The death of 29-year-old Tennessee resident Tyre Nichols after a brutal, caught-on-video beating by several police officers following a traffic stop has sparked a wave of near-universal outrage and reignited passions over police violence and race. But this horrific case also highlights the complexity of this issue: While Nichols is Black, so are the five officers who beat him. Does this mean, as some suggest, that racism in American policing is so pervasive it is internalized even by Black police officers — or that other, nonracial factors are paramount?

“THIS DONT (sic) HAPPEN TO WHITE FOLKS. And this is what BLACK LIVES MATTER MEANS,” wrote Twitter user and podcaster “Papo” in the aftermath of Nichols’ death, expressing a common sentiment. But, of course, it does. Take Deven Guilford, a white teen in Michigan shot dead by a cop after a traffic stop in 2015 under murky circumstances (Guilford complied with orders to get out of the car and lie down, but did not obey a command to let go of his cellphone; the police camera went out just before the actual shooting). Most such cases make only local headlines.

Given the very real history of state-sanctioned racist violence against Black Americans, it’s entirely understandable that the deaths of Black people at the hands of cops get more attention, while the names of white people who die under often-similar circumstances are unknown. But this inevitably creates the widespread impression that the problem is exclusively racial.

At the same time, race is a factor — but a complicated one. Black men are about 2.5 times more likely than white men to die at the hands of a police officer. How much of a racial disparity there is for women is unclear, since very few women of any racial background are victims of such killings. (A white man is about 20 times more likely to be killed by the police than a Black woman.)

Partly, these disparities are unquestionably due to higher crime rates in the Black community (which certainly reflect a history of racial oppression but are still a fact, and which mean that criminal suspects are disproportionately Black). They also undoubtedly reflect racial profiling that extends to people like Nichols who are not criminal suspects, sometimes based on unconscious assumptions and sometimes on conscious racial prejudice.

But there are also nonracial factors that shape police violence. As New Yorker staff writer David Kirkpatrick notes in a comment on the Nichols killing, police training in the United States has often promoted a “warrior” mindset based on a flawed study from the 1990s which encourages officers to see any civilian as a potential cop killer and stresses aggressive dominance as the correct attitude. While such methods have been going out of fashion, they still influence police work (with the presence of legal and illegal guns an additional factor).

Other commentators point to a lack of professionalism as a key part of the problem. American police officers often receive relatively little training — much less than in other countries (minimum hours of training are twice as high in Canada and four times as high in England), and much less than cosmetologists and plumbers. Of course, more police training means more police funding, not less.

Police brutality is an outrage every civilized society must condemn. The greater attention to it in recent years is a good thing. But if we want practical change, not passions in the social media, we need less rhetoric and more facts about what drives police violence and how to reduce it.