By
Andrew Peterson
September, 2016
Introduction
One of the central challenges of deciding how to address intractable conflict is to understand how to respond to their dynamics and complexity. Rather than being surprised by the intricate and quickly changing dynamics of a problem, or being overwhelmed by the interaction of different elements of the conflict system, we need some way of directly approaching that complexity and taking advantage of it to discover the leverage points from which the conflict system can be transformed. This is precisely what it means to engage in strategic peacebuilding [1]. Similarly, any potential intervention must be examined to determine what impact it will have on the social system. That information should then be used to design an intervention that is robust, sustainable, and able to adapt to further system changes. This article proposes one tool for trying to achieve these aims by modeling various forms of social systems, organizations, and processes, and provides some concrete examples of how this can be useful in responding to intractable conflicts.
Why model systems?
Conflicts frequently involve many different component parts, including people, institutions, resources, and so on, that interact in complex ways. These interactions can, for example, generate self-reinforcing systems, or give rise to multiple pathways that produce a given result. So it can be important to understand these processes, to visualize them and learn how they may change over time. Systems modeling has been done in the natural sciences for many years, [2] and is increasingly used in the social sciences, [3] but the basic process is also useful for everyday practitioners who seek to represent and understand the interactions of the factors they identify in a conflict mapping process. Although any such model is necessarily incomplete, it can nonetheless be a lens for looking at conflict. While it is important to understand conflicts as complex adaptive systems rather than complicated deterministic systems, a self-reflexive process of modeling such systems can help to identify the process structures that develop and to draw out the dynamic nature of such systems. Imagine, for example, that you want to explore what might be the most critical points from which to address an intractable conflict, or to understand your organization's role in a complex society and web of other organizations. System modeling is a versatile tool that can incorporate diverse strategies for use in different contexts and for different purposes.
What are Systems Models?
A system is a set of elements that relate to each other in some manner. The elements of a system can be objects, people, organizations, processes, descriptions or even ideas. The relationships between these elements can include different kinds of influence, flows of information, resources, associations, temporal relationships, or origins. Models of systems therefore try to capture these relationships in a way that gives a perspective on how the system as a whole interacts. Some of the different models that might be useful for conflict transformation include:
1. Causal models that can be used to understand how a country, or conflict, or other social system functions, particularly, what factors (or independent variables) cause change in other factors (dependent variables). Factors that are often considered as variables include the nature of changes made (e.g., increasing communication between disputing parties, undertaking joint projects, dialogue groups) and structural factors (levels of income, education, etc.) that are not a focus of intervention. Causal models are distinguished by their inclusion of variables, that is, factors that can take on different values. For example, a causal model might consider how the level of violence is affected by income and the extent of civil society. In order to be specific, however, it might focus on variables to represent these relationships, such as the number of murders in a community, income per capita, and the number of peacebuilding NGOs in operation in an area.
2. Human network maps to trace the relationships between disputants/stakeholders, spoilers, external supporters, peacebuilders, or factions of political organizations or armed groups and the ways these relationships influence the ways those actors operate.
3. Organization models that relate functional or other parts of an organization to each other and the overall purpose of the organization or the system as a whole. These can include models of processes or activities [4] or flows of information, personnel, or materials, and may suggest ways that the organization acts coherently or in conflicting fashions. It might also consider how programs have evolved over time, or provide a forward-looking perspective to organize how a project is designed to work. One specific example is the mapping of the social space of conflict for the purpose of designing a project to weave key social actors into a flexible, resilient network like a spider's web [5].
4. Decision trees are a type of system model with a specific linear flow that can be useful in analyzing the possible combinations of sequences of decisions, or perhaps the sequential actions of individuals or groups on different sides of a conflict.
5. Conceptual models can relate ideas and ways of thinking to each other [6].
6. There are no doubt many other possibilities, including the potential to relate the spatial dimension of a model to maps of physical spaces, or even graphical information system (GIS) maps.
Ideas for Exploring the Modeling of Systems
Ask five people in a room to create a model of a conflict and you will get six (or more) models. Rather than seeing this as a defect, it suggests that the activity can form a creative brainstorming activity as part of a collective envisioning process or participatory planning process that generates conversations about how people see things differently. The emphasis is perhaps not so much on doing it 'right,' but experimenting with the process, repeating it in new ways, with a new focus, or for example, by trying to create the model as one might imagine it to look from the perspective of the parties involved. Needless to say, the product is not static, but designed to be re-created periodically as the situation and one's understanding changes, and is meant not to be a test of understanding so much as a tool to "make complexity a friend, not a foe" [7]. This reflects the difference between complicated and complex adaptive systems.
All models leave something out, including the best models possible. The challenge is to be comprehensive in perspective, in the sense of having considered all of the possible elements, but not necessarily including them in the model. Good models are "parsimonious," that is, they try to capture an important set of relationships without being distracted by less important details. The key to the process of being selective is to think of the process, not as creating a shopping list of everything that might be included, but as relating key elements on the basis of an understanding of what is going on. For example a modeler might try to uderstand and model the underlying causes of conflict, or try to map out what causes conflicts systems to change in particular desired ways (sometimes called theories of change).
One way to understand how to keep things simple is to keep in mind the idea of different levels of analysis, from the broadest overall perspective that misses details to the most specific focus on details that neglects the big picture. Each of these levels of analysis is not superior as such, but is, instead, more or less useful for a given purpose. A precise drawing of a cell would not help you understand nuclear fission, any more than a model of an atom would be useful in understanding how cells divide. And neither would be particularly useful to understand global warming. Similarly, a model of how conflict functions in a family would not be very useful for understanding how conflict functions in a society, except perhaps by analogy. Other ways for breaking apart the conceptual space of a model include John Paul Lederach's distinction of four dimensions through which conflict has influence and can be affected: personal, relational, cultural, and structural, each of which might be conducive to being modeled separately.
The fact that some elements are not included should, however, not mean that they are forgotten completely. It is important to keep in mind and occasionally reflect on relevant elements that were not included, and their reasons for not being incorporated. At its worst, after all, modeling can create simplistic and myopic views that are reductionist, that seek naively to escape from complexity, rather than providing a means for engaging with it. Like the maps that were used by colonial powers, systems models are representations, and they have a certain tendency to ignore complexity and simplify cross-cutting elements as instead self-contained and monolithic. Once models are created, if they are treated as if they were objectively real, they tend to try to make the world conform to themselves rather than the other way around.
One concrete way that this happens is that we tend to give greater weight to elements which we can picture easily, while ignoring those that might have equal weight but are difficult to represent. For example, a diagram of factors influencing poverty might face the difficulty of either including gender relationships as an element, which makes it seem important only in a local way, or not including gender as an element because it is instead an intrinsic part of other relationships, such as land distribution, job opportunities, education, and so on.
Another aspect that is not necessarily easy to picture is the context in which a model operates. This can include social and cultural norms that govern the nature of relationships in the model, long-term historical trends, or simply an overall analysis of the conflict [8]. Reflecting the potential of complex systems to change, it is important to try to understand how a given model evolves and adapts over time, and how it thereby changes its relationships to these contextual factors. Finally, while a model reflects what is happening at one moment in time, this picture needs also to be related to past relationships from which it evolved and the future relationships into which it may progress. For example, a model might reflect merely one moment in a series of conflict stages. It might show a conflict intensifying, as various attributes reinforce eather other to create a conflict spiral, (escalation) and polarization. So again, care must be taken to think of models not as comprehensive, static pictures, but rather as jumping off points from which further analysis can be undertaken.
Learning from Systems Models
As the above discussion illustrates, the process of creating models itself is a central learning process. There are also specific techniques for analyzing systems models to explicitly learn from their structure. The nature of systems, in which relationships may create mutually-influencing loops, or may contain different pathways that duplicate and reinforce each other, or act in a contradictory (or self-canceling) manner, create dynamics that need to be understood in a holistic way. For example, imagine a program aimed at ending corrupt practices that are interfering with good governance and creating conflict between ethnic groups that see these practices as discriminatory. A narrow, straightforward approach might seek simply to identify the corrupt individuals and have them prosecuted, clearing the way for honest individuals. Such an approach will not work, however, if corrupt individuals make up only one element in a mutually-reinforcing system of factors (Diagram A). Perhaps the approach succeeds in its direct aims, but the structure that gave rise to that situation initially will simply recreate the problem with other individuals. For example, if the justice system fails to provide checks on the development of corruption, a regime may well elect new officials who are also corrupt. Or the public may be unable (or unwilling) to vote out officials who have become corrupt after taking office. Further, as politicians may seek to exploit ethnic identities for political gain turn to corruption as a means of maintaining group cohesion and loyalty.
A real solution involves an integrated approach to resolving the interconnected network of structures, actors, and processes that produce a given outcome. One step of this process can be to identify "points of intervention" that are most susceptible to leveraging change [9]. Critical points could be those that are by themselves necessary and /or sufficient to produce a given outcome. A systematic approach might, however, require intervention at not just one critical point, but rather at multiple points simultaneously, as suggested in the corruption example. The envisioning process in this case might involve creating a virtuous cycle that ties together the relevant elements in a novel, counteracting way. Thus an anti-corruption campaign might seek to create a coalition of political candidates, jurists, public officials, and concerned civil society groups. As this campaign gains strength, people may see joining it as a path towards greater power, and thus react quite differently than if they see it merely as a threat. Simultaneously, the institutional rules and norms that facilitated the old habits come to change to make corruption more difficult and less rewarding (Diagram B).
Since the system described in a given model is embedded in and intertwined with other systems, it is worth also considering the possibility of "spillover effects" in related systems [10]. This suggests that solutions focusing on one narrow outcome, say for example, income sources for demobilized soldiers, may have negative or positive effects on other systems, such as damaging the environment, displacing other workers, or (more optimistically) providing resources that benefit another segment of society. The impact of a given intervention must be understood, then, by reference to the broad context in which its impacts are felt.
This brings up a final role that systems models can play, which is in evaluation. Reflecting on models created early in a project can be compared to how relationships turned out in real life, thus serve as a way of judging the accuracy or biases in one's analysis. Systems models could also serve to more directly evaluate a program by mapping the relationships within an organization or society that took place over the time of the program. Regardless how such systems are used, it is useful to keep in mind the limitations of the approach, the context and contingency of any model, and responding to these challenges creatively and conscientiously. In one sense, modeling processes of conflict can be a highly technical skill, full of controversy that is the focus of disputes in academia over how to understand conflict. While these debates are enriching, one does not to be versed in them to begin to use the modeling processes presented here to better understand the systematic nature of a conflict. But it is worth remind oneself to always look for deeper understandings that can help one to more effectively assess needs and design solutions, a process to which systems modeling can become a natural and everyday part.
[1] John Paul Lederach and R. Scott Appleby. "Strategic Peacebuilding: An Overview" In Strategies of Peace. (Oxford: Oxford University Press: 2010), 22.
[2] See, for example, George J. Klir. An Approach to General Systems Theory. (New York: Van Nostrand Reinhold Company, 1969).
[3] In making the transition from positivist, reductionist understandings of the traditional self-understanding of the natural sciences, a distinction was made by suggesting that the emergent sytems of the social world should be characterized as "soft systems," a term that is useful in for seeking further research in this area. Additional resources can also be found in under the heading of 'systems engineering' that is used by governments, corporations and other organizations to organize practical activities and programs. See Peter Checkland and Jim Scholes. Soft Systems Methodology in Action. (Chicago: John Wiley Sons, 1990) and John P. van Gigch. System Design Modeling and Metamodeling. (New York: Plenum Press, 1991). For a more recent overview, see John H. Miller and Scott E. Page. Complex Adaptive Systems: An Introduction to Computational Models of Social Life. (Princeton: Princeton University Press, 2007).
[4] Peter Checkland and Jim Scholes. Soft Systems Methodology in Action. (Chicago: John Wiley Sons, 1990), 106.
[5] John Paul Lederach. The Moral Imagination: The Art and Soul of Building Peace. (Oxford: Oxford University Press, 2005), 82.
[6] Brian Wilson. Systems: Concepts, Methodologies, and Applications, 2nd ed. (Chichester: John Wiley & Sons, 1990), 12.
[7] Michelle Maiese and John Paul Lederach, "Conflict Transformation" in Beyond Intractability.
[8] "Reflecting on Peace Practice: Participant Training Manuel" CDA Collaborative Learning Projects. March 2009, p.6. Available online at: http://www.cdainc.com/
[9] Ibid., p. 8.
[10] John P. van Gigch. System Design Modeling and Metamodeling. (New York: Plenum Press, 1991), 38.
