Skip to main content
ABC News
When Proof Is Not Enough

According to the pundits, the revolution, if you would call it that, began with video. The first and foremost was the excruciating recording of George Floyd’s last moments as Minneapolis police officer Derek Chauvin1 publicly pinned the life out of him. That was on May 25, but more than a month later, the recordings have continued to disseminate. Protesters uploaded photos of rubber bullets, their wounds and their mangled faces, while journalists and other concerned members of the public aggregated footage of police brutality into lists and websites.

The compilation of evidence has seemed to jar something loose, for now. Corporations are pledging to donate millions to racial and social justice causes,2 legislators have proposed tentative yet unprecedented restrictions on the police, and the Marines and Navy have banned the Confederate battle flag3 some 150 years after the ending of the war that sparked its creation.

But is this really going to be what commentator Van Jones has called a “Great Awakening of empathy and solidarity”? And if it is, is it really appropriate to claim that video has been the catalyst? I work with civic data and teach about the power of data collection, so I want to believe that data (in the form of video footage depicting police brutality against Black people) can effect social change. Just as it is comforting to see corporate and institutional pledges as revolution, it is comforting to attribute power to the millions of glowing screens that have been called as witnesses.

Data showing racism might be useful in clarifying the things we already know to be true, but it is far more limited in terms of shifting them.

But it is precisely because of my attachment to the power of data collection that I’m unconvinced video footage can solely, or even primarily, lead to meaningful change. I know too well the stories of a century of Black Americans who have presented evidence of violence and racism only to have it summarily denied or ignored. The idea that structural racism can be proven and overcome by gathering just enough or the right kind of evidence is nothing more than a myth. Historically, it has rarely been the case.

Consider, for instance, the study that the Bureau of Labor commissioned famed Black scholar W.E.B. DuBois to complete in the early 1900s. Determined to employ sound sociological methods to disprove racist beliefs that Black people were inferior, he and a team of researchers spent three years in Lowndes County, Alabama, gathering data from 5,000 Black families (approximately 25,000 individuals). It detailed the conditions of life in the region, and was one of the largest sociological studies of rural Black life ever conducted. When DuBois submitted the final manuscript, it was a handwritten document full of charts and infographics.4 Not only did the government bureau refuse to publish the study, but it destroyed the document entirely, claiming it was rejected due to technical matters. DuBois made the case in his correspondence and autobiography, however, that the bureau rejected the document because it revealed the inconvenient political truth about conditions for Black Americans.

[Related: The States Taking On Police Reform After The Death Of George Floyd]

In the case of Sam Faulkner, an innocent 20-year-old Black man who was shot in the head inside his sister’s home by Los Angeles police in 1927,5 evidence came in the form of testimony from the other cop on the scene as well as bullet fragments. Yet this was not enough to bring about a conviction, and the officer who killed Faulkner continued to work in the LAPD for two more years.6

In 1951, the Civil Rights Congress appealed to the United Nations for help, asserting that the history of disenfranchisement, lynching, and police brutality that Black people faced in the United States was tantamount to genocide.7 The CRC’s petition8 documented years’ worth of atrocities against Black Americans but was ignored by the U.N., which at the time was heavily influenced by the U.S.

In 1969, Illinois Black Panther Party leader Fred Hampton, 21, was gunned down9 in his Chicago apartment after being sedated by an FBI Informant. A target of the FBI’s Counterintelligence Program, he was perceived to be a threat to the nation for negotiating a truce between street gangs, organizing rallies and instituting free breakfast programs for children. A coroner’s jury ruled the killing a justifiable homicide.

And even more recent incidents including video footage of police brutality have been doubted. When Philando Castile was pulled over and shot by police in 2016, the dashcam footage revealed that Castile, who had been stopped by the police at least 46 times prior for minor infractions, had followed all the instructions that officer Jeronimo Yanez had given him. Regardless, an NRA spokesperson still blamed Castile for the incident, while conservative commentator Sean Hannity criticized Castile’s girlfriend, who was seated beside him in the car, for live-streaming the interaction in the first place.

These killings, and the many more that reveal just a glimpse of how totalizing anti-Blackness can be, are part of a longer trend. It is a trend that has claimed countless more names, and still more stories. By nearly every statistical measurement possible, from housing to incarceration to wealth to land ownership, Black Americans are disproportionately disadvantaged. But the grand ritual of collecting and reporting this data has not improved the situation. American history is lined with innumerable instances of what scholar Saidiya Hartman bemoans as “the demand that this suffering be materialized and evidenced by the display of the tortured body or endless recitations of the ghastly and the terrible,” only for very little to change.

If the data hasn’t undone the bias, then surely we must acknowledge that there are deeper forces that tug the levers of change in America. I am reminded of James Baldwin’s response to the 1954 Supreme Court case that ended segregation: “Had it been a matter of love or justice, the 1954 decision would surely have occurred sooner; were it not for the realities of power in this difficult era, it might very well not have occurred yet.” Love, justice, data — alone, none have been enough.

But perhaps we have asked too much of the evidence in the first place. Or perhaps we have asked too much of those who wield evidence, and too little of those presented with it. These are two different groups. After all, evidence is not intended for the people who have been harmed — why show proof of a fire to the person it burned? In most cases, evidence is used to convince an Other of a thing that they did not encounter. Ironically, data is not very good at this.

In 1949, two psychologists, Jerome Bruner and Leo Postman, designed an experiment to test people’s responses to anomalies, or moments when they faced events that deviated from what they had expected to encounter. In the experiment, participants were shown sets of playing cards and asked to identify the cards’ color and suit. The catch was that, in addition to regular cards, the sets contained irregular “trick” cards in which the color and suit of the cards had been reversed to create incongruities (like a black three of hearts or a red two of spades).

[Related: Why Statistics Don’t Capture The Full Extent Of The Systemic Bias In Policing]

In the early rounds, the participants were quick to identify the cards, in part because they simply could not see the anomalies. When presented with a trick card like a red six of spades, they would confidently misidentify it as a red six of hearts or a black six of spades. But as they were exposed to the cards for longer periods of time, some participants began to notice that something was off. They could sense strangeness but could not determine what caused it. It was only with further exposure that some participants finally experienced what the psychologists called a “shock of recognition.” Abruptly and quite clearly, the participants were able to recognize what they had not seen before. Suddenly they could see that they had been looking at a red six of spades the entire time. From that point on, they were more easily able to identify the anomalous cards, having developed a new perception.

The conclusion: When confronted with something that does not fit the paradigm we know, we are likely to resist acknowledging the incongruity. This is because we see what we have been primed — through shared education and culture, and our own lived experiences — to see, so that new evidence that we encounter is immediately, as philosopher of science Thomas Kuhn would explain it more than a decade later, “fitted to one of the conceptual categories prepared by prior experience.” Kuhn applied this reasoning to explaining the tumultuous nature of scientific revolutions, where he argued that the conceptual categories that ordered scientific research were precisely those that made it so difficult for scientists to accept information that could challenge the frameworks they operated within. In such moments, logic and experiment alone were not enough to settle the matter. Kuhn noted, too, that the more time and effort a scientist had already invested in a research paradigm, the more resistance he or she was likely to exhibit toward accepting change. In other words, the higher the stakes, the greater the resistance.

If wider society recognizes data’s limitations, it, too, can move on from overly relying upon it as the only proxy for evidence.

You can see how this is a useful metaphor for considering the United States, racism and the role that data has historically played in unraveling the latter’s hold on the former. Data showing racism might be useful in clarifying the things we already know to be true, but it is far more limited in terms of shifting them. To those who have not experienced the ever more creative forms that structural racism can take, even when presented with evidence of racism, the world may still appear to be full of regular playing cards. This is complicated, too, by the fact that in life we face different likelihoods of encountering anomalous cards, depending on factors like the color of our skin (whiteness, of course, lowering frequency of exposure) and proximity to the affordances promised by wealth, influence and cultural/political capital. Regardless, any exposure to an anomaly card is more likely to be dismissed if it does not support the expectations of the receiver.

Of course, as in the experiment, there is the opportunity for change. Perhaps one part of what has characterized this current moment is that some sections of American society have experienced their own moments akin to when the experiment participants first squinted at the trick cards and felt that something now felt off. At some point, America will have to confront head-on the fact that the country not only has long educated its children to deny anti-Blackness and to treat any conversation of racism with silence or wariness but also has exported this worldview around the globe. For some, that point may have come.

But regardless, a luckless great many of us know that the deck has been stacked from the beginning. And because we know that no amount of shouting, pleading, calculating or visualizing will persuade those who have been educated and raised to deny this, we have put our efforts in other places.

If wider society recognizes data’s limitations, it, too, can move on from overly relying upon it as the only proxy for evidence. That which can be captured on camera is always incomplete. It is never the totality of what occurs in our lives, let alone what occurs in our communities. By considering the vast context and evidence present in the nation’s history, we can save ourselves from tacitly reinforcing the idea that structural violence matters only when it can be compressed into a form that fits what we recognize as evidence. And, in doing so, we give ourselves new frames for thinking about the many people who have died at the hands of brutality and whose deaths were not recorded. As we find a fluency in addressing the greater mass of life that is lived outside of our data, we can begin, finally, to fully address the living.

How cable news is talking about the protests — and why it matters | FiveThirtyEight


  1. He was fired from the police department on May 26 and later charged with second-degree murder and second-degree manslaughter.

  2. A total of some $1.7 billion has been pledged so far.

  3. The Army says it is considering a ban.

  4. No copy existed according to DuBois’s autobiography.

  5. Two police officers raided the home of Faulkner’s sister to search for contraband liquor but found nothing.

  6. The chief of the LAPD promoted the exonerated officer to lieutenant sergeant and fired the officer who testified for the prosecution.

  7. The U.N. General Assembly adopted the Genocide Convention in 1948.

  8. Titled “We Charge Genocide: The Historic Petition to the United Nations for Relief from a Crime of the United States Government Against the Negro People,” the paper was 239 pages long plus five appendices.

  9. Hampton and Mark Clark were killed in the raid. Four others in the apartment were wounded.

Mimi Onuoha is a Brooklyn-based researcher and artist whose work focuses on power dynamics within emerging technology and data collection. She is currently a visiting assistant arts professor for the Interactive Telecommunications Program at NYU’s Tisch School of the Arts.