When riots followed protests here in the Twin Cities in the immediate aftermath of the death of George Floyd, a rumor began going around our neighborhood. Per a Facebook post and a printed handout of the same text apparently distributed door to door, some 75,000 white supremacists were going to converge on the metro area in a single Friday night to harass people of color and incite violent clashes in the streets.

The rumor was unfounded. That Friday night was mostly peaceful — the worst chaos had come in the two days prior. And the 75,000 figure was a garbled version of a state warning that up to that many outside agitators of various (i.e. competing) extremist groups could converge and, most likely, conflict among themselves.

But before it was debunked by the passing of a comparatively quiet evening, this rumor had people terrified. Some were organizing their neighbors for all-night patrol shifts on their blocks. The Facebook post advised taking out pots and pans to raise the alarm if you saw something suspicious. Gossip spread a story of a Ku Klux Klan gathering, hoods and all, gearing up for terror at a local park. The park was empty and, later, the story retracted. But why did this all seem so plausible to so many at the time?

The answer the Very Online Discourse increasingly wishes to give is that we may be nearing a new civil war. Two Ph.D. students at the conservative Hillsdale College in Michigan have started a website tracking media discussion of this possibility across the political spectrum, featuring articles like my colleague Damon Linker's recent column on whether our country could "split up." Just a single month, the aggregators report at American Greatness, has already demonstrated civil war rhetoric is on the rise: "In June, we counted 23 articles written about the prospect of a new or cold civil war in the United States. In July, that number doubled to 46. That's no mere 'uptick.'"

Neither, however, is it proof of anything at all. Take a look at these Google Trends lines for relevant terms:

Ignore the March 2016 spike — that's about a trailer release for Captain America: Civil War. Exclude that, though, and there's a clear pattern: Every four years, in every run-up to every presidential election since 2004 (the furthest back this data goes), Americans start despairing of our national unity.

But just how significant is that despair? Let's keep in the two most popular of those terms (secession and civil war) and add another useful comparison, "electoral college." The four-year cycle is even clearer. But notice how much greater search interest there is in the Electoral College (and, presumably, its abolition) than in violently dividing the country. Searching "recount" produces a similar pattern and also generates more interest than internecine battle. Even in our most intense moments of political strife, Americans seem substantially more interested in peaceful, if dramatic, political reform than in taking up arms against their neighbors.

Yet — and maybe this is just an artifact of being part of the Very Online Discourse myself — I understand the feeling that this year is different. The tensions of 2004, 2008, 2012, and 2016 were not compounded by the strain of pandemic, its economic disasters, and months of large-scale protests and riots in cities nationwide. Only one of those elections involved Donald J. Trump, now a sitting president who likes to troll about holding office indefinitely and talk — in a manner that's not indisputably mere trolling — about refusing to accept this year's election results should he lose. And 2020 is the only year, too, where the incumbent's presumptive challenger has spoken of ordering the military to hustle his predecessor out the White House should he decline to leave.

This year we have all that layered atop of the usual concerns about debt and taxes, corruption and partisanship, competing rights and zero-sum privileges, and an apparently endless accumulation of wars, both foreign and cultural. Have we reached an inflection point that wasn't there in anxious cycles past? Should we see something more in that uptick of "new civil war" stories than the standard (and perhaps self-reinforcing) four-year pattern? Do we, as Linker asks, truly "hate each other"? Do we hate each other enough to kill?

My inclination is to say the hate is there, but also largely abstract. Negative partisanship, the academic term for loathing those selfish, short-sighted monsters on the other side, is on a long rise, even as positive partisan identification falls. In other words, someone on the left is now less likely than in the past to identify as a Democrat but more likely to outright hate Republicans — and vice versa for the right.

But such hate isn't as clean-cut as some polling suggests. Negative partisans express more distaste for particular, polarizing political figures than the other party in general, and I suspect their hatred declines again — even ceases to be hatred at all — if asked to think of specific family and coworkers on the other side of the partisan line. There's a lot of space between working oneself up while watching cable news and actually shooting or stabbing someone over politics. So far, I remain unconvinced that's a space enough Americans would cross to produce conditions fairly labeled "civil war."

Still, this mutual animosity and anger strikes me as a more compelling basis for a claim of movement toward civil war than what may be merely cyclical imaginations. In pre-Civil War Spain, children play fought as "leftists and rightists" rather than "cops and robbers," Hilari Raguer, a historian and monk who was 8 years old when the war began, has recalled. If your kids start playing violent "Republicans and Democrats," maybe sound the alarm.