Last October, THE CIVICS Innovation Hub and the European School of Politics convened an international group in Istanbul for a conversation about “trust and polarisation.” Kameliya Tomova has written a nice summary. I’ll paste the portion that mentions me below and recommend the rest as well. (Note that I was talking here about the world, not necessarily or specifically about US politics.)

Peter Levine, political theorist and civic scholar, cautioned against treating all forms of division as equivalent. “Is the problem that two sides are too far apart,” he asked, “or that one side is organised around hate and the other around love and dignity?” The answer to that question, he suggested, has profound consequences for whether and how we even attempt dialogue. Levine argued that not all polarisation reflects symmetrical extremes — sometimes one side advances exclusion while the other defends basic rights. In such cases, the work of bridging may look very different, or may not be appropriate at all.

Building on this concern, human rights and peace activist Harsh Mander warned that insisting on symmetry between “sides” can normalise authoritarian or dehumanising positions. Drawing on his experience in India, he asked: “If I say Muslims deserve to live with dignity, and that’s seen as an ‘extreme’ view, then what is the centre? Mild dehumanisation?” The language of depolarisation and how broadly it’s currently being used, he argued, risks collapsing injustice into mere disagreement if moral asymmetries are not explicitly acknowledged.

Others spoke of perception gaps — the distance between what we think others believe and what they actually do. When people have limited direct contact and rely instead on distorted signals from online spaces, they tend to assume others hold more extreme and internally consistent positions than is often the case. Peter Levine noted that quantitative research frequently reinforces this assumption by treating political identities as coherent blocks — for example, presuming that someone who holds a conservative position on one issue will do so across others. In practice, he argued, people’s views are far more fragmented and situational. These misperceptions reduce willingness to cooperate, until direct interaction or clearer information disrupts the assumed coherence of the “other side”.

See also: “People Are Not Points in Space: Network Models of Beliefs and Discussions“; US polarization in context; class inversion as an alternative to the polarization thesis