Small Wars Journal

Complexity Theory and Counterinsurgency Strategy

Share this Post


Both analysts and strategists have a tendency to categorize various security threats in rather neat categories.  In many – likely most – cases these categories simply do not fit well for areas where there is significant internal insecurity.  This particularly is the case for groups that can be called hybrids or shape shifters.  Armed groups have an unfortunate tendency to shift both their form and significance to security operations.  All too often, planners and intelligence analysts are unable to adequately track the changes among what are viewed as peripheral groups, and the significance of these changes.  Military strategy normally is focused on insurgents or terrorists, with some peripheral attention to militias or other armed groups.  Trying to unpack the significance and operations of other, non-included groups can be very difficult.  A broader form of environmental sensing – to include how intervention forces may serve as yet another actor for analysis – is critical in understanding how to achieve operational success.

As Western forces increasingly have become involved in various interventions in countries facing internal conflicts -- or in some cases, creating internal conflicts -- much more attention has been paid to complex security environments.  A number of scholars and analysts have devoted considerable attention to developing a typology of armed actors similar to terrorists, insurgents, warlords/militias, organized crime, and other groups.  The purpose of this paper is not to pick through all the definitions, particularly distinctions between terrorism and insurgency, which can become a sterile exercise.  Instead, the focus is on whether such fixed types are dangerous analytically in developing strategy. 

There are several issues that are important.   First, most military strategy is focused on insurgents or terrorists, with some peripheral attention to militias or other armed groups.  Trying to unpack the significance and operations of other, non-included groups can be very difficult.  Second, the importance of other groups is very dependent on the overall security environment, and they can be both agents and the result of this environment.  Finally, most of these movements exist and in some cases thrive in societies in which multiple stressors exist.   

The ‘others’ can be used as a catch-all category, but such groups can be critical for internal security.  (For a similar approach, although more focused on militias and criminal groups, see Steven Metz’ Rethinking Insurgencies) They can emerge (and at times recede) quickly as security threats.  Many of these groups tend to be at best secondary targets for analysis and commonly more difficult to predict.  If ideologies exist, they may be incomprehensible to outsiders.  They are more likely to emerge in already stressed environments, and the salience of particular identities can shift with environmental shifts.  Two types of groups have particular salience.  Loosely described, they are hybrids and shape shifters.

Hybrid Groups

Hybrid groups are those which show aspects of combining the basic characteristics of differing kinds of groups.  Most commonly, they might be viewed as organized criminal groups that also develop political or ideological aspirations or ‘political’ groups whose level of criminality is such that they appear to be more interested in profit than politics.  The former organizations – at times rather loosely organized – typically have been treated as common gangs until their political importance is recognized, albeit frequently rather late.  Their movement into political violence usually has caught analysts and governments by surprise.  Three good examples are provided by the Jamaat al Muslimeen (Organization [or Group] of Muslims) in Trinidad and Tobago, D Company in India, and the Jamaican gangs referred to as posses.

The Shape Shifters

Clearly, groups can change over time.  These changes can involve membership, goals, and at times types.  A key question for governments and analysts is once groups are labeled, how well can security services shift as groups morph?  At best, there likely will be a lag as analysts acquire information on violent groups.  At worst, viewing groups as static entities can lead to ineffective or perhaps counterproductive strategies in responding to internal threats.

There are other issues in group transitions.  Elements within groups can change at different paces.  For example, it appears that different Frentes (Fronts) within the Revolutionary Armed Forces of Colombia (FARC) have developed different levels of interest in politics versus the narcotics trade (discussions with Colombian military officers, Bogota, Colombia, June 2011). The mix of political versus commercial interests almost certainly will complicate efforts to deal with the group.  More generally, this is exacerbated by multiple motives of the members of armed groups.

Uncertainty and Labeling

In general, this leads to the question of whether a political version of the Heisenberg Principle applies.  In physics, the basics of the Heisenberg Principle are that the more precisely the position of a particle is determined, the less precisely the momentum is known in this instant, and vice versa; and that the measurement of position necessarily disturbs a particle's momentum, and vice versa.  In terms of internal security, do governments really know what they are looking at?  Typically, in the intelligence and analytical worlds, more attention is paid to groups at particular points in time:  the nature of intelligence collection most often involves a series of snapshots.  Although these snapshots if collected and analyzed over time with emphasis on changes observed can provide valuable clues as to shifts in groups’ behaviors and structures, it is much more common for labels to ‘stick’ to particular groups, with changes being viewed as shifts from pre-existing norms.  It can require considerable flexibility for analysts to completely re-cast their mental constructs of violent groups that might be undergoing major changes in type.  

There also are practical aspects to this issue.  For example, police services inherently focus on criminal groups.  Both through pre-existing mindsets and through bureaucratic imperatives, police intelligence analysts almost certainly will be pre-disposed to label most illegal violent groups as criminal, minimizing possible shifts to increased political activities.  Conversely, ministries of interior and militaries are more likely to focus on political and internal security, with a concomitant emphasis on ‘insurgent’ groups.  As groups shift their forms of identity, there almost always will be a lag time – and, seemingly at times, a complete failure – in agencies’ recognition of this.

The second related question with this issue is what impact does government labeling have on the behaviors of armed groups?  This issue is much more amorphous, but may play a role in groups’ operations.  Arguably, groups may shift their activities in response to government labels.  For instance, criminal groups may increase their social support and political activities in response to a label of “organized crime” in an effort to increase their legitimacy.  This certainly has been the situation in the case of the Jamaican posses, which have developed at least de facto links with Jamaican political parties.  Likewise, in Mexico, some of the drug gangs appear to have increased their social services in certain areas.  These types of activities can change the perceptions of the populace they serve, even as governments or external supporting forces continue to view the groups as solely criminal.  Conversely, insurgent groups that become increasingly involved in organized crime – such as the Colombian FARC – may lose popular support even as a government continues to base its strategy on traditional counterinsurgency strategies.    

Two other issues also must be factored into strategic and operational analysis.  The first is what David Kilcullen calls the “accidental guerrilla” for whom insurgency is more a matter of circumstance than a deep ideological support for a cause. Also, in practice, there might also be marked differences between formal ideologies – or a lack of such ideologies – versus realities on the ground.  Groups can adopt (and have adopted) any number of varying ideologies that would seem to support mobilization and recruitment for their particular goals, but this does not necessarily mean some form of pure adherence to these ideologies.  This is not to argue that most insurgent leaders are somehow hypocritical; there certainly are many (and probably the majority) who are ‘true believers.’  At the same time, however, the environment in which they are operating typically will cause shifts and modifications to the base ideology.  These shifts (if recognized by counterinsurgency forces) may have considerable significance in devising broader counterinsurgency and internal security strategies.

Complexity Theory & Armed Groups

Analysts in recent years have paid much more attention to network structures.  Clearly this has been a positive development overall, but groups’ interactions are more than networks.  A key question for analysts in dealing with environments in which more than one violent group is operating – which likely describes most countries in which there is significant instability – is how these interactions might change each group’s behavior.

This is more than a simple one-on-one relationship amenable to conventional network diagramming, although this is of course a starting point.  In many cases, ‘reverse’ links – groups in active opposition to each other – may in fact be much more important than cooperative relationships.  Each group has an impact on the security environment, and this is true regardless of whether they are directly linked.  Complex environments lead to complex groups, and their interactions can even further add to problems in analysis.

Several analytical issues revolve around viewing the complexities of multiple armed groups of multiple types within a country or regions.  The first is identifying triggers and tipping points that might cause groups to either ally or split.  The second is determining whether splits and schisms are net positives or negatives for the security forces and the security environment. Finally, the interactions of multiple groups can make analysis very difficult.  Complexity is non-deterministic, and gives no way whatsoever to precisely predict the future.  At best, probable trends and ranges of probabilities can be expected from analysis of this environment.

A key point in dealing with such groups is that their activities can serve (and decidedly have served) as a catalyst for wider violence and unrest.  This can be exemplified by the activities of the Mungiki in Kenya:

"Initially, we were seeing three kinds of violence," says Muthoni Wanyeki, executive director of the Kenyan Human Rights Commission. Disorganized violence in villages tended to rise up suddenly, but fizzle out quickly. Organized militias – with paid, motivated members – have kept the violence going and have largely led the charge in expelling minority ethnic groups by force. Police use of extreme force – live bullets rather than water cannons or tear gas – has also stirred ethnic passions.  A fourth type of violence has now emerged, as displaced people carry back stories of horror and spur on calls of revenge in communities that had previously been peaceful. "Now we are seeing a communal response in areas where it has not happened before," says Ms. Wanyeki.

The Mungiki also raise one other issue:  “armed” groups do not have to be well armed to be significant.  Any group with some level of organization – either formal or loosely networked – and motivation to use violence can shift internal security environments rather quickly.  Clearly, underlying stressors must exist, but if so, ill-armed (and perhaps poorly organized) groups can serve as a tipping point for widespread violence.  At times, this might lead to the virtual collapse of state control. 

Insurgency environments can bring many other security issues.  Phil Williams of the Ridgeway Center has shown the impact of criminal activities on security in Iraq. In some cases, insurgents of various groups have engaged in organized crime; in other cases, criminal groups have cooperated with the insurgents; in yet others, gangs have operated along parallel tracks separate from the insurgency.  In all cases, organized crime has significantly complicated restoration of security.  Clearly, in some cases – most notably the Maras in Central America and arguably the drug gangs in Mexico – organized crime by itself can constitute a national security crisis.

A Paler Shade of Red

To some degree, some of these concerns of too great an emphasis on ‘pure’ insurgent – or in the military color scheme, ‘red’ – groups have become increasingly a matter of official recognition.  Most significant in recent years (at least within the US military) has been the Major General Flynn report on his experiences in Afghanistan. Although MG Flynn’s report very correctly identifies the need for intelligence collection on the ‘white’ components of a country facing an insurgency, it can leave out significant components of the overall security environment.  Continuing the color coding system of analysis, there likely are actors that can be represented as various shades of pink.  These are groups – in some cases, inchoate or very loosely organized – that can have a major impact on the course of internal security.  Many of these groups may have few aspirations beyond either profit motive or in some cases simply providing localized security for their tribes or immediate areas, but they must be factored in to strategic analysis.

Strategy for Complex Environments

So where does this leave strategic planning for forces involved in these types of environments, whether the governments themselves or intervening forces or advisors?  There certainly are lessons to be drawn from earlier “imperial policing”, whether British or French colonial efforts or US involvement in places such as the Philippines or Central America.  C. E. Callwell, Frank Kitson, the US Marine Small Wars Manual, and various French authors certainly can be read with value.  At the same time, however, adopting their precepts wholesale is very unlikely to succeed in the age of the video cell phone and improved tools for networking and rapid communications.  Localized security threats are unlikely to remain localized, and they can become ‘viral.’ 

The major element for success is a broader understanding of the overall security environment and awareness that it probably is subject to continuing shifts.  Environmental sensing may be the most important tool in trying to track the trajectory of armed groups.  The term ‘environmental sensing’ is used quite deliberately.  A focus on ‘red’ and ‘white’ groups or ‘government controlled’ or ‘opposition controlled’ areas likely will inevitably lead to faulty analysis and consequent strategic failure.  Instead, the stress should be on varying levels of potential instability and the multiple groups (with multiple motivations) that can impact this.

The ‘government controlled-opposition controlled’ dichotomy also can lead to a false sense of progress.  It is all too easy to view areas in which there are no active insurgent groups or operations as being ‘government controlled.’  Competing power centers and groups with no particular allegiance to either the government or the insurgents can also be key players.  As such, these areas may be as important to long-term strategic goals as active insurgent areas.  This issue is closely connected with the fallacy of the term “ungoverned areas” which is common among analysts.  Where there are people, there will be some form of governance.  “Alternative governance” is a much more realistic concept.  The variety of groups that do not fit into the standard strategic templates certainly can provide their version of governance either at local or broader areas that must be factored into analysis of the operational environment.

The other aspect of environmental sensing is that intervention forces themselves become part of the environment.  Even if forces avoid the ‘us versus them’ syndrome, it is easy to overlook the ‘us and them’ reality.  By their very presence, intervention forces become enmeshed in the complexities of the local environment.  Analysis has tended to revolve around the effects of governmental or intervention force operations, with much less thought given to the impacts of the simple presence of these forces.  This is one reason why some tools such as Human Terrain Teams and other forms of cultural analysis of operational theaters have experienced problems:  they usually rely on historical patterns of what the society was like before the imposition of fresh actors.   The presence of intervention forces can (and usually does) change the dynamics of various power centers and armed groups of whatever stripe.  In a real sense, any network analysis of armed groups in an operational theater must include intervention forces as yet another actor.

The complexities within and among the various groups present (and emerging) in environments facing internal security threats almost certainly will continue to plague strategists, whether among the governments directly facing these threats or by intervention forces.  Unless analysts and planners broaden their focus to incorporate a broader picture of competing power centers to the government, long-term stability and success are very unlikely.  Successful strategy will rely on incorporating analysis both on interactions among a broader range of actors and their changes over time.

Categories: complexity theory - COIN

About the Author(s)

Lawrence Cline, PhD, is a contract instructor with the Counterterrorism Fellowship Program, Center for Civil-Military Relations, Naval Postgraduate School.  He is a retired US Army Military Intelligence officer and Middle Eastern Foreign Area Officer, with service in Lebanon, El Salvador, Desert Storm, Somalia, and Operation IRAQI FREEDOM.



Wed, 03/21/2012 - 11:07am

Good stuff. I've been working up a paper on Complex Interdependency (CI) theory as it relates to current ops. If you haven't already please see Rober Keohane/Joseph Nye's works re. same.

I firmly believe that we (DoD, et. al) suffer from the observer effect as well.