The most profound division may be over the nature of bias itself. Now that frank prejudice is ostensibly out of bounds, the country finds itself in murkier territory, arguing about the kind of bias that is less obvious and intentional. While some people (mostly on the left) puzzle over the lessons of studies like "Seeing Black," others (mostly on the right) feel blamed for what they see as an imaginary problem.

Sigmund Freud popularized the concept of the unconscious in the early decades of the 20th century, describing a model of the mind in which some cognitive processes lie beneath the surface, waiting to be "discovered and translated into conscious form." Beginning in the 1930s, Freud's followers in psychoanalysis theorized that along with fleeting thoughts and feelings, attitudes including prejudice and stereotyping could take root in the unconscious.

In 1998, the psychologists Anthony G. Greenwald, Mahzarin R. Banaji and Brian Nosek began a groundbreaking project to test these ideas empirically. They used an "implicit association test," which measured the speed of responses for associating positive and negative terms ("terrific," "lovely," "evil," "hurtful") with black and white faces, by hitting keys on a computer. More than five million people have taken this test online to date, and the researchers have found that most whites and Asian-Americans more quickly associate positive qualities with white faces than with black ones. So do more than 30 percent of African-Americans.

This type of bias stems from the human instinct to order the world, by sorting its pieces into familiar groups. Often, automatic associations are morally neutral, like the link between "doctor" or "nurse" and "hospital." Instantly connect "doctor" to "he" and "nurse" to "she," however, and the links become more loaded. Still, your choices don't necessarily reveal that your true self is sexist. They express the influence of stereotypes, but they're not an endorsement. "It's not a comment on your character," says Phillip Atiba Goff, one of the psychologists who conducted the experiments in "Seeing Black," with a team led by Jennifer Eberhardt of Stanford University.

In the realm of politics, however, that distinction is easily lost. "Implicit bias is a problem for everyone," Hillary Clinton said at the first presidential debate in September. She was answering a question about the police, and for a second after responding she paused, perhaps nervous that she would be misinterpreted as painting them as racist. "I think, unfortunately, too many of us in our great country jump to conclusions about each other," Clinton finished. It took two days for Trump to twist her words into the accusation "that everyone, including our police, are basically racist and prejudiced" — in other words, you're guilty and you just don't know it.

At the vice-presidential debate a couple of weeks after Clinton and Trump's exchange, Mike Pence expressed incredulity that the shooting of Keith Lamont Scott, a black man in Charlotte, by a black police officer could be an example of the same phenomenon. "Enough of this seeking every opportunity to demean law enforcement broadly by making the accusation of implicit bias every time tragedy occurs," he said.

Law-enforcement officials do not necessarily see it Pence's way, however. In a speech about race and bias at Georgetown University last February, the F.B.I. director, James Comey, frankly acknowledged that for many police officers, "the two young black men on one side of the street look like so many others that officer has locked up. Two white men on the other side of the street, even in the same clothes, do not. The officer does not make the same association about the two white guys, whether that officer is white or black."

Over the summer, the Department of Justice announced that it would start implicit-bias training for 28,000 personnel, including F.B.I. agents, prosecutors and immigration judges. Many police officers already take implicit association tests about race and crime, says Noble Wray, a former police chief in Madison, Wis., who leads the Policing Practices and Accountability Initiative at the Justice Department. "We try to help officers understand that if you're going to deal with your implicit biases, you have to be introspective. You have to ask yourself questions about how you're interacting with people."

The stakes of implicit bias may be particularly high in policing, but that's hardly the only area of public and private life where bias has consequences. In the last decade, researchers have repeatedly found a discrepancy among men and women in letters of recommendation for faculty positions in the sciences and in teaching evaluations across various disciplines. Last summer, Sara B. Pritchard, an associate professor of science and technology studies at Cornell University, ran through the evidence for this kind of gender bias and proposed giving women a scoring bonus for teaching evaluations across the board.

"Critics will no doubt argue that such a policy would give female faculty a distinct and unfair advantage," Pritchard wrote in an online post — and indeed, on a Fox News panel, Tucker Carlson called her idea "insane." One of his fellow panelists was Cabot Phillips of Campus Reform, a conservative group that describes itself as exposing "bias and abuse" on college campuses. He cited a study that found that university hiring committees preferred women for positions in the sciences — when they were competing with identical male candidates. In other words, when women, too, had stellar recommendation letters and teaching evaluations, universities would snap them up to address the gender imbalance in science departments.

Phillips, who had just graduated from Liberty University, said that his favorite college professor was a woman. He meant to undermine the case that gender bias is real. But he was actually exemplifying the effect of working with people across lines of race and gender when they are your equals or betters.

This presidential campaign, which began with a wall as its chief symbol, has driven a wedge into the body politic. Polls show a widening split in how men and women plan to vote, and the gap between white voters and those of color is even larger. We have to confront that rift. The work of knitting the country back together means looking inward, to understand the biases we didn't ask to have, and then outward, at people we didn't fully see before. The question before us is how many Americans are willing to do that.

Continue reading the main story