Seeing Machines

VIRILIO Vision machineIn a series of posts on photography Trevor Paglen provides some ideas that intersect with my own work on Militarized Vision and ‘seeing like a military’.  First, riffing off Paul Virilio, Trevor develops the idea of photography as a ‘seeing machine‘:

‘Seeing machines is an expansive definition of photography. It is intended to encompass the myriad ways that not only humans use technology to “see” the world, but the ways machines see the world for other machines. Seeing machines includes familiar photographic devices and categories like viewfinder cameras and photosensitive films and papers, but quickly moves far beyond that. It embraces everything from iPhones to airport security backscatter-imaging devices, from electro-optical reconnaissance satellites in low-earth orbit, to QR code readers at supermarket checkouts, from border checkpoint facial-recognition surveillance cameras to privatized networks of Automated License Plate Recognition systems, and from military wide-area-airborne-surveillance systems, to the roving cameras on board legions of Google’s Street View” cars.

What’s more, the idea of seeing machines I’m sketching out here isn’t confined to the imaging devices and systems I’ve described in broad strokes. The definition extends to include the images (or data) produced by such imaging systems, the digital metadata associated with those images, as well as additional systems for storage, archiving, search and interpretation (either human or algorithmic). Finally, and crucially, seeing machines encompasses not only imaging systems, search, and storage capacities, it encompasses something a bit more abstract, namely the “styles” or “practices” of seeing that different imaging systems enable (i.e. the difference between what a view camera and an automated license-plate reading camera “want” to do and how they see the world differently).  Crucially, the definition of photography I’m proposing here encompasses imaging devices (“cameras” broadly understood), the data (“images” being one possible manifestation of that data) they produce, and the seeing-practices with which they are enmeshed.’

In a subsequent post on Geographies of Photography Trevor then links these seeing-practices to what he calls the production of space (and what I now prefer to think of as performances of space), and uses the example of the Reaper to illustrate what he has in mind:

What exactly is a Reaper drone? In essence, it’s a camera attached to a remote-controlled airplane. Sometimes it carries missiles. What’s particular about a Reaper drone (and other drones in its larger family, including the Predator and the Sentinel) is that airplane, pilot, navigator, analysts, and commander don’t have to be in the same place. The aircraft might be flying a combat mission in Yemen by a pilot based in Nevada, overseen by a manager in Virginia, and supported by intelligence officers in Tampa (geographer Derek Gregory has written about what he calls “Drone Geographies.”) The drone creates its own “relative” geographies, folding several noncontiguous spaces around the globe into a single, distributed, “battlefield.” The folding of space-time that the Reaper drone system enables is a contemporary version of what Marx famously called the “annihilation of space with time,” i.e. the ability to capitalize on the speed of new transportation and communications technologies to bring disparate spaces “closer” together, relatively speaking.

I think that’s more or less right: these new, networked political technologies of vision have been instrumental in the production of a non-linear and discontinuous battlespace, threaded by wormholes that connect one site to another.  But, as I’ll try to show when I eventually get to my post on Uruzgan, the process is far from seamless, the folds are more fragile than most of us realise, and the discontinuities and ruptures are as important as the connections for the administration of military violence.

Blurring the lines

In the short-form version of ‘The everywhere war’ (DOWNLOADS tab) I emphasised the blurring of the lines between the increasingly paramilitary but nominally civilian CIA and the US military, and for the last several years I’ve been including this slide in most of my presentations about CIA-directed drone strikes in Pakistan (and I’ve been very careful to use precisely that description: ‘CIA-directed’):


Today’s Guardian (online) carries a report that lends support to these claims and concerns:

17th Reconnaissance SquadronA regular US air force unit based in the Nevada desert is responsible for flying the CIA’s drone strike programme in Pakistan, according to a new documentary to be released on Tuesday.

The film – which has been three years in the making – identifies the unit conducting CIA strikes in Pakistan’s tribal areas as the 17th Reconnaissance Squadron, which operates from a secure compound in a corner of Creech Air Force Base, 45 miles from Las Vegas in the Mojave desert.

Several former drone operators have claimed that the unit’s conventional air force personnel – rather than civilian contractors – have been flying the CIA’s heavily armed Predator missions in Pakistan, a 10-year campaign which according to some estimates has killed more than 2,400 people.

The film is Tonje Schei‘s documentary Drone, which has its premiere tomorrow.  You can read an interview with her about the drone wars here.  In an overlapping interview for Pakistani media, she explains:

DRONE investigates the human consequences of the US drone war. Through unique access to voices on both sides of this new technology, DRONE offers new insights into the nature of drone warfare. DRONE juxtaposes the realities of drone victims in Waziristan to drone pilots who struggle to come to terms with the new warfare. The film covers diverse and integral ground from the recruitment of young pilots at gaming conventions and the re-definition of “going to war”, to the moral stance of engineers behind the technology, the world leaders giving the secret “greenlight” to engage in the biggest targeted killing program in history, and the people willing to stand up against the violations of civil liberties and fight for transparency, accountability and justice.

You can watch a clip on Youtube, which I’ve also embedded here, in which Chris Woods (senior reporter at the Bureau of Investigative Journalism) explains why this blurring of the lines between the CIA and the military matters:

Schei’s original source was Brandon Bryant, a former USAF sensor operator who had already gone public with his own account of the traumatic business of targeted killing (see also here and here).  He decided to add to his testimony when the Obama administration proposed transferring control of the targeted killing program from the CIA to the military, a plan that has faced Congressional opposition:

“There is a lie hidden within that truth. And the lie is that it’s always been the air force that has flown those missions. The CIA might be the customer but the air force has always flown it. A CIA label is just an excuse to not have to give up any information. That is all it has ever been.”

Bryant’s account has apparently been corroborated by another six former crew members, who claimed that the 17th transitioned to its ‘new customer’ in 2004.

Sudden-Justice_webChris Woods provides much more in Sudden Justice: America’s secret drone wars, forthcoming from Hurst at the end of this year, but – for now – here is what I said in ‘The everywhere war’ in 2011 (and I can now say much more in The everywhere war!):

These considerations radically transform the battlespace as the line between the CIA and the military is deliberately blurred. Obama’s recent decision to appoint Panetta as Secretary of Defense and have General David Petraeus take his place as Director of the CIA makes at least that much clear. So too do the braiding lines of responsibility between the CIA and Special Forces in the killing of Osama bin Laden in Abbottabad in May 2011, which for that reason (and others) was undertaken in what Axe (2011) portrays as a ‘legal grey zone’ between two US codes, Title 10 (which includes the Uniformed Code of Military Justice) and Title 50 (which authorises the CIA and its covert operations) (Stone 2003). The role of the CIA in this not-so-secret war in Pakistan thus marks the formation of what Engelhardt and Turse (2010) call ‘a new-style [battlespace] that the American public knows remarkably little about, and that bears little relationship to the Afghan War as we imagine it or as our leaders generally discuss it’.


Drones, militarized vision and civilian casualties

I’m just back from the AAG Conference in Tampa, and there’s a lot to catch up on.

First, an art installation in Pakistan called #NotABlugSplat that reverses the paramilitary gaze and ‘targets Predator drone operators sitting thousands of miles away who refer to kills as BugSplats.’  Now they’ll see on their screens the face of a child who lost her parents and two young siblings in a drone strike.


It’s a collaboration between Pakistani and American artists, working with Reprieve and the Foundation for Fundamental Rights, who also designed it ‘to be captured by satellites in order to make it a permanent part of the landscape on online mapping sites.’


It’s an arresting project – but if you scroll through the comments that followed gizmodo‘s report you’ll see that ‘hope’ and ’empathy’ remain dismally distant for many people.

LEWIS Drone strikes in PakistanPerhaps some of them would benefit from a new report for the Center of Naval Analyses (CNA) by Larry Lewis, Drone strikes in Pakistan: reasons to assess civilian casualties.  I’ve noted his (largely classified) work on Drone strikes and civilian casualties in Afghanistan before, and in this – unclassified – report he leverages what we know about US military drone strikes in Afghanistan to address the cross-border attacks directed by the CIA.  Lewis makes two key points.

First, he notes that the US government’s claims about civilian casualties for its supposedly covert operations in Pakistan are significantly lower than ‘nearly every other estimate available’.  (En route, he draws attention to something that is usually overlooked in these calculations: under International Humanitarian Law, ‘the burden of proof is to determine whether a casualty is a combatant’, and where in doubt the casualty must be regarded and recorded as a civilian).  Based on his previous work in Afghanistan, Lewis suggests three overlapping reasons – apart from a disinclination to tell the truth – that ‘complicate the estimation process’:

  • An irregular enemy –  it is exceptionally difficult to distinguish combatants from civilians in irregular warfare, and this is exacerbated by combatants ‘co-locating with the local population’;
  • Misidentifications – ‘US forces mistakenly believe civilians to be enemy combatants’; I’m not sure how this is different from the first, but Lewis provides two examples that suggest he has in mind specific rather than general characteristics: mistaking men digging drainage ditches for militants burying an IED, for example, or assuming all those in close proximity to an engagement were involved (‘guilt by association’);
  • Battle Damage Assessments (BDA) based on aerial surveillance – determining the consequences of an air strike without ‘boots on the ground’ is likely to be defective

Of all of these, Lewis suggests that it is misidentification that is likely to be ‘the basis for the majority of civilian casualty incidents’ and cites the case that I discussed in detail in Tampa: the strike carried out by two attack helicopters following persistent surveillance from a Predator of a ‘convoy’ of three vehicles in Uruzgan province in Afghanistan in February 2010.  I’ll post my version of events shortly, since I think it is a mistake to collapse this episode into a monotonic ‘Predator vision’; there were other eyes in the sky [see the image below], and – still more significantly – military vision is not a uniquely technical process (which is why the concept of visuality is so important) and in this case involved different interpretations of the Full Motion Video Feed from the drone by different people at different locations. In short, there was a de-centralized, distributed and dispersed geography of militarized vision that was never resolved into a plenary (still less totalizing) frame.

AC 130 Gunship Imagery Afghanistan.001


That said, Lewis’s second point is about process not platform.  He has no truck with claims like Avery Plaw’s – ‘Where civilian casualties cannot be avoided they must be minimized.  This is what drone strikes do’ – because they mistake ‘platform precision for a comprehensive process that minimizes civilian casualties’ and are in fact ‘contradicted by operational data’.  He cites his earlier analysis of 2010-2011 data from Afghanistan, which ‘showed that several forms of attack, including engagements by manned air platforms, were less likely to cause civilian casualties than drone strikes’ (my emphasis; see my earlier discussion here). In his view, then, ‘minimizing civilian casualties is less a matter of platform or ordnance selection as it is using an approach that considers factors that lead to civilian casualties and then effectively takes them into account.’

The point is sharpened by Mark Gubrud in a response to a report from Charli Carpenter at the Duck of Minerva:

‘…drones use the same targeting pods and precision-guided weapons as the manned platforms they replace; in fact, the quality of imagery from the drones is degraded by the limited bandwidth and frequent interruptions of satellite links, as well as the transmission delay which can frustrate last-moment aborts. On top of the “soda straw” vision as compared with low-flying aircraft in close air support, these factors mean that, if anything, the drones are actually inherently less discriminate.’

Again, all of these factors were in play in the Uruzgan attack: degradations and interruptions of both video and audio transmissions were of critical importance – again, see the image from my Tampa presentation above – but even more significant was the way in which the military field of view expanded, contracted and even closed at different locations as the episode unfolded.  More to come.


I’ve changed the header for this site; the original was cropped from a famous Frank Hurley image – perhaps the most famous Frank Hurley image – showing Australian infantry moving up to the front near Hooge in October 1917:


I’ve written about Hurley before, and it turns out that this may be one of his many manipulated images: see Bob Meade‘s detective work here (though see the cautionary notes in the comments too).

While I was doing some image research for my presentation at the Association of American Geographers in Tampa, I discovered the parallel, contemporary  image I’ve substituted.  Here is the uncropped image, taken by Mark Doran, which shows soldiers from the 7th Battalion Royal Australian Regiment at sunset at Tarin Kot (Uruzgan, Afghanistan) in 2013:

Sunset shadows in Uruzgan


Imag(in)ing drones


The USAF has at last published its RPA Vector: Vision and enabling concepts, 2013-2048, outlining its projected future for Remotely Piloted Aircraft: you can download it here.

I’ll be working my way through this in detail in the next several days, but scanning its 100 pages my eyes were drawn to Figure 4 on p. 18.  Noting that the Air Force is transitioning to an all-Reaper flight, the images is captioned ‘MQ-9 firing an AGM-114 Hellfire.’


As James Bridle noted, the image is everywhere; it’s also nowhere: a fake.

At first, the feeling was just unease. Staring at it for some time, seeing it endlessly reproduced across the web and in print, it began to seem unreal, a fiction, too smooth, too perfect. But that’s an effect of drones: they always appear otherworldly…

Of course, it’s not just that. The Canon Drone is indeed entirely unreal. A close inspection, and comparison with other Reaper images, including 09-4066, bears this out almost immediately. The level of detail is too low: missing hatches on the cockpit and tail, the shape of the air intake, the greebling on the fins and body. That ‘NY’ on the tail: it’s not aligned properly, it’s a photoshop. Finally, the Canon Drone’s serial, partly obscured, appears to be 85-566. The first two numbers of USAF serials refer to the year an aircraft entered service: there were no Reapers back in 1985 (development didn’t even begin until 2001).

The Canon Drone does not exist, it never has. It is computer generated rendering of a drone, a fiction. It flies over an abstracted landscape…

In fact, as Alexis Madrigal revealed last year, the image is a computer rendering produced by Michael Hahn:

“I had never seen an image of a drone actually firing a missile so that is what I decided to create,” he said. And suddenly, everyone else, who also had never seen a drone actually firing a missile, had a way of seeing with their own eyes.

Strange then, that the US Air Force – which surely has seen countless drones firing countless missiles – should resort to a computer-generated, photoshopped image.  And an inaccurate one at that.

Endless War and the machine

When I wrote ‘Seeing Red: Baghdad and the event-ful city’ (DOWNLOADS tab) I was intrigued by the way in which the US military apprehended the city as a field of events:

‘In Baghdad, these security practices performed a continuous audit that compiled reports of events (Significant Activity Reports or SIGACTS) and correlated the incidence of ‘enemy-initiated attacks’ and other ‘enemy actions’ with a series of civil, commercial and environmental indicators of the population at large: moments in the production of what Dillon and Lobo-Guerrero call a generalized bio-economy.’  

The animating core of the system was the SIGACT  – shown below – and these were eventually fed into a single reporting and analysis platform, the Combined Information Data Network Exchange (CIDNE).


‘The primary transcription of an event, its constitution as a SIGACT, with all its uncertainties and limitations, was transmitted downstream to be digitized and visualized, correlated and ‘cleansed’, so that it could be aggregated to show trends or mapped to show distributions.  All the systems for SIGACT recording and analysis interfaced with visualization and presentation software, which was used to generate ‘storyboards’ at every level in the chain.’

The chain as it was constituted in Iraq in May 2006 is shown below; CENTRIX (top left) is the Combined Enterprise Regional Information Exchange System that provided information exchange across the US-led coalition;  CPoF (scattered across the centre field) is the Command Post of the Future, a distributed GIS system I discussed in the original essay that provided a command-level visualization of the battlespace as a field of events (a system that has since been upgraded multiple times); and at the centre right you can see the key automated data exchanges to and from CIDNE:

MNF-I Event Reporting, May 2006 (Wortman)

Since I wrote, scholars have used SIGACT reports much more systematically to analyse the connective tissue between ethno-sectarian violence and the ‘surge’ – see, for example, Stephen Biddle, Jeffrey Friedman and Jacob Shapiro, ‘Testing the surge’, International Security 37 (1) (2012) 7-40; Nils Weidmann and Idean Saleyhan, ‘Violence and ethnic segregation: a computational model applied to Baghdad‘, International Studies Quarterly 57 (2013) 52-64 – to explore the political dynamics of civilian casualties – see, for example, Luke Condra and Jacob Shapiro, ‘Who takes the blame?  The strategic effects of collateral damage’, American Journal of Political Science 56 (1) (2012) 167-87 – and to conduct more general evaluations of counterinsurgency in Iraq: see, for example, Eli Berman, Jacob Shapiro and Joseph Felter, ‘Can hearts and minds be bought? The economics of counterinsurgency in Iraq’, Journal of political economy 119 (4) (2011) 766-819.

I’ve been revisiting these modelling exercises for The everywhere war, because they require me to rework my essay on ‘The biopolitics of Baghdad’ (though not, I think, to change its main argument).  I’m struck by the idiom they use – my critique of spatial science written in another age would have been substantially different had it been less preoccupied with the detecting of spatial pattern, had its methods been applied more often to issues that matter, and had its architects been less convinced of the self-sufficiency of their methods.

But I’m also struck by the idiom of the SIGACT itself.  We’ve since become much more accustomed to its staccato rhythm through Wikileaks’ release of the Afghan and Iraq War Diaries, whose key source was CIDNE.  Again, these have been visualised and analysed in all sorts of ways: see, for example, herehere, here and here (and especially Visualizing Data and its links here).  The image below comes from Andrew Zammit-Mangion, Michael Dewar, Visakan Kadirkamanathan and Guido Sanguinetti,’Point-process modelling of the Afghan War Diaries’, Proceedings of the National Academy of Sciences 109 (31) (2012) and shows the time-space incidence of events recorded in the Diaries (here I suspect I’m channelling half-remembered conversations with Andrew Cliff….)


And here, rather more prosaically, is another version – by Drew Conway and Mike Dewar – that provides a time-sequence of the mounting intensity of the war 2004-2009 (for Danger Room‘s discussion and gloss, see here):


Now David Pinder has kindly drawn my attention to an extraordinarily suggestive essay by Graham Harwood,Endless War: on the database structure of armed conflict’ over at rhizome, and to the art-work that is it subject, which together have returned me to my original interest in the ‘event-field’ of later modern war and the automated interactions between its data platforms.

Graham’s central question is deceptively simple: ‘How does the way war is thought relate to how it is fought?’ SIGACTS populate the digital battlespace with events and invite a calculative and algorithmic apprehension of the field of military violence.  To show what this means, Graham and his partner Matsuko Yokokoji (who together compose YoHa: English translation ‘aftermath’) joined with his Goldsmith’s colleague Matthew Fuller to produce an intriguing artwork, Endless War.

It processes the WikiLeaks Afghan War Diary data set as a collection of analytic viewpoints, both machine and human. A software-driven system, Endless War reveals the structure of these viewpoints by using N-gram fingerprints, a method that allows sorting of the text as an anonymous corpus without having to impose predetermined categories on it. Presented as a gallery installation, the system includes a computer that processes the data in real time, projections of the results, and coil pick-up microphones on the central processing unit that sonify the inner working of the machine.

The torrent files released by WikiLeaks in 2010 are the residue of the system that created them, both machine and human. They seem to hint at the existence of a sensorium, an entire sensory and intellectual apparatus of the military body readied for battle, an apparatus through which the Afghan war is both thought and fought.

You can get a sense of the result from vimeo’s record of the installation at the Void Gallery in Derry:

This is a video, obviously, but Endless War isn’t a video.  As the artists explain in a note added to the vimeo clip:

Just as an algorithm is an ‘effective procedure’, a series of logical steps required to complete a task, the Afghan War Diary shows war as it is computed, reduced to an endless permutation of jargon, acronyms, procedure recorded, cross-referenced and seen as a sequence or pattern of events.

Endless War is not a video installation but a month-long real-time processing of this data seen from a series of different analytical points of view. (From the point of view of each individual entry; in terms of phrase matching between entries; and searches for the frequency of terms.) As the war is fought it produces entries in databases that are in turn analysed by software looking for repeated patterns of events, spatial information, kinds of actors, timings and other factors. Endless War shows how the way war is thought relates to the way it is fought. Both are seen as, potentially endless, computational processes. The algorithmic imaginary of contemporary power meshes with the drawn out failure of imperial adventure.

This computational assemblage (think not only of the cascading algorithms but also of the people and the handshakes between the machines: a political technology ‘full of hungry operators’, as Graham has it) is performative: it is at once an inventory – an archive – and a machine for producing a particular version of the military future. Graham again:

‘A SigAct necessarily retains evidential power that reflects its origin outside of the system that will now preserve it, but once isolated from blood and guts, sweat and secretions of the theatre of action (TOA), the SigActs are reassembled through a process of data atomisation. This filter constructs a domain where the formal relation, set theory, and predicate logic has priority over the semantic descriptions of death, missile strikes, or the changing of a tank track and the nuts and bolts needed…

This system of record keeping can be seen as a utopia of war. It is idealized, abstracted, contained; time can be rolled back or forward at a keystroke, vast distances traversed in a query, a Foucauldian placeless place that opens itself up behind the surface of blood-letting and hardware maintenance and the ordering of toilet rolls. A residue that casts a shadow to give NATO visibility to itself. As the ensemble of technical objects and flesh congeal, they create an organ to collectively act to rid itself of some perceived threat—this time from Al Qaeda or the Taliban—faulty vehicles, bad supplies, or invasive politics. This organ also allows NATO’s human souls to imagine themselves grasping the moment, the contingency of now. All of the war, all of the significant events, all of the time, all of the land, coming under the symbolic control of a central administration through the database, affording governance to coerce down the chain of command.’

This is a much more powerful way of capturing – and, through the physical installation itself, conveying – what I originally and imperfectly argued in ‘Seeing Red’:

‘… optical detachment is threatened by a battle space that is visibly and viscerally alive with death; biopolitics bleeds into necropolitics. And yet the Press Briefings that are parasitic upon these visualizations move in a dialectical spiral, and their carefully orchestrated parade of maps, screens and decks reinstates optical detachment. For even as the distancing apparatus of the world-as-exhibition is dissolved and the map becomes the city, so the city becomes the map: and in that moment – in that movement – Baghdad is transformed into an abstract geometry of points and areas and returned to the field of geopolitics. And as those maps are animated, the body politic is scanned, and the tumours visibly shrink, so Baghdad is transformed into a biopolitical field whose ‘death-producing activities [are hidden] under the rhetoric of making live’ (Dauphinee and Masters 2007: xii). In this looking-glass world bodies are counted but they do not count; they become the signs of a pathological condition and the vector of recovery. These processes of abstraction are, of course, profoundly embodied. This is not algorithmic war, and behind every mark on the map/city is a constellation of fear and terror, pain and grief. For that very reason our disclosure of the infrastructure of insight cannot be limited to the nomination of the visible.’


I’m putting together my presentation on ‘Seeing like a military‘ for the AAG Conference in Tampa next week, but – prompted in part by my interest in forensic architecture (see also here and here) – I’ve also been thinking about other ways of seeing (perhaps ‘re-viewing’ would be better) military violence.

2014_cover_publication_forensis_imgsize_SSo I’ve been interested to read a report over at rhizome on Forensis, an exhibition and installation curated by Anselm Franke and Eyal Weizman at the Haus der Kulteren der Welt in Berlin, on ‘Constructions of Truth in a Drone Age’:

Any act of looking or being looked at is mediated by technology. This is true of any scientific process too, where each tool or method of looking is developed with a purpose in mind which influences the data that it produces. This is precisely what forensic investigation reveals: not only the reality of an event, but also the intention of a viewing mechanism and the political weight of that intention once made visible. Representations of warfare illustrate this as successfully as any art object.

As part of the exhibition Forensis, now on view at Haus der Kulteren der Welt in Berlin, Forensic Architecture and SITU Research investigate drone strikes in situations where state-mandated degradation and pixelation of publicly available surveillance footage is a legal regulation rather than a visual constraint, and drones are designed to evade the digital image. Missiles are developed that burrow through targeted buildings, leaving holes that are smaller than a low resolution pixel. Attacking at “the threshold of visibility,” the legal, political, and technical conditions equally attempt to remain invisible. The job of forensics is then to recover them.

Anselm and Eyal traffic in the roots between forensics and the Roman forum, which they envisage as a ‘multi-dimensional space of negotiation and truth-finding in which humans and objects participated together in politics, law and the economy.’  The underlying argument of Forensis, the report suggests, is that ‘the object of forensics should be as much the looker and the act of looking as the looked-upon’ – which will be precisely my point in Tampa.