Since 1991, when Rodney King’s beating was captured on video by a civilian named George Holliday, the medium’s use as legal evidence has increased immensely. Understanding the functions and consequences of this shift demands a deep consideration of video’s relationship to temporality and memory. For years, artists and media theorists have been pursuing this question. While Dan Graham’s video delay works of the 1970s are of foundational importance, this line of inquiry runs all the way to current art practice. Paying particular attention to evidence of police brutality, this essay reflects on video’s alteration of our ability to recollect cause and effect. How, I will ask, has video technology changed our ability to memorialize the “moments” in which decisions of exigent importance are made?
The reception of video art has been buffeted by historical shifts in the role video plays in the broader culture, for instance, as surveillance, entertainment media, or “home video.” Currently, video is framed by its role as evidence, both official and amateur, surveillance and sousveillance. While Sony originally marketed its cameras as a way to archive the ephemeral moments of family life, it is now true, as columnist Gil Troy recently opined, that “[d]uring the ’90s, America was becoming Surveillance Central, the most photographed nation ever, thanks to proliferating personal devices and institutional security cameras. Today, we are all George Holliday, armed with digital cameras embedded in our cell phones, ready to videotape history.”1 But while the fifty-six baton strikes captured by Holliday’s camera seemed to render the police’s guilt self-evident, an all-white jury disagreed. One has to wonder, as Curtis E. Renoe wrote in a 1996 essay on the power dynamics of expert testimony, “what the verdict would have been without the video.”2 Renoe continues, offering an explanation for the verdict that is counter-intuitive to the belief that video images are intrinsically delivering of justice:
The videotape […] which at first seemed to be so damning to the defense is miraculously transformed into their most important piece of evidence demonstrating innocence. The ability to take the videotape out of the context of its production, stressing some parts while deemphasizing or eliminating others, provides a way in which to incorporate this record into the framework of expert analysis. Without the video, there would be nothing with which the experts could work, no visual record, thus no way to make the beating “make sense.”3
The indeterminacy of the image, its role as a meeting point of materiality and interpretation, has been intensively investigated by artists. Not a video work, but a film, John Smith’s well-known Girl Chewing Gum (1970), in which the artist narrates footage of passersby on a Hackney street corner, as if he is directing extras, as if he has total agency and control of the image, is paradigmatic here. Since the late 1970s, consumers have been able to use VCRs in order to stop, fast-forward, and rewind footage, just as the defense did in the King case. In the 1990s, shortly after King’s beating, artist and theorist Tetsuo Kogawa deftly observed that this ability to manipulate video gave the medium a performative dimension.4
The recent influx of videos of police brutality has reminded us that while cameras can capture bodily gestures, they fall short of capturing the hyper-speed—not to mention invisible—process of human decision-making. In this gap between video’s abilities and failures, lawyers and “video experts” function as performers of the video material, in the sense of “performing” video technology as forwarded by Kogawa. Controlling the flow of the video, and interlacing it with expertly crafted narratives, these lawyers and experts take on the aura—in the eyes of jurors—of firsthand observers. This is a kind of courtroom sleight of hand, which gives uncertain images the authority of impartial and readable documents.
A similar phenomenon attends “hindsight” evidence. Hindsight is the knowledge you—in a legal context, perhaps in the role of defendant, witness, judge, or jury—have of an event based on subsequent experience or information. It connotes a perspective or knowledge you did not have at the time of the event. In courtrooms, hindsight may include an important fact, such as the revelation that the victim was reaching for a wallet, not a weapon—facts that may never enter the proceedings if deemed to be without probative value in relation to a decision made in the moment (i.e., without future knowledge or hindsight knowledge) and therefore inadmissible. My speculation is that video itself is a powerful form of hindsight and the mechanics of hindsight are being structurally affected by video. Even the word “hindsight” evokes qualities of visuality, temporality, and memory. Writers from Paul Virilio to Ina Blom have contended that video produces (and does not simply “reflect”) a sense of duration. In art, a similar postulate has been modeled by Graham, in works like Opposing Mirrors and Video Monitors on Time Delay (1974). In popular perception, however, these complex relationships between video technology and experience have been overshadowed by a positivist notion of telepresence as “real time,” and television as a constant “flow.”
The psychological and legal term for the effect of hindsight on our retroactive appraisal of events is “hindsight bias,” first defined by Ruth Beyth and Baruch Fischhoff under the poetic title, “‘I Knew It Would Happen’: Remembered Probabilities of Once-Future Things.”5 As legal scholar Maggie Wittlin has recently reiterated in the pages of the Columbia Law Review, research overwhelmingly demonstrates that “after learning about an outcome, people view that outcome as more predictable than it was.”6 With the introduction of video comes a perceived ability to rewind time. In her essay “Hindsight Evidence,” Wittlin lays out the case for allowing hindsight evidence in court, especially with respect to allegations of excessive force at the hands of police. Currently, in the United States excessive force trials do not permit hindsight evidence:
excessive force is analyzed under the Fourth Amendment’s “objective reasonableness” standard, which looks at whether the [police] officer’s actions were reasonable in light of the circumstances, independent of her subjective intent. “Reasonableness” is “judged from the perspective of a reasonable officer on the scene, rather than with the 20/20 vision of hindsight.” Police may use deadly force if “they have probable cause to believe they must protect themselves or another person present from imminent death or serious physical injury.” Many courts deem evidence that the plaintiff was unarmed irrelevant.7
This difference in policy amounts to a difference in temporality, applied to police versus regular citizens claiming self-defence. Wittlin’s point is similar although neutrally worded:
In criminal homicide prosecutions where the defendant claims he was acting in self-defence because he believed the victim was drawing a weapon, courts almost always allow evidence that the victim was not actually armed. Courts have deemed this evidence relevant and admissible for over a century […]. The problem is that courts fail to apply the same reasoning to excessive force cases. By applying a sound theory of hindsight evidence, courts can avoid this inconsistency.8
In order to come to grips with the behavior of hindsight vis-à-vis temporality, it is necessary to consider the forces of duration and memory, which are related but distinct. While memory refers to the veracity of testimony, duration refers to the temporal conditions in which “hindsight” and decision-making unfold. Artists and media theorists have made valuable contributions to this analysis. Case in point is art historian Blom’s 2015 book The Autobiography of Video: The Life and Times of a Memory Technology. The book is one of the most astute investigations of video aesthetics, as it engages with the medium’s “organic interior.”9 Blom is writing explicitly about analog video and what is recently relevant in video evidence is digital video, however I have applied her theories without further elaboration in relation to analog vs. digital (this would take another essay). For Blom, video is memory and duration.
Blom writes of this video memory not as a form of storage, but as a dynamic and constitutive force, through which the past is perpetually reformed. Following Wendy Chung, Blom traces this miss-association of memory as storage to the “neuronal” metaphors that became attached to early computing. “Algorithms need to read and write values,” she explains, “and this is the simple reason why the digital paradigm reinterprets memory as storage. Hence, the dynamic and generative dimension of memory was suppressed in favor of a mode of thinking preoccupied with safekeeping a stabilized stock of pure information, reinforced by the constant references to more permanent archival media, such as documents, files and folders.”10 It is easy to understand why the latter, more familiar description of memory has been more accepted by courts, as it is exactly these stable printed documents that the court is well versed in archiving and retrieving.
Following her postulation that memory is not a container for information, but a perpetually emergent process, Blom is interested in “feedback” video works, which link video to “constant generation” rather than a stable dead past.11 In these early video works, cameras focused on monitors playing their own live feed produced constantly changing patterns. If—as Blom’s work suggests—video is a perpetual reactivation and recomposition, an important question arises: Should we be resisting the reduction of our concept of hindsight to a notion of memory as a non-fluctuating repository of information?
Because it is tacitly understood that narratives in video evidence are mutable, and because the trial transcript remains a literary form (with no hyperlinks to images or videos), images do not directly enter the record but must be testified to by the makers, witnesses, and experts.12 As Renoe writes of the King trial, the videotape “cannot be simply entered into the record and be left to speak for itself as an ‘objective’ document, even though this was largely the rhetorical strategy the prosecution chose throughout this case.”13 Video produces experts whose expertise is bolstered by their command of the time of the video, performing it by requests for rewinding, stopping, replaying the image, and re-authoring by way of a narrative voice-over.
In recent years, citizens have been pulled into this process, as the duty to gather evidence has been incorporated into daily life. While individuals are spied on by surveillance cameras, they also surveil one another. In September 2015 the American Civil Liberties Union released a smartphone app by which a person can immediately send them footage which they feel is evidence of a transgression of civil liberties—primarily police violence. The “mobile justice” app is available in eighteen states and the “stop and frisk” app is available in New York state to “monitor police activity and hold the NYPD accountable for unlawful stop-and-frisk encounters and other police misconduct.”14 At the same time, many police forces are committing to wearing body cameras, with a study undertaken by the Toronto Police concluding that such cameras “provide the unbiased, independent account of police/community interactions.”15 In turn, a 2016 University of Cambridge study documented a “93 per cent decrease in complaints made against officers” wearing cameras, compared to the previous year.16
Despite these limited successes, video evidence becomes necessary in cases where violence is not prevented. In these cases, though, bystander video is often subject to intense skepticism, arising from the ease of manipulating digital images, and the difficulty of establishing their provenance. Recently, however, these questions of authenticity are compounded by questions of readability. It is especially difficult to adapt our ability to “read” images to the relation between the pictures presented and the subjective perceptions of those involved in a given incident. Video’s ability to depict intent or fear is inarguably lacking. And despite the power of Hollywood cinema to create believable dramas built around clearly defined villains and victims, intention and innocence, the camera does not see “into” an accused person’s intentions, which are so often central to judgment and justice.
Arguments (which may amount to differing ontological and ideological positions) around an image’s perceptual vs. psychological truth-value in relation to media technologies abound. We can be sure that shaky police body cameras will increasingly be aided by image stabilization technology, in order to provide more “reliable” pictures—meaning more reliable in appearance.
Camera operator and media theorist Jean-Pierre Geuens has argued that radically different worldviews are represented by the shaky handheld cameras of 1960s cinema verité and the steady-cam systems developed in the 1990s that dissolve human perception into a smooth Euclidian space, a space of homogeneity reminiscent of video games and characterized by “continuity, infinity, and uniformity.”17 Geuens finds this to be a totalized capitalist space that claims perceptual verisimilitude for itself but which he contrasts to cinema verité: “What was occurring, as the [camera] operator followed an actor, turning swiftly around a corner, and even running up a flight of stairs, was the recreation of some of the sensations experienced by a human being undertaking these actions. […] Although the handheld camera cannot reproduce these physiological reactions per se, it can—through its ups and downs, bumps and goggles—nonetheless visually echo the excitement that accompanies the physical endeavor. This, in turn, was ‘recognized’ by the audience as more real, more alive.”18 Following Geuens’ argument, the question becomes: Are these smoothing effects more faithful to human visual perception or do they sterilize and commodify the image?
Geuens’ findings are echoed by those of Florian Fessel and Neal J. Roese, who studied the tendency of forensic video animations used in court to increase hindsight bias. Their studies show that the slickness and cohesion of digitally recreated traffic incidents produces a false sense of coherence, abnegating the confusion and contingency characteristic of day-to-day perception. In other words, “the more that individuals can make sense of an outcome by integrating outcome knowledge into a coherent explanation, the more certain they become of the outcome’s a priori predictability.”19 It is likely that the ability to control video clips—pausing, fast-forwarding, rewinding—produces a similar authoritative effect in videos to that produced by the false coherency of animated reproductions.
Our ability to comprehend present time has been embroiled with video since the medium’s beginning. Video’s urgency, its production in the moment, and its links to daily life practices and rhythms, intertwine it into the lived moment. Although Blom is discussing early, pre-digital video art when she describes video as memory, there are important resemblances between that art and current video evidence, which is likewise raw and unedited, perpetually streamed and replayed—in other words, always unstable and in flux.
Blom cites Nam June Paik’s suggestion that history should be supplanted by a new inscription of time named “videorie, in which new time technologies would replace historiography.”20 Likewise we might replace the word “hindsight” with “videvidence,” which would denote the influence that video technology has exerted on our ability to remember. While such speculations fixate on the human testimonial, the fact is that digital evidence is “read” by machines long before it is seen and testified to by humans. Artist Trevor Paglen has written that although post-structuralists and artists working and reworking the matter of representation consider that the link “between meaning and representation is elastic [based on the semiotic theories that predominate in art interpretation] this idea of ambiguity simply ceases to exist on the plane of quantified machine-machine seeing.”21
Paglen’s observation does not only add another layer to the problem of interpreting digital imagery as evidence, it also leads us to an ideological function in the technology itself. Machines designed to “read digital images,” he argues, have become “extraordinary intimate instruments of power that operate through an aesthetics and ideology of objectivity, but the categories they employ are designed to reify the forms of power that those systems are set up to serve.”22 For instance, the continuous association of digital images with the power to identify faces and therefore individuals enforces a positivist identitarian (in Theodor W. Adorno’s sense of the word) or we might say “extractionist” understanding of the image. That is, each object/individual is thought to be incontrovertibly and wholly equivalent with its image. The “extraction” of a face from the crowd for purposes of identification is seen not as a contradiction of the structure of the image but as something the image “provides.”
There is a gap between belief in the power of images, and the actual cultural changes wrought by them. But with each newly released video of an abuse of power, the feeling is enforced that the camera proffers truth, while embodied witnesses cannot be trusted. Additionally, perhaps more than the voices of victims of police brutality, these images have managed to stir widespread righteous anger. Still it seems unclear whether these videos are producing actual change in the prosecution of excessive force, or just becoming fodder for the endlessly repetitive information fixes so endemic to our era.
When Laquan McDonald was shot by a Chicago police officer sixteen times, there was potential footage from eight police cars and seven cameras from the local Burger King (whose footage had a remarkable time gap during the event after the police came to collect it as evidence). Of these videos recording the event, just one was released (under duress and legal challenge) to the public by Chicago’s justice system almost a year later.23 After citywide protests, four more videos were released, missing their audio, and provoking yet more questions.24 The reality of cameras is that (visually at least) they always frame out much more than they show. Documentarians count on a shooting ratio of approximately thirty shot minutes to one useable minute. Would thirty cameras present during the 45 seconds that passed between Officer Van Dyke exiting his car and the last bullet fired have given an accurate picture?
Through freeze-frames, replays, and visual and audio enhancements, we will learn about abuses of power and the quick-trigger anxiety of authority. We may find that police are in the same videographic time frame as citizens—that their decisions should not be severed from outcomes. But if we understand video memory only as storage memory, we will ask for the truth of the picture by subjecting it to tests, calling for more cameras, and interrogating it for facts. We may fail to notice that the image guides and even limits the scope of discussion. Regarding this point, legal theorist Cornelia Vismann wrote in 2001 about ways in which the authority of video subtly overtakes the (previously theatrical/textual) script of trial proceedings. Regarding the presence of video in the courtroom, she writes, “[a]bove all, it inserts words into the mouth of the inquisitor who can hence only repeat what is displayed on the monitor—a consequence that seems to radicalize what the metaphor of theatre evoked from the beginning. The emergence of video assumes the functional place of the script for the theatre or place of justice.”25 Each interrogation inscribes the image as truth. Even if subject to gross prejudices and distortions, the image is called upon to function as a readable repository, and the task of justice becomes to interpret what has been seen by a camera. Can video art’s engagement with temporality offer insights into the videographic force behind hindsight’s (potentially) increasing acceptability? If video time is not a measure but a “constitutive force” can we use video as a memory that is not too late?26 Virilio commented that video has provided an unceasing, sleepless, open eye since the first surveillance camera was installed. Now in the form of “personal” devices, video and its digital moving-image correlates form a continuum of sorts. Does hindsight of one event produce the foresight of the next in collective media memory?
In artworks, “the technical conditions under which the past is conserved in the present” can be modeled so as to be considered at length, without the framework of conclusions and consequences essential to the legal system.27 As Blom—parsing Henri Bergson in the context of video—writes, “memory is quite simply a force of retention at work in all perception and sensation, carrying past materials across the temporal divide that installs itself even in the articulation of the syllables of a single word.”28 Video is memory force passing image matter back and forth across the barrier between events and memory, foresight and hindsight.
–Judy Radul, February 2017
Originally published 23 February 2017 in Hearings: The Online Journal of Contour Biennale.