Tribal Epistemology

Here's a fact that should unsettle anyone who thinks of themselves as tolerant: your unconscious bias against people from the other political party is roughly fifty percent stronger than your unconscious bias against people of other races. That's not some fringe finding — it comes from applying the same Implicit Association Test methodology that established the modern consensus on unconscious racial bias, except this time aimed at political affiliation.¹

Scott Alexander's essay "I Can Tolerate Anything Except The Outgroup" is one of the clearest articulations of what this means. The core argument: what we call "tolerance" is usually just indifference toward people we don't think about much. Real tolerance — the kind that requires actual effort — is directed at your outgroup. And your outgroup isn't who you think it is.

Proximity Makes Enemies

The naive theory of prejudice says we hate people who are different from us. The actual psychology says we hate people who are almost the same as us — close enough to be in competition, different enough to be identifiable. Freud called it the narcissism of small differences. The Nazis didn't hate the Japanese (their outgroup by distance and culture) — they hated German Jews, who were nearly indistinguishable from other Germans except for a few identifying markers.

Alexander maps this onto American political life and finds the same pattern. Blue Tribe members (educated coastal liberals) don't viscerally hate al-Qaeda or ISIS. Those groups are too alien, too distant to trigger real outgroup emotions. The Blue Tribe's actual outgroup is the Red Tribe — conservative, religious, rural Americans who look like them, speak the same language, and live in the same country. And vice versa.

The evidence for this is everywhere once you notice it. When Osama bin Laden was killed, the "intelligent, reasonable" people Alexander knew couldn't bring themselves to celebrate. When Margaret Thatcher died, the same people gleefully shared "Ding Dong, The Witch Is Dead." Bin Laden was more monstrous by any rational measure, but Thatcher was the outgroup. And outgroup triggers emotion that rational evaluation of distant evil simply doesn't.

The Dark Matter Analogy

One of the essay's most striking images is the "dark matter" metaphor for political segregation. Alexander, living in a Republican congressional district, calculates that the odds of not having a single creationist among his 150 closest acquaintances — in a country where 46% of people are creationists — is roughly 1 in 10^45. That's the probability of randomly selecting one specific atom from all the atoms on Earth.

This isn't just a bubble. It's something much stranger: two populations occupying the same physical space but sharing almost no social connections. They drive the same roads, shop at the same stores, but they might as well be made of different kinds of matter. The segregation isn't imposed by geography or law — it emerges spontaneously from thousands of small lifestyle choices about what to eat, what to watch, where to shop, what to find funny. California Pizza Kitchen vs. Cracker Barrel. NPR vs. talk radio. Each choice is individually trivial but collectively they produce 10^45-level separation.

This suggests that tribal identity is more fundamental than political beliefs. "Republican" and "Democrat" are the visible tips of much larger icebergs of cultural affiliation. Harvard might be 80-20 Democrat to Republican, but it's probably 99-1 Blue Tribe to Red Tribe. The sorting happens on implicit markers — tastes, aesthetics, speech patterns — not on explicit political positions.

The Tolerance Trap

The essay's deepest point is about the structure of tolerance itself. If you only "tolerate" people you already agree with, you aren't being tolerant — you're being indifferent. Real tolerance means extending respect and kindness to your actual outgroup, the people who make your skin crawl. For Blue Tribe members, that means conservatives. For Red Tribe members, that means coastal liberals.

Almost nobody does this. Instead, we perform tolerance toward groups that are safely distant (Blue Tribe members ostentatiously supporting groups that are, for them, not actually outgroups at all) while directing genuine hostility at the near-enemy. The result is a political landscape where people who pride themselves on tolerance are often the least tolerant of all — not because they're hypocrites, exactly, but because they've misidentified who their outgroup is.

This connects to the toxoplasma of rage — the tendency for the most controversial, divisive cases to become the most viral. Toxoplasma works precisely because it activates outgroup psychology: the cases that go viral are the ones where each tribe can see the other tribe being maximally wrong. The rage isn't about justice; it's about tribal signaling.

What Belief Congruence Tells Us

The research literature on "belief congruence theory" — the idea that shared beliefs matter more than shared demographics for forming in-groups — has been finding the same thing for decades. As early as 1967, studies showed that people were more willing to form friendships across racial lines than across ideological lines. The finding has been replicated scores of times since.

This doesn't mean race doesn't matter. It means that belief — or more precisely, the cultural tribe that belief proxies for — is an even more powerful sorting mechanism. And because tribe is less visible than race, people don't notice themselves doing it. You notice when your neighborhood is racially segregated. You don't notice when your social circle is ideologically segregated, because everyone around you seems like a reasonable person with reasonable views.

The implication for voting systems and political design is sobering. The problem isn't just that our electoral systems produce bad outcomes. It's that the electorate itself is partitioned into groups that don't share information, don't share experiences, and don't share a model of reality. No voting system can fix a population that has sorted itself into non-communicating tribes.

I'm not sure there's a solution here. Alexander doesn't propose one either — the essay is diagnostic, not prescriptive. But the diagnosis is sharp enough to be useful: the first step toward actual tolerance is recognizing who you actually can't stand. It's probably not who you think.

When Truth Becomes Party Line

Orwell saw the same phenomenon in 1942, without the IAT data, and described it more vividly than any modern researcher has managed. Writing about the Spanish Civil War, he noticed that atrocities were believed or disbelieved "solely on grounds of political predilection." The left believed every Franco atrocity and denied Republican ones; the right did the reverse. When the political landscape shifted, yesterday's proven atrocity became today's propaganda lie, "merely because the political landscape has changed." The same people who mocked patriotism in 1933 were denouncing pacifists as Trotsky-Fascists by 1937.²

What makes Orwell's account devastating is that he's not describing stupidity or ignorance. He's describing the intelligentsia — educated, well-read people who swing from "war is hell" to "war is glorious" with no sense of incongruity. The mechanism is tribal allegiance operating on factual beliefs. Once you know which side someone is on, you can predict with near-certainty which atrocity stories they'll believe and which they'll dismiss, regardless of evidence. "The truth, it is felt, becomes untruth when your enemy utters it."

The deepest worry in the essay isn't about any particular war — it's about what happens to the concept of objective truth itself. Nazi theory explicitly denied that "Science" exists — there is only "German Science" and "Jewish Science." Orwell saw this not as an aberration but as the logical endpoint of tribal epistemology: "a nightmare world in which the Leader, or some ruling clique, controls not only the future but the past." He wrote this in 1942. Two years later he started writing Nineteen Eighty-Four.²