Rhetoric and the Making of Algorithmic Worlds
Open Access
- Author:
- Johnson, Jeremy David
- Graduate Program:
- Communication Arts and Sciences
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- January 24, 2018
- Committee Members:
- Rosa Eberly, Dissertation Advisor/Co-Advisor
Rosa Eberly, Committee Chair/Co-Chair
Anne Demo, Committee Member
Stephen Browne, Committee Member
Stuart Selber, Outside Member
Michele Kennerly, Committee Member - Keywords:
- Rhetoric
Algorithms
Networks
Agency
Power
Kosmos
Facebook
Google
YouTube - Abstract:
- Algorithms are key for guiding human attention in a chaotic sea of information. Increasingly, algorithms shape social interaction and political change by arranging material and symbolic phenomena. This dissertation studies the patterns that emerge as algorithms and humans operate together in networked ecologies. Centrally, this project asks how algorithms exhibit agency and how human power intersects with algorithmic arrangement. The project opens by revisiting ancient Greek writings, many of which feature kosmos, roughly translated as “order.” Kosmos’s social and material patterns complement posthumanism and new materialism, which assert that the world unfolds through human and non-human interactions. I argue that algorithms exhibit agency in arranging kosmos, though their control is circumscribed by human power. The project then traces algorithmic agency and power in three platforms: Google search, YouTube, and Facebook. Google’s algorithms are responsible for ordering relevance, ranking the best results at the top of its search pages. Humans “Google bomb” to alter algorithmic arrangement, manipulating sites to the top of search results pages for political purposes ranging from defaming Rick Santorum to promoting Holocaust denial. On YouTube, algorithms judge whether content should remain on the site, removing or demonetizing both copyrighted works and “offensive content.” Unfortunately, as LGBTQ+ members of YouTube have found, this algorithmic judgment can damage communities and endanger lives. Finally, on Facebook, algorithms infer racial identity and offer “ethnic affinity advertising” that can target individuals perceived to align with racial categories such as “African American” or “Hispanic.” Facebook also employs algorithms for moderation, with both machine algorithms and human-executed algorithms searching for hateful and offensive speech. Problematically, Facebook’s inference schemes and moderation practices stunted the growth of #BlackLivesMatter and #Ferguson, and otherwise disproportionately harmed black communities, instead of protecting them from hate speech and discrimination. This dissertation traces algorithmic rhetorics through attention to the ethical dynamics of algorithms, to power and discrimination in algorithmic systems, and to the material arrangements of public discourse and performance. Critical attention, public awareness, and swift action can mitigate some of the harms of algorithmic arrangement. This project offers a critical move in developing rhetorical critiques of algorithms by reflecting on the socio-political impacts of algorithmic world-making.