Of making many books there is no end ~ Ecclesiastes 12:12.
Higher education is undergoing a major upheaval at present, due to COVID-19, #BlackintheIvory on twitter, and Black Lives Matter movements and protests world-wide. Higher education is under scrutiny, as it ought to be. We must rethink and revise and rebuild higher education. So anyone who is teaching/learning in higher education these days needs to be aware of the content of these two books, Algorithms of Oppression and Data Versus Democracy, available to us for some time but of special urgency now, the spring and summer of 2020. Many schools are now regularly using algorithmic proctoring software and plagiarism detection tools, and the use of these tools has only increased over the last several months.
Our pedagogical and social lives are often guided by and, too often, controlled by what we can find on the internet. And what it can find about us.
Algorithms are not constructed or developed in a pure and magical space, filled only with light and good will. Conditions for construction of algorithms should be much better known. For example, algorithms, from the get-go, have been developed by (mostly) white men, sometimes ones with the emotional age of twelve-year old boys. One must ask where are the missing women creators and BIPOC creators of algorithms? We are all affected by what algorithms let us see; everybody is thus under the sway of algorithms.
These two excellent books, laying bare much-needed history of and information about algorithms, should be read together and separately and then, later, thought over together and separately. The authors bring us timely warnings. Will we heed them?
First book designed and guaranteed to frighten us.
Safiya Umoja Noble. Algorithms of Oppression: How Search Engines Reinforce Racism. Pp. 229. New York: New York University Press, 2018.
I read Safiya Umoja Noble’s Algorithms of Oppression when it was first published. Since then it has gained a world-wide following. Upon re-reading it, I am more convinced than ever that it is a crucial text for guiding us through the mess we are in. This book is consistently well-argued and well-documented, bringing clear evidence of how algorithms are unconsciously and deliberately biased. Her language is easy to grasp, and her examples are often disturbing and at times horrific. But the reader should press on, as each chapter is a gem in and of itself.
Chapter One, “A Society, Searching” treats the structure of how the internet works, how search engines come up with their offerings and rankings. And the author shows, with blinding clarity, how contextualization is much needed. She notes that such context is needed for all search engines, but particularly for the omni-present Google Search. Her own google search for “black girls” in 2012 led her to a detailed analysis of the problems with search engine results. She notes that 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” (65). While she notes that things are marginally better now, she also indicates that we have a long, long way to go.
Chapter Two, “Searching for Black Girls,” adroitly shows “how women, particularly Black women, are misrepresented on the internet in search results and how this malfeasance is tied to a longer legacy of White racial patriarchy” (107). The path to this conclusion is a dismaying, terrible, even horrifying one, illustrated with cogent examples that she provides and with the data she adduces.
Everyone in higher education might want, as I did, to engage in the google searches that Noble undertook. When I did a google search at the end of 2019 for “black girls,” the first three sites that emerged were:
1. Best Black dating sites
2. #blackgirls hashtag on instagram
3. Hot Black Girls (45 pictures)
When I googled “black women,” I found “Hot Black Women” as the fifth result, with the first being “Black Online Women Meet Singles Near You.” The results I found for “white girls” and “white women” were also depressing but not nearly as awful. When I googled “black boys,” I was offered a lot of sites about hair in the top results. When I googled “black men,” the three top links were to dating sites and the fourth was to amazon.com. Sigh.
Chapter Three, “Searching for People and Communities” is a brief analysis of a particular case, Dylan Roof, who did a google search for “black on white crime” in 2015. He was taken directly to the site for the Council of Conservative Citizens, a White supremacist site known and condemned as a hate group by the Southern Poverty Law Center. Roof, a White man now sentenced to death, drew horrific conclusions and subsequently murdered nine worshippers in the Charleston Emanuel African Methodist Episcopal Church, a noted symbol for “African American freedom in the United States” (110). This chapter shows, in detail, how Roof was [mis-]led by the algorithms of Google Search, noting, for example, that his search did not lead him to FBI crime statistics on violence, which point out that most crime against White Americans is committed by other White Americans. Further, the phrase “black on white crimes” did not lead him “to any experts on race or to any universities, libraries, books, or articles about the history of race in the United States and the invention of racist myths in service of White supremacy” (115). The invidious results of Roof’s search, cross-verified by the author, clearly led to his White on black crime.
When I was drafting this review, 2019 was drawing to a close and national newspapers were reporting that Grafton Thomas has been charged with federal hate-crimes for attacking Hasidic people with a machete in the home of a rabbi in Monsey NY. His cell phone records show he searched at least four times for “why did Hitler hate Jews” and he also searched for “synagogues near me.” The Washington Post notes, “Saturday’s stabbing was the 13th anti-Semitic incident in three weeks in New York state,” (31 December 2019). In June 2020, when I entered “George Floyd” into my search on Chrome, the first item was the Wikipedia account, “Killing of George Floyd,” as is the second item. The third line offers up the horrific videos of his murder.
Again, I think readers of this book would probably want to undertake google searches themselves as we all try to work toward greater understanding of the invidiousness of algorithms offering us what we are supposed to think about. In my end of year 2019 results for a search of “black on white crime,” I was offered these links:
- Black on White crime video results
- White on Black Crime v. Black on White Crime: New Statistics
- Black versus White Crime: White Privilege Isn’t Real (linked to an article published 20 November 2019, hence an item heavily read just before I searched)
- Black on White Crime - Image Results
- Amazon.com Black Racism, White Victims (Second Edition)
- White supremacists’ favorite myths about black crime rates
Sigh indeed.
Chapter Four, “Searching for Protections from Search Engines,” begins with examples of what we now call “revenge porn” or its baby brother, which I call “gotcha porn.” The right to be forgotten, and not be dredged up forever via the internet, is under scrutiny in social areas and in legal circles in the USA. Noble focuses on the “struggle over power, rights, and notions of what constitutes freedom, social good, and human rights to privacy and the right to futures unencumbered by the past” (121). And one of her key points is this: “a difficult aspect of challenging group versus individual representations online is that there are not protections or basis for action under our current legal regime [in the United States]” (122). Data retention is a serious business. It is very profitable for commercial companies. We need to figure out how to do such retention responsibly and humanely rather than automatically in bulk and forever, as we are doing now. Right this minute. I think it is up to all of us to figure out how and when to destroy data, to banish it from searchability. We are well-accustomed to paper-shredding those documents that no longer serve a purpose or that should be preserved; surely there be digital data that also needs to be sunsetted.
Chapter Five, “The Future of Knowledge in the Public” discusses library classification systems and their similarity to search engines. Noble makes clear her belief that Critical Race Theory “as a stance in the field” (of LIS, Library and Information Science) would work for the good of all. Her telling discussion of the struggles over library subject heading terms “illegal alien” and “Jewish question” is part of a consideration of the problem in classifying people, to the detriment of people not in the majority of groupings. Her short history of the misrepresentation in classifying people points to the ground-breaking work of Sanford Berman, who inveighed against the Library of Congress Subject Headings as containing “Western racial bias” (1971 and following), and to the more recent work of Hope A. Olson. Noble briefly discusses search engines much smaller than Google. Her assessment, for example, of Blackbird illustrates how these can be so useful (151). Her conclusion calls for “public search engine alternatives, united with public-interest journalism and librarianship, to ensure that the public has access to the highest quality information available” (152).
Chapter Six, “The Future of Information Culture.” Noble asserts, “Access to high-quality information, from journalism to research, is essential to a healthy and viable democracy” (153). Noble’s brief discussions of how advertising and commercial interests have adversely affected journalism, how media stereotypes mask unequal access to life in the USA, and how corporate takeover of public-interest information have adversely eroded professional standards is a very, very chilling one. Noting that “the higher a web page is ranked, the more it is trusted” (155), she points out that there is no fact checking, that the legitimacy of these sites “is simply taken for granted” (155). And this chapter also brings warnings about a so-called monopoly on information and on why public policy matters. Her brief discussion of the digital divide rightly calls out the “too narrow” framework that focuses “on the skills and capabilities of people of color and women, rather than questioning the historical and cultural development of science and technology and representations prioritized through digital technologies, as well as the uneven and exploitive global distribution of resources and labor in the information and communication ecosystem” (160).
The “Conclusion” calls for using black feminist technology studies in doing Internet research. She recounts an extensive interview/conversation with the pseudonymous Kandis, a small business owner whose local African American hair salon, was adversely impacted by the rise of the Internet. Word-of-mouth used to send patrons to her but with the advent of reviews on yelp, she lost business. And among other issues, Kandis explains how yelp’s costs harmed her business significantly (177). And in commenting on the apparent normality of seeing just a few possible results when doing a search, Noble writes, “this ‘normal’ is a direct result of the way that human beings have consciously designed both software and hardware to function this way and no other” (179). Noble’s call for new tools, new studies should be heeded.
Noble’s poignant “Epilogue” to this volume was written after the results of the 2016 presidential election in the USA were known. Noble wryly comments that she argued “’for years about the harm toward women and girls through commercial information bias circulating through platforms like Google,’ [but] no one has seemed to care until it threw a presidential election” (183). Many of us in higher education are extremely worried about potential tampering with the 2020 US presidential election. Noble and her twitter followers sometimes tweet about the serious issues arising with this looming election. She and they should be followed, IMHO, as we say.
Second book guaranteed to scare the bejeesus out of us.
Kris Shaffer. Data versus Democracy: How Big Data Algorithms Shape Opinions and Alter the Course of History. Pp. 120. Apress, 2019. Distributed to the book trade by Springer+Business Media, New York.
This excellent, brief book brings us timely warnings. Will we listen? Kris Shaffer explains cogently and clearly why we should be concerned about big data. He offers two sections: “the propaganda problem” and “case studies.”
As an Introduction to Part One, “The Propaganda Problem,” Shaffer notes that the human cognitive system, “the interaction of the brain and the body, memory and the senses—took its more-or-less modern form tens of thousands of years ago” (xii-xiv, using Eric Wayman as a source). He provides a brief, percipient discussion of cognitive hacking and where we might be headed, (pp.xiv-xvii).
In Part One, “The Propaganda Problem,” Shaffer discusses the sheer “information abundance” that we find flooding into our daily lives, laying out general “economic, cognitive, and technological backdrop for the emergence of those algorithms” (3). These pages make the reader very uneasy, for Shaffer goes beyond the headlines and brief commentaries found in our reputable daily newspapers. Each of his first 3 chapters, “Pay Attention: how information abundance affects the way we consume media,” “Cog in the System: how the limits of our brains leave us vulnerable to cognitive hacking,” and “Swimming Upstream: how content recommendation engines impact information and manipulate our attention,” ends with a percipient summary of his discussions.
As a segue into Part Two, “Case Studies,” he offers hope. “But embedded in some of these case studies—and the way the public, governments, and/or platforms have responded to them—is hope as well” (44).
In “Case Studies,” we move into territory that is frequently frightening. Chapter 4: “Domestic Disturbance: Ferguson, Gamergate, and the Rise of the American Alt-Right” lays out with precision how important twitter has been for these 3 nodes of contention in our culture. Shaffer notes that protests and rallies over the murder of Michael Brown were on twitter and Vine for two days before national/mainstream media started carrying the stories (47). As I was writing this commentary in December 2019, twitter was recounting how Courtney Milan, a Romance writer, has been judged unethical by the Romance Writers of America organization. There were tweets in my feed for days before the national and international media picked up the story. And of course in 2020, the horrors of the murder of George Floyd were abundant in the twittersphere before the national media jumped on the story.
Next, Shaffer’s discussion of the viciousness and outrages of GamerGate is especially horrifying to this reader, who got set on fire (or “flamed” as it was then) many a time back in “the olden days” when making comments in those long-forgotten Chat Rooms. Regularly excoriated, I was usually called “dude” since my name did not reveal my gender. To the cost of all of us, we know that GamerGate has inflicted untold misery on undeserving women who dared to game. These days, one’s gender nearly always gets identified and it is no surprise that all women, POC, LBGTI people, those who are not white males get it in the neck. Or get trolled. Or get doxxed. Etc. Is it likely to change? Probably not, because Shaffer’s poignant discussion of the rise of the alt-right shows its firm grip. This chapter also sketches issues raging in fall 2019 and spring 2020 on twitter that have surfaced in discussions about medieval literature, the use of the term “Anglo-Saxon,” and the concept of “RaceB4Race.”
Chapter 5, “Democracy Hacked, Part One,” lays out Russian interference in the 2016 presidential election in the United States and concomitant stirrings up of trouble in Ukraine, in the Baltic, and by Fancy Bear (the name given to a Russian military intelligence team [the GRU] known as APT28). Recalling the illustrative film, War Games, (1983), Shaffer lays out a hacker’s mindset quite cogently, noting that Fancy Bear chose email in 2016 for spear phishing. When being spear phished, the recipient of an email is told there is a problem and then is supposed to click on a link in the email to “provide your personal login details—including your username and password—to solve the problem” (79). This way madness lies, as many know to their cost. Shaffer also discusses the focus on Clinton’s emails, especially when FBI Director Comey announced he was reopening an investigation into her use of a private server. Shaffer says, “the GRU’s release of kompromat on Clinton and the DNC meant that Russian got to set a significant portion of the agenda of the presidential election” [italics Shaffer’s] (82). This chapter continues with additional analysis of the incredible current-day extension of what in Nixon’s day were called “dirty tricks.” These include creation of fake people on Facebook, ads placed on Facebook, and messaging directed at “three general communities—right-leaning Americans, left-leaning Americans, and African Americans” (86). According to Shaffer, these operations are on-going.
Chapter 6, “Democracy Hacked, Part 2,” focuses on rumors, bots, and genocide in the Global South and reading it also makes one want to tear one’s hair out and set it on fire. Evidence comes from Egypt, Brazil, the Philippines, Myanmar, and the Latin American elections of 2018. Shaffer’s summary for this chapter includes this bleak statement: “the problem of online disinformation is bigger and more diverse than many in the West realize,” (p.108). He says that the problem is global, that it is a human problem, that it is “fueled but not created—by technology,” (p.108). One must do more than merely sigh at this state of affairs.
Chapter 7, the conclusion, rehearses and summarizes Shaffer’s findings throughout. One of his cogent cautions about disinformation is that “no amount of individual fact-checking will solve this problem. This is first and foremost a behavioral issue that can only be seen and solved with the analysis of data on the large scale: cross-community, cross-platform, over long stretches of time” (113). Shaffer thinks it is possible to restore integrity. Tellingly, the final word, that word being “good,” in this book offers hope, as Shaffer concludes, “The problem likely won’t go away anytime soon. But that shouldn’t stop us from working to minimize its effects and harness the power of new technologies for good” (115). Hope. Good. Let’s hope we find ways forward that address the enormous problems before us.
I got my first computer in 1957, a GENIAC kit, and I still have it. Algorithms for the GENIAC and the ENIAC are of little interest today. I went on to do research in the late 1960s on 100+ texts of Shakespeare’s *A Midsummer Night’s Dream.” Fretting about algorithms still was not my top priority. Rather, I was focused on making my coding in FORTRAN IV for the IBM 360-75 actually work and behave and do what I wanted it to. In the end, I was successful. But then technology started leaping and bounding and leaping and bounding some more. I have watched the developments over decades and with advances, I have become much less sanguine about whether these machines and their algorithms are going to accomplish what we intend and what we expect. As I write these words in June 2020, twitter and instagram are afire with recommendations (especially to white people) to read some crucial books, including White Fragility by Robin DiAngelo and How To Be an Anti-Racist by Ibram X. Kendi and many others. It will be useful to anybody in higher education to read and absorb those books, of course. The books by Safiya Umoja Noble and Kris Shaffer that I discuss here are also crucial in helping us see and understand our current swirl of information.
Additional useful readings from the many books:
Bowen, William G. in collaboration with Kelly A. Lack. Higher Education in the Digital Age. Princeton & Oxford: Princeton University Press, 2013. A book by people who are old-school educators but who have interesting things to say.
Criado-Perez, Caroline. Invisible Women: Exposing Data Bias in a World Designed for Men. Harry N. Abrams. 2019. A book that now has gained a world-wide following and is essential reading. She covers many topics, beginning with how snow removal from sidewalks and streets discriminates against girls and women. And her work on PPE is superb.
Davidson, Cathy. The New Education: How to Revolutionize the University of Prepare Students for a World in Flux. Basic Books. 2017. A classic. Needed now more than ever.
Davidson, Cathy. Now You See It: How Technology and Brain Science Will Transform Schools and Business for the 21st Century. Penguin. 2011. A classic book, needed now more than ever.
Morris, Sean Michael and Jesse Stommel. An Urgency of Teachers: The Work of Critical Digital Pedagogy. Hybrid Pedagogy Inc. 2018. Thoughtful commentary by two visionaries, who have studied higher education for two decades and who have much to teach all of us.
Reidsma, Matthew. Masked by Trust: Bias in Library Discovery. Library Juice Press, 2019. Note: though aimed at librarians concerned with library discovery systems, the contents of this volume will be of interest to anybody who searches for knowledge and wants to know how we find it using machines these days.