Image credit: Irina Baturina/ Shutterstock

The following text is excerpted from an essay first published in Social Research: An International Quarterly. It is part of the journal’s issue Photography and Film as Evidence.

On September 23, 2016, the New York Times released a cellphone video of the police shooting death of Keith Lamont Scott in Charlotte, North Carolina. Police claim that they saw him holding a gun. (In the days after his death, the police claimed to have “recovered [a gun] from the scene.” Scott’s wife disputes that he was holding a gun or even owned one. She also disputes the official police account of the shooting, which was attributed to the only Black officer on the scene, whom Scott claims witnessed the event from a distance.) Scott had been sitting in his parked car outside his apartment complex, as he often did when waiting for one of his children to return home from school. Seeing the police suddenly surround Scott’s parked car—they arrived there to serve a warrant to another person—his wife repeatedly told the police that her husband did not have a gun and presented no threat. The cellphone video of the shooting provided by the family to the New York Times included an introductory slide and subtitle: “It was recorded by his wife, Rakeiya Scott.” I pressed pause, feeling unable to watch. I had heard that the police had shot and killed a Black man in Charlotte, that his wife had been there, pleading with the police. But I had not known who recorded the video. I was stopped in my tracks by the matter-of-fact annotation: Scott had recorded the murder of her own husband as she tried to prevent it.

It took some time before I could come back to the video. Immediately my thoughts returned to Diamond “Lavish” Reynolds, a Black woman in Minnesota, who just two months before had also recorded the police murder of her beloved, Philando Castile, during a traffic stop. Reynolds’s video was immediate in its dissemination, livestreamed on Facebook Live. The shooting of Castile by police officer Jeronimo Yanez was not captured on video; there is no visual evidence of the event made by a bystander. Instead, Castile’s dying moments were captured as Reynolds narrated the immediate aftermath of the shooting, the event still happening. There was no rescue she could have attempted. Even though Castile was alive and Reynolds’s daughter sat in the back seat, the effort immediately and of necessity turned to testimony. The camera—but also narration, putting into words the event that was still unfolding—afforded Reynolds and her daughter some measure of protection against the officer’s gun, still aimed in the car. Reynolds perhaps imagined that if she herself did not survive the event, the recording would have already been seen by the public; her voice would testify from within it.

It was an unpredicted use of what was then a new medium, which has since banned violent images and scrubbed Reynolds’s recording from the site.

Facebook’s censorship of such documented violence points to a boundary in public life: death is not meant to be livestreamed. As a medium, livestream posits the desire for imagined community and life. If there was no one who could intercede, it was because the authorities who might respond to such a call were the ones responsible. The very fact of the video’s existence posits a fissure in the imagined community’s legal and ethical condition of possibility. With the family intimacy between witness and witnessed, the determination of such recordings exceeds those of bystander videos that, since Rodney King, have saturated the media sphere with images of police brutality. Their determinations are both personal and temporal. These are family documents and records that are thrust into the public sphere, the loved one displayed as citizen. The temporality of smartphone technology makes an unspeakably horrific demand upon these women as “witness and participant”: to testify to and gather audiovisual evidence of a crime against a loved one while it is taking place. Such speech on behalf of the beloved places a series of demands upon viewers as they are also listeners. These demands, if attended to, mark a critical impasse in existing discourse surrounding viral videos of Black death, discourse that, since Elizabeth Alexander’s seminal “Can You Be BLACK and Look at This?”, primarily concerns looking or “not looking,” or what media theorist Alexandra Juhasz calls, in an essay on her decision not to watch Reynolds’s video, “surfeit images.” Such a decision is largely consistent with the paradigm of visual violence proposed by Susan Sontag in On Photography, wherein she claims that looking at photographs of war not only cannot make one more ethical but also inures one to violence. Sontag recovered an ethical impulse in a refusal to look (paradoxically, in a book that studies images in detail), a counterpart to the commonplace argument that if there were more bodycams, these crimes would become more avoidable; in other words, in order to see less, we must first see more.

Speaking of Rakeiya Scott’s video, Charlotte Police Chief Kerr Putney acknowledged that neither the police bodycam footage nor the family footage shows Keith Scott’s hands. “The video does not give me absolute, definite visual evidence that would confirm that a person is pointing a gun,” Putney said. Reynolds’s video begins in medias res and it does not capture the events leading up to the shooting. Neither video offers conclusive visual evidence that would legally exonerate or incriminate the killers. The videos’ very audibility and emotional proximity to the event—made by women who are both witnesses and kin—raise questions about the meaning of the videos’ existence, the purpose of their dissemination. These questions countermand the initial impulse not to look or to excise snuff. Race is no doubt a visible artifact, the visual field being what Judith Butler has called a “racial formation … hegemonic and forceful.”

What in the videos demands to be heard? “Hearing, too, is central to witnessing,” Alexander writes. “Sounds … haunt the mind as much as visual images.” While the video of King being beaten is largely remembered as silent, the police repeatedly hurled slurs at King, audible in a video that was muted by lawyers who slowed the footage to play it frame-by-frame before the jury in an effort to make King appear more menacing. “In this regard,” Alexander continues, “the freeze-framed Simi Valley videotape, stripped of a soundtrack in which falling blows and bystanders’ screams are audible, disallowed the possibility that the sounds of terror could imprint themselves on the jury’s mind.” The silence is an artifact of a juridical desire to secure a racialized field of vision.

If the jury had heard the video, perhaps they would have seen it differently. With an audio artifact, we are in the realm of an interpellation or call: the sound of terror is an address. One experiences the call as being for someone, sonority being the relational, reciprocal ground of democratic politics in general. A voice is “structurally for the other,” writes Adriana Cavarero. In the cases of Reynolds and Scott, the sounds of terror are an oblique address, alongside the voices that narrate the unfolding events, which are a direct address. As an address, then, these videos carry a particular kind of evidentiary and ethical charge.

As civic documents in the United States, these videos arrive in a contentious public sphere that has inherited multiple traditions of witnessing. There is a long history of being forced to witness a loved one’s state-facilitated murder and narrating its outrage, and such testimony is foundational to Western letters. The Gospels culminate in the execution of Jesus as his followers witness the spectacle. The text harnesses the evocative power of narrative to transmit the message and arouse belief. In the Phaedo, Socrates’s followers report his dying words “so imaginatively that the reader feels he is actually present.” The transmission of the message as conviction lies in the temporal effect of narrative discourse whose power is to produce belief, not simply in the ethical character of the narrator, but in the narrator as present to the scene of the execution. Juridical traditions in the United States inherit these techniques, where to “testify” in a court of law never loses its doctrinal sense of conviction through physical and emotional proximity to the crime…

In the Oxford English Dictionary, “testify,” from the classical Latin testificari, means to bear witness, proclaim. “Testimony,” however, makes less plain the oral character of the evidence: “personal or documentary evidence or attestation in support of a fact or statement; hence, any form of evidence or proof.” As testimony, one can offer a statement or tangible object. Yet, the verb “testify” retains its oral/aural charge, partly through its Christian origins wherein to testify is to speak the truth, as in John: “We speake that we knowe, and testify that we have seen.” The meaning of transmitting witnessing to those who were not there is to reinforce the social bond: through listening to the testimony of the other, I “see” the crime, and in “seeing” it, I become a witness. In becoming witness, I become party to the demand for adjudication. When speech is addressed to the polis and pertains to a crime, the story bonds citizens to the actions undertaken by others. Thus, I may feel outraged at a crime against another because I feel it directly related to me. The change in person, then, is both temporal and ethical…

When Reynolds began recording and streaming live, Castile was already mortally injured … and could no longer speak… She became Castile’s voice for him. The officer ordered her to leave the car and then to get down on the ground. Her phone fell to the ground, and while the lens was pointed at the blue sky, the camera continued to record and capture audio. Though viewers continue to see, the frame is so limited as to no longer offer visual evidence in any traditional sense, showing no crime and no perpetrator. The power of the wound and the voice nevertheless remains constant.

Julie Beth Napolin is Associate Professor of Digital Humanities at Eugene Lang College. Her monograph, The Fact of Resonance: Modernist Acoustics and Narrative Form (Fordham UP, 2020), was shortlisted for the Memory Studies Association First Book Award.