Over the past few years, we have become increasingly concerned about the rise of political polarization on campuses and in society. Not only does it make us angrier and dumber, and elevates the voices of the most narcissistic and Manichean among us, it also makes it hard for society to govern itself and solve our most important and difficult problems—problems like COVID-19 and climate change.

Motivated by this concern about polarization, we convened an informal discussion group in fall 2019 at the University of Colorado Boulder (CU Boulder)—a group which included several local members of Heterodox Academy, some faculty, staff, and community members, and two CU Regents—one a Republican and one a Democrat. One of the staff members was Pilar Sattler McQuillan, the Assistant Director of the CU Dialogues Program. The CU Dialogues Program—within CU Engage in CU’s School of Education—facilitates conversation approaches on campus aimed at building empathy and understanding across differences.

Our group initially decided to host a half-day dialogue event in the fall, before the election, where we would invite ideologically affiliated groups (e.g. the college Democrats and Republicans) and lead them through a dialogue of the sort that CU Dialogues specializes in. We did a test run of such a dialogue in one of Matt Burgess’ upper-level classes in early Spring 2020, but then COVID-19 threw a wrench in our plans for an in-person Fall event.

Instead, we decided to host an online weekly dialogue series (via Zoom) on Fridays at noon throughout the Fall semester. Each week we had a new discussion prompt, usually selected based on a mixture of the science of reducing intergroup conflict and current events, through which we focused on drawing out participants’ curiosity, humility and empathy. For example, the first week, our topic was “What experiences, ideas, and aspects of your background and/or upbringing have shaped your political worldview?” The second week’s topic was: “What is an issue or position that you have changed your mind on recently (and why)?” Two weeks before the election—and motivated by observations that both campaigns, to varying degrees, used fear-based messaging—our topic was: “What are we each afraid of, when it comes to the election and/or issues being debated in the campaign?” The week of the election, our topic was simply an election debrief.

We maintained a fairly simple and consistent dialogue structure each week: we started with a brief overview and reminder of our group expectations and norms. Each week we began with a 5-minute introduction of the topic, then—if there were more than a dozen participants—we broke into randomly assigned breakout rooms of 4-5 participants each for roughly 40 minutes of discussion. For the last 15 minutes we reconvened as a larger group to discuss any general thoughts or insights from the smaller groups. The participants also started a Google Drive to share general notes and resources from each week’s dialogue.

We had a few norms, which we initially developed in consultation with Pilar and the CU Dialogues program, and then shared with the group (on our shared Google Drive) for feedback, discussion, and revision. We felt that while it was important to set up norms for these discussions and make clear that behavior that fell outside of these norms would result in being removed from the Zoom call (and potentially future dialogues), we didn’t want the “rules” to feel overbearing to the point of limiting people’s willingness to share openly, listen attentively, and respond with humility and empathy. We never had an issue with these norms being followed. The norms we established, and which all participants actively discussed and agreed to, were:

• • We welcome all perspectives and viewpoints.
  • Our objectives are to better understand the issues and where each of us is coming from, not to "win" a debate on any particular topic, nor to have a struggle session.
  • Curiosity and humility: Each of us will likely encounter ideas that we disagree with. In a dialogue, our first instinct should be curiosity—if we don’t agree maybe we can at least try to understand where the other person is coming from. And when we do want to express disagreement, that’s ok, but we should criticize ideas, not each other as people.
  • We will try to separate normative questions, descriptive questions, and personal experiences. We will listen to each others' experiences with empathy and non-judgement. On normative questions, we will leave room for a diversity of values and perspectives ('polyvocality', as CU Dialogues calls it). We will confront descriptive questions with reason, facts, and data to the greatest extent possible.
  • We will be precise in our points (precision is often the opposite of polarized discourse pathologies—e.g. sweeping generalizations, ad hominem attacks, etc.)
  • What we discuss stays in the room. Certainly, we can share broad ideas—and solutions, if we come up with any—with others outside the group afterwards, but we won't be sharing sound bytes (e.g. "so and so said this").
  • We will be mindful of how we share speaking time, and make sure that we are listening to each other and no one person is dominating the discussion.

So how did the dialogues turn out? We had 80 participants sign up for our mailing list, and we had 10-20 participants at each dialogue. A core group of 8-10 almost always attended. Participants in these dialogues were diverse in every way. They came from a very wide range of roles on campus—including undergraduates, graduate students, alumni and members of the local community, staff, faculty, senior administrators, Regents and Regent candidates, and a former U.S. Congressperson. For each dialogue, we had roughly equal numbers of women and men; members of the LGBTQ community; veterans; and people from a wide variety of racial and ethnic backgrounds.

We do not directly ask our participants to identify their politics, but based on the discussions, we’d say there was a political mix, with viewpoints and perspectives that ranged from Trump-supporting and Trump-skeptical Republicans and conservatives, to centrists and liberals with viewpoints that ranged from center to farther left. There was perhaps a slight skew towards center-left among participants. We initially had some trouble getting consistent participation from conservatives—especially those supporting Trump—but, with some targeted outreach mid-semester, we think we were able to secure more regular participation from some (again, we can’t say with absolute certainty since we didn’t specifically ask who espoused any particular political ideology or viewpoint). We cannot say for sure why this was a challenge initially: perhaps it was sheer numbers (Republicans are severely underrepresented among graduate students, staff, and faculty); perhaps we had been advertising in the wrong places; or perhaps there was a lower level of trust, given the campus setting. These are things to pay attention to as we move forward.

What did these dialogues accomplish? A few observations stand out. First, participants mostly seemed to appreciate the dialogues, as evidenced by feedback from some of them, and by the fact that there were regular participants from a variety of roles on campus. Second, we never had anything remotely close to a heated discussion, though we have no doubt that participants of all stripes encountered ideas they disagreed with at some point throughout these dialogues. Third, early in the semester, a couple of participants expressed concern that there were not enough participants sharing conservative viewpoints and perspectives, but these concerns abated somewhat later on. Fourth, we can think of very few instances in which individual participants indicated or appeared to experience a significant change in their opinion. Together, these observations suggest to us that what participants were getting out of the dialogues was an opportunity to try to better understand people with opposing views, even if they didn’t want or expect to find these views convincing. Clearly there is a demand on campus for spaces that facilitate empathy and understanding on challenging issues, even among people with strong and contrasting opinions.

It is also noteworthy that the group received (much-appreciated) support and encouragement from university leaders from department-level to other key campus leaders (and generated zero controversy). This shows that viewpoint diversity, open discussion, and constructive disagreement can and should be core--mundane even--aspects of a campus climate, and shouldn’t be seen as edgy or provocative.

There was enough continued interest that we decided to resume these dialogues in the spring (for those in the University of Colorado system and surrounding communities, the sign-up link is here). We joked that we will continue these dialogues until we solve polarization, or when we retire, whichever happens first.

Two key aspects of the dialogues that seem transferable to fostering viewpoint diversity, open discussion, and constructive disagreement in the classroom are the norms (listed above) and the aspect of repeated interactions in manageably sized groups. Both seemed important to establishing a core level of trust (which participants occasionally remarked on) that gave participants the confidence to both share their thoughts openly and listen to each other with curiosity and humility. The dialogues also tried to strike a balance between being structured enough to encourage participation and feel substantive, but also open-ended enough to allow conversations to move tangentially when constructive and not feel pressured. For instructors who want to try something like this in the classroom, we suggest setting the right tone from the beginning, and keeping the topics relatively simple and open ended.

A final observation from these dialogues that really stuck with us is that Twitter really isn’t representative of the breadth nor diversity of people’s feelings, viewpoints and perspectives, nor are the heated, vitriolic sentiments expressed there. In reality, there is so much more nuance and context and diversity of real, lived experiences and viewpoints, many of which are little-expressed on social media or simply don’t rise to the surface as “trending”. In short, Twitter isn’t real life. We’ve of course been aware of the evidence that the views of frequent Tweeters are relatively unrepresentative and extreme, and we saw how the Biden campaign continued to outperform expectations, and ultimately win, with a strategy of explicitly ignoring Twitter. But it was another thing to see it firsthand: to feel like we were talking to a somewhat representative cross-section of our campus, and feel that the discourse was a lot more reasoned, nuanced, respectful, and less narcissistic. In summary, the dialogues have been a refreshing experience that gives us considerable hope for the future of college campuses.