Skip to Main Content

Identifying Unheard Voices: Exploring Bias

The following guide provides an overview for identifying underrepresented voices and communities within research.

"While we often think of terms such as "big data" and "algorithims" as being benign, neutral, or objective. The people who make these decisions hold all types of values, many of which openly promote racism, sexism, and false notions of meritocracy..."

-Sofia Noble, Algorithms of Oppression (2)

The following page provides users with resources related to exploring the production and replication of bias across a variety information spaces, including:

  • Within the operation of algorithms.
  • Within the organizational and descriptive practices of librarians.
  • Within archives, western material culture, and associated institutional appraisal and collecting practices. 
  • Within the everyday of information searching and acquisition. 

Exploring Bias Across the Information Continuum

"Google Search is an advertising company, not a reliable information company. At the very least, we must ask when we find [these] search results, is this the best information?"-- Dr. Sofia Noble (5)

What are algorithms, and how do they replicate and reinforce socially-constructed bias?

algorithm, n.: Mathematics and Computing. A procedure or set of rules used in calculation and problem-solving; (in later use spec.) a precisely defined set of mathematical or logical operations for the performance of a particular task (OED, 2021). 

algorithm: "...a set of step-by-step instructions that computers follow to perform a task..." (Brookings Institute, 2019).

algorithm: "...is often paired with words specifying the activity for which a set of rules have been designed. A search algorithm, for example, is a procedure that determines what kind of information is retrieved from a large mass of data" (Merriam-Webster, 2022).

While many have heard the term algorithm, few outside of the fields of computer science or mathematics understand their operation on a technical level, and even less recognize the broader, societal significance of algorithms beyond their role in presenting information to consumers in ever-efficient and timely ways. Subsequently, there remains a dearth of non-technical research which critically explores the broad social impacts--positively or negatively--of algorithms on individual and collective identity & self-determination. Even less work still which attempts to understand how algorithms could possibly reproduce the social conditions of domination, alienation, and marginalization given their assumed value neutrality. And yet, scholarship is increasingly turning an eye towards the manner in which sociotechnical information systems--like algorithms, mobiles, or social media--contribute to the maintenance of traditionally repressive or exclusionary structures of power. Organizing under an emerging banner of 'critical information studies', this field's emergence is punctuated by the publication of Dr. Sophia Noble's far-reaching work, Algorithms of Oppression.

 

Algorithms of Oppression (S. Noble, 2018):

At the onset of this ground-breaking work, Dr. Noble challenges readers to think about, "[t]he near-ubiquitous use of algorithmically driven software, both visible and invisible to everyday people" which "demands a closer inspection of what values are prioritized in such automated decision-making systems" (1). As such, the core of Noble's research are the dominant narratives, values, and societal assumptions--both explicit and implicit--about group and individual identity which are enmeshed and reproduced within search systems (like Google) and concomitant search results. 

In this way, Noble's work "probe[s] the results that are generated by Google on a variety of keyword combinations relating to racial and gender identity" as a means of "engaging [in] a commonsense understanding of how power works...by seeing and discussing these intersectional power relations", scholars and critics will have a "significant opportunity" of "changing the processes of control" and to "transform the consciousness embedded in artificial intelligence, since it is in fact, in part, a product of our own collective creation" (29). Importantly for Noble is the acknowledgment that "search happens in a highly commercial environment", and that "a variety of processes shape what can be found" such that "these results are then normalized as believable and often presented as factual" (25).  

Turning to the role of Google in regard to the reinforcement of bias and dominant social values, Noble notes that "search results reflect the values and norms of the search company's commercial partners and advertisers and often reflect our lowest and most demeaning beliefs, [and] because these ideas circulate so freely and so often that they are normalized and extremely profitable. Search results are more than simply what is popular. The dominant notion of search results as being both "objective" and "popular" makes it seem as if misogynist or racist search results are a simple mirror of the collective. Not only do problematic search results seem "normal", but they seem completely unavoidable as well, even though these ideas have been thoroughly debunked by scholars" (36). As such, "Google's dominant narratives reflect the kinds of hegemonic frameworks and notion that are often resisted by women and people of color. Interrogating what advertising companies serve up as credible information must happen, rather than a public instantly gratified with stereotypes in three hundredths of a second" (24). 

Here, the dominant narratives Noble is interested in deconstructing relate to the kinds of information--and biased perspective--conveyed by Google search results about Black girls & women: "[T]he commodified online status of Black women's and girl's bodies deserves scholarly attention...[I]n essence, the social context or meaning of derogatory or problematic Black women's representations in Google's ranking is normalized by virtue of their placement, making it easier for some people to believe that what exists on the page is strictly the result of the fact that more people are looking for Black women in pornography than anything else. This is because the public believes that what rises to the top in search is either the most popular or the most credible or both" (32). 

Noble. (2018). Algorithms of oppression: how search engines reinforce racism. New York University Press.

Physical edition available in the Gordon Library General Collection,  ZA4230 N63 2018.

eBook edition available through ProQuest eBook Central via the Gordon Library website.

How biased are our algorithms? | Safiya Umoja Noble | TEDxUIUC

TEDx Talks

"What do our algorithms say about our society? In this talk, social scientist Safiya Umoja Noble investigates the bias revealed in search engine results and argues why we have to be skeptical of the algorithms we rely upon every day.

Algorithms of Oppression: A Conversation with Dr. Safiya Umoja Noble

From the Harvard Carr Center for Human Rights Policy:

"Towards Life 3.0: Ethics and Technology in the 21st Century is a talk series organized by the Carr Center for Human Rights Policy featuring prominent scholars, business and technology leaders, public interest technologists, and activists who address the ethical and rights implications of the impact of Artificial Intelligence on society and human life. The title of the series draws inspiration from the title of Max Tegmark’s book, Life 3.0: Being Human in the Age of Artificial Intelligence."

Algorithmic Bias Explained: Institute for Public Policy Research (UK)

"Algorithms risk magnifying human bias and error on an unprecedented scale. Rachel Statham explains how they work and why we have to ensure they don’t perpetuate historic forms of discrimination"

 

Similar to that of Google and other proprietary search platforms, libraries have long negotiated the specter of bias within information retrieval and management systems. In particular, there exists a rich tradition of scholarship in the discipline of Library and Information Science related to understanding bias, both within cataloguing and descriptive arrangement practices. The following excerpts were taken from the University of North Carolina at Chapel Hill Conscious Editing Guide, which was developed by archival and library staff at UNC as a means of rectifying the biases ensconced in the legacy language which describes university collections and holdings (like, for example, the UNC Southern Collection, which houses many documents and ephemera related to the antebellum south in the US). 

From the UNC Conscious Editing Guide: 

Who decides what language is used to describe resources and how do changes happen? "The Library of Congress Subject Headings (LCSH) are the main subject terms that libraries across the country use as a controlled vocabulary for describing their materials. LCSH dates back to the late 19th century and is maintained by the Library of Congress, though other libraries can propose new headings and  changes to existing headings through the Subject Authority Cooperative (SACO) Program.

Who does the LCSH work? "LCSH is based on the concept of “literary warrant,” meaning that a subject heading can be established only when it is needed to describe a book or another resource that a library is cataloging. It is arranged hierarchically, with most terms having one or more broader terms, narrower terms, or related terms (for instance, “Women violinists” has the broader terms “Violinists” and “Women musicians”)".

What are some of the issues and/or critiques of the LCSH? "Because LCSH is such an old vocabulary, and because it reflects our white supremacist and patriarchal culture, it is rife with problems. Many terms are marked by race and gender only when the race is nonwhite or the gender is something other than male. [For example] “Women violinists” is an established subject heading, but not “Men violinists”; there generally is an established term for men only in cases of predominantly female professions, such as “Male nurses.” Similarly, “African American lawyers” is an established subject heading, along with many other professions, but only very seldom is there a corresponding term for white people; again, often the term exists in LCSH only when the profession is predominantly nonwhite, such as “Caregivers, White.” These lacunae result in part from a lack of literary warrant—seldom does someone write explicitly about male violinists or white lawyers."

How are libraries adapting? "The fight to change LCSH terminology is not new. Sanford Berman, a former librarian at the Hennepin County Library in Minnesota, is a longtime critic of LCSH and its use of biased and archaic terminology. His book was first published in 1971 and argued that LCSH terminology “can only ‘satisfy’ parochial, jingoistic Europeans and North Americans, white-hued, at least nominally Christian (and preferably Protestant) in faith, comfortably situated in the middle- and higher-income brackets, largely domiciled in suburbia, fundamentally loyal to the Established Order, and heavily imbued with the transcendent, incomparable glory of Western civilization” (Berman, 1993, p. 15)". 

Critical Information Studies Related Resources:

  • Berman, S. (2014). Prejudices and antipathies: A tract on the LC subject heads concerning people. Jefferson, N.C: McFarland and Co. 
  • Leckie, G. J., Given, L. M., & Buschman, J. (2010). Critical theory for library and information science: Exploring the social from across the disciplines. Santa Barbara, Calif: Libraries Unlimited.
  • Leung, & López-McKnight. (2021). Knowledge justice : disrupting library and information studies through critical race theory (Leung & J. R. López-McKnight, Eds.). The MIT Press.
  • Pagowsky, & McElroy. (2016). Critical library pedagogy handbook (Pagowsky & K. McElroy, Eds.). Association of College and Research Libraries, a division of the American Library Association.

Selected Academic Articles with a Critical Literacy Focus: 

Journals with a Critical Focus:

In the Library with a Lead Pipe began as a peer-reviewed blog and repositioned itself as an open peer-reviewed online journal in 2012, featuring professional literature by librarians, administrators, library support staff, technologists, educators, and community members. Published continually. Open access.

InterActions is edited and managed by UCLA graduate students and supported by funding from the UCLA Graduate Students Association and the Graduate School of Education and Information Studies. The journal is "committed to the promotion of interdisciplinary and inclusive scholarship. [The] publication’s authors foster an open and critical dialogue with readers and colleagues through applying diverse social justice frameworks to the discussion of pressing issues in the fields of education and information." Published twice a year. Open access.

The JCLIS is a peer-reviewed online journal of scholarly work (research, literature reviews, interviews, commentaries, book or exhibition reviews) that “queries and critiques current and prevailing paradigms in library and information studies, in theory and practice through critical approaches and perspectives that originate from across the humanities and social sciences.” Published quarterly. Open access.

The JRL is a peer-reviewed online journal publishing research, reports, commentary, and reviews about critical library and information theory and practice. Published on a continual basis. Open access.

PL is the journal for the Progressive Librarians Guild, historically published twice a year. The journal is currently “in transition” and has not published an issue since Winter 2017/2018. Open access.

Though not explicitly dedicated to critical librarianship, ULJ is a peer-reviewed journal publishing research on urban libraries and the relationship between urban librarianship and topics of interest to critical librarians, such as urban studies, social justice, critical race studies, gender and work, and sexuality and identity within librarianship. Published continually online, collected into two issues annually. Open access.

Archival Silence (n.) - the unintentional or purposeful absence or distortion of documentation of enduring value, resulting in gaps and inabilities to represent the past accurately. -- Dictionary of Archival Terminology, Society of American Archivists

Minding the Gaps and Silences in the Archive

While the core function of an archive is undoubtedly the preservation of collective memory--individual, phenomenological, institutional--archives do not preserve all history or histories in their entirety. Consequently, there remain multitudes of voices, perspectives, and experiences that are absent from traditional, non-community oriented repositories. The reasons for these gaps, silences, and whispers include:

  • The unmitigated dominance of a traditional western worldview, which prioritizes and centers white, heteronormative, male views, which, by extension, become more frequently represented in historic records and primary source documents.
  • Records created by those with 'status' (ill-defined) in society have been afforded much more influence and clout in the west, and as a result, are granted a much greater chance of surviving and being preserved for enduring value.
  • Archives staff have made decisions over time about who and what is deemed important enough to be preserved, and these decisions have influenced the scope and character of historical record.

Strategies for addressing archival gaps and silences in your research:

  • Acknowledge that archival silences exist as your starting point.
  • When evaluating historical records, ask yourself:
    • What is missing?
    • Whose perspectives are not represented?
    • What could be the reasons for these archival silences?
    • Check if you can find evidence for the missing perspective elsewhere (in another archives? in the community?)
  • Critically reflect upon archival silences and address them in your research.

Adapted from the USC Special Collections and Archives and Michaela Ullmann.

Archives Have the Power to Boost Marginalized Voices | Dominique Luster | TEDxPittsburgh

"Archivists have an important job — a job that has the ability to save or erase an individual's history or even the history of an entire people. Dominique Luster works to build a historical view that includes marginalized voices and conscious language. In this talk, she shares lessons of this as put in motion with her work archiving the iconic photography of Charles "Teenie" Harris". 

Confronting Biases: Their Impact on Collections, Research and Scholarship,  University of Arkansas Libraries

Cornell University Library, Rare and Distinctive Collections (RAD) Hour: When WHITE Libraries Happen to BLACK Rare & Distinctive Collections: Finding Pathways from Marginalization to Narratives of Empire (2020).

Research related to Archival Silence, Bias, and Remediation 

From the Visual Capitalist; the following infographic discusses some of the more common forms of Cognitive Biases observed and discussed by researchers and social psychologists within the context of political engagement and behavior.