Blurred Visions

In January of this year, the writer Teju Cole published ‘Seven Short Stories About Drones’ via Twitter. Each tweet contained the first line of a famous novel, followed by a sentence or two inspired by reports of drones in the news: ‘Mrs. Dalloway said she would buy the flowers herself. Pity. A signature strike levelled the florist’s.’ Cole’s short missives provided a surgically precise commentary on the US government’s drone programme. On The New Yorker’s blog a month later, he discussed his motivations in more explicit terms: ‘Drones are discriminate, expensive and brutal. And yet they are insufficiently discriminate: the assassination of the Taliban chief Baitullah Mehsud in Pakistan in 2009 succeeded only on the seventeenth attempt. The 16 near-misses of the preceding year killed between 280 and 410 other people. Literature fails us here.’

Does art fail us here, too? We see few statistics and even fewer images of the United States’ use of unmanned aerial vehicles (UAVs). How can artists and others visualize that which we can’t see? And what kind of alternative views can be offered? 

In the US, both the Air Force and the Central Intelligence Agency (CIA) operate drone programmes, with an estimated 7,000 UAVs active as of 2011. While the Air Force has used them primarily to aid ground troops in Iraq and Afghanistan, the CIA’s controversial covert programme deploys drones in Somalia, Pakistan and Yemen to perform targeted killings of suspected members of terrorist organizations. Remotely piloted aircraft range in size from small devices, such as the hand-held Raven, to the Global Hawk (nicknamed the ‘flying albino whale’), which can hover at 18,000 metres for more than 35 hours at a time. The MQ-1 Predator, which is armed with two laser-guided Hellfire missiles, is the cia’s favoured drone and has been the most deadly UAV to date. These aircraft are primarily operated remotely from bases in the US, several thousand miles away from their targets in the battlefield, via a satellite link.¹ The country now has more trained drone pilots than bomber pilots ­– nearly 1,300 as of February 2013 – and by 2015 the military expects to have more unmanned aerial vehicles than manned ones.

Of course, drones are not only weapons. Many UAVs are not armed at all, and can be used to carry out non-military missions. Drone technology provides us with a new way of seeing, and a vast new typology of imagery – one in which the line between voyeurism, surveillance and weaponized vision is dependent on who operates the technology, and for what purposes. The imagery provided by autonomous machines is, as Harun Farocki described it in his 2000 film Auge/Maschine I (Eye/Machine I), ‘operational’. Farocki’s work examined what was then a proto-drone vision: the new camera-equipped warheads deployed during the Gulf War in 1990–91. The footage from these missiles, which was broadcast live on primetime US television, showed the weapons zeroing-in on their targets, then turning to static upon impact. ‘These images lacked plasticity,’ Farocki observed. ‘The human scale was missing.’ Indeed, his words seem to presage the characteristics of today’s drone vision, which also appears to lack ‘plasticity’. What we know of the technology indicates that it is objective and panoptic. As Farocki suggested, ‘These images are devoid of social intent. They are not for edification. Not for reflection.’

What the public knows and sees about today’s newly expanded and expansive drone technology is limited and obfuscated by those who control it. The US drone programme remains veiled in secrecy, government crosstalk and contradictory reports. Though it has been operating since 2002, the CIA did not publicly acknowledge its drone activities until last year. As of March 2013, the government has refused to release the classified memos that contain the legal justifications for their targeted killings. Images of UAVs in the media are typically confined to officially sanctioned press images of drones glinting in the sun on the runway, or flying innocuously over an indeterminate landscape. On his Tumblr, One Visible Future, artist and writer James Bridle recently reported that the most widely circulated image of a drone – a Reaper dropping a missile while in flight – is actually a 3D rendering. Meanwhile, images of drone strikes and their consequences are even scarcer. In ‘The Predator War’, published in The New Yorker in 2009, Jane Mayer noted: ‘No videos of a drone attack in progress have been released and only a few photographs of the immediate aftermath of a Predator strike have been published.' 

Most disturbingly, the public has no access to definitive data about the number of total drone attacks, their locations, justifications or the number of casualties of either militants or civilians. The UK-based Bureau of Investigative Journalism (BIJ), an independent journalistic outlet, estimates that there have been as many as 400 confirmed cia drone strikes in Pakistan, Somalia and Yemen, and as many as 3,500 people killed as of March 2013. It also claims that, in Pakistan alone, up to 500 of these deaths have been credibly reported as civilians. Similar organizations – such as the New America Foundation, The Long War Journal and Pitch Interactive – have all made efforts to visualize the details of the CIA’s covert drone strikes via charts, graphs and interactive maps and graphics. But they all admit that the data is hard to verify because of conflicting reports on the ground and vague information from the US government – a policy of obscurity perhaps best visualized in the US Air Force’s recent removal of previously published statistics of drone strikes in Afghanistan.²

In October 2012, Bridle launched Dronestagram: an ongoing feed on Instagram, Tumblr and Twitter that provides satellite images from Google Earth of the approximate sites of CIA drone strikes shortly after they occur. Along with the BIJ’s database, Bridle’s project is one of the most comprehensive attempts to map these events. Each satellite view on Dronestagram is accompanied by a caption about the circumstances and number of casualties, which Bridle draws from records provided by the BIJ, Wikipedia and other sources. Perhaps the most radical element of the project is the counter-intuitive platform Bridle has chosen for it – Instagram, a medium that is defined visually by its unique ‘filter’, and is more closely associated with nostalgia-tinged images of dogs and beach holidays.

The posts that appear on Dronestagram almost weekly (sometimes more often) range from images of dense villages to deserted mountain roads. For the most part there are no traces of human presence or physical destruction. We get a sense of the abstraction created by the distance and flattening effect of the drone’s-eye view. Though Bridle’s project is a deliberate attempt to render these drone strikes visible, the images and captions can’t help but remind us how little definitive proof and clarity we have about these locations and their circumstances. A typical caption, like the one accompanying an image of barren mountainous terrain on 10 March 2013, reads: ‘2–3 killed by a strike in Northern Waziristan on the Afghan/Pakistan border, riding horses or motorbikes. Identities unknown. Rescue work was reportedly delayed as drones hovered over the area after the strike.’ 

The Center for Land Use Interpretation (CLUI) maintains an online database of ‘Notable Drone-Related Sites in the USA’. CLUI mimics the bureaucratic opacity of the government; its aesthetic and activities often seem to mirror the government office that would provide this information, if such an office existed. The CLUI database uses satellite images from Bing and Google Maps to pinpoint the locations where drones are tested, launched, stored and developed. In the main, these aerial views picture beige desert landscapes, dotted with a few nondescript hangars and intersected by freshly paved runways. Though the views themselves are remarkably sharp, they open up a suggestive, hypothetical space – a way of imagining the consequences of this infrastructure. The relationship of the CLUI database to Dronestagram’s feed can read like one of cause and effect. 

One of the ‘Notable Drone-Related Sites’ is Creech Air Force Base, the primary base from which drones are operated, located in a desert landscape strikingly similar to the landscapes that the drones are surveying more than 11,000 kilometres away. Creech is sited on the Nellis Range Complex, an area about the size of Switzerland, not far from Las Vegas, Nevada, which is reserved for classified military operations. As the satellite image of Creech reveals, beside the hangars and runways, a small city of provisional structures has sprung up. Here, drones are piloted by civilian contractors or Air Force pilots who monitor their movements and surveillance on up to 14 different screens. The pilots and sensor operators are housed inside locked shipping containers with no windows, kept at a cool 15 degrees centigrade, working shifts that can last up to 12 hours.³

Nellis is also home to the Tonopah Test Range, the distant lights of which artist and geographer Trevor Paglen photographed in his series ‘Limit Telephotography’ (2005–ongoing). Paglen photographed the bases from public land at a distance of up to 32 kilometres, using equipment normally employed for astronomy and astrophotography adapted to his camera. Canyon Hangars and Unidentified Vehicle (2006) pictures an indistinct horizon of blurred structures, wavering in the heat that rises from the desert ground. The image replicates the feeling of straining to see, of reaching the limits of human vision. ‘When you push your eye that far, you see the places that you’re trying to see, but you also begin to see what the limits of your own vision are,’ Paglen explained in an interview. ‘Colours start to collapse, light starts to collapse, images really collapse. For me, this collapse of vision is a metaphor for looking at this secret world; the more you look at it, the less distinct it is.’⁴   

We experience something similar when looking at Paglen’s images of drones in the sky above the Nellis Range, Untitled (Predator Drone) (2012) and Untitled (Reaper Drone) (2010). To photograph the UAVs, Paglen wakes up before dawn and drives out along a desert road to shoot their training exercises with a wide-angle lens. Anyone looking at Paglen’s photographs without knowing what to look for would see a picture of a yellow or pinkish sky at sunrise, brushed with wispy cirrus clouds. But then we strain to see the aircraft that Paglen’s titles indicate are there, which only emerge as indistinct black specks swallowed within the colourful abstract swathes of a Romantic skyscape. Though he could have easily rendered the drones more clearly with another lens, he chooses to place them at the limits of our vision. 

These images don’t tell us anything about drones per se, but the gesture of pointing his camera toward the thing that is theoretically watching him is a symbolic one. While the drone hovering above Paglen’s head at a distance of several kilometres could zoom in on the make of his camera, he can see only a vague outline of the vehicle itself. The scale of Paglen’s photographs of drones makes explicit the symbolic distance between automated drone vision and our vision. Paglen’s practice is grounded in his work as a geographer and his fundamental belief that secrecy must have a visible, spatial component on land. As he explained in a 2012 profile in The New Yorker: ‘If you’re going to build a secret airplane, you can’t do it in an invisible factory.’ His extensively researched publications are a counterpoint to the willing abstraction and artistic cues toward the sublime that his images suggest. His writing, in that sense, forms the framework that props up the photos on the wall. But Paglen is equally ambivalent about both approaches. He admits to being convinced of ‘the necessity of an objective observation’, and, at the same time, ‘the impossibility of a truly objective observation’.⁵ ‘I quickly end up in situations where the question is, “How do I point to, engage with, and represent something that I don’t quite understand?” The answer often has to do with trying to represent that epistemological-political gap or in-between space, or that moment of incomprehension.’⁶

Paglen’s blurred and unresolved images could also be read as metaphors for drone vision. If the satellite image used to suggest the existence of WMDs before the 2003 Iraq invasion proved one thing, it’s that there can be no ‘clear picture’. And even drone vision, despite the advances in technology, can be blurred, abstract or unresolved. Pilots and officials must extrapolate their knowledge of 3D forms from distant 2D views. Most UAVs are equipped with a sensor that can see targets in detail from an altitude of up to eight kilometres or more, although the image it returns is often black and white, and always flattened and pixelated, arriving via a two-to-five-second delay. The sensor is equipped with infrared technology that allows its operators to spot things like the tip of a lit cigarette, the warm heads of soldiers camped out in sleeping bags, or newly upturned soil. But drone pilots have also testified that it can be difficult to determine women from men, a camera from a rifle, or children from ‘young men of military age’. With some sensors, once the operator zooms in, the overview of the surrounding area is lost – an effect that pilots liken to ‘looking through a soda straw’. When a target is identified, the operator deploys a laser beam (sometimes referred to as ‘the light of god’), which guides the Hellfire missile to the target. Hence the government often praises this military technology for its ‘laser-like’ or ‘surgical’ precision. 

But increased vision and accuracy does not always equal definitive knowledge. Along with the euphemism of the ‘surgical strike’ is the similar-sounding but completely antithetical ‘signature strike’ – the CIA’s chief means of targeting. Signature strikes are based on previously analyzed ‘suspicious’ patterns of behaviour – for instance, a group of what looks like military-age men convened at night – a strategy that has resulted in the US Air Force and CIA accidentally hitting wedding parties and funerals. As one CIA official confided: ‘Believe me, no tall man with a beard is safe anywhere in Southwest Asia.’⁷   

Like any surveillance technology, the intelligence it provides is only as accurate as the human vision of those interpreting it. A structure that emerges in the shadowy darkness on a monitor still needs to be deciphered and analyzed correctly. Which may be even more difficult when that imagery is a composite of different screens and sources of intelligence. The most advanced sensor – the Autonomous Real-Time Ground Ubiquitous Surveillance Imaging System (argus-is) – is the world’s highest pixelated camera, providing a 1.8-gigapixel video stream. It can simultaneously zoom in on 65 specific sites while maintaining an overview of 40 square metres, from 5,000 metres height, while tracking every moving object in its visual range. But even this technology functions by creating a mosaic of composite images – each of which comes from 368 tiny cameras and computer chips identical to the ones found in smart phones. 

Omer Fast’s 5,000 Feet is the Best (2011) represents the human contingency of drone vision by mixing a factual account with a fictional story, combining the aesthetics of Hollywood film, television news-magazine exposés and ‘war porn’. The film’s script is largely derived from an interview Fast conducted with a former drone sensor operator, who he found through placing an ad on Craigslist (after several unsuccessful attempts to obtain an interview through the official US Air Force channels). The title refers to the ideal surveillance altitude from which, according to the pilot’s account: ‘I can tell you what type of shoes you’re wearing.’

Fast’s film is made up of three repeated sequences with slight variations, in which actors play the roles of the artist and the operator. In the confining corridors and poorly lit room of a Las Vegas hotel where the interview takes place, the pilot-actor is shifty and uncomfortable, plagued by persistent noises in his head, a result of ‘virtual stress’. In between each of these sequences, Fast inserts excerpts from the real drone operator’s interview, with his face blurred to suggest the authenticity of this confidential testimony. As the anonymous source describes the process of operating a drone – which he likens several times to playing a video game – we see gradually widening aerial views of a boy riding his bike in a suburban Las Vegas neighbourhood, along with other views of middle-class American towns.

The final chapter of 5,000 Feet is the Best weaves the actor’s fictional anecdote with the drone operator’s account of a uav strike during the war, accompanied by visuals of a suburban family on a car trip that slowly morphs into a warlike terrain. We begin to see black and white aerial drone views of the family being stopped at a checkpoint, where two men are eventually struck by a Hellfire missile. The chilling effect of Fast’s work is the way he uncannily brings drone vision close to home, in order to visualize events taking place in landscapes we have no direct access to, or that we can’t easily imagine.

Fast’s film is both allegorical and a quasi-journalistic account of what drones and their pilots can and can’t see. We are given access to insider information, but we also question what we learn from the anonymous informant, filtered as it is through Fast’s fractured and authored interpretation. But Fast also deliberately reflects doubt on the truthful possibilities of his own artistic approach. The most crucial exchange between himself and the actor playing the drone pilot is repeated in each of the three chapters: ‘But you’re not a real pilot,’ Fast declares. ‘So what?’ he replies, ‘You’re not a real journalist.’ 

The 2012 book that accompanies 5,000 Feet is the Best includes even more revealing, candid and graphic excerpts of the original interview transcript. The pilot’s testimony seems at times uncensored: he speaks of seeing arms and legs flying off victims, and a head rolling down the street. But he also boldly admits to the limitations of the technology and the potential for human error: ‘Things can go horribly wrong. Like, you know, you gotta let off a missile and you have no clue where it went. That’s one of the instances we all kind of dread, you know. When it lands fifty feet away because you moved all of a sudden.’ 

But the public rarely hears about these incidents, or sees evidence of their aftermath. And even those views would be subjective. As Eyal Weizman pointed out in his 2012 essay ‘Forensic Architecture: The Thick Surface of the Earth’: ‘Attempting to read and interpret the event from the images of the trash and rubble left behind is a highly indeterminate process.’ Earlier this year, CLUI hosted an exhibition at their Los Angeles space entitled ‘Down to Earth: Experimental Aircraft Crash Sites of the Mojave’. The show, in the typical informational style of CLUI, displayed 11 aerial photographs of spots on the earth (or sometimes no traces at all) of crash sites around Edwards Air Force Base, more than 70 years’ worth of launches and tests with experimental aircraft. CLUI culled the photographs from an archive created by ‘X-Hunters Aerospace Archaeology Team’, a group of hobbyists who have been documenting these incidents for 25 years. The images, like those on Dronestagram, tell us little except what the surface of the desert terrain looks like: a smattering of dry lakebeds, shrubs, cactuses, rocks. But they are also, as CLUI’s newsletter puts it, ‘accidental monuments to one of the most advanced forms of technology and human endeavour’.

CLUI’s practice is an important model of individual agency – amateurs and artists, like Paglen, attempting to point to, or trace, the blank spots in our vision. Efforts are being made to add to the growing archive of drone images that might be used for symbolic, rather than ‘operational’, purposes. George Barber’s video The Freestone Drone (2013), for instance, stars an unmanned aerial vehicle, which the London-based artist anthropomorphizes with a high robotic voice and human thoughts and emotions. The Freestone Drone ‘never obeyed orders’, so he decides to fly over New York, sending back useless images of Starbucks cups and old sofas. If we are to imagine, as Farocki foretold in 2001, ‘a war of autonomous machines’, Barber posits the possibility of a self-aware machine, who chooses to use ‘poetry to describe what it [is] seeing’.

The spectrum of attempts to visualize what we don’t know about drones acknowledges the shortcomings of any possible representations of those exclusions. These artists’ views rupture the myth of a foolproof surveillance technology by adopting illegibility and opacity as their own visual strategy. In an age of the highest visual resolution, artists are making images that are perpetually unresolved – pointing toward what we might not see or understand. These artists’ efforts create a symbolic, second-order evidence, which might work around the ‘lack of plasticity’ that Farocki identified. Such renderings remain open to interpretation, because they operate within a symbolic system where obscurity can also be plastic, or revealing. After all, it is a political act to show things that aren’t otherwise visible to the public. These artists ask hypothetical questions about what the impact would be if we could all train our vision on the things that are being hidden from us – the very things that can see us at the same time.