zenpundit.com » Blog Archive » Kilcullen on COIN “Persistent-Presence” vs. “Repetitive Raiding”

Kilcullen on COIN “Persistent-Presence” vs. “Repetitive Raiding”

The Accidental Guerrilla: Fighting Small Wars in the Midst of a Big One by David Kilcullen

I purchased a copy of The Accidental Guerrilla, intending to read it last summer but, being buried under my own academic course work, I was forced to put it aside until recently. I am not finished yet but I can say that Col. Kilcullen has written a seminal, if idiosyncratic, work on the theory and practice of counterinsurgency – no doubt why some reviewers found The Accidental Guerrilla be difficult book to read, one that “…could be like a junior high school student’s attempting “Ulysses.” Or were aggravated by Kilcullen’s format through which he enunciated a more nuanced understanding of the war and COIN than they found politically tolerable. Most readers in this corner of the blogosphere  will find The Accidental Guerrilla an intellectually stimulating book from an author well grounded in the realities of Iraq and Afghanistan, who is the leading theorist of counterinsurgency today.

I would like to take a look at one section where Dr. Kilcullen discusses the merits of “presence” vs. “raiding” in the context of road-building operations in the Kunar and Korengal vallies of Afghanistan by American troops under, successively, LTC. Chris Cavoli and LTC. Bill Ostlund [p. 96]:

Cavoli contrasts this “permanent-presence” methodology with the “repetitive raiding” that has characterized operations at some other times and places. He argues that persistent presence is essentially a “counterpunching” strategy that relies on a cycle of defense and counterattack, in which the presence of the road and Coalition forces protecting and interacting with the population draws the enemy into attacking defended areas, causing him to come to the population and the government – the opposite of the “search and destroy” approach in which security forces “sweep” the countryside looking for the enemy within the population, as if for a needle in a haystack, and often destroy the haystack to find the needle. More particularly, search and destroy operations tend to create a popular backlash and contribute to the “antibody response” that generates large numbers of accidental guerrillas and pushes the population and the enemy together. The persistent-presence method avoids this.

My Comments: 

The context that Kilcullen is writing here is a tactical one but the conceptual conflict of “presence vs. raiding” scales up easily to one of strategy and engages ( or should engage) consideration of how you want to position yourself at the mental and moral levels of war. Colonel  John Boyd, in Patterns of Conflict recommended principles to create strategies and tactics that would: 

  • Morally-mentally-physically isolate adversary from allies or any outside support as well as isolate elements of adversary or adversaries form on another and overwhelm them by being able to penetrate and splinter their moral-mental-physical being at any and all levels.
  • Pump-up our resolve, drain-away adversary resolve, and attract the uncommitted.
  • Subvert, disorient, disrupt, overload, or seize adversary’s vulnerable, yet critical, connections, centers, and activities that provide cohesion and permit coherent observation-orientation-decision-action in order to dismember organism and isolate remnants for absorption or mop-up.
  • Operate inside adversary’s observation-orientation-decision-action loops, or get inside his mind-time-space, to create a tangle of threatening and/or non-threatening events/efforts as well as repeatedly generate mismatches between those events/efforts adversary observes, or anticipates, and those he must react to, to survive

Abstractly, Kilcullen’s “persistent-presence” has superior strategic qualities – it isolates and demoralizes the enemy and daunts the latently hostile while connecting our side to the population and “pumping up” the morale of allies and sympathizers. The initiative is seized and control of the battleground is determined. Most of the time, this is an advantage, so long as the chosen ground is also tactically defensible, unlike, say at Dien Bien Phu. When Julius Caesar was carrying out his conquest of Gaul, he often divided his legions for their winter quarters, even though this entailed some risk, because doing so reinforced the political spine of Rome’s local allies in tribes of uncertain loyalty and intimidated the malcontents or secured the population against  raiding by still hostile Gauls or Germans from across the Rhine. Caesar did a lot better in Gaul than did the French in Indochina.

The problem, is not Kilcullen’s theory of COIN, which seems to me to be solidly based upon his empirical observation and deep experience in counterinsurgency warfare. Nor is tactical execution by American troops the issue either; while the US/ISAF have had successes and failures, the principles of COIN seem to be widely understood, if not always perfectly implemented. The dilemma is at the intermediate level of “state building”, one Kilcullen’s primary strategic goals in Afghanistan, that is supposed to support the progress made in the villages by COIN operations.  

On COIN specifically, Boyd wrote:

Counter-guerrilla campaign  


  • Undermine guerrilla cause and destroy their cohesion by demonstrating integrity and competence of government to represent and serve needs of people-rather than exploit and impoverish them for the benefit of a greedy elite.*
  • Take political initiative to root out and visibly punish corruption. Select new leaders with recognized competence as well as popular appeal. Ensure that they deliver justice, eliminate grievances and connect government with grass roots.*
  • Infiltrate guerrilla movement as well as employ population for intelligence about guerrilla plans, operations, and organization.
  • Seal-off guerrilla regions from outside world by diplomatic, psychological, and various other activities that strip-away potential allies as well as by disrupting or straddling communications that connect these regions with outside world.
  • Deploy administrative talent, police, and counter-guerrilla teams into affected localities and regions to: inhibit guerrilla communication, coordination and movement; minimize guerrilla contact with local inhabitants; isolate their ruling cadres; and destroy their infrastructure.
  • Exploit presence of above teams to build-up local government as well as recruit militia for local and regional security in order to protect people from the persuasion and coercion efforts of the guerrilla cadres and their fighting units.
  • Use special teams in a complementary effort to penetrate guerrilla controlled regions. Employ (guerrillas’ own) tactics of reconnaissance, infiltration, surprise hit-and-run, and sudden ambush to: keep roving bands off-balance, make base areas untenable, and disrupt communication with outside world.
  • Expand these complementary security/penetration efforts into affected region after affected region in order to undermine, collapse, and replace guerrilla influence with government influence and control.
  • Visibly link these efforts with local political/economic/social reform in order to connect central government with hopes and needs of people, thereby gain their support and confirm government legitimacy.


  • Break guerrillas’ moral-mental-physical hold over the population, destroy their cohesion, and bring about their collapse via political initiative that demonstrates moral legitimacy and vitality of government and by relentless military operations that emphasize stealth/fast-tempo/fluidity-of-action and cohesion of overall effort.


* If you cannot realize such a political program, you might consider changing sides! 

Arguably, we cannot realize this kind of political program without a) significantly altering the political culture of Afghanistan which is historically exceptionally hostile to an efficient, centralized state, and b) getting a better set of clients to run the state. Or, c) changing our objectives to ones that are realistic for our time frame, resources and national security interests.

Hamid Karzai is our more humane version of Barbrak Karmal, equally incompetent but more corrupt. Frankly, having stolen the last election and forfeited whatever legitimacy he had in Afghan eyes, Karzai is now a net negative on our efforts and by extending the reach of his government, we alienate every villager and tribesman with whom his officials come into contact. If we are serious, then we should either abandon state-building in Afghanistan and concentrate all our efforts on localities until we secure al Qaida’s destruction in neighboring Pakistan or we should remove Karzai from power and find more effective clients. We need to choose.

If a piece of territory, be it province or nation-state is of no particular intrinsic value to the national interests of the United States, it becomes hard to justify, except upon exigent humanitarian grounds – say, intervening to stop a genocide – a “permanent-presence” COIN operation that lasts for years. It might be better in such places if determined enemies, who are likely to be state supported or at least tolerated non-state actors, faced swiftly dispatched “repetitive raiding” but in a more robust form more properly termed a “punitive expedition“. The the infrastructure that makes the territory militarily useful is systematically and thoroughly destroyed, along with any enemy combatants who assemble to contest the field. Raids, other than neatly targeted assassinations, should not be cruise missile pinpricks but destruction on a scale that General Sherman would find recognizable

Is state-building in Afghanistan and appeasing Pakistan’s military elite our primary national objectives in this war?

If our interest in a regime’s survival is vital, then by all means dig in with a “persistent-presence”. If not, then scale down to a more appropriate level of response.


Dr. Kilcullen has a new book out, Counterinsurgency.

10 Responses to “Kilcullen on COIN “Persistent-Presence” vs. “Repetitive Raiding””

  1. Chicago Boyz » Blog Archive » Kilcullen on COIN “Persistent-Presence” vs. “Repetitive Raiding” Says:

    […] at zenpundit.com […]

  2. Rob Paterson Says:

    I cannot see how with a government like K’s that there can be legitimacy? Is maybe the whole idea of a "State" the issue? A State is merely a funnel for corruption and an avoidance system for authority.But a regional tribal network would have legitimacy – we may find the same is true everywhere as we see the "States" unable to cope with what is going on – Greece – even the UK??Are we seeing the end of the state itself as a workable concept?

  3. Abu Nasr Says:


    Spot on analysis. The biggest issue I have with modern COIN is that it is modeled after post-colonial era communist insurgencies. The third phase or the "build" phase works when the conflict can be boiled down to two competing economic systems. This is not the case with the conflicts we find ourselves in today.

    Both the Iraq and Afghanistan "insurgencies" share the common theme of competing justice systems. One side asks for fair treatment from their government, we respond with an internet cafe. One side has a tribal land dispute, we respond with new power generators. These counteractions do not make sense.

    One of the questions I ask my students is the following: "If you went to court against your rich neighbor and were in the right, would you rather go to an American court or a Taliban court?" The fact that most say "Taliban" highlights precisely what is wrong with our current doctrine.

  4. Stephen Pampinella Says:

    Mark,Thank you for finally nailing down how COIN is really a vehicle for state-building, particularly among the state’s security institutions.  I’m not entirely optimistic about state-building in Afghanistan because of Karzai either. But also, I don’t think the only options are either pulling a Diem or leaving (at least not yet). We’ll all know more after the Kandahar operation as well, which may serve to empower local centers of power in the province that can rival Karzai’s brother.On further thought, this would also necessitate that the CIA stop working with Karzai’s brother – that relationship alone has consistently undercut the foundations of a strong Afghan state (this was predated by our support of Gul Agha Sherzai, who also undercut legitimate government). Practicing COIN on the one hand and pro-warlord strategy on the other works at contrary ends, and we can only hope someone in DC has resolved this contradiction.

  5. zen Says:

    Hi rob,

    Afghanistan does not have a modern history of strong state rule. At the turn of the century, the "Iron Emir" Abdur Rahman Khan was a strong ruler by Afghan standards but it wasn’t a modern Weberian bureaucratic nation-state that he presided over. King Amanullah’s ("the socialist king") modernizing efforts were rejected as apostasy (what finally did him in was a photograph of his Queen in "immodest" Western attire in Europe). King Zahir Shah reigned with a light hand and was therefore popular but accomplished very modest gains, a secularized intellectual urban middle-class in Kabul that supported his overthrow by Daoud, who in turn was toppled by the Parcham-Khalq Communists in Afghanistan’s military which led to decades of war.  It s not an ideal place for state-building even in peacetime.
    Thank you. I agree with your comments – our prescriptions are not bad, nor unwelcome but they do not redress the core grievances of Afghans and the Karzai regime is an engine for increasing their aggravation. Even when Afghan government officials are acting in good fath, locals are apt to assume any decision is the result of bribes while Taliban justice is seen as, if cruel, predictable and relatively incorruptible.
    hi Stephen,
    Unfortunately, I suspect Karzai’s brother has out manuvered the US – i.e. succeeded in getting his Washington backers to put pressure McChrystal and put the Kandahar offensive on hold and/or whittle away at the plan. A best the Karzais are triangulating, at worst Ahmed Karzai works in tandem with the Taliban against the US/NATO:


  6. Joseph Fouche Says:

    This is the inevitable result of leaving a man with a cape and a Texan in a room together for too long.

    Tragedy is the inevitable result.

  7. Stephen Pampinella Says:

    So true.

  8. Cameron Schaefer Says:

    I agree with Rob, I often wonder if our obsession with building a functional "state" is blinding us from other, more realistic methods for gaining our objectives.  Is the prophecy of the decline of the state coming true?  If so, the people working for the state will probably be the last to know.  Afghanistan’s only operable source of governing has been at the tribal level.  As Zen pointed out, it’s not as if a strong, legitimate central government once ruled effectively and we’re simply trying to recreate that…we’re trying to invent something that has never existed. Many troops on the ground, the ones who actually interact with people at the local level, understand this and have been trying to train individual tribes to defend themselves from the Taliban, however their efforts have been stymied from above as the leadership continues to point to Karzai and Co. as the path for Afghanistan’s salvation, not farmers with rifles.  However, one tribe at a time takes time…perhaps more time than we now have, but building an effective central government will take much much more time and likely never happen."Repetitive raiding" is a curious option that wouldn’t require large, long and persistent presence.  I’m curious though why you say it must be a General Sherman style destructive rampage and not targeted assassination?  I realize errant Hellfire’s fired from drones make us more enemies in areas we employ them, but wouldn’t "burning down the forest" do that even more?

  9. seydlitz89 Says:

    Someone remind me again why exactly do we need a new theory of COIN? 

    What about David Galula?  OK, 1960s and a lot has changed . . . but in terms of being clear, concise and Clausewitzian . . . looking at things from his perspective tells a lot.

    I haven’t read Kilcullen – can’t comment on him at all, but still the simple question remains.  Why, and then why now? 

    When we’re broke and divided at home? 

  10. zen Says:

    Hi Cameron,
    My pardon, my writing was unclear. I am fine with "targeted assassination" where that is feasible. It would often be a better choice and, certainly, more cost-effective. It is also legal if it was used and framed as military retaliation for an attack. I’m not so hot on using it preventatively. That’s not a good reputation to have in the international community. But where a state or quasi-state ( like Hezbollah) is behind an attack on American citizens, exacting systemic and structural punishment creates a future deterrent. A state might be less likely to sponsor a terror group’s attack if they are fairly sure that PACOM would then sail by and take out their refineries, power grid and telecommunications systems.
    Hi seydlitz89,
    You should read Kilcullen, if for no other reason, than he is no armchair theorist, but one who drew empirical conclusions from where bullets were flying in very different states, not unlike Galula.
    " Why, and then why now? "
    I am not a good person to argue for state-building in Afghanistan as I do not believe in it except for instrumental purposes ( quid pro quo benefits for local support). The Afghans do not want a state and by all measures, seriously distrust the one they have even more than either the Taliban or US forces. Our national interest is in Afghanistan  1) physically liquidating and politically discrediting al Qaida, Afghanistan is just a big launch pad to raid Pakistan

Switch to our mobile site