Reem Hilu interviews Wendy Hui Kyong Chun, Simon Fraser University’s Canada 150 Research Chair in New Media in the School of Communication, professor of Communication, and director of the Digital Democracies Institute. Chun discusses her work that historicizes the utopian and dystopian framings of digital media and network technologies, focusing on the nature of community, or its absence, online. Throughout the conversation, Hilu and Chun sought to maintain a dialogue between interpretations of contemporary mediated communities online and longer histories through which media have worked to organize connections and associations.
Wendy Hui Kyong Chun is Simon Fraser University’s Canada 150 Research Chair in New Media in the School of Communication, professor of Communication, and director of the Digital Democracies Institute. This conversation took place on April 20, 2023, and opened the symposium “Pleasure, Danger, and the Long History of (Social) Media” at Washington University in St. Louis. Like this special issue, the symposium sought to explore how media have long functioned in creating and shaping communities, even before the emergence of what we now commonly call “social media.” Chun has consistently done interdisciplinary work that maintains a rigorous historical commitment. Her work, including the most recent book Discriminating Data: Correlation, Neighborhoods, and the New Politics of Recognition (MIT Press, 2021), is enlivened by an ability to think through utopian and dystopian impulses in digital media in surprising ways and ways that hold open the possibility for difference and change while still maintaining historical awareness. Chun often complicates the expected values assigned to concepts like freedom, control, security, and vulnerability—and in this interview, pleasure and danger as well.
The interview explores the dynamics of pleasure and danger and utopia and dystopia through discussions of the nature of community, or its absence, as it is formed online. Throughout, we sought to maintain a dialogue between interpretations of contemporary mediated communities online and longer histories through which connections and associations have been formed around and through media. In other words, we hoped to understand the continuities and differences between mediated communities online and mediated communities in other contexts.
One of the important distinctions that Chun makes in thinking about mediated groups is that between community and homophily. The organization of groups around media texts offers a way to better understand what distinguishes the two. Chun offers the example of how network scientists have used the music of Wu Tang Clan, or rather the online likes or dislikes of this music, as a divisive “like” that can predict identity and classify users into distinct yet internally homophilic groups. Yet, as she notes, Wu Tang Clan and their work are the result of a multiplicity of influences and meanings. This difference of perspective, seeing Wu Tang Clan as a data point to categorize users or seeing it as a knit of multiple meanings and influences is a significant one. This is one example through which we can see how analysis of contemporary social media and online networks benefit from being put into dialogue with media studies approaches that consider other media forms such as films, television, and music as resources through which to create communities. When consumers are united around shared media texts, this does not have to result in homogenous groups, as previous mediated communities have shown us. The subsequent essays in this issue, in fact, provide many more examples of the identitopias constructed around media that form communities and not only homophilic associations as they point to the way that these communities are brought together around shared media resources even while being a multiplicity. Additionally, Chun’s discussion of the way that the dystopian novel Neuromancer (William Gibson, 1984) was misread as a source for utopian framings of the internet shows how the meanings of media and the work that they do in shaping technologies and communities is not self-evident or singular.
In addition to considering the role of media in defining and shaping communities or clusters, this conversation also helps highlight some of the important differences in conceptions of media consumers, or so-called users, as they are theorized by media scholars or by network scientists. Chun provides insight into the way that consumers are conceptualized in the discourses of network science, not only as users but also as the potential members of agitated clusters. Here too, it is difficult to disentangle pleasure and danger when discussing mediated communities and other groups online. The intense affects that create these associations threaten to overwhelm and do harm within the community and beyond. This interview addresses this entanglement while also exploring alternative sources through which pleasure and community might flourish. Chun’s background across the humanities and engineering provides a perspective on how framings of media technology and community travel between fields and how we can find potential in this exchange for thinking about new forms of community. The interview has been condensed and edited for clarity.
Can you talk more about how the framing of utopia and dystopia has worked in histories of the internet?
I find fascinating the temptation to see the early internet as utopian and what we have now as dystopian. We need to push against this story because the internet has always been both at once. For example, cyberspace—what the internet was called when it erupted into the public imagination during the mid-1990s as a virtual space that fixed all our political and economic problems—stems from really dystopian cyberpunk fiction. The world that Neuromancer describes is not happy. That the internet made cyberspace utopian reveals not only how science fiction gets misread but also how it gets projected onto and against existing technologies.
We need to question the tendency to say “back then, things were utopian, now we have dystopia” because this rewrites history. During the cyberspace era, a lot of people pointed out that these bodiless dreams of cyberspace as a realm of “only minds,” dissolving racism, sexism, and ageism, was bunk because it poorly understood lived experiences, bodies, race, and sexuality. It also came from this really overriding desire to have technology solve our political and social problems for us. As soon as you frame technology as a solution to political and cultural problems, you automatically have a double utopian-dystopian situation because it provokes this dream of “somehow, it would be better”—and this always fails and often makes things worse.
A clear recent example of this is COMPAS, which is used by some courts to predict the likelihood of reoffending and therefore sentencing. It’s been sued for discriminating against racial minorities. But it was introduced while Obama was president as a way to solve the problem of discrimination by judges because the idea was that suddenly, these programs, unlike humans, could not see race.
Again, this coupling of the utopian and dystopianism is linked to this desire to have technological solutions to our political problems.
Do you feel that it has become more common among those developing these technologies to think of this coupling of dystopian and utopian or does technological utopianism remain widespread?
Technological utopianism amongst tech folks remains widespread, even as it waxes and wanes. At the turn of the century, when the dot-coms exploded and 9/11 happened, articles like Steven Levy’s “Tech’s Double-Edged Sword” appeared.1 All of a sudden, it went from Cisco Systems advertisements, which basically said “yeah people riding camels in the desert have access to the internet” to “oh sh*t, people riding camels in the desert have access to the internet.” What’s considered to be utopian and/or dystopian, though, has changed. Early on, when the internet became a public medium by being sold to private corporations, fears around cyberporn dominated. In the moment of reflection after the dot-coms and Y2K, and 9/11, anxieties around terrorism dominated. Now artificial intelligence and machine learning capture utopia-dystopia.
Related to this thread of dystopia and utopia, you discuss in your work how fear of difference often motivates online activity where users interact with others that are essentially similar. Since you bring up AI, can you share your thoughts on the potential in plumbing our relationship with AI for models of how to be more open to difference when we encounter it?
Absolutely. First, we need to work across disciplines and address how the problems with technology, as they currently exist, aren’t simply because technology development ignores the social sciences—the problem isn’t that technology exists in a bubble outside the human sciences—but rather because it really poorly embeds problematic concepts from the social sciences as defaults. Consider concepts that ground network science such as homophily—the idea that similarity breeds connection, that network neighborhoods should be filled with people who are alike. Homophily emerges from studies of US racial segregation. But these studies of US housing projects and residential segregation in the mid-1940s also pushed back against homophily as natural. They coined both the terms homophily and heterophily. They asked: under what conditions does homophily emerge or not? The backstory of course behind these studies was a battle over public opinion and funding—at that time over 80 percent of the public supported public housing; private developers in contrast wanted government-backed mortgages for whites-only properties. We know who won.
Also, the concept and mathematics of correlation, which grounds “Big Data,” comes from twentieth-century eugenics and desires to create a “eutopian” future by manipulating what were allegedly unchanging human traits. We need to think through the history of these concepts not so we can condemn anyone who uses correlation as a eugenicist, but rather so we can think through them differently and more richly. Correlation is also key to the humanities, and indeed to thinkers such as Lacan, Freud, and many literary critics who frame correlation as a metaphor or metonymy, as something other than linear similarity.
Correlation as it exists in networks now also reveals how we’re all intertwined. The recommendations that you receive online, they’re not due to you alone. They seem personalized—“Wendy, you might want to buy Ian’s latest book”—but this recommendation is not solely based on who I am and what I’ve purchased but rather on the neighborhood, filled with others like me or others like Ian’s book, they’ve put me or it in. Crucially, they seek the likes that are most divisive. You like a lot of things but your most valuable like is one that divides, that creates a clean divide between two different groups.
To give you an example that both works and doesn’t work, an early article claimed that Wu Tang Clan is one of the strongest markers for being a heterosexual male. It’s not because all heterosexual males like Wu Tang Clan but because it was viewed as something that only heterosexual men tend to like. The goal is to find these divisive likes in order to put you into certain categories. Which, if you push on them, actually open up different ways of understanding our co-relations. Wu Tang Clan is fascinating because they bring together Hong Kong, Jacques Cousteau, East Coast rap. Correlation, which seems to be all about restricting you into smaller corners, can also be used to understand how you’re related in other ways.
Responding to this idea about recommendations and your example of Wu Tang Clan, it strikes me that there are some comparisons to longer histories of distinction and taste built around media preferences and media consumption or fandom. Previously, these were distinctions or media communities that formed around media taste and shared media experiences that people had to work to develop in different ways. Do you think that digital media and recommendation systems change how communities can form around these types of shared experiences because of the new ways that people arrive at or experience a text (for example, through recommendation algorithms)?
That’s a great question. It gets to the difference between community and homophily, because homophily reduces community—all the work that goes into making community—to an individual tendency or preference. “You just naturally want to be with people like you.” In fact, it’s often just referred to as an ecological fact. But anyone who studies ecology knows that birds of a feather don’t always flock together, that there’s incredible diversity within ecology. As an aside, one thing that fascinates me is the bizarre ways in which ecology has come to explain media interactions and behavior—and how it’s usually presumed to be more virtuous or environmentally sensitive. Ecology is “1984 for Bambi”—surveillance is a good thing in ecology. Ecology also presumes resources and scarcity, since it draws inspiration from economics. Ecology is the economics of nature. So, the fact that we view it as somehow going to save the world is really strange to me.
To return to homophily, when homophily is justified in terms of ecological reasons, it’s the return of the economic in natural clothing. Again, the difference between homophily and community is precisely everything that needs to go into building community: infrastructure, communal engagement, etc.…
I wanted to shift to discuss the related theme of pleasure/danger in the long history of social media that this special issue is also exploring. Thinking about other media before and along with the internet—film, television, comics, novels—there is a tendency to characterize these as offering a pleasure that is overpowering and addictive—even to the point that they are perceived as dangerous. This can receive more attention than analysis of their positive pleasures or ability to support community. In online spaces do you feel that there’s a tendency to focus more on the harms to users or communities over some of the ways that community is fostered online? Here, I am also thinking of the work of scholars in Black digital media studies like Raven Maragh-Lloyd or Sarah Florini.2 There is a framing that interprets online experience as segregated communities built on something that may be similar to what you call “comforting rage,” but there is another framing, potentially a counterpoint, that addresses enclave publics, media communities online such as Black Twitter where Black cultural discourses can proliferate. Along these lines, I’m wondering if you have thoughts about how online experience may be interpreted through this pleasure/danger dynamic?
Yeah. Here, I think through André Brock’s work and how he’s been thinking through the libidinal economy of the internet, as well as putting the internet in conversation with The Green Book (Green, 1941) and various technologies key to the long histories of Black Americans negotiating spaces.3 I think this is absolutely key. When you think through pleasure/danger, there’s so much there that’s rich and wonderful. I think you’re right, there’s the addictive thing that becomes overwhelming.
Just think of the idea of the user. I was always fascinated when I was a girl in engineering. I thought, “Why are we users?” Users—it’s this dream of addiction? But also, in the playground when you were called a “user,” that was not a nice term—it was somebody who used other people. So what is this fantasy? By examining this double meaning, though, we can understand how we’re used as we use—that was very much the early part of my work. Part of the goal of calling us users was addiction, wanting these technologies to be so addictive as a way to overwhelm some of these other dynamics at play. Neuromancer is all about addiction and drugs—that’s part of its appeal. As an aside, I have to admit, as a girl in engineering, I was always really confused by the sexual metaphors and the sexual terms embedded within network engineering. I remember when I finally figured out what a male to female connector was. I was like, “I can’t touch my computer anymore.” I had this moment of like, “I can’t do this.” And then also—how is this really male to female?
The question of sex and sexuality—why it was used to describe technical connections—and how concepts do or don’t translate between disciplines drove my dissertation, “Sexuality in the Age of Fiber Optics.”4 Sex and sexuality encapsulated this utopian-dystopian pair: dreams of tele-sex, Lawnmower Man (dir. Brett Leonard, 1992), cyberporn. When US senators suddenly discovered cyberporn (through printouts) and created legislation, I was like, “porn has been on the internet a very long time.” Why now this scare over porn? Why was it that certain versions of sex and sexuality were used to understand contact? At the same time this was going on, sex and sexuality was being explored very positively, and this is also part of the beginnings of the web. So, what was this space of danger and pleasure and how could we understand it as a space that could be opened up and explored in interesting ways?
I’d like to pursue these ideas about gender and sexuality that you bring up and how it is so embedded in discourses about the technology. In Updating to Remain the Same, you demonstrate how discourses about “safety” and vulnerability online are often framed in gendered terms in which teenage girls became a central figure.5 I’m wondering if you might talk about how the teen girl continues to serve as a focal point for theorizations of the pleasures and harms of digital media or how that has shifted?
You’re referring to the last chapter of Updating to Remain the Same. I discuss note card videos, in particular Amanda Todd’s. (I now live in Vancouver, and she lived just outside of Vancouver in Port Coquitlam.) She was bullied, after a topless picture of her was shared with her schoolmates and teachers, and she committed suicide. There was a long trial to prosecute the person from Amsterdam who was deemed responsible for tormenting her. At first, the authorities reacted by saying, “Well, you should get offline. If the internet isn’t a safe space for you, then you should get off of it.” There’s a long history to this. Shilpa Phadke, Sameera Khan, and Shilpa Ranade’s Why Loiter? about the streets in Mumbai is so important because it asserts that, no, women need to have access to public spaces.6 When young girls and people of color are safe in public spaces, then we’re all safe.
Along these lines, Kara Keeling and I are writing something together that takes on our impoverished notions of privacy and publicity online. When privacy becomes reduced to only Facebook knows what you’re doing and they’re not going to tell anyone else, then that’s not privacy. It’s a form of property that does not protect people. If we think instead of the right to be in public—public as a space young women belong in—then we can build spaces in which they and others can be vulnerable but not attacked.
Shifting focus somewhat, your work often traces how the logics of discrimination play into the development of media technologies, but also your stance is not that we just get rid of these technologies, that there’s still some potential there. I’m wondering if you might talk to us a bit about how you conceive of these traces of a potential for a different world or different media that can be found in the technologies that we have today?
That’s such an important question and it gets to the core of networks. Networks are only possible because of the spaces in between lines and nodes. If you didn’t have gaps you wouldn’t have a network—you would have a dense mass. But these empty spaces aren’t empty—within them are thriving populations, without whom these connections wouldn’t be possible. If you think back to the housing survey that developed the term homophily, this concept depended on emptying out relations to produce clean homophilic relations. Specifically, they surveyed a biracial housing complex and they found that white liberals overselect white liberals, and white illiberals overselect white illiberals (you were considered liberal if you thought housing projects should be biracial and the races got along; illiberals thought the opposite; ambivalents thought that the projects should not be biracial but that both races got along). Now, in order to make this point—and this is where homophily emerges—they threw out all the responses of the Black residents, and they threw out the response of the largest category of white residents within the housing complex—the white ambivalents. But without these people who were elided, these connections we see could never be possible.
If we start thinking with the spaces that appear to be empty but really make what appears possible, then things open up in really interesting ways. Lisa Nakamura, Grace Hong, and I are writing a piece addressing the rise of anti-Asian sentiment and linking it to the longer history of anti-Asian racism, in particular the use of sentiment analysis in Japanese internment camps. When you start tracing these concepts, you go back to the populations that were literally captured and necessary for these concepts to emerge. Sentiment analysis actually first came from studying women workers—five women workers who were literally secluded in this one room for five years. They had medical exams, reported their diets. It was a very strange and famous experiment at the Hawthorne Western Electric plant, which allegedly proved the importance of worker sentiment.
If you trace out these concepts, which have become so embedded within our algorithmic structures, and you retrace populations that have been captured, you discover other interpretations and possibilities. These women in the Test Room at Hawthorne, for example, viewed themselves and were viewed as minor celebrities in the factory, who could influence things. Because of their actions, other workers received breaks. But the experimenters from the electric company and Harvard Business School viewed themselves as reading the authentic key and sentiment from these people. The same thing is true with the Japanese internment camps. What we’ve been trying to think through—in particular what I’ve been trying to think through increasingly—is the ways in which asking someone how you are, “How are you?” (people ask this all the time but you don’t care)—how asking someone how you are but not caring, asks people to double themselves. It creates the latent and the manifest.
If we think through Instagram, everything online is this combination of what’s latent and manifest, and that’s arguably what sentiment produces and analyzes. If you realize this, then you have a very different way into understanding what social media is and does. We’re actors not rats (although that’s not quite right because rats are actors too).
When data gets thrown out, is this because it represents material that is harder to instrumentalize? Is there a sense of why certain things get thrown out and others don’t?
Well, in the case of the housing study, it was because they wanted to see homophily. They massaged the data in all sorts of ways—in the archive, you see the chart in which they finally managed to make it appear and they’ve starred it. But we have to realize that data gets thrown out all the time—it has to for networks to emerge. In that study, they got rid of the white ambivalents because they viewed these as unstable, because the pull of tolerance/intolerance was so strong. I always think about this in terms of the classic iron filing experiment, in which you take magnets and put them underneath a bunch of filings, and then all of a sudden, what emerges is actually a network. You have these clusters at either side, and what emerges is space.
Each cluster is agitated though because the filings also repel each other. But, because they’re so attracted to their opposite—or so hate their opposite, if you think through social media—they cling together. So, homophily creates very angry spaces filled with people who are alike (think of turbulent family vacations). So, it is not simply that data gets thrown out, but also things have to get moved around in order for these patterns to emerge.
Relating this idea of anger and repulsion back to that of pleasure and danger and the impression that media can have an overwhelming power or overwhelming affect, are there ways to think about affects other than anger and agitation as ways to build connections and look toward the future?
We’ve been thinking through indifference because heterophily—opposites attracting—also misses the importance of the spaces that touch us all. Heterophily is also something corporations are investigating because they know the limitations of homophily. They’re the first people to say our recommender systems are limited by homophily. For us, what matters are infrastructures, or things that you don’t care about, as things that connect us.
The example I always give is everything you have to forget in order to ride the R5 bus, which is an express bus in Vancouver. You have to forget a lot of things to ride this bus. But riding it is a form of caring, nonetheless—caring for public infrastructure and the planet. We think about indifference as a form of care that builds the infrastructures that connect. And again, for us, that’s the space that connects us. The space without which there couldn’t be connection. And they’re actually the largest part of any network.
Steven Levy, “Tech’s Double-Edged Sword,” Newsweek, September 23, 2001, www.newsweek.com/techs-double-edged-sword-152095.
Raven Simone Maragh-Lloyd, Black Networked Resistance: Strategic Rearticulations in the Digital Age (Berkeley: University of California Press, 2024); Sarah Florini, Beyond Hashtags: Racial Politics and Black Digital Networks (New York: New York University Press, 2019).
André Brock Jr., Distributed Blackness: African American Cybercultures (New York: New York University Press, 2020).
Wendy Hui Kyong Chun, “Sexuality in the Age of Fiber Optics” (PhD diss., Princeton University, 1999).
Wendy Hui Kyong Chun, Updating to Remain the Same: Habitual New Media (Cambridge, MA: MIT Press, 2016).
Shilpa Phadke, Sameera Khan, and Shilpa Ranade, Why Loiter?: Women and Risk on Mumbai Streets (New Delhi: Penguin Books India, 2011).