‘Ubiquitous googling’ (Ridgway 2021) with keywords is a new media habit (Chun 2016) and Google serves a window into users’ thoughts, interests and desires. Maintaining its intellectual property rights (IPR) through patents (Pasquale 2015:98), corporate secrets such as Google’s proprietary algorithms are incredibly valuable and ‘guarded, like missile codes’ (Noble 2018b). Although Google promotes itself as a neutral purveyor of information (Introna and Nissenbaum 2000, Vaidhyanathan 2011, Castells 2013, Gillespie 2010, 2012 and 2014, Crawford and Gillespie 2014, Noble and Roberts 2015, Pasquale 2015, Tripodi 2017) and likes to come across as an objective platform, with its commercial search activities it ‘prioritizes results predicated on a variety of factors that are anything but objective or value free’ (Noble 2018:65).
With ‘paid search’, Google collects money for ads to be displayed when certain terms are searched, yet in 2012, of the 83% who state that they use Google, ‘[m]ost people surveyed could not tell the difference between paid advertising and genuine [organic] results’ (Noble 2018:35). Even if users believe they know the difference, there is a ‘blurred line’ between the carefully calculated presentation of organic results and ads and most users stay above the ‘fold’ (Introna 2016, Lewandowski 2017:22). Often users view the content they see (including ads) as trustworthy and many enjoy receiving ‘personalised’ results, which are sometimes useful. Transient as well as opaque, what are the criteria determining search results and how can they be captured?
Latanya Sweeney’s screenshotting showed how Google Ads with a greater percentage of the word ‘arrest’ appeared more often for black identifying first names in searches’ (2013:34), including her own. Safiya Noble unearthed how Google’s ‘algorithmically crafted web search’ delivered racism and sexism as the first results with ‘black girls’ through paid advertising (2018:5), based on her own searching and screenshotting. These examples of discrimination within machine learning search algorithms (Chun 2019:64) reflect how data sets are attached to bodies as ‘biopolitical implications’ related to ‘gender, race, sexuality, class, disability’ (Cifor et al., 2019).
The ‘Feminist data manifesto’ and subversive movements informed by feminist, critical data and critical archival practices (D’Ignazio and Klein, 2020) combat oppression and increase understanding of data sets (Thylstrup 2021). These data feminist practices could also be considered interventions in design justice that bridge algorithmic accountability with participatory research, art and activism. Although screenshots are often considered aesthetic, depending on what is captured and for what reason, they are documentary, evidentiary but also surveillant. Screenshots capture the ‘gaze of the search engine’ (Noble 2018:71,116) yet an intersectional lens of analysis also exposes not only the technical but some of their socio-political workings and repercussions.
In this presentation/workshop I put forth a method of ‘critical feminist screenshotting’ to visualise the search results of black-boxed algorithms, offering an alternative ‘way of seeing’ (Berger 1972; Cox 2017). Together with a group of interested participants in an open and caring format, (re)search results will be captured and critiqued, in order to reflect on the critical making. Combining praxis and engaged theory in a socially-engaged learning environment, the workshop intends to find out collaboratively: Why do we receive the search results we do? how can we understand them? and in which ways can we interpret their effects? Group exercises include investigating locative data, whether one is personalized and to what extent, based on individual data collection.
By revealing some of the back-end invisible design/workings of Google algorithms, which are dependent on the entered key word or phrase, as well as the personal choices or the user (computer, browser, search history, location, etc.), the attendees will gain a greater comprehension on how Google search organises search results, how to decipher them and why their results differ from others through a ‘critical feminist screenshotting’ method.
