UC Berkeley School of Information Assistant Professor Niloufar Salehi is part of a multidisciplinary research team that won a Facebook Research Award to study how Muslim Americans adopt counter-narratives online, empowering and giving voice to an often denigrated population. The request for proposals sought projects focused on online content policies, specifically around bullying and harassment, and fairness in global enforcement. This project was awarded $100,000.
The team is comprised of Mariam Asad, who holds a Ph.D. in Digital Media from Georgia Institute of Technology, and focuses on how technology design can offer opportunities for civic participation through grassroots/community-based initiatives, Nazita Lajevardi, an Assistant Professor in the Department of Political Science at Michigan State University, who focuses on issues related to public opinion and political behavior through the lens of religious and racial identity, Roya Pakzad, the founder and director at Taraaz, an organization working at the intersection of technology and human rights, and Professor Salehi, who focuses on social computing, participatory and critical design, human-centered AI, and more broadly, human-computer-interaction (HCI).
We spoke with the researchers to learn more about the project and what they hope to achieve.
Why did you decide to focus on counter-narratives?
Salehi: Anti-Muslim hate groups have been on the rise in recent years and hostility and abuse is rampant in online platforms. In early work, we came across two important dynamics that appeared to be happening simultaneously. First, there was organized and pervasive targeting of Muslims. Second, there was an ongoing struggle by Muslim Americans to try to take control of their narrative online by crafting counter-narratives. The goal of this project is to understand how Muslim American counter-narratives are shaped online, who amplifies them, how sustainable they are, and how their impact differs.
For instance, #muslimsreportstuff was started as a way to counter the pervasive narrative that Muslim Americans know things that they are not telling everyone else, so this narrative of Muslims as outsiders. It went viral on Twitter after a presidential debate and was covered in NPR.
Pakzad: It’s been frustrating to see social media platforms’ lack of accountability when it comes to protecting Muslims online. Journalists and Muslim advocacy groups often don’t have sufficient resources to report every single hateful incident to the platforms, or when they do, they often receive no response or a very late response. Platforms have been over enforcing — and on different occasions under enforcing — their content moderation practices. In some cases, as part of what platforms frame as “counter-terrorism” efforts, Muslims’ accounts and content have been taken down for no reason, often because content moderation algorithms don’t understand the context behind the speech and act loose-handedly.
On the bright side, during our preliminary interviews with Muslim American journalists, political candidates, and advocacy groups, we noticed that counter-narrative strategies, sometimes as simple as creating hashtags or as formal as publishing an online statement, have shown promising results in nullifying dehumanizing narratives about Muslims. When we compared platforms’ haphazard content moderation practices with counter-narratives created by Muslims themselves, we found that the latter helped Muslims to take control of their own narratives and might have a more long-lasting effect. Does that mean we think platforms should be hands-off? Absolutely not. But we believe that by studying these counter-narratives and their impacts, we will also help platforms to think more creatively and humanely about their policies and practices.
Why Muslims in particular?
Lajevardi: Muslims have been increasingly stigmatized in the sociopolitical sphere, on- and off-line. Hostility and negative attitudes towards U.S. Muslims are pervasive and growing: in 2017, the number of anti-Muslim hate groups rose for the third straight year and the Council on American Islamic Relations (CAIR) recorded 2,599 anti-Muslim bias incidents taking place across the nation. Especially worrisome for the status of US Muslims is the large degree to which hostility has also manifested on online platforms; for instance, the 2018 election saw coordinated activities on Twitter to track Muslim candidates and utilize Islamophobia to mobilize voters against them. This climate, moreover, has negatively affected U.S. Muslims and Arabs who at least temporarily reduced their visibility in public spaces, both online on Twitter and offline in public spaces.
Asad: Muslims are one of the most vilified groups — state-side, certainly, but also globally — despite there being about a billion of us from different ethnicities, cultures, beliefs, and politics. As a result, Muslims have been the target of different kinds of abuse, from individual attacks to state-sanctioned violence (post-9/11 stop and frisk, the Rohingya refugee crisis). In this way, Muslims are an unfortunately good barometer for the wide spectrum of harassment that happens and, consequently, normalized and even internalized. By understanding the harms done to Muslims, we can also understand the kinds of harms that have been, currently are, and will be done to other marginalized groups.
What are you doing as part of this research?
Lajevardi: This project consists of two distinct steps.
First, we are conducting interviews with Muslims of varying backgrounds who have social media accounts. These are public figures, activists, representatives of NGOs, journalists, and academics. The interviews serve to help us understand the types of online harassment that individuals and organizations face, but also to evaluate the actions and tools, such as hashtags, they have used to counter-narratives and hate speech on online platforms.
The second step then examines the various hashtags that have been employed to evaluate when these have been used actively or retroactively in response to events or hate speech.
What do you hope this project will achieve?
Asad: In some respects, my hopes for this project are very humble, which is to say I hope that this work can help introduce folks to the nuances of Muslim American life, and find in these mundanities some connections to their own experiences. In another sense, my hope is that this work can be a catalyst for folks to get some glimpses into the complexities and concerns that so many of their neighbors, peers, and friends have had to internalize for so long without having done anything wrong. The goal here is not to victimize all the members of one faith, but to reveal the myriad ways folks have been unfairly punished and castigated in very material and sustained ways for belonging to the same religion as some arbitrary strangers completely unrelated to them who have done harmful things (which is to say, every religion).
Salehi: For me, this work ties deeply into my ongoing research on ways to address online harm based on principles of restorative and transformative justice. I see hate speech as one of those harms. This mostly happens when people are targeted because of their race, ethnicity, religion, sexual orientation, etc. by strangers online. I hope that this work will help unpack this type of harm and help us identify what types of support are helpful to address these harms. I remember when we started this work I interviewed the director of a prominent Muslim advocacy group and asked them about reporting harmful content to platforms, they just said: yeah we don’t use it, and moved on. I was so surprised. We spend all of this time on how to do content moderation well, and it seems like some of the most at-risk people are not even using it. Trying to understand what strategies they do use is what motivates this project for me.
Pakzad: Additionally, I hope for this project to go beyond academic conferences and journal websites; I hope this research serves as a tool to help activists, Muslims advocacy and policy groups in their day to day efforts toward countering dehumanizing speech and taking control of their own narratives.
How will you share the findings of the research?
Lajevardi: We are hoping to communicate our findings in at least two ways: 1.) through a public-facing policy report that we will share with participants from the first step of our research and with public outlets, such as news programs; and 2.) through an academic article
Asad: Given the interdisciplinary nature of our team, we anticipate our results to be widely applicable across our different research communities, from computer scientists to HCI designers to human rights practitioners, policymakers, and legislators. In addition to our research publications, we hope to disseminate our work through more accessible and public-facing artifacts, such as whitepapers, panels, and media outlets, as well as more informal channels, such as discussions within our respective communities and social circles.
If you want to learn more about this project or get in touch with the group, check out the project website: ourmuslimvoices.com.