zenpundit.com » Blog Archive » Legibility at War

Legibility at War

(by Adam Elkus)

Apropos of a conversation I had with infosec provocateur The Grugq last night and a previous conversation with Nick Prime, a short comment on this piece on US covert aid to Colombia (mostly quotations from others):

Only then would Colombian ground forces arrive to round up prisoners, collecting the dead, as well as cellphones, computers and hard drives. The CIA also spent three years training Colombian close air support teams on using lasers to clandestinely guide pilots and laser-guided smart bombs to their targets. Most every operation relied heavily on NSA signal intercepts, which fed intelligence to troops on the ground or pilots before and during an operation. “Intercepts .?.?. were a game changer,” said Scoggins, of U.S. Southern Command. The round-the-clock nature of the NSA’s work was captured in a secret State Department cable released by WikiLeaks.

In the spring of 2009, the target was drug trafficker Daniel Rendon Herrera, known as Don Mario, then Colombia’s most wanted man and responsible for 3,000 assassinations over an 18-month period. “For seven days, using signal and human intelligence,” NSA assets “worked day and night” to reposition 250 U.S.-trained and equipped airborne commandos near Herrera as he tried to flee, according to an April 2009 cable and a senior government official who confirmed the NSA’s role in the mission.

The piece mainly focuses on the use of intelligence, precision weapons, and targeting to kill off key FARC leaders, even if the quoted paragraph talks about a drug lord. I included it because it was the most succinct summary of the methodologies used. What the piece really shows is the exporting of “industrial” counterterrorism and counter-guerrilla targeting methods pioneered in Iraq by special operations forces. These methods differ from older ones in Vietnam in their speed and technological sophistication, and they differ from the “Killing Pablo” mission the sheer scale of the problem (a resilient insurgent group, not just a drug kingpin). And it’s based in large part on metadata, as Jack McDonald argued:

For me, the importance of Prism, and like efforts, isn’t the question of government invasions of privacy, but rather the ability of the government to use violence against a population. Regardless of the strategic end, analysis of metadata allowed the American government to pull apart Baghdad bomb networks in a way that would have been far more difficult without it, if not impossible. If a couple of thousand special forces soldiers could do that in a foreign country, think what the same capability could do in a domestic context. This capability, I think, is what re-writes the social contract in favour of the government. The reason for this is that it alters the latent balance of violence between the state and the population. I think, however, that this takes place alongside another changing relationship, which is the balance of violence between individuals and the population. In the security/liberty debate we tend to focus on the former, sometimes forgetting the latter. We don’t like big states because they can oppress us, but at the same time, these days, individuals can do that, too.

The NSA caper is one outgrowth of the increasing legibility of social systems (in one respect) that the rise in graph analysis technologies, databases, and improved intelligence collection techniques brings. Here’s a bit on legibility, with Venkatesh Rao riffing off James C. Scott’s Seeing Like A State: 

The book is about the 2-3 century long process by which modern states reorganized the societies they governed, to make them more legible to the apparatus of governance. The state is not actually interested in the rich functional structure and complex behavior of the very organic entities that it governs (and indeed, is part of, rather than “above”). It merely views them as resources that must be organized in order to yield optimal returns according to a centralized, narrow, and strictly utilitarian logic. The attempt to maximize returns need not arise from the grasping greed of a predatory state. In fact, the dynamic is most often driven by a genuine desire to improve the lot of the people, on the part of governments with a popular, left-of-center mandate. Hence the subtitle (don’t jump to the conclusion that this is a simplistic anti-big-government conservative/libertarian view though; this failure mode is ideology-neutral, since it arises from a flawed pattern of reasoning rather than values).

The book begins with an early example, “scientific” forestry (illustrated in the picture above). The early modern state, Germany in this case, was only interested in maximizing tax revenues from forestry. This meant that the acreage, yield and market value of a forest had to be measured, and only these obviously relevant variables were comprehended by the statist mental model. Traditional wild and unruly forests were literally illegible to the state surveyor’s eyes, and this gave birth to “scientific” forestry: the gradual transformation of forests with a rich diversity of species growing wildly and randomly into orderly stands of the highest-yielding varieties. The resulting catastrophes — better recognized these days as the problems of monoculture — were inevitable.

The picture is not an exception, and the word “legibility” is not a metaphor; the actual visual/textual sense of the word (as in “readability”) is what is meant. The book is full of thought-provoking pictures like this: farmland neatly divided up into squares versus farmland that is confusing to the eye, but conforms to the constraints of local topography, soil quality, and hydrological patterns; rational and unlivable grid-cities like Brasilia, versus chaotic and alive cities like Sao Paolo. This might explain, by the way, why I resonated so strongly with the book.  The name “ribbonfarm” is inspired by the history of the geography of Detroit and its roots in “ribbon farms” (see my About page and the historic picture of Detroit ribbon farms below).


22 Responses to “Legibility at War”

  1. Lynn Wheeler Says:

    a side point from recent B&B mini-meeting was that western TV influence on the rest of the world might also because it would make their response to various stimulus more predictable (with predictable on the path to controlled)

  2. J.ScottShipman Says:

    Hi Adam,
    The question of legibility came up in my strategy post a couple weeks ago; you may find of interest: https://zenpundit.com/?p=29998

  3. seydlitz89 Says:

    Interesting post.  Thanks.  A couple of comments:
    First, the use of metadata doesn’t provide context, from what I understand, just identifies a shared means.  Say I’m an Iraqi and I phone my vegetable seller who’s involved in the resistance, unknown to me.  Say I phone him three/four times a week.  Say I’m only wondering about what fresh vegetables he’s going to be selling the next day . . . I get waxed along with the vegetable seller and a bunch of other people . . . just another “operational success” in the strategic disaster known as the invasion of Iraq.  My death then sold as part of the larger “surging to victory meme”, which of course few in the US is going to question.  Of course Gian Gentile’s pretty much trashed that meme, but I think you get my point, especially since there are plenty of other “targets” out there . . .  So the basic problem is that metadata really doesn’t discriminate in terms of social action: the various types of value rational and the various types of instrumental rational social action, rather only the means in which one communicates and the frequency . . . one would have to actually listen to every conversation to gleen what was actually going on, but where’s the whiz-bang in that?  Especially when there’s next to no penalty for mistakes . . . but then of course I come from a Humint background . . .
    Second, “legibility” is a very interesting concept.  How much of our recent strategic failure is due to “illegibility”  . . . Notice that is not a question.  This concept also works within the same society through time.  What is comprehensible to us today about our own society, is not necessarily what was understood by our society in the past.  I’m reading John Jay Chapman’s William Lloyd Garrison and his take on the Civil War is so different from that of today.  He mentions a book written by John C. Wise titled The End of an Era and assumes that anyone wishing to understand the South prior to and during the Civil War would have to read this book, but how many have even heard of Wise, let alone Chapman or even Garrison today?  Do we not in effect require people of our own time to (re)interpret the past, that is provide “legibility” even of our own collective past?    

  4. Aelkus Says:

    I think you undersell the use of network analysis. There are some very sophisticated methods out there and they were employed during the Surge. Read Lamb’s work on HVTs and interagency team for more. That said, SOF were only one part of the overall military effort. And the US military effort was likely secondary to thet temporary pause of the Iraqi civil war produced by the shift in the correlation of forces to the Shia side. Of course, the limitation of network analysis is simply that quality of data matters. Notice that most network analysis academically is done with social media data, which is abundant and of a relative high quality. When one is operating in a contested environment, the ability to collect data already presumes some measure of control that is bought and paid for with blood. 
    The subject of legibility is interesting. The type of legibility that Scott talks about is not really understanding of a complex system’s composition, range of behaviors, and dynamics. Rather, it really just refers to how much of the social system can be processed by a bureaucracy. That relates to the network analysis discussion. You can pull metadata and kill off a bunch of leaders of an insurgent group. But you may never really understand how the actual social system that produces them functions. That’s just a recipe for “mowing the grass.”
    In Los Angeles, my hometown, a great example of this was the disruption of the Black Power movements in the 1960s by the police and FBI. By fragmenting a group that provided a sense of identity and order, it generated a series of autonomous trust groups within the greater LA area. These groups eventually devolved into the powerful gangs that we know today — a process described in the documentary “Bastards of the Party.” 
    Hence legibility is not necessarily understanding, but rather just machine-readability.  

  5. Aelkus Says:

    Also on surge…..I am aware of your sympathies for Gentile. But at the same time the Trinity teaches us that explaining war does not involve fairly simple answers. In particular, one massive problem with our understanding of the Iraq War is the (1) lack of historical or social-scientific investigation of the Iraqis themselves (2) lack of access to the Iranian archives and documents. The latter in particular could very well completely change our understanding of the war. In the interim Douglas Ollivant’s explanation is the most convincing to me, since it puts the sectarian civil war at the forefront but allows a causal role for US military power. http://www.newamerica.net/sites/newamerica.net/files/policydocs/Ollivant_Reinterpreting_Counterinsurgency.pdf

  6. Aelkus Says:

    I doubt that watching Western TV makes people more predictable. The only thing it often tells us: where strange stereotypes about Americans come from. I don’t blame them. I’ve never visited Hong Kong and if I took HK cinema literally I’d think that the island has a worse crime rate than Mexico. 
    Thanks for the link to the thread. I read your entry but not the comments. Fascinating stuff.

  7. seydlitz89 Says:


    “social media data”

    This seems to be mostly a domestic political question at present.  I guess we’ll have to see how it comes out in the courts  . . . 

    From a Clausewitzian perspective it seems firstly as a means of targeting some potential political movement among the citizen population.  In whose interest exactly is developing this capability?  The people or the state?  The citizenry of course have an interest in security, but theirs is separate from that of the state, the age-old problem. Also, I would say that such a system would work best among a population like that of the US, far more than in the ME or elsewhere.  A place when where real social action is at a minimum, where people instead are atomized, separated, reduced to focusing solely on themselves.  Communicating individually electronically.  Social action?
    Tactical and strategic.  Hardly anybody thinks strategically any more. It is nice talking to people who do. 
    In terms of US social norms, it’s taken a long time to get here, where we are now.  Perhaps our capabilities finally fulfill our “needs”, not like we’d ever be sure.   

  8. Lynn Wheeler Says:

    much of world complaints about corrupting influence of western TV have been about the commercials … starting more than 20yrs ago could find large percentage of people on the HK streets talking on their cellphones (US taking while to catch up).

  9. T. Greer Says:

    So i think there is one aspect of Scott’s concept of legibility that is missing from this discussion. Scott does not see legibility as a tool of understanding so much as a tool of control. In order to control the system they are supposed to govern the guys up top have to simplify the system until the thing measured comes into accord with the (much less complex) measurements. 


    The limits of legibility are seen very clearly in this example – as Seylidtz points out, collecting and organizing metadata ensures that the social context of the relationships recorded will be lost. While I do think he underestimates the power of metadata, this part of his critique is spot on. NSA metadata reduces individuals to nodes in a network. These nodes, in turn, are seen in a very simplified fashion – in the case of the Columbians, to be bombed or not. 


    What interests me is what happens next. How does this attempt to make terrorists organizations, insurgencies, or entire societies legible change them? This is Scott’s broader point: you cannot impose measurements and categories on systems in order to exercise control without forcing them to conform to your system. Systems do not just become legible – they are forced to take a more legible shape. Measurement itself is a form of control.


    And if war is an attempt to impose control… well then, all conflict must be one form of “legibility at war.”


    It is a very interesting way to think of things, isn’t?


    P.S. Also, as Seylidtz suggests, the basic job of a historian is to create a narrative capable of explaining the past – in essence, making it more legible. These narratives have all the faults that befall those who try to impose their narrative on the present. Think about it:  everytime historians talk about how the “aristocracy/plebeians/scholars/soldiers/communists/[enter preferred social group here] did this” or “felt that” the historian has indulged in a dangerous simplification that cannot truly capture what the hundred, thousands, or millions of individuals thus described individually did or felt. But historians go ahead and do it anyways because if they provided the necessary details to actually describe what happened they would “bog down the narrative” and not be left with a coherent story to tell. They must distort the past to make sense of it. The only difference between the historian and the policy maker is that the policy maker has the power to distort things now, while the historian only distorts memories of what has already passed. 

  10. Lynn Wheeler Says:

    as an aside, 20yrs ago we were asked to do a system that would take the metadata from all credit card transactions in the US for running 18month window and do profiles for target marketing (there were something like dozen or more privacy organizations doing quarterly audits that nothing was leaking). things have since gotten significantly more sophisticated (with little or no privacy oversight)

    a DBMS conference at the time, one of the men-in-black made some comment about having to stop using credit cards because we would know more about them than they knew about us. At the time, a modest sized system could handle couple hundred thousand transactions/minute … now (non-cluster) modest system for tpmC is 8.5M (factor of 40-50 times, and collections of systems handle significantly larger).
    disclaimer, in past life Jim was co-worker:

  11. Aelkus Says:

    T. Greer, 
    Actually, Rao is very clear that legibility is not about understanding as much as machine readability — but I think that gets lost here in this discussion.  

  12. Aelkus Says:

    Also, in terms of historical narrative the interesting thing that I see is more or less this: historians make implicit choices about several aspects of describing a system: 

    Level of causal explanation and privileging of explanation. The distinction between psychological, group, polity, multi-polity, global, etc. 
    Which events are judged as essential to describe and which are judged as random/irrelevant. 
    Whether or not a property of interest is scale-invariant or dependent on scale.  
    Causal mechanism (e.g. moving backwards in time and space)

    In terms of imposing legibility, to some extent that’s a universal human property. From heuristic shortcuts in problem solving to the way we can solve the frame problem to the millions of assumptions we make everyday about the world without any real sound justification. 

  13. T. Greer Says:

    “In terms of imposing legibility, to some extent that’s a universal human property. ”


    Correctly said, I think. I have thought about this a lot over the last few months and increasingly I see this as a central challenge to all human endeavor and a driving force of most human action.


    I cannot fully claim this insight as my own. Sometime back a few years LynnR, then blogging under name Jospeh Fouche, wrote a blog post that was both incredibly effective and remarkably concise. The centerpiece of this post was an image of two circles – one very large, labeled “the universe.” The other a mere dot, labeled “your brain.” The lesson was that no matter what you did, you could never fit the universe in your brain. Any method or narrative you use to make sense of the world will miss the whole picture. 


    Legibility is simply this concept applied to political control.

  14. Lynn C. Rees Says:

    The root word for “legible” in Late Latin means “something readable”. That roots earlier Latin root not only means “to read” but also gave rise to the Latin word for law now reflected in the English “legal” and “legislate”. This is not coincidental: its ultimate PIE root means “to gather”.


    One meaning of the PIE root for the English word “read” is “to count”. Not surprising: reading and writing were created for accounting, a primordial pillar for any architecture of control. The first written records were tally marks for recording tribute. Writing and reading were important milestones in the progress of human domestication. Ultimately, you ended up not only with people who were increasingly readable but people who were increasingly self-documenting. The metadata collection that works the best is the one conditioned into your mind. To think is to squeeze. To squeeze is to grip. Every other human consideration flows from those two.


    Ultimately, the word “legible” is a foreign obstruction that obscures the English speaker from the real feature: readability as a tool for keeping the brother down. That’s what Billy the Bastard wanted: those whom the gods would destroy they first make speak Latin.

  15. zen Says:

    Very interesting post Adam – several things caught my eye:
    What the piece really shows is the exporting of “industrial” counterterrorism and counter-guerrilla targeting methods pioneered in Iraq by special operations forces. These methods differ from older ones in Vietnam in their speed and technological sophistication,
    McNamara had hoped to seed the Ho Chi Minh trail with electronic sensors in operation Muscle Shoals. He was ahead of his time but the idea of persistent surveillance was there, properly as a tool of war against an armed enemy ( vice the preference of our current rulers to have the NSA focus on monitoring the American people – I don’t have a problem with NSA surveillance of foreign states, that’s why you have intelligence agencies)
    The domestic equivalent to this idea was the “Huston Plan” proposal that came out of the Nixon White House a few years later and was named after a junior Nixon staffer who had the misfortune of presenting this power-grab wish list to an interagency commitee chaired by the venerable and ominous J. Edgar Hoover. The FBI Director for Life was not about to allow the White House or another agency in the IC to end run the FBI’s near absolute monopoly on domestic surveillance legal and illegal and he ruthlessly shredded each point of the Huston plan while mispronouncing the aides’ name in a new way each time. 
    Nixon was forced to retreat after that debacle, given Hoover’s objections ( and more importantly, Hoover’s private archives which vanished from his house while the Director was on his deathbed) but everything in the Huston Plan and more has been implemented since 9/11, first in the Patriot Act and then by NSA lawyers and FISA judges turning the Patriot Act language into an unrecognizable pretzel.
    There will be some walking-back judicially and legislatively of day to day practice of the surveillance state and some marginally greater degree of oversight but compelling the state to voluntarily render US citizenry less “legible” will require several more scandals timed near an election. Probably one reason for the desperation to treat with Snowden and arrange some kind of “deal” though there are perfectly legitimate and vitally important reasons to debrief him. The mere fact that the NSA could not be trusted to use operational and personnel procedures to keep their own most vital secrets safe is a good reason that they should not be trusted with ours ( that and creating that kind of database presents a real risk to national security when hacked or penetrated by spies and walk-in defectors)

  16. Grurray Says:

    Regarding Igloo White and McNamara – they may have been ahead of their time or on the right track. You’re a better judge than me. However, something memorable I read awhile back, a bit snarky, but still relevant to the debate 
    From Col Harry Summers in On Strategy: A Critical Analysis of the Vietnam War:
    When the Nixon Administration took over in 1969 all the data on North Vietnam and the United States was fed into a Pentagon computer -populations, gross national product, manufacturing capability, number of tanks, ships, and aircraft, size of the armed forces, and the like. The computer was then asked,
    “When will we win?”
    It took only moments to give the answer:
    “You won in 1964.” 

  17. Lynn Wheeler Says:

    Boyd would say he claimed that McNamara line would never work … possibly why he was sent over for tour in command of spook base (Coram biography also lists spook base as $2.5B windfall for IBM). lots of detail here … gone 404 but lives on at the wayback machine

    folklore includes large number of air attacks on elephants … similar signature to large troop movements

  18. Lynn Wheeler Says:

    as an aside … note author’s editorial at bottom of piece (“Public Benefit vs. Private Image” … after “Other High Technology Assets”)

  19. prbeckman Says:

    On ‘legibility’ and, as T. Greer pointed out, how a government’s need for legibility impacts society: 
    I’m currently reading “Uncle Sam Wants You: WW1 and the Making of the Modern American Citizen” and the first chapter tells the story of the Selective Service Act of 1917 and its consequences. At that time citizen’s were not very ‘legible’ to the federal government and this was a great challenge for implementing mass draft registration. Other than the Census Bureau, the Post Office and telephone directories the federal government had no idea who its citizens were and who was actually eligible for selective service, so “voluntary registration was necessary because the state did not have the ability to locate or identify all those who lived under its authority.” Documentation like driver’s licenses, birth certificates, voter registration were not ubiquitous, “In 1917, however, the average American man lacked most of these documents; some carried none at all.”  

    So legibility was very low and it was the need to achieve legibility that would leave a significant impact on society. 

    T. Greer: “you cannot impose measurements and categories on systems in order to exercise control without forcing them to conform to your system. Systems do not just become legible – they are forced to take a more legible shape.”

    Exactly right. The federal government wanted to conscript millions of eligible men, but it had no information about those men and it lacked the institutions and money to gather that information so it depended upon private, voluntary organizations to fill the gaps. This is where the culture of voluntary associations reveals its dark side. The army’s estimate suggested that perhaps 3 million men never registered at all. This illegibility was a great dilemma and that’s where voluntary associations came in. Americans of this era are famous for their prolific creations of associations of every kind. You would think that would be a good thing except that they too often veered into vigilantism. These organizations were populated by people who weren’t themselves eligible for the draft, but they saw it as their duty to ensure that those who were eligible weren’t shirking. Organizations were formed all over the country, the most prominent being the American Protective League which counted 250,000 members. In 1917 and 1918 the APL and these other organizations, in collaboration with federal, state & local gov’ts, ran “slacker raids” to try to find those men who were eligible but who hadn’t registered. The accounts of these raids are frightening. The raids varied in size but they culminated in a massive operation in New York City on September 3-5, 1918:

    “The APL later estimated that somewhere between twenty thousand and thirty thousand men participated: city police, government agents from the Department of Justice, more than two thousand soldiers and one thousand sailors, and thousands of American Protective League operatives. For three days they scoured the city’s streets and public places interrogating somewhere between 300,000 and 500,000 men. A man who lacked a draft registration or classification card found himself escorted by these self-appointed authorities to the nearest police station.”

    They surrounded the “exits and entrances of every train, ferry, subway” station, “cordoning off whole blocks and interrogation men on the street. Later they raided theaters, saloons, billiard parlors, and boarding houses. Sailors wandered through the city’s restaurants moving from table to table inspecting the cards of diners.”

    All the consequence of trying to achieve ‘legibility’. And it would have an impact on concepts of citizenship, changing how citizens interacted with their government. The WW1 period was the transition era from the “illegible,” “wild and unruly forest”-era of citizenship to one that has taken on “a more legible shape.” Our surveillance technology and methodology have gotten out ahead of our thinking on the endstate of our 21st century liberal democracy. We’re due for a period of reflection.

  20. AdamElkus Says:

    I often find myself thinking about this when I hear calls for a peacetime “societal” draft. See this great criticism, for example: http://20committee.com/2013/12/25/on-conscription-and-military-effectiveness/. Intellectuals love to talk about how society must be disciplined, organized, etc but few really are willing to reckon with the coercion that this actually requires. Or the authoritarian overtones of the idea that society must be regimented and cohesive for us to have sound politics.

  21. prbeckman Says:

    Adam, that’s right, they are only imagining positive outcomes and not considering potential negative outcomes. And it’s not only intellectuals, some of the more recent calls for a draft have been coming from people like McChrystal, Barno, Eikenberry. I find it a little disturbing when current or former military personnel point their fingers at civilians and say “There’s something wrong with YOU” and then propose militarizing citizenship as a solution. 

  22. AdamElkus Says:

    I think the problem with Igloo White (beyond the goofy name, even for military standards) was the lack of reflection about the problem it was attempting to solve. Also note that in the Algerian War, the fence/mine/patrol system the French developed was vastly more effective and much less high-tech — but the war was still lost anyway….

Switch to our mobile site