People living and working on complex systems, which is pretty much all of us, find ourselves baffled and inspired in equal measures by their unpredictable behavior. Complex systems, be they storm systems (environmental), the endocrine system (biological), or the dancefloor at your office Christmas party (social), can be impossible to predict, let alone control. As thinkers such as Easterly and Taleb argue, we should treat with great scepticism anyone who tells you that they can.
Our theoretical and informal models to understand them are all simplifications of reality, and can be misleading when not used cautiously. As the number of factors that influence the behavior of any system increases, the more complex its ontology becomes, and the less closely our models represent reality. The further we get from simplicity, the weaker our diagnostic ability becomes, and the blunter (and potentially counterproductive) our tools for intervention. Just think Syria, or the financial crisis, or any relationship you've been in, particularly those with 'complicating factors'.
So what can systems thinking offer us in the face of complexity that other models don't? Systems thinking is more of a conceptual tool box than a model. The advantage of its concepts, when used humbly and in conjunction with other ideas and methodologies, is to better appreciate the non-linear nature of reality. This helps us think and plan more realistically, instead of assuming that change follows linear cause and effect rules (see the primer). The world is round; we should stop pretending it's flat.
Systems thinking also offers practical tools, particularly systems mapping, which is a visual thinking process to consider and model the causality of a given system. It allows us to delineate causal factors and their interdependencies, identify what actions have highest potential (or are potentially futile) in effecting change, and predict both intended and unintended consequences of our actions. It's used by researchers and practitioners alike for analysis and as a tool to stimulate better informed planning.
A primer in systems mapping (and thinking)
At the heart of systems mapping is the question of why. Why did C happen? Was it because of A, or B, or A and B? What could be stopping C from happening? How are these factors related? In systems speak, it's a means of analysing (and visualising) a system to discern the causal relationships between these factors and the emergent properties their interaction gives rise to. A system can be anything set within defined boundaries. Most examples provided here relate to social systems. The emergent properties of complex systems are outcomes of system behaviour that are 'more than the sum of their parts' insofar as they are not caused by simple cause and effect relationships, but emerge from the complex interaction of many factors.
Many of the relationships between causal factors and system properties are non-linear, meaning that they don't follow the immediate cause and effect relationships that many of our mental and theoretical models often assume. Some factors and relationships cause more of themselves, such as the weapons stocks of two competing countries caught in an arms race (reinforcing feedback). Other factors balance each other out or help to keep a system in equilibrium, like the countervailing effects of heating and cooling on temperature in a thermostat-controlled room (balancing feedback). Time delays and disproportionality between causes and effects are common. The so-called Butterfly effect is an example of the latter, borrowed from chaos theory. Peter Coleman et al introduce the concept of attractors, which help explain how complex social systems stabilise with particular dynamics.
Systems maps can be created by one or a few analysts by collecting data, deducing the causal relationships between factors that govern system behavior, and visualising this interpretation. Quite sophisticated maps can be developed based upon traditional data collection methods and specialized approaches such as chronology of events mapping. Important dynamics within maps can be identified, allowing us to theorize the key patterns that make systems resistant to change, or potential leverage points where effort could be targeted to achieve desired results. Kumu has bought this work online, making it cleaner, more powerful, and easier to collaborate on.
Open source your analysis: participatory systems mapping
Increasing collaboration in any analysis and planning endeavor is useful. Two (or more) heads are usually better than one when you are faced with complex problems, as long as you can manage the tensions that multiple viewpoints create. By having more perspectives involved in the articulation of a system's dynamics we guard against risks that analyses are subjective or incomplete. As Rumi's fable of the blind men and the elephant alludes to, it's foolish to think that one person (or a group of people sharing similar perspectives) can grasp the complex totality of any given situation. Instead of pretending we can, including more stakeholders in your mapping process helps to develop more objective, holistic understanding of system dynamics.
Participatory mapping processes are by intention an inclusive process of shared knowledge creation. To align people’s thinking contrasting perspectives sometimes need to be disrupted first. This is achieved by introducing additional complexity into peoples' understanding of why things happen. It often generates disagreements, which should be expected because sometimes things don't happen in the same ways for the same reasons, or because people have adopted different narratives to render complexity comprehensible. If people are open minded and the evidence is compelling enough, participatory mapping processes expose new information that destabilizes these narratives, opening the opportunity for new and shared understandings to bed in.
This is a powerful tool for resolving disagreement and building a foundation for collective action. Participatory analytical and planning processes allow people a higher degree of autonomy than what is typically afforded in hierarchical social systems, unlocking the intrinsic motivation that compels people to most effectively implement change strategies. Even when we can coerce people to agree with our reasoning, like when we order them to (military) or pay them to (business), we typically achieve better results when people really believe in what they are doing.
Working with conflict-affected communities in remote areas in South East Asia does not lend itself to high tech approaches, so our mapping is typically done on large sheets of paper with marker pens and post it notes. There’s no reason that participatory approaches can’t combine offline processes and discussions with online tools for discussion and visualisation, which is where Kumu offers advantages.
How to conduct a participatory mapping process
A complete guide to data collecting and participatory mapping for systemic action research can be found in Danny Burns's recent book. This methodology goes beyond analysis to include action, which is often of more interest to practitioners than researchers.
The short of participatory mappings is to:
1. Select a diverse group of stakeholders. These should be people who should have a stake, experience, and/or direct knowledge of the system being analyzed, and represent as many 'parts' of the system (or value chain, or causal chain..) as possible. They should be available to participate in several iterations of data collection, mapping, discussion, and action. New participants might be added to the group as the research and mapping evolves because they have needed expertise or abilities to implement the emerging plan.
2. Collect your data. This can be drawn from the same sources as traditional research. In systemic action research, data is drawn from the experience of people living in the systems we are trying to understand in the forms of personal perspectives, stories and recollections. These are recorded in detail during a 'systemic inquiry' process.
3. The mapping process uses the same people who have collected the data. Data is mapped to identify and visualize the specific factors (elements) that influence how the system behaves and to articulate the causal links between these elements. Links between elements can be understood as causal chains that may exist in the present, past or future. Associated factual information or references to actors in that system (people or organizations) is included on the maps, which can cover entire walls. The emphasis is on broadening participants' thinking about what influences a given situation, rather than getting the analysis perfect. Participants should embrace the complexity of the system, adding as much detail into their maps as possible.
4. Once the initial 'big picture mapping' in completed, attention can turn to the identifying important causal patterns, which might be subjects of clarifying research or re-mapped in isolation. These patterns offer unique insights into why that system behaves the way it does, or what we could be done to enact a desired change in system behavior.
These patterns might be system archetypes, which are causal dynamics that can explain system dysfunction. Or they might be more basic, such as reinforcing or balancing feedback loops, which are causal dynamics within a system that serve to change or maintain or change its properties. Sometimes we might notice unintended consequences of well-intentioned actions.
5. Plan collective actions. Actions to change the system are selected and elaborated by the group based upon discussions around leverage points - typically system elements or relationships of system elements where action can be focused to realize the desired change. As Donella Meadows description alludes to, some of these may be visible within the map (e.g. information flows or feedback loops), though some of the most powerful may not be (e.g. rules, goals or the paradigm that defines the system).
6. Take action. Participatory processes are often the basis for collective action taken by the group. After taken action, maps might be revisited or mapped one more to incorporate data gathered through the experience of implementation.
We would love to hear about your application of these tools. Please send any questions or feedback to firstname.lastname@example.org