The first meta-analysis of police-community interactions, as recorded by body camera footage, shows officers using less respectful words and phrases with black community members than with white community members.
The researchers, from Stanford University, used artificial intelligence to detect language that had been deemed either more respectful or less so in over 36,000 examples of police speaking to community members. The breadth of the data made available by the artificial intelligence was quite novel.
Researchers hope that having such thorough data will give society the ability to highlight behavior trends like microaggressions that police officers might not even be conscious of, as study co-author Dan Jurafsky notes: "computational linguistics offers a way to...detect the way that everyday language can reflect our attitudes, thoughts and emotions." It may also be helpful, researchers say, in avoiding blame-oriented conversations because the data points from a study that large are so impersonal. With few emotional obstacles to behavior change, policy makers could then focus on implementing awareness-building programs for officers.