Sophisticated image manipulation tools have done for photographs what the printing press did for the written word. Images no longer have any pretense of objectivity; instead documentary practices collapse once more into the realm of illustration. Now is a time when the image is a literary device, whose verisimilitude hinges on belief rather than reality. Neural networks and render engines may summon any number of photoreal depictions, freed from the time and space constraints of camera apertures. But technology alone is not to blame for the fact that “disinformation”, once a coordinated military tactic, is now a daily aspect of political discourse. Every form of representation has always necessitated paring down the noise of real life, choosing what to include and what to exclude. Such value systems, rather than reality itself, produce the effects of realism.
In the wake of numerous protests against police brutality, increasingly catalyzed by footage depicting abuses of power, police forces have been widely equipped with body cams to ensure greater transparency. These cameras, positioned on the exterior of a bulletproof vest, provide a “cops-eye” perspective to events otherwise captured by bystanders’ smartphones. But unlike civilians, police officers are only accountable for what they perceive in “realtime”—evidence gathered in hindsight is irrelevant when judging whether or not the officer had probable cause to use excessive force. What matters is the perceived rather than actual threat. As with other modes of surveillance, body cam policies are not intended to capture truth. They are meant to gather predictive data.
While law-makers prescribe additional surveillance measures, law enforcement has begun circulating widespread media campaigns warning about the rise of deepfakes: fictitious yet photoreal imagery with the potential to distribute “malicious propaganda”. In order to generate convincingly realistic imagery, a kind of forecasting must occur. The software must extract general principles from large data sets, and apply them to virtual circumstances. So, many images of a given person can be analyzed and extrapolated to produce new images of the same person in made-up circumstances. A similar procedure was at play in the late-19th century, when mugshots were standardized with the paired side-view/front-view photos of criminal offenders. Once standardized, this data set began to yield recurring patterns. But because the data set was already biased in its inclusions and exclusions, the emergent patterns only affirmed a pseudo-scientific basis for racist policing practices.
These images become dis-informative when removed from their original context: no longer seen as a simulated model of a possible world, they now appear as the thing itself. Such simulations—whether legal, economical, or political—come to produce their own realities. What results is an image without a screen, a narrative that needs no context, perception without hindsight. The problem is no longer differentiating real from fake, but actual from virtual. Already it is clear that more transparency does not lead to more accountability. These circumstances require us to think of imagery beyond the question of real or fake, and more along the lines of power and belief. As every aspect of an image becomes available for authorial control, as fluid as any text, then they no longer suffice to be seen, but to be read.
Suspended Animation explores some of the grammars of this “postphotoreal” condition, both novel and ancient in scope. Images are peeled back, showing layers of wireframe scaffolding and shader passes. Disparate camera technologies augment and obscure the empirical gaze. Omniscient surveillance fulfills the role left vacant by a watchful god. But as long as conviction rules representation in both aesthetics and politics, then true or false, right or wrong, it does not matter what an image looks like as much as what the viewer is looking for.
- Maggie Wittlin, “Hindsight Evidence,” Columbia Law Review 116, no. 5 (June 2016): 1327–94
- Federal Bureau of Investigation, “Malicious Actors Almost Certainly Will Leverage Synthetic Content for Cyber and Foreign Influence Operations” (10 March 2021)