
In the opening days of America’s ongoing air campaign in Iraq, many critics and pundits decried the admitted absence of a strategy guiding the Obama administration’s response to the emerging threat posed by the Islamic State of Iraq and the Levant (ISIL). However, in the last two weeks there has been a rapid evolution towards the development of both a policy, and an ensuing strategy, to apply American (and allied) military power to the aim of degrading and destroying ISIL. President Obama has now announced a coalition strategy, premised largely on the support of NATO allies and key partners in the region such Saudi Arabia, the UAE, and Qatar – who have sent warplanes to bomb ISIL in Syria.
Before we can talk, however, about the strength, weakness, existence, or nonexistence of American strategy we must first correctly describe its character. Strategy is defined and disciplined by the various possibilities, which the political process sets through policy. Specifically, we argue that limitations on war make American strategy one of “control,” and analyze both positive and negative implications of trying to achieve control over ISIL.
The Finite Spectrum of Policy
As this spectrum of acceptable policy options is finite, it follows that the range of possible strategies permitted by that spectrum are similarly limited. Hence we can learn much about American strategy simply through a process of elimination. We cannot take bellicose language about the “degradation and destruction” of ISIL at face value. To understand why, one need only look at all of the options the Obama administration has excluded.
First, it seems obvious that another large-scale counterinsurgency campaign is, at the moment, out of the question. While the military wants to preserve the option of medium to large-scale ground operations, Obama does not want to “eat soup with a knife.” Even as policymakers use extremely bellicose rhetoric when talking about ISIL, American actions do not suggest President Obama truly believes ISIL is the worst of the worst.
If the Obama administration truly thought ISIL was the worst of the worst, they would be open to options that might more directly benefit the likes of Bashar al Assad or an expansionist Iran. Yet they would prefer that striking ISIL and other terror groups not benefit Assad. The administration would also think nothing of accepting risk in other theaters of engagement – Europe and Asia – to wipe out ISIL, down to the last black flag-waving man. But are they actually taking risks to properly resource a fight to “destroy” ISIL?
Presuming the administration wanted to completely offshore the effort, they might also arm the Shiites and look the other way as they use U.S. support as an excuse to finally achieve sectarian victory. The U.S. has, after all, made far more morally depraved bargains to destroy our enemies. The need to defeat the Axis was sufficient cause to temporarily transform mass murdering tyrant Joseph Stalin, into kindly ally “Uncle Joe.” But policymakers and the public do not live in the desperate world of 1941, and thus are unlikely to tolerate a U.S. policy that enables such sectarian horrors.
Although experience has taught defense observers to be cynical about so-called “red lines,” all of the possibilities we have enumerated represent political, strategic, moral, and legal lines that American policymakers are unwilling to cross.
Control: A Heuristic for Limited Wars
So far we have defined American policy and strategy by what we believe it is not. We caution that the understanding of strategy currently articulated by national security elites is capable of evolution and change. “Limited” wars in Korea and Vietnam, after all, escalated tremendously in practice. But let’s accept, for the sake of argument, that the assumptions framing current policy and strategy have at least some staying power.
So what is the United States actually prepared to do? One cannot deduce this purely through process of elimination. In addition to laying out what we think the United States won’t do, we have to articulate what it likely will do, and why it will do it.
First, we will observe that in the absence of clear articulation, the policy and strategy must be assumed to be a messy agglomerate of the goals and actions that – between the lines – the U.S. is actually considering. One might argue that the balancing of ends, ways, and means is the key to operationalizing strategy and providing organization for the application of force. But ends cannot be computed simply by solving for ways and means. Ends are precisely the product of that messy agglomerate that current Obama administration policy statements and elite discourse lay out.
Since the start of the Syrian civil war, American policy has focused on the containment and, to an extent, management of potential escalation, keeping the conflict at an “acceptable” level of violence, and more importantly preventing it from spilling beyond Syrian borders. In Iraq, the United States has sought, at minimal commitment, to shore up the Iraqi government against sub-state adversaries. The rise of ISIL, which operates in both theaters of engagement, increasingly focuses U.S. policy on its containment within its original areas of operation.
ISIL’s military success and shocking brutality have altered the equation, but have not transformed it completely. Though the objective of preventing ISIL from spreading beyond its original areas of operation is now overcome by events, it seems that both the administration and many of its critics still share a similar conception of the ultimate strategic ends. Despite Kerry and others’ talk of wiping out “wickedness” and “evil,” even hawkish policymakers that deride so-called “pinprick president” responses envisioned nothing more expansive than a Kosovo-style bombing campaign coupled with the use of special operations forces.
The development and actions of ISIL have threatened American interests and impugned American honor; therefore, it must be confronted. Policymakers should take both threats seriously, and acknowledge their interplay. On a purely realpolitik level, ISIL destabilizes regional governments and raises the threat of another safe haven from which attacks can be planned and recruits radicalized. However, Americans also have an emotional drive to see ISIL punished for its brutalization of American citizens and flouting of American power and prestige. Both of these factors have produced a tentative will to use violence against ISIL.
However, this will is not without conditions. At the very least, the U.S. is certainly willing to try to contain ISIL from further advance; at most, it will seek to cut it down to size. It is difficult to discern any American, or allied, desire and will sufficient to achieve the complete destruction of ISIL.
While American politicians speak in terms of “destruction” (at a high end) and “degradation” (at a low end), we feel these terms are imprecise and ill-chosen. Destruction implies a goal American policymakers are apparently unwilling to actually pursue. So let us cease pretending otherwise. Degradation is a variable, not a goal. One aims for both quantitative and qualitative degrees of degradation depending on the actual goal at hand.
The term that policymakers seem to be reaching for is, in fact, “control,” whether they realize it or not. A certain level of control violently achieved will satisfy both American security concerns and restore American dignity and self-image. And a certain level of control denied will prevent ISIL from expanding any further, or, after the process of disciplining is achieved, prevent it from growing back again.
Strategic Assessment
How can policymakers understand control?
In the words of strategic theorist J.C. Wylie, the aim of war is either the assertion of some measure of control over the enemy, or the denial of the adversary’s desired measure of control. Accordingly, the range of options for American strategy and policy that current politics allow could be placed within two broadly defined categories: those which seek to assert some degree of control and those which seek to deny the adversary’s desired degree of control. In this case, denying ISIL’s desired degree of control would entail capping its size and power at some qualitatively defined scale. Attaining our own degree of control involves violently cutting them down to size.
Perhaps the first thing to understand is that control is not a discrete process. Thermostats are the simplest examples of classical control systems. Once a thermostat is set to a desired temperature value, it monitors observed temperature values for discrepancies with the desired value. The goal of the control system is to minimize the error between the desired temperature value and the observed temperature value. Control systems run continuously, using feedback from the environment as input and altering the environment as output.
Cynically speaking, policymakers mostly ignored ISIL until its power passed a threshold value. They would like to reset ISIL’s strategic power down to a desired value that would reduce the perception of an imminent threat to the West. But control, as Wylie perceived it, differs in some critical ways from simple control theory models like the thermostat. To control ISIL, the U.S. would need to both reset ISIL’s power down to the desired value and prevent them from passing the threshold again. While correction of room temperature by the thermostat is automatic and effortless (and thus can be maintained as long as the thermostat has access to electrical power), each “correction” of ISIL’s power value is both costly and durative.
Still, one thing that both Wylie’s conceptual model and the thermostat have in common is that control does not sustain itself. Rather, it is continuously maintained over time. We have no special insights regarding ISIL’s resilience and capacity for self-regeneration, but in a worst-case scenario it might take a substantial amount of time to finally decay to a point in which it is endogenously incapable of passing the American threshold of threat perception. In the meantime, the U.S. will need to maintain constant vigilance to keep it in the desired “degraded” state that our military efforts ideally would produce.
Another complication is that ISIL itself and the Iraqi and Syrian polities are dynamic entities. Fluctuations in their power and capacities are unlikely to look linear if plotted on a graph. For example, ISIL’s military rise was – up until fairly recently – slow and belabored. Then, suddenly, the collapse of the Iraqi army enabled a meteoric rise in ISIL military power and territorial sway. Assuming that the strategy to confront ISIL is successful (and there are any number of possible reasons why it might not be), a core problem for security planners will be maintaining the successful outcome.
A potential virtue of the control-based conception of policy and strategy, however, is that it does not deny the uncomfortable fact that speaking of “exit strategies” or asking, “tell me how this ends?” is naïve and misleading. The long-run consequences of the Iraq War and the Syrian Civil War suggest continued regional instability that ultimately benefits ISIL. This in and of itself does not necessarily predict U.S. intervention, as policymakers may opt to accept risk in order to extricate the U.S. from expansive commitments. However, policymakers’ fear of the political consequences of terrorist attacks suggests that both brutal provocations (the beheadings of Western journalists and aid workers) and militant territorial control constitute a bridge too far. When both occur simultaneously, a threshold is crossed and some form of intervention is inevitable.
If we can assume these political dynamics will remain stable over time, then consciously embracing the control concept can help discipline and structure U.S. policy and strategy towards ISIL.
Command and (Out of) Control?
We do not pretend to possess any kind of superior insight that would help a policymaker practically untangle the twisted knot that both ISIL’s advances and the implications of our analysis suggest. However, if defense decisionmakers’ thinking indeed fits the conceptual model we have outlined, they ought to use it to better structure the strategies they seek to carry out, taking both advantages and risks into consideration. The first step to overcoming ISIL is not deceiving ourselves about the nature and character of the war that we want to fight.
Adam Elkus is a PhD student in Computational Social Science at George Mason University and a columnist at War on the Rocks. He has published articles on defense, international security, and technology at CTOVision, The Atlantic, the West Point Combating Terrorism Center’s Sentinel, and Foreign Policy.
Nick Prime is a PhD candidate in the Department of War Studies, King’s College London.


Great stuff gentlemen – I am left to wonder whether under your terms of control, you believe that one force can really control another force – when the latter is a non-nation state actor who is willing to take ruthless actions while the former is not. I have long held to the idea that to defeat our enemy, we must be willing to beat them under their rules, not our own.
War has evolved from the gentlemanly skirmish lines of history into a melee which makes trench warfare of WWI and the booby traps of the Viet Cong look like a traipse through the woods. How does a President who I believe actually believes that ISIL is the worst of the worst, convince a nation that not only do we need to commit ground forces to take the fight to them, but that the rules of engagement need to come off – and that the network news can kiss the ass of the President when we start dragging insurgents out into the street and beheading them. I don’t intend to make light, but as an Egyptian I met in my travels once said to me “When you defeated Saddam, you lost the war the minute you were nice to the iraqi people – You needed to treat them like he did until you could gradually release the liberties.”
In short, I don’t believe you can exert control on an enemy, unless you can be as ruthless as they will be – and to do so rapidly escalates this into a game I am not sure we want to play. Great stuff again guys – really drove me to do some thinking on the topic
Jason, that’s a great question. The real key to the entire idea of this control theory is the recognition that it exists in any number of forms that are differentiated by their degrees of severity/comprehensiveness. The trick then is to determine the degree of control you want to achieve, and at the lower end concepts like containment, coercion, and influence fit within the conception of control we expressed. The value of control then is that it places them all on a scale, like the metric system for strategy, that lets you appreciated what the levels are and what you should expect each degree to require.
Now, to actually answer you’re question: I think that a more comprehensive level of control over a non-state group, is extremely difficult, but possible. Other measures short of that are certainly possible and more achievable, especially when it comes to denying them control of an area or means of achieving their aims.
Coming back to this, I think that too that gaining control over another agent also has to do with certain levels (somewhat borrowed from Daniel Dennett) of understanding.
Most people settle in their everyday lives instinctively impute intentions to people and things (“the computer isn’t working” or “Joe did this because he was upset with me”). Of course, “work” is an anthropomorphism — a computer doesn’t work because it is not a human laborer. And we have no idea why Joe was upset with us, we just have a feeling that it was because of something we did.
All of these are crude attempts not just at causal explanation but also at prediction. The computer isn’t working, therefore we can know with a degree of certainty that it will continue to malfunction unless we do something about it. Similarly, we might predict that Joe — if he truly is upset with us — might be mollified by an appropriate gesture of contrition.
The reason I bring this up is because one core element of Barry Watts’ “Regaining Strategic Competence” and Freedman’s new book is the idea that even in the best case we may not be able to convincingly show that a given action results in a desired strategic effect. Yet to avoid “analysis by paralysis” we guess anyway because we have no other choice.
This brings me to the main point:
Control necessitates some degree of predictability, however crude. At a high level, control is a fuzzy and imprecise process because our abilities to predict will never be sufficient to actual need. The less ambitious the objective, the greater the potential for developing — over time — a way of predicting the opponent’s ability to compete with us in the area we truly care about.
The tradeoff, and it is a rather steep one, is that lowering expectations also means less strategic payoff. Now, it might be argued given the resources we are willing to invest and the red lines we are prepared to cross in Iraq it may be that this is irrelevant — if we were never going to aim bigger in the first place, are we really suffering that much of an opportunity cost? But the problem still remains.
Another relevant point, made in this paper, is that opponents learn about each other over time. http://slantchev.ucsd.edu/courses/pdf/slantchev-apsr2003v97n4.pdf
Opponents ultimately coordinate their expectations about what each is prepared to achieve, and do so after getting a more informed perspective about their respective chances of military victory. Perhaps that may be viewed as an arena for control as well, given that both battlefield displays and negotiations do not necessarily give unambiguous information.
Thank you, gentlemen, for a quality look at the broader strategy question. I found the article original and the ideas soundly stated, and the reader benefit from a cool hard restatement of commonsense. The complex arena that is international politics, alas, typically lacks honest clear articulation. Politicking in democratic societes and the practice of realpolitik go together like fish and ice cream: they don’t. Expectation management is an underappreciated ingredient of the statesman’s art.
The inference of control is how the strategy enables one to gain control of the opponents political order – in this case – the application of military force via a “broad coalition.”
The stated objective of “destruction” of ISIS is a stretch at best;”containment” will be relative, and control will probably ultimately depend on the short term and short lived expediency gained from temporary alliances among this broad coalition of players.