Intelligence Preparation of the Urban Environment:
Adopting a Complex Adaptive Systems Approach

by Captain Colin Papuschak

Introduction

Military operations in urban environments present unique challenges that demand a comprehensive understanding of their interconnected components. As a result of the density, dynamism and interconnectedness of data points, the core components of the urban environment— human terrain, physical terrain and infrastructure—interact in seemingly unpredictable ways that often frustrate military forces. Peacekeeping operations in Port-au-Prince and Mogadishu, counter-insurgency operations in Fallujah and Marawi and major combat operations in Bakhmut and Sievierodonetsk are poignant examples of urban environments that posed enormous challenges to all combatants. In the 2015 Canada’s Future Army series, the Canadian Army Land Warfare Centre (CALWC) conceived that rapid urbanization compounds poverty and insecurity of health care, water and energy, especially in developing regions, increasing tensions and the prospect for armed conflict.^{Footnote 1}  Close Engagement: Land Power in an Age of Uncertainty reinforced this concern, stating that “not only are urban operations expected to become more frequent, but they may well pose the highest degree of challenge owing to their human, environment, geographic and physical complexities.”^{Footnote 2} 

To navigate this complex landscape, intelligence preparation takes on paramount importance. More than any other environment, urban operations demand intelligence with greater targeting precision to discriminate between adversarial actors and innocent civilians and to reduce collateral damage. This is partly due to the sheer density of urban populations and the tactics employed by some adversaries to use civilian populations as camouflage and concealment. However, the ability to deliver greater targeting precision is exasperated by the complexity of the urban environment.

The urban environment must be analyzed as a complex adaptive system with an approach responsive to the demands of a given campaign theme. While CFJP 2-1.1 Intelligence Preparation of the Operational Environment promotes taking a “systems perspective” of the operating environment, this article contends that this perspective must be taken a step further in urban environments, moving beyond a general systems theory to complexity theory for deeper understanding.

Understanding the Urban Environment

Many attempts have been made to better understand the nature of urban environments, often turning to metaphors and natural sciences to make sense of their deep complexities.^{Footnote 3}  Cities have often been described as “complex adaptive systems,” and this article advocates including this concept in the intelligence preparation of the urban environment.^{Footnote 4}  Complex adaptive systems are complex with a dynamic network of interactions. However, the behaviour of the ensemble may not be predictable according to the conduct of the individual components.^{Footnote 5} They are adaptive, which implies that the individual and collective behaviour mutate and self-organize, thus corresponding to a change-initiating micro-event or collection of events.^{Footnote 6} 

This is not the first effort to employ complexity theory and complex adaptive systems for intelligence preparation of the operating environment. Tom Pike, Eddie Brown, and Piotr Zagorowski have developed and advocated for a “Complex Intelligence Preparation of the Battlefield (IPB).”^{Footnote 7} They employ fitness landscapes and agent-based modelling to better understand how the choices available to actors within the social–political–economic system (or fitness landscape) are constrained by the system and how each actor employs their fitness functions (or competitive advantage) when interacting with their social–political–economic landscape. Complex IPB also restructures the traditional four-step IPB process into a seven-step process. There is merit in this proposal, but if Pike’s contribution of “Complex IPB” is regarded as disruptive to well-established military planning processes, this article represents a more evolutionary approach to integrating complex adaptive systems into the existing process of intelligence preparation of the operating environment.

Complexity Theory and Complex Adaptive Systems

There is a notable difference between the “systems perspective” proposed in CFJP 2-1.1 and the complexity theory from which the idea of complex adaptive systems is derived. CFJP 2-1.1 describes a system as “an interconnected or interrelated network, group, or chain of functionally, physically, and/or behaviorally related interdependent elements that form a functional whole.”^{Footnote 8}  While not explicitly stated, this description leans heavily on general systems theory, a reductionist framework that focuses on understanding systems by breaking them down into their individual parts and studying how these parts interact with each other.^{Footnote 9}   Where CFJP 2-1.1 does describe systems as “complex,” it still does so in terms of general systems theory.^{Footnote 10} 

Like general systems theory, reductionist frameworks are regarded by some as inadequate due to the complexities of the modern world. Complexity theory “expands on the reductionist framework by not only understanding the parts that contribute to the whole but by understanding how each part interacts with all other parts and emerges into a new entity, thus having a more comprehensive and complete understanding of the whole.”^{Footnote 11}  In other words, complexity theory considers the emergent properties of a system that arise from the interactions between its parts and offers complex adaptive systems as a means for understanding systems like these.^{Footnote 12}  This approach is better suited for understanding systems like stock markets, ecosystems, and cities that exhibit unpredictable behaviour and cannot be fully understood by studying their individual parts in isolation.

Complex adaptive systems generally refer to dynamic, open systems that can self-organize their configuration by exchanging information, energy, and other resources within their environment can transform these resources. Organic interactions within and between systems occur as the system’s components learn to adapt, making the system very dynamic.^{Footnote 13}  Complex adaptive systems tend to transform into new stable states—a characteristic known as emergence—and any description of the system cannot be objective, complete or permanent.^{Footnote 14} 

Haiti’s tumultuous capital city, Port-au-Prince, offers a ready example of the urban environment as a complex adaptive system. The city faces unique and serious problems—an estimated 80% of the city is under the control of gangs that have grown significantly stronger in recent years by establishing alliances with other armed groups, creating some seven major gang coalitions and approximately 200 affiliated groups, according to a recent United Nations report.^{Footnote 15}  Increasingly, there are fears of an emerging civil war during an enduring political crisis since the assassination of President Jovenel Moïse in 2021.^{Footnote 16}  In the sections below, Port-au-Prince will serve as a case study for understanding the characteristics of complex adaptive systems that are relevant to intelligence preparation of the urban environment.

Emergence

Emergence refers to the arising of new properties or characteristics in a system that cannot be reduced to the sum of its parts. It is a phenomenon where new patterns, structures or behaviours arise from the interactions between the components of a complex system. These interactions are self-organizing with no external direction leading them. These emergent properties are often unpredictable and cannot be understood by simply analyzing the system’s individual components. Instead, they result from the collective behaviour of the system as a whole.^{Footnote 17} 

The concepts of emergence and self-organization are evident in the case of Port-au-Prince.^{Footnote 18}  There has long been a relationship between Haiti’s political elite and the country’s non-state armed actors. The seeds of this equation emerged as early as 1959 when President Dr. François “Papa Doc” Duvalier, deeply mistrustful of the military due to an attempted coup the year prior, organized a private militia known as Tontons Macoutes (“Boogeymen”), which maintained order by terrorizing the populace and political opponents. The Tontons Macoutes reportedly murdered between 30,000 and 60,000 Haitians. Countless more were sexually and physically assaulted and tortured with their violent excesses outliving the Duvalier dynasty’s demise in 1986.^{Footnote 19} 

After being removed by a military coup in 1991, then-President Jean-Bertrand Aristide returned to power in 1994, disbanded the military and began arming his political supporters in the slums of Port-au-Prince. Haiti’s political and economic elite frequently turned to street gangs to intimidate their opponents, offering them financial support, weapons and protection from law enforcement in return for their loyalty. But these relationships appear to have ruptured after the 2021 assassination of Haitian President Jovenel Moïse, the erstwhile benefactor of the infamous G9 family of gangs led by Jimmy “Barbecue” Chérizier.^{Footnote 20} 

Chérizier was not content to find a new political sponsor after Moïse’s death. He preferred to upend the established order by attempting to assert dominance over the political elite and rival gangs. He carved out yet more territory and eventually controlled Port-au-Prince’s port facilities that were used to extort politicians by throttling port activity.^{Footnote 21}  By controlling the port facilities, Chérizier’s G9 no longer relies on previous patron–client relationships with the political elite. A new stable state is emerging with new relationships, and Chérizier is attempting to position himself as the leader of a revolutionary movement.^{Footnote 22}  This example illustrates the self-organization of the system’s components in the absence of external direction and the emergence of new patterns of behaviour amongst these components, leading to a new stable state or equilibrium.

The gang activity in Haiti cannot be generalized so simply. Complex adaptive systems are not analyzed assuming linear causality or even multiple causality. They employ a connectionist perspective wherein the interactions among and between elements are viewed along with the system’s adaptiveness to environmental forces.^{Footnote 23} 

Adaptiveness

Adaptiveness refers to a system’s ability to change its behaviour or structure in response to changes in its environment or internal conditions. This allows a system to cope with uncertainty, complexity and unpredictability and to evolve into new states or configurations. Adaptiveness is also related to learning, as a system can use feedback and information from its interactions to modify its actions and expectations.

Recently, in Port-au-Prince, vigilante attacks by residents targeting suspected gang members have grown into a movement named Bwa Kale, Haitian Creole for “peeled wood,” a crude metaphor for phallus, insinuating merciless male dominance. Increasingly frequent vigilante attacks suggest internal adaptation within the system in response to the dangerous uncertainty posed by gang violence.^{Footnote 24}  That noted, Haiti has a long history of vigilantism, and it would be misleading to characterize the current Bwa Kale movement as something new. In contrast to previous movements, it is unique in terms of the environmental conditions that led to its emergence.

As Haiti’s political crisis continues with no legitimate legislative or executive leadership and no prospect for democratic elections, the actors within the system are adapting to a new political landscape where gangs are the de facto authority in much of the country. The state’s security apparatus, embodied in the Police Nationale d’Haiti, itself beset with corruption, is one of many armed actors and arguably not the strongest.^{Footnote 25}  It would be misguided to characterize this system as “ungoverned,” as is sometimes the tendency when describing “feral” cities and failed states (more on this terminology below). Even where the central government is incapable of providing public goods, local governance is provided through a complex mix of informal structures provided by gangs, vigilante groups and other local and international civil society actors providing social services and security.^{Footnote 26} 

The Port-au-Prince urban system has adapted to the ongoing political, social and economic crises that have gripped the nation. One can reasonably anticipate that any foreign military intervention in the city would result in new adaptations, whether a peace support operation or a non-combatant evacuation operation. This is largely because the system’s components react to the intervention of a new agent and the behaviour of the system adapts. However, because of the nonlinear nature of complex adaptive systems, it would be challenging to predict the exact contours of these adaptations leading to the next characteristic of complex adaptive systems.

Nonlinearity

Nonlinear dynamics is a branch of mathematics that studies systems whose behaviour is not directly proportional to the input.^{Footnote 27} Small changes in the initial conditions can result in large and often unpredictable changes in the system’s behaviour. Nonlinear systems are often complex and can exhibit a wide range of behaviours, including chaos, bifurcations and multiple stable states.

The decision of some Port-au-Prince residents in April 2023 to attack and execute suspected gang members on their own appears to be setting in motion greater momentum of vigilantism in the city. While it is difficult to predict the evolution of the Bwa Kale movement and its effects, a small change in the system represented by the April attack may yield large changes in the system’s future behaviour. Some regional experts are already concerned that the movement may expand into criminal economies.^{Footnote 28} 

One potential way of mapping nonlinear dynamics in urban environments is to study second- and third-order effects. As Michael Miller observes, “While we may not be able to predict the cause-effect behaviour of inter-related complex systems with precise certainty, we can try to understand the nature of the elements that will interact.”^{Footnote 29} Mathematical models and simulations can help solve problems of causality in complex adaptive systems, but applying these models to intelligence preparation and operational planning can be problematic. Often there is just not enough time to establish proper conditions and define the variables. Miller proposes a systematic methodology for mapping cause-effect chains (Figure 1).

This linear model is simple enough to employ in time-constrained situations like major combat operations to support branch and contingency planning and can be expanded where the campaign theme requires deeper analysis. As noted above, complex adaptive systems are not analyzed assuming linear causality or even multiple causalities. Still, for intelligence preparation and operational planning, linear causality is a necessary first step in the analysis. Afterwards, an expanded cause-effect chain might begin to look something like a cause-effect web (Figure 2), where feedback loops and disproportionality characterize causal relationships:

Several probabilistic causation models can be employed to represent causal relationships in a system and make predictions about the behaviour of a system.^{Footnote 30}  Given the time and expertise required to run these models, they are more suitable for peace support and counter-insurgency operations with a robust intelligence enterprise.

Often described as action-reaction cycles, nonlinearity and feedback loops characterize all human interactions. The rising Bwa Kale movement in Port-au-Prince is caused by citizen frustration with gang violence and the police’s inability or unwillingness to address it effectively. An outcome of executing suspected gang members is the immediate drop in reported kidnappings and killings attributed to gangs in some communities.^{Footnote 31}  A second outcome, not yet present but possible under the right conditions, is increased gang retribution. A feedback loop would emerge when gangs react to attacks by targeting members of the Bwa Kale movement, which will affect how the vigilante movement responds in the future. This kind of bi-directionality can potentially create a self-reinforcing feedback loop, where causal interactions reinforce and perpetuate a certain trajectory—in this case, continued violence between gangs and vigilante groups.^{Footnote 32}  

Components of the Urban Environment

The urban environment is often understood as a triad of human terrain, physical terrain and infrastructure. Complexity theory lends itself to analyzing human terrain and infrastructure. Despite the crucial importance of understanding the physical terrain of an urban environment, complexity theory does not offer a pathway for improving analysis, despite shortfalls in the doctrinal approach to urban terrain analysis.^{Footnote 33}  The application of complexity theory to the analysis of the information and cyber domains is more likely to render a better analysis. However, these domains are geographically agnostic and not necessarily unique to the urban environment. Despite being an inviting avenue for future research, this discussion would be too broad and detailed for this article. The following sections will discuss the employment of complex adaptive systems for understanding an urban environment’s human terrain and infrastructure.

Human Terrain

A unique feature of urban environments is the density and diversity of human terrain that results in complex interconnections within the city itself and beyond (i.e. diaspora communities). Not only are residents part of the human terrain, but also foreign non-governmental organizations and intergovernmental organizations. A crucial task for the intelligence function is to discriminate between hostile actors and innocent civilians on the battlefield.

The ASCOPE-PMESII^{Footnote 34}  matrix is commonly employed as a method of human terrain analysis for counter-insurgency and peace support operations. However, as the pendulum has swung toward preparation for major combat operations, this tool is no longer taught in detail in officer or non-commissioned member (NCM) training.^{Footnote 35}  That noted, the deductions derived from an ASCOPE-PMESII matrix are of substantial military value in urban environments. Still, as depicted in CFJP 2-1.1, the ASCOPE-PMESII matrix has critical shortfalls that must be addressed to be employed effectively.

The ASCOPE-PMESII matrix was developed by the United States Marine Corps (USMC) during the counter-insurgency campaigns in Iraq and Afghanistan to support civil affairs. The matrix is the coversheet of an often-overlooked estimate process. Doctrine writers in the Canadian Armed Forces declined to integrate this estimate process fully, instead opting to include an expanded PMESII to the ASCOPE map (Figure 3).^{Footnote 36}  If ASCOPE-PMESII was still taught in detail in officer and NCM training, this might not be a concern, but the result is a sub-optimal tool that encourages cataloguing information into discrete categories that fail to determine their interconnectedness rather than enable a systems approach.^{Footnote 37}

Figure 3: PMESII Political to ASCOPE Map, as depicted in CFJP 2-1.1 Intelligence Preparation of the Operating Environment.

If utilized as an estimate (Figure 4) in conjunction with network analysis, the data points identified in the matrix become factors for analysis in the estimate and nodes within a network. The analyst moves beyond simply cataloging and mapping salient factors of the urban environment to assessing their interconnectedness and impact on operations.

Figure 4: Example ASCOPE-PMESII estimate page for the Political Variable.

Returning to Port-au-Prince helps illustrate another important point about human terrain in the urban environment. Heretofore this article has painted a picture of a city blighted with chaotic criminal violence, a “feral” city without order. But this is not the whole story, and it leads to an inaccurate and distinctly negative view of the city’s inhabitants. This narrative also fails to consider that people often turn to gang membership to fulfill genuine security, and social and financial, needs.^{Footnote 38}  Intelligence preparation in urban environments must understand individual and group interests, the cultural, economic and security drivers of those interests, and how individuals and groups perceive their interests in relation to others.

To their credit, Canadian doctrine writers included in B-GL-323-004-FP-003, Counter-Insurgency Operations, the concept of a spectrum of relative interest to overcome the traditional “enemy–friendly–neutral” paradigm and the obfuscation it creates (Figure  5).^{Footnote 39}  Individuals and groups rarely fit neatly into such categories, and common descriptors like “pro-government” or “anti-government” fail to properly relate to the myriad interests that an individual or group may hold.

A methodical analysis of group interests leads to a better understanding of the complexities of the human terrain by determining a group’s primary and secondary interests. Primary interests are core to a group’s survival and might be considered “red flag” interests. For instance, if local groups cooperate with external forces primarily out of aligned motivation, then “red flag” interests would be those that, if crossed or otherwise violated, would mean that cooperation must cease no matter the benefits of cooperation or the alignment of objectives. The opposite could occur if a group’s primary interests are satisfied by friendly forces out of aligned motivation. Assessing and mapping these interests offer insights into points of vulnerability, such that an external force can avoid events that turn erstwhile allies into adversaries and gain cooperation from groups that might otherwise be classified as hostile.^{Footnote 40}

In Port-au-Prince, the Global Initiative Against Transnational Organized Crime estimates the number of active gangs at 95.^{Footnote 41}  Statistically, it is unlikely that each of these gangs would be an enemy to a foreign peace support intervention, nor would it be desirable for a foreign military force to find itself fighting all these groups. Understanding their interests provides an avenue to exploit common interests—even amongst groups that would otherwise be categorized as hostile—to achieve mission success. It also helps understand the areas where a military intervention might be able to find common ground with the various vigilante groups, several of whom already informally collaborate with state security forces in Haiti.^{Footnote 42}  Therefore, human intelligence and civil-military operations take on disproportionate importance for intelligence preparation of the urban environment.

CFJP 2-1.1 already provides helpful tools for understanding systems––network analysis chief among them. Invoking characteristics of complex adaptive systems—emergence, adaptiveness and nonlinearity—will help the analyst understand how relationships adapt and evolve by introducing a new element, such as a foreign military intervention.

Infrastructure

There are several ways to understand infrastructure in academia and in a military context. As an umbrella term, infrastructure can be conceived as physical and social. Physical infrastructure comprises the basic physical structures and facilities needed for the operation of a society––bridges, water treatment plants, electrical substations and the like. Without these structures, daily life and economic activity would halt. Social infrastructure is often understood as comprising the policies, resources and services that ensure people can participate in productive social and economic activities.^{Footnote 43} 

CFJP 2-1.1 identifies the importance of infrastructure analysis but does not sufficiently differentiate between infrastructure and services enabled by that infrastructure. It tends to conflate the two, arguing that “[a]n accurate portrayal of the infrastructure status will potentially prevent or help eliminate humanitarian crises.”^{Footnote 44}  While this is undoubtedly true, this interpretation of infrastructure does not acknowledge the systemic interconnectedness of components that enable services, of which (physical) infrastructure is just one.

For military analysis, and in the absence of a precise doctrinal definition, this article proposes a definition that can be integrated more easily into the language of complex adaptive systems. Infrastructure––physical structures like bridges, schools, hospitals and electrical substations––are one of several components that enable the provision of services that allow daily life and economic activity to occur. In this sense, the International Committee of the Red Cross (ICRC) makes a clear distinction between “critical infrastructure” and “essential services,” which is helpful for military analysis. Critical infrastructure is necessary for the functioning of an essential service whose damage or destruction significantly impacts the delivery of the service. Essential services are services indispensable to the survival of the population, such as water and sanitation, electricity, health care and solid-waste disposal.^{Footnote 45}

Expanding on the Red Cross’s definition, services may include not only essential services but also social services, such as education and childcare, economic services like unemployment insurance and worker’s compensation, or emergency services like fire and ambulance services.

All services, essential or otherwise, depend on the function of at least four interconnected components:^{Footnote 46}

• People (operators and technicians);
• Hardware (infrastructure and equipment);
• Consumables (raw materials and finance); and
• Regulation (policies, governance).

Invoking complex adaptive systems in the urban environment gives a basis for understanding the dynamic and adaptive interdependencies between these components, especially in conflict zones. None of these components are designed with urban warfare in mind. Social, political and economic dysfunction will disrupt the adequacy of these components for their demand. Even systems that are badly disrupted will adapt in unique ways for the service or some semblance of it to be delivered. Employing nonlinear causal analysis allows the analyst to understand how even a relatively minor disruption in any of these components can result in disproportionate impacts on the overall system.

Infrastructure is too narrowly understood in military analysis, and the ASCOPE-PMESII matrix by itself does not adequately reveal these interconnections. Adopting complex adaptive systems enables one to better acknowledge these components and how these components and systems will evolve and adapt, especially in the context of urban warfare.

Used properly and to understand relationships, the ASCOPE-PMESII matrix can reveal the myriad ways in which the human terrain interacts with physical infrastructure and the services it enables. Whether hostile or not, all population groups in an urban environment will use, exploit, defend or attack infrastructure and the services it enables. How these acts occur may seem imperceptible to a foreign force, highlighting the relationships and interdependencies between population groups and urban infrastructure. For example, in Port-au-Prince the G9’s occupation of the Varreux fuel terminal in 2022 offers an obvious example of how population groups can interact with a city’s infrastructure in evolving ways.^{Footnote 47} 

But infrastructure does not simply enable the provision of services––it also provides the foundation for economic activity whether legal, illegal, or in the gray space between. Various attempts have been made to represent these activities graphically. A popular method is to view these activities as flows (metabolic flows using the language of urban metabolism) of people, goods, capital and information.^{Footnote 48}  Mapping these flows geospatially and through network analysis diagrams, considering where they originate and terminate, why they exist, what nodes they use, how they are linked, who controls them, and who depends upon them, will reveal how they impact military operations. As described above, a causal analysis will show how military operations impact these flows.

Understanding Complex Adaptive Systems

At first glance, one disadvantage of implementing complex adaptive systems in intelligence analysis is the apparent contradiction between the responsibility to provide predictive analysis and the explicit inability of complexity theory to predict outcomes. Traditional correlation and cause-and-effect analysis do not adequately describe complex adaptive systems because of emergence and nonlinearity. In truth, analysts have never been able to provide absolute certainty in their assessments regardless of the analytical approach used. Implementing complex adaptive systems does not introduce more uncertainty into analysis. Instead, it provides a stronger theoretical basis to identify the sources of that uncertainty than the current “systems perspective” employed in doctrine. An advantage of implementing complex adaptive systems in the intelligence preparation of urban environments is overcoming the language of “feral cities.” Dr. Richard Norton coined the term in 2003 to describe an urban environment that 

“is now a vast collection of blighted buildings, an immense petri dish of both ancient and new diseases, a territory where the rule of law has long been replaced by near anarchy in which the only security available is that which is attained through brute power… [T]his city would still be globally connected. It would possess at least a modicum of commercial linkages, and some of its inhabitants would have access to the world’s most modern communication and computing technologies. It would, in effect, be a feral city.”^{Footnote 49}

The phrase has been employed to describe cities like Port-au-Prince or Mogadishu, where the rule of formal law is weak relative to informal governance structures, where social services have broken down and security in various forms is uneven at best. It describes the same conditions the CALWC does in Canada’s Future Army, where rapid urbanization compounds poverty and insecurity of health care, water and energy, especially in developing regions, increasing tensions and the prospect of armed conflict.

Dr. Norton acknowledges that his selection of the phrase is controversial and provocative, although the term “feral” has biological and zoological utility. However, common and uncareful application of the term to describe human populations risks imposing cultural and racial biases. Employing the language of complex adaptive systems, on the other hand, provides an analytical framework to describe the same phenomena without risking the implicit inclusion of cultural and racial biases that might undermine sound intelligence analysis.

Conclusion

The tools provided by CFJP 2-1.1, its systems perspective and network analysis, can be expanded upon with the help of complex adaptive systems to understand urban environments better. Expanding the use of certain technologies, including artificial intelligence, will help process the vast quantities of data that characterize urban operations and overwhelm analytical capabilities. It is already commonplace to invoke complex adaptive systems to describe the urban environment, but applying its characteristics to intelligence analysis—specifically emergence, adaptiveness and nonlinearity—can better support military operations. Effective targeting that discriminates between adversarial actors and innocent civilians and avoids collateral damage depends upon it.

About the Author

Captain Colin Papuschak joined the Canadian Armed Forces in 2009 as an infantry officer with the Loyal Edmonton Regiment, where he participated in numerous urban operations exercises. He has held various positions at the Western Area Training Centre and 3rd Canadian Division Headquarters. In 2020 Capt Papuschak transferred to the Regular Force as an intelligence officer and was previously posted to 3rd Battalion, Princess Patricia's Canadian Light Infantry (3 PPCLI), as the battalion's intelligence officer.

This article first appeared in the October, 2024 edition of Canadian Army Journal (21-1).