By Ioana B. Jucan, Roopa Vasudevan, Anthony Glyn Burton, Tong Wu, and Yuguang Zhang
"Left and Right" draws on research supported by the Social Sciences and Humanities Research Council.
Introduction
How is the construction and performance of political identity shaped by algorithms and technological platforms? How can we use online theatre performance to expose and push against prescribed political identities that fuel polarization? Can we repurpose technological tools that sustain machinic homophily and media manipulation towards this end? What possibilities do online spaces hold for live theatre performance and audience engagement? The following collection of short video essays addresses these questions by engaging with Left and Right, Or Being who/where you are, an online devised performance presented by the Brown Arts Initiative and Re-Fest, an art and technology festival, in 2021. A ninety-minute, online performance, Left and Right used the digital platform ohyay to bring together audience members from Brazil, Canada, Germany, Portugal, Romania, South Korea, Turkey, and the United States in order to feel and think through political divides and polarization through a transnational perspective.
The concept of the performance emerged out of performance scholar and artist Ioana Jucan's work with the Beyond Verification research team of the Digital Democracies Institute at Simon Fraser University, led by media scholar Wendy Chun. The performance was developed in collaboration with Beyond Verification researchers, as well as an international team of artists and creative technologists working across different disciplines. Left and Right aimed to both extend the team's cross- and inter-disciplinary research through embodied practice and to disseminate the team's findings to a broader audience. In this way, Left and Right participates in research-creation and sits within the tradition of Practice as Research/Performance as Research, which—as scholars Owen Chapman and Kim Sawchuk express—offers "an epistemological intervention on the level of academic methodology" that "partakes of the spectacle of the work of art and its demonstration of alternative frameworks for understanding, communicating, and disseminating knowledge."1
In this way, Left and Right explores both the algorithmically-sustained, emotion- driven production of polarization and the possibilities that the online space, mediated bodies, and machine-generated actors hold for live performance. The video essays below serve to build on and disseminate our findings in working collaboratively on this performance. As members of Left and Right's creative team, the authors of these video essays present their thoughts and reflections through a video essay format in an attempt to show rather than tell: to put the concepts and issues they tackle into performative play with the aid of the audio-visual medium in a way that gives the readers/viewers a richer experience of both Left and Right itself and of the theoretical argument and critical-creative space explored by the performance.
Left and Right, Or Being who/where you are: A Summary
Left and Right featured conversations on so-called hot-button topics and complex realities centered around the COVID-19 pandemic and immigration. These conversations were performed by both human actors and bots—computer programs that simulate a human activity (conversation, in the case of this performance). The world of the performance is centered around three characters, whom the actors introduce in the opening scene:
Maitê Stédile (played by Brazilian multimedia artist and creative technologist Marcela Mancino): "She's 22 years old, Brazilian, and was studying acting in the US until the pandemic hit. She is a white, cisgender woman, proudly bisexual, feminist, antiracist, vegan, and zero waste. She is a big fan of Djamila Ribeiro, Sabrina Fernandes, and Grada Kilomba."2
Sean Lin (played by US theatre artist Patrick Elizalde): He "became friends with Maitê through their mutual involvement in college theatre. He is 24 years old, Christian, a 2nd generation Taiwanese-American from Flushing, NY, and is now working at a top consulting firm in Boston."3
Mara Stan (played by Romanian actors Andra Jurj and Fabiola Petri): She "is a 29-year old Romanian woman, secretly bisexual, Eastern Orthodox by birth. She is a psychiatrist, but wants to become an actor."4 She represents the "Exhausted Majority."5 She had met Maitê and Sean at the international theatre festival in her Romanian hometown the year before the start of the COVID-19 pandemic.
Within the performance's narrative, the characters reconnect online right after the outbreak of the COVID-19 pandemic in the United States. Together, they share their lived experiences, as well as their media-informed opinions on politicized topics, such as the relevance of the lockdowns. Through their online interaction, Maitê, Mara, and Sean negotiate and reckon with their digitally mediated (dis)connection, as well as their physical and ideological distance through dialogue, choreographed movement, and role-playing. They also interact with the audience and two bots, Kimberly and Nick. The latter were trained on news articles addressing COVID-19, immigration, and climate change from the start of the pandemic to the end of 2020; these articles were handpicked to correspond to the wings of the political spectrum: that is, they were articles that a "devoted conservative" (Kimberly) and a "progressive activist" (Nick) would read.6 These two political identities were constructed based on the profiles outlined in the report on Hidden Tribes: A Study of America's Polarized Landscape by More in Common.7 At different points in the performance the human and machinic actors inhabit and push against the political profiles outlined in the More in Common study.
Video Essays
I. Performing a Political Turing Machine (Ioana B. Jucan)
The following video functions as an expanded "trailer" for Left and Right, Or Being who/where you are, featuring key moments from the performance that activate and/or push against political "archetypes," defined by media scholar and Beyond Verification team member Melody Devries as patterned political identities, or "scripted ways of being (political)."8 Left and Right draws on the methodology of "spotting the homophilic avatar" developed by Devries, who theorized the homophilic avatar as a political character produced by homophilic networks subsequently embodied by users.9 As she argues, the concept of the homophilic avatar can help us "better conceptualize political identities as not personal, stubborn or unalterable, but rather the result of relational, performative processes that occur over time and with technology."10 Online action on different social media platforms is both "predictable and performative, … articulating patterned political identities" that further inform and prescribe "even miniscule, mundane political actions."11 For Left and Right, Jucan worked with Devries and the Beyond Verification team members to repurpose this methodology toward an artistic end. Specifically, the human actors inhabited these political archetypes and pushed against them in different ways: for instance, by showing the ways in which such archetypes are constructed; by diverging from the scripts that undergird these archetypes; by using a transnational lens to destabilize them (a "communist" or a "liberal" within the Romanian or Brazilian context is quite different than in the US context); and by creating their own labels and label-defying identities and inviting the audience members to do the same. As for the bots, the subsequent video essays by media artist and computer programmer Roopa Vasudevan and media theorist Anthony Burton (both of whom worked on the development of Kimberly and Nick) outline the ways in which these machinic actors both participate in and complicate the programmed production of polarization. View "Performing a Political Turing Machine" below:
II. Just Bots (Roopa Vasudevan)
"Just Bots" is an overview of the machine-generated bot characters in Left and Right, or Being who/where you are—as told by the bots themselves. Bots Nick and Kimberly introduce themselves to the viewer, explain their purpose in the performance, and offer examples of the questions and content they contribute to the larger performance experience. By posing questions related to authenticity and believability, and inviting further engagement with them in an interactive web portal, the bots allow us to question whether or not we can easily dismiss automated or algorithmic statements and content as simply "manufactured" when they serve to trigger affective reactions in the actors and the audience.
The video essay functions as an expository clip, outlining the ways in which the bots operate within the performances. However, by "letting" Nick and Kimberly speak for themselves, the video also points to questions of authorship and authenticity that pervade Left and Right as a whole. By engaging with their constructed personas outside of the context of the performance itself, the piece affirms Nick and Kimberly as characters with histories and voices, whose statements generate authentic impact in those who interact with them—and raise the question of how we determine who and what qualifies as an authentic or believable source of information:
III. Discourse and Feedback: What Bots Talk About (Anthony G. Burton)
This video is an archaeological analysis of the techniques used to construct the bots featured in the performance, one that works toward a critical perspective of machine learning and its applications. It traces the processes of collecting a training set for the performance's bots, detailing the opacities that emerge in the construction of this dataset. "Opacities" here refers to those parts of algorithmic techniques that would be colloquially characterized as "black boxed", which means that they are viewed only in terms of their inputs and outputs, their inner workings remaining opaque. While we are able to understand such opacities on the basis of their input and their output, we are unable to introspect their functionalities. By taking an archaeological approach to the end-goal of producing our training dataset, this video illustrates not only what happened behind the scenes, but also illuminates the various considerations that developed over time in both handling the problems that we came across and synthesizing these encounters into the argument of the final performance. This video suggests that the accessibility, perceived ease of use, and informational richness of any dataset that attempts to make a claim over something as humanist and felt as political valence simultaneously covers up a series of assumptions, techniques, and shortcuts that complicate the representational claims expressed by the data. View "Discourse and Feedback: What Bots Talk About" here:
IV. Mediated Bodies on Digital Stages (Tong Wu and Yuguang Zhang)
Using Left and Right as its main case study, this video offers a reflection on the creation of theatrical projects in digital environments, with a focus on aspects of design. It is also an artistic reflection on the idea that one coexists with his/her/their digital doppelgänger. The video is divided into three main chapters: "The use of image/video-based visual manipulation for character presentation"; "Constructing the digital stage"; and "Human and AI as co-creators." Each chapter can be seen as a building block toward the final "digital theatre" that interaction designers Tong Wu and Yuguang Zhang are envisioning, culminating in a theatre that transforms the solid bodies on a physical stage into mediated bodies on a digital stage.
The video is organized around a methodology within performance that Wu and Zhang have developed. Although still in an early stage, this methodology gestures toward innovative ways that theatre artists can use the digital medium toward theatrical expression and experience, rather than as merely a drop-in replacement for physical productions that are forced to go remote and online. The aim of this methodology is to maximize the affordances provided by a digital and interactive theatre through its online immediacy, its participatory nature, its unique visual language constructed via close-up webcam feeds and various digital assets, and its fluidity in terms of live digital content creation and manipulation. By connecting the performers and the audience live online via digital forms and AI algorithms, Wu and Zhang explore new techniques to create theatrical storytelling experiences that are organic to the digital environment in which they are contextualized. They hope that these techniques can introduce a unique aesthetic that responds to the era of digital systems and the COVID-19 pandemic, ultimately broadening the capability of theatre-making to go beyond physical restrictions. View "Mediated Bodies on Digital Stages" here:
Footnotes
1. Owen Chapman and Kim Sawchuk, "Research-Creation: Intervention, Analysis and Family Resemblances," Canadian Journal of Communication 37 (2012): 23.
2. Patrick Elizalde, Andra Jurj, Marcela Mancino, Fabiola Petri, Ioana Jucan, and Melody Devries, Left and Right, Or Being who/where you are (unpublished script).
3. Ibid.
4. Ibid.
5. The notion of the "Exhausted Majority"—comprising the middle sectors of the political spectrum (the traditional liberals, passive liberals, politically disengaged, and moderates)—is borrowed from the study Hidden Tribes: A Study of America's Polarized Landscape by Stephen Hawkins, Daniel Yudkin, Miriam Juan-Torres, and Tim Dixon (New York: More in Common, 2018), available at https://hiddentribes. us/media/qfpekz4g/hidden_tribes_report.pdf. The findings of this study informed Left and Right. We used the study with the permission of More in Common, but More in Common was not a contributor or partner to the performance, and the performance in no way represents the views of More in Common.
6. The bots are programs that are fed a dataset of news articles collected from sources that express a range of political views from the wings of the political spectrum. They then generate new content by using the sentence structure, vocabulary, and styles of the articles within the dataset as a model.
7. Hawkins, et al., Hidden Tribes.
8. Melody Devries, Ioana Jucan, and Alexandra Juhasz, "Authenticity, Performativity, and Performance." Paper presented at the "Beyond Disinformation: Authenticity and Trust in the Online World" workshop organized by the Social Science Research Council (2020).
9. Melody Devries, "Archetypes vs. Homophilic Avatars: Re-defining Far/Right Facebook Practice," Canadian Journal of Communication. Special issue on "Politics, Communication and Alt-Rights" (forthcoming).
10. Ibid.
11. Ibid.