Complexity, the root of networked warfare

Note: This short essay is part of the Series by emmi and Frank entitled, “Adapting to Transcend Networked Conflict: How complexity is our biggest asset” which can be found in full here.

A central reason that networked technology is causing such a disruption is that it goes against assumptions that generations of people took to be common sense.

The major conflicts that defined the 20th century were exemplified by command and control structures, both in terms of managing actual soldiers on the ground and in terms of the economy that supplied the war effort. Such an approach was also applied to the largely invisible conflict that defined the Cold War – the mechanisms of control and communication that operated the nuclear arsenal of both the US and USSR and things like inter-imperialist counter-intelligence warfare operated along such lines.

However, even at the height of high modernist planning, decentralization still had its virtues. The go to example is the Vietnam war, which saw a small pre-industrial state defeat a global superpower.1 Part of why the Vietnamese won was due to how the US army operated. The US army assumed that it could fight the war through a “rationalized” statistical approach that assumed that the actors involved were operating along the lines of economic rationality. But these assumptions were fatally wrong. As a 2006 paper by Michele Chwastiak put it:

“[T]he U.S. leaders … thought they were fighting a limited war with an economically rational adversary. They … assumed that victory was assured as long as resources were employed in an efficient manner and statistics were improving. They further presumed that the U.S. soldiers could be manipulated through incentives and controls into producing high body counts. The Vietnamese, however, were not fighting a limited war but rather a total war, and the U.S. soldiers finally refused to conform to their instrumental identity as killers. As the war dragged on and a large percentage of U.S. ground troops became unwilling conscripts, the soldiers began to rebel against the identity imposed on them through the discourse of PPB and turned their training on the institution that put them in Vietnam. … [Ironically] the U.S. soldiers would contribute almost as much as the Viet Cong towards the defeat of the U.S. war effort.” 2

There was also the fact that even when the terroristic CIA eventually figured out that the Vietnam War was lost and tried to tell the executive branch, the hierarchies of the military were too ideologically committed to let the message pass. As Tim Wiener described it in Legacy of Ashes:

“The suppression and falsification of reporting on Vietnam had a long history. In the spring of 1963, John McCone had come under enormous pressure from the Pentagon to scuttle a pessimistic estimate that cited “very great weaknesses” in the government of South Vietnam—including poor morale among the troops, terrible intelligence, and communist penetration of the military. The CIA rewrote that estimate to read: “We believe that Communist progress has been blunted and that the situation is improving.” The CIA did not believe that. A few weeks later came the riots in Hue, followed by the burning Buddhists, and the plotting to do away with Diem. 

The pressure never stopped; the president’s new national security adviser, Walt Rostow, constantly ordered the CIA to produce good news about the war for the White House. Whose side are you on, anyway? Rostow growled. But on the same day that Helms squared the circle, he also sent a brutally honest CIA study to the president. ‘The attached paper is sensitive, particularly if its existence were to leak,” Helms’s letter to the president began. “It has not been given, and will not be given, to any other official of the Government.” The very title of the report—”Implications of an Unfavorable Outcome in Vietnam”—was explosive. “The compelling proposition,” it said, was that “the U.S., acting within the constraints imposed by its traditions and public attitudes, cannot crush a revolutionary movement which is sufficiently large, dedicated, competent, and well supported. . . . The structure of U.S. military power is ill-suited to cope with guerrilla warfare waged by a determined, resourceful, and politically astute opponent. This is not a novel discovery.”


Never had there been a war where more intelligence was placed in the hands of commanders: captured enemy documents, brutal interrogations of prisoners of war, electronic intercepts, overhead reconnaissance, field reports brought home to the Saigon station through the blood and mud of the front lines, careful analyses, statistical studies, quarterly syntheses of everything the CIA and American military commanders knew. Today an old torpedo factory not far from the Pentagon houses eight miles of microfilm, a small part of the archive of American intelligence from the war.

Never had so much intelligence meant so little. The conduct of the war had been set by a series of lies that the leaders of the United States told one another and the American people. The White House and the Pentagon kept trying to convince the people that the war was going well.” 3

While there were many factors that went into the loss in Vietnam, this inability to adjust to facts they received played a significant part. To understand why this caused so many problems, let’s turn to the premier 20th century theorist on ambiguous conflict, colonel John Boyd.

Boyd’s signature idea is the Observe, Orientate, Decide, Act (OODA) loop. Originally developed to describe the process of decision making by fighter pilots engaged in combat, it was then generalized and has now become a staple of strategic thinking writ large. 4

The steps in the OODA loop are pretty simple to understand. The Observe step is where an actor takes in information from the outside world. The Orient step is where the actor either constructs a new model of reality or modifies its previous one to better reflect the information it has received. The Decide step is where the actor decides which course of action it should take going forward. The Act step is where it carries out the decision.

The central part of the OODA loop is the Orient step. The process of orientation is what separates the OODA loop from trivial self-correcting feedback systems. Reorientation involves changing your model of reality, the filters by which they receive information, the process by which they make decisions and the approach they take when it comes time to act.

Such an approach is necessary when operating within any complex environment that cannot be fully modeled. You cannot understand a complex environment from first principles. As such you need to adopt a trial-and-error approach wherein you construct and test models of reality to see how they fare.

Things become even more complex when you have adversarial actors facing off against each other. Combatants now not only have to keep up to date with the environment but also keep up with each other. This is amplified in hybrid warfare scenarios (conflicts that blend conventional and irregular warfare) where the line between combatants and civilians is intentionally blurred.

When you have two or more intelligent actors facing off against each other, each actor can influence the decision-making process of the other. When you successfully do this, you “get inside” the OODA loop of your opponents and can disrupt their decision-making process. Those who have a disrupted decision-making process are less effective at acting in the world. Things become especially complicated when agents start to cooperate with each other, which enables things like the building of trust and deception. However no matter how complicated things get the agents involved will still be moving through OODA loops when they make decisions.

What many people take from OODA loops, or updating collective decision making, is that you should move through the OODA loop quicker than your opponents. And while there is some value to be had in speed, what matters is the Orient step. Even if you can quickly Observe, Decide and Act, if you can’t escape from your poor model of the world then your actions will be ineffectual because they don’t accurately reflect reality. 

Diagram of the OODA loop.

This is why Boyd is particularly concerned with making adversaries ‘fold back in themselves’. An actor folds inside itself when it finds the world to be too overwhelming and so retreats to established models instead of trying to keep pace with the world. By closing themselves off from the world, they are now vulnerable to unforeseen shifts in the environment because they can’t register them and thus can’t appropriately respond to them.

As such the model of conflict envisioned by Boyd puts just as much emphasis on psychology, the environment, and the moral fiber of the combatants as it does on the material capacity that they bring to the fight. How you organize is just as important as what you organize.

Further complicating matters is the fact that each person can run through their own OODA loop and share what they uncover with other people. This increases the iteration speed significantly by allowing individual actors to share information between themselves such as through a big protest group message. Hence, successful approaches or useful insights can quickly be spread among the network.

This form of organizing poses significant challenges to hierarchical forms of organization such as governments and corporations because hierarchies justify themselves on the basis that those in charge know best through greater awareness of the whole game. But in an environment in which things are constantly changing and enemies are trying to out-think you, bureaucratic approaches are often less effective because they can’t adapt to change and can be gamed by intelligent and agile adversaries.

When faced with this fundamental uncertainty, what you want are organizations, networks and ultimately individuals that are at “the edge of chaos”. The edge of chaos is a phase transition where systems break out of rigid, predefined behaviors, while also avoiding the sort of dissipative chaos that would dissolve the system. This can be seen in the way that groups of people use local knowledge of a protest scene to route around outsiders who don’t live there (even if that’s the state). As M. Mitchell Waldrop describes it:

“Right in between the two extremes [of order and chaos], at a kind of abstract phase transition called “the edge of chaos”, you also find complexity: a class of behaviours in which the components of the system never quite lock into place, yet never quite dissolve into turbulence, either. These are the systems that are both stable enough to store information, and yet evanescent enough to transmit it. These are the systems that can be organised to perform complex computations, to react to the world, to be spontaneous, adaptive, and alive.” 5

Ultimately what is required to operate at the edge of chaos with any efficacy is the willingness to relinquish control, the ability to build trust and maintain trust with others and engage openly with novelty. Or to put it another way, it rewards autonomy, solidarity, curiosity and cosmopolitanism.

The fact that certain organizational forms require not just theoretical understanding, but also the nurturing and exercising of values and interpersonal trust goes a long way in explaining why the theories were never seriously adopted by the US military. Antione Bousquet’s book The Scientific Way of Warfare is a good reference on this topic and has an entire chapter dedicated to explaining the implications complexity theory has for conflict and then unpacking why the US military failed to seriously integrate such organizational insights. As he writes:

“Rather than constituting a decentralised organisation which can operate on the basis of limited and dispersed information, as in the case of al-Qaeda, the US military is developing armed forces which are dependent on large volumes of accurate information to take their decisions and act in unison. This information is to be acquired, processed and distributed through an overarching “system of systems” that has been an ambition of the military since General Westmoreland. For NCW [Network Centric Warfare] advocates, this is the “entry fee” to the brave new military world they promise and has prompted the Pentagon to earmark $200 billion or more in expenditure for the acquisition of network hardware and software over the next decade. However, reliance on this elaborate infrastructure and the skills and habits it will likely breed may in fact prevent troops from ever operating autonomously where only local or partial awareness is available. This point is all the more crucial when it becomes clear the information infrastructure will be the Achilles heel of any such army and that there exists a number of means to effectively disrupt both the hardware and software of electromagnetic equipment.” 6

Part of the failure comes from an inability. But there’s also the fact that those in charge interpret the ideas of theorists like John Boyd in a self-serving fashion. As Bousquet writes:

“[W]hile advocates of NCW frequently refer to Boyd and the OODA “loop”, their main concern seems to be to defeat the enemy through an acceleration of the decision cycle, as opposed to the ability to act unpredictably and creatively adapt in response to contingencies.

Rather than utilising gains in the speed of information-processing and distribution to increase the time available for orientation and thereby the ability to adapt to changing circumstances and surprise adversaries, NCW persists with a rigid cybernetic understanding of warfare that risks rendering US military operations utterly predictable to a competent opponent.”

Attempts at decentralization and the granting of autonomy to subordinates are challenged by the natural inclination of those in charge to want to micromanage those on the ground. People seek communication mechanisms like water flows into a larger container so they automatically adapt. So the US military and intelligence agencies regularly lose despite vast economies of scale in violence and resources. As Bousquet writes:

“[M]uch will depend on the ability of the US military hierarchy to show appropriate judgment and resist the temptations of centralisation and micro-management when it is counter-productive. The historical record in this respect inspires little confidence.” 7

The failure for the US military to successfully adopt organizational approaches based on complexity theory speaks less to the utility of such theoretical models and more to what the actual imperialist purpose of the US military is. The fact that the military failed to adapt to new organizational possibilities because of the necessity of retaining control. 

It’s a testament to anarchist insights into how power functions that such dynamics of self-sabotage were predicted in fiction decades prior in works like Ursula LeGuin’s The Dispossessed. In one scene she has anarchist Shevek debate a conservative over the purpose of the military and why it fails to adopt more effective organizing principles:

“Atro had once explained to him how this was managed, how the sergeants could give the privates orders, how the lieutenants could give the privates and the sergeants orders, how the captains . . . and so on and so on up to the generals, who could give everyone else orders and need take them from none, except the commander in chief. Shevek had listened with incredulous disgust. “You call that organization?” he had inquired. “You even call it discipline? But it is neither. It is a coercive mechanism of extraordinary inefficiency — a kind of seventh-millennium steam engine! With such a rigid and fragile structure what could be done that was worth doing?” This had given Atro a chance to argue the worth of warfare as the breeder of courage and manliness and weeder-out of the unfit, but the very line of his argument had forced him to concede the effectiveness of guerrillas, organized from below, self-disciplined. “But that only works when the people think they’re fighting for something of their own — you know, their homes, or some notion or other,” the old man had said. Shevek had dropped the argument. He now continued it, in the darkening basement among the stacked crates of unlabeled chemicals. He explained to Atro that he now understood why the Army was organized as it was. It was indeed quite necessary. No rational form of organization would serve the purpose. He simply had not understood that the purpose was to enable men with machine guns to kill unarmed men and women easily and in great quantities when told to do so.” 8

The US military failed to adapt to the new realities of conflict envisioned by figures by Boyd because the sort of decentralization and autonomy that the model demanded would almost certainly result in spillover for the rest of society. Once the most conservative institution in society (the military) moved to a model that privileged autonomy and independence, other institutions like workplaces, community organizations, schools, etc would probably see similar shifts. To use Boyd’s language, the US military (and by extension the rest of society) folded in on itself because the process of orientation would have resulted in significant social upheaval.

Unfortunately, something similar happened with the left. The ‘capitalist realism’ that has predominated much of the radical left since the 80s was the result of similar epistemic closure. So much was invested in organized labor, in social democratic states, and actually-existing-socialist states, that when these vehicles for change fell apart or were defeated, many leftists were left without a fallback strategy. To put it in Boydian terms, the left folded back in on itself and struggled to take in new information. 

We think that it’s this epistemic paralysis that explains the impotence of the left in the last few decades. While the loss the labor movement was a tragedy, we think that those on the left who believe that there can be no progress until we resurrect the labor movement simply don’t understand the world today. We think that the technological shifts we have experienced, the possibilities in the future and the theoretical insights of fields like complex systems theory offer a far more solid basis to build a project of universal human emancipation than centralized industrial production and Marxist dialectics.

To understand why, let’s examine how these technologies and insights could potentially spell the doom for the archetypal industrial era organization – the Westphalian nation state.

  1. Per capita income adjusted for inflation at the time for North Vietnam was $1000 a year, whereas the US was roughly $18,000. Obviously the actual situation was far more complicated due to the flaws of GDP as a metric and the material support provided by China and the USSR to Vietnam. Nevertheless it demonstrates the material difference between the two belligerents. GDP data sourced from the Maddison Project Database, version 2018. Bolt, Jutta, Robert Inklaar, Herman de Jong and Jan Luiten van Zanden (2018), “Rebasing ‘Maddison’: new income comparisons and the shape of long-run economic development”, Maddison Project Working paper 10
  2.  Rationality, performance measures and representations of reality: planning, programming and budgeting and the Vietnam war, Michele Chwastiak, 2006
  3.  Legacy of Ashes, Tim Weiner, pg 268-269h (you can get it as a free audiobook on the Internet Archive).
  4.  See for example Chet Richard’s Certain to Win: The Strategy of John Boyd, Applied to Business
  5. Complexity: The Emerging Science at the Edge of Order and Chaos, M. Mitchell Waldrop, pg 293
  6.  The Scientific Way of Warfare: Order and Chaos on the Battlefields of Modernity, Antoine Bousquet, pg 231
  7.  Ibid pg 232-233
  8. The Dispossessed, Ursula LeGuin, pg 305-306.