Edward Luttwak's "Counterinsurgency Malpractice"

I spent a few hours recently, reading Edward N. Luttwak's article in Harper's Magazine, "Dead End: Counter-Insurgency as Military Malpractice", and carefully thinking over his argument. It was a pleasant holiday from the reality of war here in Baghdad, and a reassuring reminder that there are still havens of calm (like CSIS, where Dr Luttwak is a Senior Fellow) where one can consider issues thoroughly and arrive at firm conclusions. From my viewpoint, here in Iraq, things somehow never seem quite so black-and-white.

Professor Luttwak is a famous defense policy expert, with publications on the Roman Empire, nuclear strategy, coups d'état and globalization, among others. He is not a specialist in counterinsurgency, but his opinions carry much weight and we should all welcome his recent foray into the field. I hope he will forgive this précis, but in essence he argues that "insurgents do not always win, actually they usually lose. But their defeats can rarely be attributed to counterinsurgency warfare, as we shall see". The means he argues are most effective (but does not himself advocate, of course) are wholesale reprisals and "out-terrorizing" the insurgents.

The first part of the article is a critique of the new counter-insurgency manual, FM 3-24. As Dr Luttwak acknowledges, he is actually critiquing the so-called "straw-man" draft (leaked onto the internet in early 2006) not the final version published last December (about three months before his Harper's piece appeared). This is a pity, because the official version differs substantially from the straw-man he critiques.

For example, Dr Luttwak criticizes the draft's focus on legitimacy as a means to popular consent, suggesting that coercion makes up for lack of consent: "government needs no popular support when it has obedience". If he had seen the final version he may have been more comfortable with these words (from Chapter 1):

All governments rule through a combination of consent and coercion...Legitimacy makes it easier for a state to carry out its key functions.... "police states" typically cannot regulate society or can do so only by applying overwhelming coercion. Legitimate governance is inherently stable; the societal support it engenders allows it to adequately manage the internal problems, change, and conflict that affect individual and collective well-being. Conversely, governance that is not legitimate is inherently unstable; as soon as the state's coercive power is disrupted, the populace ceases to obey it.

Some other comments seem based on incomplete information. Dr Luttwak observes that "better government...is certainly wanted in France, Norway and the United States but obviously not in Afghanistan or Iraq, where many people prefer indigenous and religious oppression". This differs markedly from my field experience in both countries. One night last summer, a member of a tribe that sits astride the Afghan-Pakistan border told me: "you want to bring us 'democracy' at the national level, but we already have democracy within the khel. What we want from the government is security, honor, justice and prosperity. If anyone offers us those things, we will fight for him to the death. If democracy only brings elections, what use is it?" In my experience, this is a fairly widespread view in both Afghanistan and Iraq. It reflects cultural differences in the way social groups conceive of the state, and varying notions of democracy and legitimacy. But ordinary Iraqis and Afghans have told me repeatedly that they want effective government that is responsive to people's needs. This is legitimate government, defined through their (emic) terms of justice, honor and responsiveness, not our (etic) categories of parliamentary elections and English common law.

Dr Luttwak argues that "the vast majority of Afghans and Iraqis, assiduous mosque-goers, illiterates or at best semi-illiterate, naturally believe their religious leaders" (who, Dr Luttwak suggests, incite violence with claims that America seeks to destroy Islam and control oil resources). Again, this is at variance with field observation. In fact, neither Iraqis nor Afghans are particularly assiduous mosque-goers. And religious figures are prominent on all sides of both conflicts, in moderate and extreme political groups; there is an extremely wide range of clerical opinion, ranging from quietism through support for democratic government, to extremism. More fundamentally, in these societies, religious faith is not a function of ignorance and credulity, as Dr Luttwak implies, but a widespread cultural norm that infuses all social classes, political orientations and education levels. Indeed, this is one criticism of FM 3-24 that others (including Ralph Peters) have raised, arguing that it minimizes the role of religion. Again, this is a valid criticism of the leaked straw-man, but was addressed in the final version.

In his discussion of intelligence work, Dr Luttwak draws a sharp distinction between intelligence problems and political problems, criticizing the manual for insufficiently addressing political issues. In my view, this is a fair point (though, again, the final version includes much more discussion of political factors than the unofficial draft). The U.S. government is currently preparing a whole-of-government counterinsurgency handbook, which will address political issues more fully.

Dr Luttwak's point about sectarian tendencies in security forces is useful. I am unsure when Dr Luttwak was last in Afghanistan, so he may have more recent information than me. But my impression of the Iraqis I work with now, and the Afghans I worked with last year, is that though such tendencies are real, the presence of U.S. forces, joint operations, and partnership of police with soldiers, provide "checks and balances" that mitigate them. And the growing professionalism of security forces is reducing these tendencies, particularly in Afghanistan. That is, in fact, the point of the Joint Security Stations, which resulted in a large drop in sectarian violence in areas of Iraq where they have been applied. So we may end up making some progress on this issue.

We have also had some success addressing Dr Luttwak's concerns about language proficiency. He writes of the "astonishing linguistic incapacity" of the U.S. military, comparing unfavorably the number of Arabic students at defense language institutes with historical training efforts in Chinese and Japanese. Again, he may not have the latest information on this. According to the Defense Language Institute, by 2005 there were over 300 Arabic instructors and more than 1000 Arabic students in training—a ten-fold increase since before 9/11. Language skills are still too rare in the field, but I have seen steady improvement in the past few years. The new manual pays a great deal of attention to developing and improving cultural and language capabilities.

More importantly, Dr Luttwak implies that only indigenous forces can do counterinsurgency, because only they have the local knowledge, cultural understanding and language ability to work effectively. I could not agree more: imagine how Californians would react if the Iraqi military tried to police the streets of Los Angeles. Commanders in Iraq and Afghanistan also agree, and have made major efforts all along to train local forces to do the job themselves. In addition, the partnership of Iraqi or Afghan forces with U.S. units in the field today gives commanders immediate access to cultural and linguistic understanding from local partners, while U.S. forces reinforce them (Iraqis use the term thabit) and help them ensure that their behavior accords with legal and human rights standards. Local forces are crucial, which is why the manual emphasizes partnership so heavily.

Another benefit of working alongside locals, right in the heart of an operating area, is that it encourages the population to report on the insurgents. Dr Luttwak argues that "many of the local inhabitants certainly know who the insurgents are...but they are not telling". Actually, they now are. Every U.S. unit that has applied the new manual's approach, living among the people and focusing on their security, has reported a flood of unsolicited tips from the population and volunteering of information about insurgents.

Having knocked the stuffing out of the straw-man, Dr Luttwak suggests an "easy and reliable way of defeating all insurgencies everywhere": essentially, to "out-terrorize" insurgents through reprisals, mass executions, and collective punishments. He cites German forces in the Second World War, claiming this approach was standard, "and very effective it was too in containing resistance movements with very few troops". Again, I beg to differ. One German officer on the Eastern Front remarked: "the German Army in Russia was like an elephant attacking a nest of ants. The elephant will kill millions. But in the end the ants will eat him to the bone". German methods in Yugoslavia, Greece and Russia proved extremely counterproductive. And as Barbara Tuchman argued in The Guns of August, earlier German brutality against civilians in Belgium and France in 1914 helped provoke worldwide revulsion, contributing to eventual American intervention and German defeat in the First World War. (These are historical observations, of course, and do not in any way impugn modern Germany or today's Bundeswehr).

The methods Dr. Luttwak mentions are thus not a prescription for success, but a recipe for disaster. As he quickly admits, U.S. and Coalition forces would never consider such methods for a moment. And this is just as well, since this approach does not work. The best method we know of, despite its imperfections, has worked in numerous campaigns over several decades, and is the one we are now using: counterinsurgency. I admit (and have argued elsewhere) that classical counterinsurgency needs updating for current conditions. But the Nazis, Syrians, Taliban, Iranians, Saddam Hussein and others all tried brutalizing the population, and the evidence is that this simply does not work in the long term.

Dr. Luttwak's final point is one of his strongest. He argues that there is ambivalence in the United States approach, that America is "—to start wars because of future projected threats...—to conquer territory or even entire countries, and yet is un—to govern what it conquers, even for a few years". What he calls "the critical ambivalence of occupiers who refuse to govern" is indeed worth discussing, though it is properly an issue of political will, strategic culture and national character, rather than counterinsurgency technique. As a colleague said to me in Iraq last year, "we need to either make a serious effort to govern these people, or get the hell out". But Iraq and Afghanistan now have sovereign governments; and we (with many other countries) are helping these governments to do exactly that—make a serious effort to govern their people effectively. And we have no plans for permanent presence: as the President said, we will stay as long as we are needed, and not a day longer.

Overall, I found Professor Luttwak's viewpoint fascinating, and a thought-provoking addition to our ongoing professional discussion, but ultimately not quite convincing. Perhaps that's just me—things do tend to look different, and more complicated, from here in the field. But I would encourage people to read both the Harper's piece, and the actual final version of FM 3-24, and make up their own minds. On-the-ground facts (like language improvements, partnering with Iraqi forces, the drop in sectarian violence, joint operations, and improved governance) are also worth taking into account.

None of this means we will automatically win in Iraq and Afghanistan; unlike some pundits, most of us with a professional background in counterinsurgency always expected this to be a forbiddingly difficult undertaking. Our approach remains fundamentally field-oriented and evidence-based, and clearly there are still many things that could go wrong, dark days ahead, and extremely demanding and complex challenges to be faced, as in every war. Admittedly, also, we waited a long time before adopting our current approach, which inevitably makes things harder. And as John Nagl so convincingly demonstrated, our counterinsurgency doctrine will undoubtedly need ongoing and agile adaptation as things develop. But to suggest that counterinsurgency is intrinsically flawed, that it is "malpractice", or that wholesale slaughter and oppression are the only viable solution (albeit one that we will never apply) is an argument that the facts simply do not support.

Dr. David Kilcullen is Senior Counter-Insurgency Advisor to the Commanding General, Multi-National Force Iraq. Though authorized to post at SWJ Blog, these are his personal views, do not reflect those of any government or organization, and have not been screened or vetted.


SWJ Editors Note: Small Wars Council discussion: Edward Luttwak - Counterinsurgency as Military Malpractice

Your rating: None


Wow, this was a blast from the past. Four summers ago almost to the day. Lots of water under the bridge since then, lots of good discussion on SWJ too (some of it heated, some of it angry, but in it all much to learn from)

For whatever it is worth, this response by me to Kilcullen's post was probably one of my first on SWJ. Some of it as a re-read it i still stand by and think is correct, some of it not, and after four years would modify or state that it was wrong.

I still think I had it right about Kilcullen's lack of objectivity and that he was obviously an early proponent and constructionist of the Surge Triumph Narrative as he claims the Surge (and his purported key role in it as he eventually tried to show in his book) to be the primary causative factor for lowering violence.

But I think that I over-stated too early the argument about dogmatism especially with this post in how it prevented creative action on the ground during the Surge. Doug Ollivant in his recent interview with Octavian on SWJ seems to me to have it right when he says that the actions taken during the Surge at the tactical level reinforced the more important conditions that were driving violence down and may have sped up that process. I still stand by my arguments about continuity rather than discontinuity operationally within pre surge and surge armies, but to argue that FM 3-24 had taken away creativity at the tactical level as I did in this post now reads to me to be wrong.

I also dont think that Coin became the matrix-like box that completely controlled the army as I started to suggest here and developed as an argument more fully in other places. To be sure Coin became dominant in the army, but that dominance may turn (or is turning out now) to be quite fragile. It is one thing to argue that Coin has dominated the Army's intellectual climate (which I think it has), but an Army is more than just its intellectual climate and consists of many other things that are mitigating against the intellectual dominance of coin.

Just some thoughts and reflection after four years...


I question Dr. Kilkullens current objectivity and detachment as a scholar and analyst since it apparently conflicts with his current position as senior advisor to the commander of Multi-national forces Iraq, General David Patreaus; unfortunately for the sake of objectivity and detached analysis this position appears to have turned him into an advocate of the surge.

Some comments in Dr. Kilkullens response to Edward Luttwaks article are indicative of his loss of objectivity and detachment concerning the war in Iraq and the surge. Specifically, his hyper-optimistic appraisal of sectarian tendencies in the Iraqi Security Forces reads like an "official" report from and MNF-I spokesperson. He uses the reduced numbers of dead bodies on Baghdad streets as an indicator that the tactical use of joint security stations in neighborhoods along with combined operations with Iraqi Security Forces is working to reduce sectarian killings. An alternative explanation, however, exists. It is that most of the mixed areas in Baghdad have already been taken over by shia and the number of sunnis in these areas left to be killed has also been reduced thus explaining the reduced number of dead bodies on the streets. A number of comments made recently by lower enlisted soldiers and NCOs who patrol the streets to embedded reporters suggest this possibility. An objective and detached analysis of this problem would have recognized that possibility.

I commanded an Armored Reconnaissance Squadron in West-Baghdad in 2006. I worked closely with National Police, Iraqi Police, and the Iraqi Army. I dealt first hand with the ongoing civil war that is occurring in Iraq. In Ameriyah in 2006, I tried this surge plan in microcosm by concentrating my entire Squadron in Ameriyah: the result, it did not work because the underlying political conditions that have produced civil war had not changed.

What is really going on here is that the American Army has been placed into a dogmatic box that is defined by our current counterinsurgency doctrine. Our approach in Iraq now with this surge is not at all creative and innovative as pundits, parts of the media, and advisors like Dr. Kilkullen are portraying it. Instead the surge is dogmatic and archaic and does not demonstrate an operational and tactical approach that realistically takes into account actual conditions on the ground.

We have sadly turned our army into a counterinsurgency-only force with a current doctrine that is wrongly and mechanically built on "lessons" from past counterinsurgency operations. Want Proof? Go read David Galulas "Counterinsurgency Warfare" and tell me if it does not read just like the Armys new Coin manual. Further proof is a hypothetical that could be turned easily into fact: If I had a smart cavalry or infantry Sergeant First Class and First Lieutenant who just returned from Iraq and gave them a copy of Galulas lengthy essay with specific temporal references to time and place removed and asked them what they had just read they would say it is a summary of the Armys new Coin doctrine as they just got done practicing in Iraq.

The worthiness of Edward Luttwaks article is that it fundamentally questions the theory that our counterinsurgency wars are based on. Others who have been seduced to the point of dogmatism by our new doctrine should think hard about what Luttwak has to say.

In Defense of Luttwaks Argument

There is purpose in plain language. Yes, Luttwaks use of the draft version of FM 3-24 is curious, but nothing in the later version changes Luttwaks basic point. Its a misreading to focus a critique of him around the idea that he is attacking a straw-man argument. His point is that the idea of there being a kind of specialized war called counterinsurgency is mistaken. Dr. Kilcullens response misses the larger point of the article. Luttwaks argument is that the idea of a special kind of war called counterinsurgency is beside the point. Why is this search for the elusive key to counterinsurgency a dead end? Because what the long standing history (Roman, Ottoman, German, Spanish, French, American, etc.) of conventionally trained and organized armies successfully defeating insurgencies tells us is, to use the words of both the draft and final versions of FM-3-24, that the key to war in general is generating "enough violence to achieve political consequences." That is, in Luttwaks plain language, they out-terrorized the terrorists, they made a "wasteland and call[ed] it peace."

Hes saying that the idea counterinsurgency, whether from the spring 2006 manual or its present version, is the real straw-man, whatever the details. This is not because military violence is not capable of achieving political consequences, it most certainly is. But the kind and amount of violence which is required to quell an insurgency are of such a type as to be beyond the capacity of "the armed forces of the United States or of any other democratic country" to generate. [See Gen. Rupert Smiths The Utility of Force, or Gil Meroms How Democracies Lose Small Wars for more in-depth examinations of this point].
This is not an indictment of the U.S. military. It is a recognition that the chief success of Western militaries in creating the modern international environment has been to create an order wherein the use of violence to affect political change has been so de-legitimated that the very strategies used in the past to defeat insurgencies cannot be used today.

The first Gulf War was predicated on the idea that Saddams use of force to overthrow the existing Kuwaiti government was so illegitimate that it required the "international community" to step in and reverse the conquest. The Spanish government maintains the same line in justifying its use of force on the Basque insurgents. Its the same for the Russians in Chechnya, the Georgians in Abkhazia and South Ossetia, or the British in northern Ireland.
The premise of the counterinsurgency school is that it can somehow escape this double-bind. Luttwaks point is that no amount of on-the-ground experience can be collated, edited, and analyzed to produce a strategy for using military force which is both successful in defeating insurgents and consistent to the idea that the contemporary world, or at least the democratic one, rests on a foundation of politics which explicitly rejects the idea that violence is a legitimate tool of politics. There is either pure Hobbesian military-political effect, or moral consistency. There cannot be both.

Hence his use of hyperbole to highlight the absurdity (to the modern, democratic mind) of the ancient tactics used to defeat insurgents. Luttwak knows full well that the punitive approach of the ancients will not and cannot be effectively adopted in the world we live in today. Even if soldiers from democratic powers could somehow be taught to override the basic principles of their upbringing, the military effectiveness of their brutality would almost instantly make any political outcome in the conquered territory illegitimate. This is the genius of insurgency. It uses the international political system set up by the states of the militarily dominant states as a means to defend its political rights even as it is impotent militarily. Kilcullens cite from FM 3-24 makes this clear.

All governments rule through a combination of consent and coercion... Legitimacy makes it easier for a state to carry out its key functions... . "police states" typically cannot regulate society or can do so only by applying overwhelming coercion. Legitimate governance is inherently stable; the societal support it engenders allows it to adequately manage the internal problems, change, and conflict that affect individual and collective well-being. Conversely, governance that is not legitimate is inherently unstable; as soon as the states coercive power is disrupted, the populace ceases to obey it.

Coercion is expensive. It is of only limited effect, and then only in extreme conditions. The solution for governments is to rest their claim to govern on an ideology which is legitimate to enough of the population so as to make the use of the states coercive power a limited occurrence, and to rely instead on citizens to follow the law because they are internally so compelled, not externally forced. The illusion of counterinsurgency is that somehow a foreign-constituted force will be able to make its governance legitimate internally. But the idea of the national-state and the tropes of nationalism which sustain it make this a virtual impossibility. So the reality is a double-bind - admit that the philosophical claims of nationalism are as arbitrary, and that the current international order which is predicated upon them is too. Or else hold the line in favor of the insolubility of current national borders under the idea of the absolute right of sovereignty of existing governments no matter how odious their origins.

If illegitimate governments lose control "as soon as the states coercive power is disrupted," then it would seem that the only route to establishing a stable government would be to find a way to make foreign occupation legitimate (something Kilcullen admits with his California example is a near impossibility), or else to make sure the coercive power of the government is not disrupted. How do you do that? You engage in the kind of punitive raids Luttwak lists in his historic examples. The current occupations in Iraq and Afghanistan have been exercises in the disruption of the occupying powers coercive force. Luttwaks argument is that no amount of strategic or tactical innovation can enforce the writ of an occupying power in the contemporary world because the ideology of the democratic states forsook the legitimizing logic of force.

Hence his use of hyperbolic examples. The hyperbole of the examples he gives serves to highlight the absurdity of the double-bind the U.S. finds itself in. The argument of the counterinsurgency school misunderstands the militarys role in shaping political institutions and it overemphasizes the universality of Western political logic. Hence the idea that counterinsurgency is a kind of military malpractice. It is malpractice because it conflates the use of force with the achieving of political goals which are inconsistent with the tool of war making, at least in the modern world. It is, then, a dead end because it asks of the American military something that it cannot give. Not because it is not a capable force, but because what is asked of it is impossible to achieve through a force of arms that is anything less than brutal.

What he leaves unsaid in this piece is that the reasons for the wars were so feeble that the American democracy could not sustain a righteous anger capable of unleashing the necessary brutality. And its just as so because the wars themselves are not worth the cost of such an anger. That the American military has implicitly acknowledged as much in its insistence on finding a workable counterinsurgency is to its credit, even if such a search is a dead end.

I think another problem with Luttwak's thesis is his disregard for historical facts in order to support his central thesis that insurgencies are only put down through brutal means.

One of his points referred to the Romans and their use of "sticks and carrots". In order to support his innaccurate thesis, he totally skipped what those "carrots" were and focused completely on brutal examples of punishment.

A large part of the Roman empire did not require constant military action because those nations who were member states had distinct economic advantages trading with Rome and throughout the empire. For many, Roman legions represented security and Roman law represented the most equitablle system they had been subjected to. Under what other nation could a merchant sue for recompense on a broken contract?

Further, private citizens saw the advantage in joining the legions and gaining citizenshi, ability to travel in relative security and in the trickle down effects of the economic system.

To disregard those realities and focus solely on Roman brutality as the indicator by which insurgencies are destroyed or kept from occuring is completely disingenuous.

The question has always been about what benefits the individuals. More importantly, what benefits the middle class and the upper glass merchants or land owners since it is these groups who historically bankroll insurgencies and employ the masses?

A second, rather disingenuous point, was the reference to Joseph Bonaparte and the constitution he tried to install. Mr. Luttwak paints Napoleon Bonaparte as a kind of "liberator" in a backhanded attempt to equate his conquests with the activities of the United States.

Again, he loses points for disregarding historic facts in order to make his central thesis fit into the dialogue.

While it may be true that Catholicism may have been an over all uniter of people in Spain as well as Joseph Bonaparte not being "Spanish", the main problem is that no one was fooled into thinking that Bonaparte and his army were "liberators". He was well on his way to conquering half of Europe. Informatin might have moved a little slower in those days, but we are not talking about complete ignorance of current events.

Secondly, when Bonaparte's army entered Spain, they were an army living off of foraging and paying their soldiers with gold coin of whatever realm they happened to have conquered and raided the banks, treasury, the wealthy and the middle class. No peasant living along its path would have imagined the army that took his pigs, chickens, cows, wheat or daughters or sons as "liberating". There were certainly no clerks following in the army's path compensating these folks for their losses.

Luttwak was trying to make similar comparisons to the reaction of Iraqis to our own occupation. Yet, he used the incorrect anaology of ideology being the focal point and binder of the insurgency, causing the peasants to fight against their own interests. As if that ideology was the strongest motivator.

When, as both the roman and bonaparte conquests clearly show, the biggest interest of the citizens are benefits and security. If it does not look beneficial or act beneficial, no piece of paper is going to change their mind, neither a religious ideology.

Third, Luttwak leaves out a very important piece of French/spanish history: the British arrived in Spain to support the "insurgency" and fight bonaparte directly.

Fourth, for a closer analogy, had the Spanish Priests and Landowners instituted the second inquisition against the peasants and other commanders in the middle of their rebellion against Bonaparte, the entire affair might have had a different outcome.

Just ask al Qaeda in al Anbar.

In closing, Luttwak's points were historically incomplete, thus underminding his primary thesis that only brutality and terror can put down brutality and terror.


I appreciate your words. Like you I found Dr Luttwak's take challenging and interesting. You have chosen to focus your response on his dissection of FM 3-24 as well as his and many others, I believe unfounded, issues with the inherent violence of Islam.(To me this has more to with tribal culture than Islam because I have encountered such behaviour elsewhere in circumstance unrelated to Islam). I found your reply compelling and I also don't agree with what seems to be his 'well it all seems a bit hard so dont try approach. Maybe thats just my military upbringing. However, the issue that resonated most with me and where I do believe he has a most important point is on the need to allow systems to dissipate energy.

All systems have inherent energy and this energy is a product of the elements, relationships, issues and forces in play inside the system. If a system is in a high state of turmoil the energy flows inside the system are likely to be enormous. Containing such a system and bringing it back to close to equilibrium requires a number of steps not least working out how to bleed the excess and harmful energy out of the system. Sometimes this means letting things run their course rather than truncating them with equal or greater force or walls for that matter! Unlike chemical systems biological or societal energies cannot be contained for long, eventually they adapt and erupt often in unpredictable ways and directions. Dr Luttwaks thesis is that we may have to let some of the forces created from years of cultural development, repression, starvation under sanctions, ignited by poorly crafted war and sectarian conflict, run until the uncontrollable hatred and enmity starts to dissipate in revulsion and exhaustion.

Is this moral? Is this just? No it is not but neither is the sacrifice of countless lives in a doomed attempt to control the uncontrollable so that flawed policy can be sustained. Also in a purely strategic sense how much blood and treasure can the West bleed trying to maintain a status quo that was always unsustainable. The forces you are grappling with are inevitably leading to a new strategic construct in the Middle East and possibly the World, thinking that after all that has gone before, the US and its Allies are in control of this looks like hubris. I am becoming very pessimistic that despite our operational and tactical abilities our chances of success are not improving due to strategic imbalances. I have a niggling and deeply worrying feeling that in the 'Global Insurgency their strategists maybe better than ours.

Good luck mate. Come home safe.