Linguistic models of police-civilian interactions during traffic stops reveal racial biases in the politeness of law enforcement.

Figure 1. A California Highway Patrol officer makes a traffic stop on southbound Highway 85 in Cupertino. In a 2017 study of interactions from 981 traffic stops in Oakland, participants ranked police officers’ speech with black drivers as significantly less respectful than their speech with white ones. (Image Credit: Wikimedia Commons)

For the past couple decades, and very prominently in the past few years, police brutality has been a prominent topic in the media, spurring controversy and protest in multiple communities. Countless episodes through many decades, like the beating of Rodney King or shooting of Philando Castile, have become viral, shaking the nation as the images of brutality are etched into our minds as we stand in solidarity with the victims. The scenes have been labeled blatant acts of racism, and the frequency of these events have stirred distrust of the nation’s police force. These fatal altercations have brought the issue of police brutality to the forefront of U.S. policy. In investigating these cases, the debate then becomes whether the police are acting in self-defense or prejudice, prompting the question of whether or not oppression systematically exists within interactions between police and civilian.

Researchers at Stanford studied this exact issue. In a 2017 study led by Rob Voigt, researchers show that officers treat African American citizens with significantly less respect than Caucasian citizens even in common traffic violations.

Using body camera footage of police officers from Oakland, CA, the researchers gathered true verbal interactions from 981 stops of both black and white drivers during the month of April 2014. After the conversations were transcribed, they were presented to random participants who were asked to rank how respectful, polite, friendly, formal, and impartial the sentences were on a scale from 1 to 4. The participants were given an initial phrase from the driver and the reply from the police officer, thereby providing them with context for both sides.

Participants had high correlation in the similarity of their rankings for each scale (ranging from Cronbach’s alpha = 0.73 to 0.91, where 1 is the highest possible), proving that pure wording has strong signaling ability of the dynamics of the interaction. Further, it was shown that the interactions between policemen and black drivers were perceived as significantly less respectful than that with white drivers in all five categories. The researchers then built a linguistic model based on the almost 40,000 utterances they had between officers and civilians, which agreed with the ratings that were given by humans in the previous study.

Controlling for the severity of the offense, officer race, civilian race, civilian age, and civilian gender, they found significant racial disparities in the respectfulness of officers between interactions with white civilians and interactions with black civilians.

Figure 2. Several sample police statements’ respectfulness scores. Blue words are correlated with more respect, while red words are with less. (Image Credit: Voigt, R. et al., 2017)

Their linguistic model pulled out distinguished words or phrases that had a high impact on whether an interaction was considered respectful or not. Informal titles, like “my man” or “bro”, were much more common in interactions with black drivers and were highly common in sentences considered disrespectful. Similarly, the phrase “hands on the wheel” is the phrase that the model found to be the most disrespectful, and it was much more commonly used by policemen in interactions with black drivers. On the other hand, addressing the driver’s safety was significantly more common in interactions with white drivers and was a sign of respect in utterances.

In addition, they found that formality was not significantly varied between interactions with officers and white or black community members. Thus, this disproves the hypothesis that formality could be mistaken for disrespect in the interactions.

The researchers also accounted for the possibility that there could be just a few extreme officers who showed large differences in respect. They proved, through a distribution model, that it was not the interactions of a select few, but rather those of all of the officers, that showed this disparity in respect.

By utilizing the robust database of police body camera footage, the researchers dove into the everyday interactions between officer and community, proving that, even in these normal, outwardly civil conversations, racial disparities are continually propagated. Even in the interactions that don’t break out into violence and death, ones that don’t make the nationwide headlines, racial prejudice is at work.

These findings may seem insignificant to some, as, put brutally, “what doesn’t kill you makes you stronger.” And these conversations are just exchanged words, an interaction that comes and goes, another life event in a day of normalcy. So we might just let it happen, allow law enforcement agencies to deliberately act against civil rights laws under the pretense of smallness, brush the issue aside in exchange to focus on something seemingly of grander scale. But history repeats itself, and small issues never remain small for long.

These small-scale forms of aggression are the building blocks for something much greater. Society has always been convinced of it, and these numbers prove it: blacks are treated worse. The subtlety of this treatment is no excuse. Micro-aggressions are aggressions nonetheless, and a measurable level of unwarranted disrespect toward one racial group more than another is a blatant form of racism. It can be argued using this research article that officers, no matter their race, are prone to be distrusting and wary of blacks. Even when responses by the civilians are taken into account, the utterances by officers were still considered disrespectful. In a situation where complete strangers are entering conversations with preconceived notions about each other, self-defense is a potent force, and words that connote disrespect act as the weapon.

It’s in these interactions that animosity is generated, that the distrust of our law enforcement officials can be continually perpetuated until, one day, mutual disdain turns into life-threatening altercations.

As a nation, America has made great progress in the notion of ensuring the equality of all citizens. That being said, there are still subtle but potent forms of expression and interaction that contribute to the segregation of our society, that prevent us from forming a state in which citizens can feel safe in their own skin.

To see the importance of this—to be wary of the smallest forms of racism—is to understand how prejudice is still being perpetuated into the current age, and this understanding is the first step to solution.