ALGORITHMS OF OPPRESSION by Safiya Umoja Noble

ALGORITHMS OF OPPRESSION

How Search Engines Reinforce Racism
Email this review

KIRKUS REVIEW

How Google and other search engines represent marginalized people in “erroneous, stereotypical, or even pornographic ways.”

Noble (Information Studies/UCLA; co-editor: Emotions, Technology, and Design, 2016, etc.) was drawn to her subject in 2011, when her Google search on the keywords “black girls” brought up a black pornography site as the first hit. Her subsequent research has led her to conclude that such web searches yielding racism and sexism as the first results reflect “a corporate logic of either willful neglect or a profit imperative that makes money from racism and sexism.” Google has since changed its algorithm for the “black girls” search, but the author has identified and writes here about many other instances of search engine “recklessness and lack of regard” for women and people of color—e.g., a 2016 Google Images search for “gorillas” that produced photographs of black women. Arguing from a black feminist perspective, Noble says such search findings “increasingly lead to racial and gender profiling, misrepresentation, and even economic redlining.” She notes that contrary to the popular belief that Google is a public resource, the search engine is a commercial enterprise—an advertising agency—that “biases search to its own economic interests.” As a result, she writes, the company often prioritizes powerful or highly capitalized industries and interests. Also, due to the lack of diversity in Silicon Valley and the general lack of people with an understanding of racism and sexism, search engines fail to carefully analyze the potential impacts of their products. Whether by neglect or deliberation, girls’ identities are often “commercialized, sexualized, or made curiosities.” As Noble writes, “intention is not particularly important.” Meanwhile, pornography and other businesses work to maximize their search results. Other topics covered include Google’s monopoly on information and the need for regulation. Jargon limits the book’s accessibility, and a chapter on the views of search engine officials is curiously lacking.

A distressing account of algorithms run amok.

Pub Date: Feb. 20th, 2018
ISBN: 978-1-4798-4994-9
Page count: 256pp
Publisher: New York Univ.
Review Posted Online:
Kirkus Reviews Issue: Jan. 1st, 2018




SIMILAR BOOKS SUGGESTED BY OUR CRITICS:

NonfictionWE ARE DATA by John  Cheney-Lippold
by John Cheney-Lippold
NonfictionWEAPONS OF MATH DESTRUCTION by Cathy O'Neil
by Cathy O'Neil
NonfictionDATACLYSM by Christian Rudder
by Christian Rudder