federal Theater Project, eventually paving the way for such a wide variety of influential music we hear today. Women continued to fight for their rights, continuing the struggle which had initially beginning in the ‘20s, and still carries a lasting impact today, especially with jobs which had opened during the New Deal and set in place the norm of women being employed in similar spaces as men, such as is seen today. The pressures of racism and sexism however, have still not relented nearly as much as we think, and the output of search engine algorithms in present day, as noted especially in the beginning of Noble’s article, only fuel these pressing notions. Removing such significance from these digital algorithms and search engines, perhaps not so much as within popular and commercial culture, but at least in modes of historical research, can be fueled in some ways by moving back to non-digital sources.

Bibliography:

Noble, Safiya. Algorithms of Oppression: How Search Engines Reinforce Racism . New York, NY: New York University Press, 2018.

O’Neil, Cathy. “The Truth About Algorithms: Cathy O’Neil.” Vimeo, October 17, 2018. https://vimeo.com/295525907.

Ramsay, Stephen. “Algorithmic Criticism.” A Companion to Digital Literary Studies, 2008. http://www.digitalhumanities.org/companion/view?docId=blackwell/9781405148641/9781405148641.xml&chunk.id=ss1-6-7.

Algorithmic criticism is about as varied and hard to define as algorithms themselves. There are multiple lenses one can look through when critiquing algorithms: the conformist lens, the scrutinising lens, the feminist lens, the black lens, the black feminist lens, the scholarly lens, and so on. There are as many lenses for examining algorithms as there are algorithms, I feel. One thing all those perspectives can agree on is that one should always examine the established order. This goal of defiance appears to me as a lofty goal, especially when all the institutions of the web are like the monolithic ad-buildings of Blade Runner, almost primeval in their permanence.

All this is terribly reminiscent of the aims of the Popular Front movement. Just how the socialist coalition strove to battle the status quo of capitalism, modern activists strive to battle the status quo of ad agencies and complacency with those ad agencies. With the potency of folk singers and labour activists, the Popular Front encouraged alternative thinking about our institutions, tacitly urging that Americans must interact with them and mustn’t turn a blind eye to their presence. Had ‘the communist-oriented International Legal Defense (ILD)’ chosen not to come ‘to the aid of the “Scottsboro Boys”’, yet another injustice would have been carried out and America would have been further proved as a nation that does not care to challenge its status quo. Instead, the plight of the Scottsboro Boys and efforts of the Popular Front’s law firm ‘became a national symbol of continuing racial prejudice in America and a rallying point for civil rights–minded Americans.’ (Cochran et al.) ¹

The Popular Front’s stance against capitalism and racism lent a stark model for activism and interrogation of what we know (as provided by institutions). The institutions of today are perhaps more furtive, but no less insidious. Advert agencies, in service of the status quo, now implement algorithms that serve to uphold complacency to such a point that ‘the clicks of users… mean that representations of women are ranked on a search engine page in ways that underscore women’s historical and contemporary lack of status in society’. (Noble) ¹⁹

Safiya Noble, in the defiant and questioning tradition of the Popular Front activists, offers some ideas as to how to subvert and interrogate our institutions. She advocates for literacy at a basic level of Internet citizenship, maintaining ‘that one need not know the mechanism of radio transmission or television spectrum or how to build a cathode ray tube in order to critique racist or sexist depictions in song lyrics played on the radio or shown in a film or television show.’ (Noble) In other words, we should place faith in our intuition and judgment over the intuition and judgment of erroneous, as Cathy O’Neil terms it, ‘opinions embedded in math’. (O’Neil) ^{20 21}

Algorithm criticism is an important process to remember when engaging with sets of data. There are questions surrounding algorithms that could impact the data sets that are shown. Who created the algorithm? Why did they choose to do it this way? Who does this algorithm directly effect? These are a few things to consider when attempting to understand why a data set exists the way that it does. In Noble’s Algorithms of Oppression we see that there is a need to understand the power that creates these algorithms ²⁰; while a private entity can create the physical code that is the algorithm, it is dependent upon the consumers and fellow google-searchers to undermine the oppressive nature of online algorithms.

Feminism in the 1930’s proved more relevant than ever after the New Deal. The political and societal understanding of the time was that the men were the sole breadwinners meanwhile the women were to tend to the traditional house duties ²² Feminism here challenged the status quo in an attempt to show the world and the American society that women were indeed capable to provide for their families; they wanted to show that they were beneficial to society outside of the household. Today feminism still promotes the equality of women but in a different sense. No longer are women expected to remain at home to raise children but they are ever present in the workplace. An issue that feminism continues to tackle today is the gender wage gap in which women are being paid less than men even though society says that they are equal and capable of the same jobs.

WPA activists share similarities with modern-day activists in the sense that they are fighting social norms by pushing boundaries. Referring to the American Yawp‘s section XII and my previous paragraph, we see that the feminist were fighting gender norms and stereotypes by forms of protests. The modern day digital activists are challenging society’s modern day stereotypes. For example, a simple google search provides a laundry list of predicted results before even finishing your search entry. These activists want to change these predictions through understanding algorithmic code and there is an attempt to move away from the racist or sexist search tools by changing the way society accepts these results.²³

The Modern Era is a living testament to the history of the 1930’s. The Long Civil Rights Movement and the impact of feminism from the 1930’s remain prevalent to social and political movements of today. The Long Civil Rights Movement set the stage for Martin Luther King Jr.’s walk, and emphasized the importance of the set up. There are decades of set up before there is actual change within a society. Then to the modern Era those changes continue to develop and grow into movements like the Me Too movement and the Black Lives Matter movement. However, these movements and changes in society and political policy can be severely hindered by technological algorithms.

Algorithms essentially create a pattern of habit and keep people in it. An Algorithm is “Opinionated math”. This habitual box keeps the system from growing and changing with the needs of the times. On top of that people are not allowed to know how the Algorithms work. If Algorithms from the 1930’s were used in the same way they’re used today in systems then the definition of rights would have never changed for the better. There is no room for growth in a closed system.

There is no room for change unless there is revision of the algorithms.

Civil Rights and Feminism were a critique of the behavioral “algorithms” of the 1930’s, and if people treated behavior in the same way people treat technological algorithms of today there would have been no change. We take our history for granted and don’t take notes.

But there is a call for the reevaluation of how and why algorithms are written and administered. People treat algorithms as TRUTH rather than what they really are, “Opinionated Math”. Math and technology is only as good as the people behind it. In the same sense they are fighting for the value of equal treatment and American values of freedom. They’re fighting in very different ways but for the same ideas. Different times same issues. Algorithmic Criticism is fighting for substantial growth, Today we are taking our knowledge of humanity and showing people the human behind the mask of math and Technology. Algorithmic Criticism reintroduces the humanities into the sphere of technology. I believe part of the reason we have had such an issue with the revision of Algorithms is due to deification of math and Technology and the disregard for the place of humanities within the system.

The Long Civil Rights Movement (LCRM) had a big impact on the 1930s. In Jennifer Ritterhouse’s ” Discovering the South: One Man’s Travels Through a Changing America in the 1930s. “. She does a great job describing the LCRM. In this introduction, she mentions Jacquelyn Dowd Hall. She “outlined a LCRM framework that is longer, broader and deeper.” ²³. The Long Civil Rights movements in several ways. It took root in structural changes and political possibilities of the last 1930s. These changes accelerated during World War two, as well as stretching far beyond the South. It was also reshaped by the Cold War and was continuously and extremely contested. There are many Algorithm critics and they have many reasons to be critics. As Louise Matsakis’ says in the article “What does a fair algorithm look like” he mentions many downsides. For example, “algorithms can be sexist, racist and perpetuate other structural inequalities found in society. “²⁴. However, algorithms are not under any obligation to explain themselves. This causes much other bias. Many machine experts, technology companies, and event governments have called for more fairness and transparency in algorithms. As scholars have mentioned, many people struggle to agree on what constitutes a fair explanation for these algorithms. Matsaki also mentions that even lawmakers have not been able to decide what rights citizens should have when it comes to transparency in algorithmic decision-making. Another reason algorithms do not seem to be fair is because they offer more achievable resources to wealthier people, younger and people, and men. Therefore, women, elders, and people who make less money have fewer options for resources. Matsaki gives a great example of this. He says “A woman might need to lose far more weight for a health care AI to offer her a lower premium rate than a man would. We could participate better in digital media by creating a simple understanding for everyone. Many people do not think about certain codes for algorithms, just the particular reasons for the decision. In my opinion, if people were educated on this topic than they could become more involved in digital media. I had trouble finding connections between WPA-era activists and contemporary digital activists. I can say that the activists are both passionate about what they are or were trying to achieve. WPA’s main goal was to employ job-seekers and carry out public works projects. While digital activism focuses more on the internet and digital media as key platforms for mass mobilization and political action. They do both have a connection to the political side of things but other than that since they occurred at times it is complicated to find several key connections.

Ritterhouse, Jennifer. “INTRODUCTION: The Same Journey, Writ Small, That the United States Was On.” In Discovering the South: One Man’s Travels through a Changing America in the 1930s, 1-19. CHAPEL HILL: University of North Carolina Press, 2017. http://www.jstor.org.mutex.gmu.edu/stable/10.5149/9781469630953_ritterhouse.4.

Louise Matsakis’, “What does a Fair Algorithm Actually Look Like”, accessed October 10, 2019, https://www.wired.com/story/what-does-a-fair-algorithm-look-like/?