Today, I’ll be exploring the concept of “complexity” as a kind of synthesis of ideas from our discussions on functional behavior, systems thinking, emotion, intuition, and predictive modeling. Centrally, in the cross-disciplinary field of Complexity Theory, there is this idea of a “complex system” which describes things like organisms, ecosystems, societies, market economies, languages, or the internet. Human partnerships often seem to function in line with many of the principles of complex systems—at least when viewed at the behavioral level. To that extent, we may be able to usefully apply the insights of Complexity to our decision making in our own relationships. As engineers of our relational environments, we are inevitably faced with tasks of complex systems management, so it might be good to understand the mechanics of such systems and best practices when dealing with them.
Like the functional lens post, I’m intending this partly for its direct applicability, and partially as a theoretical foundation for further discussion—so just fair warning: there’s going to be even more explaining than usual.
Complexity
The abstract “quantity” of complexity is what I would be referring to if I asserted that “humans are more complex organisms than ants are.” That business of something being more or less complex represents something about how many different parts and kinds of parts a human has than an ant—in technical terms representing something like the amount of information necessary to fully describe the thing. Returning to the metaphor of evolution, it’s worth noting we generally understand evolution to have progressed towards increasing complexity over time, when we look at the complexity of the organisms present at any given point on the timeline.
As for defining a complex system, (to paraphrase Melanie Mitchell): a complex system is a particular kind of system composed of a network of components with no central control and simple principles of interaction, which give rise to complex collective behavior, sophisticated information processing, and adaptation (a mechanism of which learning and evolution are examples). Where systems are generally understood in mechanical terms (like the chained parts of an internal combustion engine)—complex systems are typically described as networks of interconnected nodes abstractly representing an intricately interwoven web of functional relationships between system components.
Properties of Complex Systems
Let’s make a working assumption that human relationships do represent “complex systems” on some meaningful level, and talk about what that might mean, before coming back to the question of what “level” that might be, exactly. I’m hoping this will be useful because complex systems work in some particular ways, so that we might know what to expect when interacting with them.
I’ll be going to go through a number of these “characteristics of complex systems,” while keeping in mind a few different questions: Do we think our relationships (and our minds, while we’re at it) display this characteristic? If so, how? At what “level of analysis?” If our relationships are this way, what does that mean for how we might go about making them work better?
- Nonlinear Determination: In complex systems, there are not “proportional” relationships between causes and effects. Popularly known as the “butterfly effect” in consideration of the complex system of global weather patterns, large events can cause negligible changes, or tiny ones cause total restructuring of the system. This is a main contributor to the fact that complex systems are technically unpredictable while also being patterned—organized but not ordered, or more dramatically often described as existing at the boundary between order and chaos. More practically, the future behavior of complex systems is difficult to predict, with longer time horizons generally decreasing predictive accuracy. Inversely, the idea of retrospective coherence refers to the fact that in looking at the history of a complex system, it often makes sense that events played out the way they did, while offering no certainty about the future.
- Understanding our relationships this way means realizing that “effort” and “outcome” will generally not have a direct correlation, and that it is probably unrealistic to expect to know, in advance, what our future together is going to look like. While we tend to hold expectations of ourselves and our partners that we have “proportional” or “rational” responses to one another’s behavior, the actual way we respond tends to be determined by a broad range of contextual factors about which we never have complete information.
- Emergence: In complex systems, higher order behavior emerges from the relatively “simple” behavior of rules of interaction between of the system. A commonly-cited example is the phenomenon of flocks of birds appearing to move as if they are one organism, which behavior results from relatively simple rules “followed” by the individual birds making up the flock. “Emergence” is the behavior that the whole system exhibits that appears to be “designed” or “controlled” “intentional” or “motivated” in the absence of any central control.
- The behavioral patterns and cycles in a relationship often “emerge” over time without conscious design—this is both the magic and the frustration of relationships. We often hear people speaking about their romantic relationships “I never even imagined something like this existed” implying that the experience of the relationship was not designed intentionally. On the flip side, negative behavioral structures can often “emerge” leading us to say things like “I don’t know what happened.” The small behavioral interactions, repeated over time, taken as a whole, result in the relationship displaying behavior often outside of the intention of either partner.
- Self-Organization: Using a city as an example, even without the influence of top-down “control” from a planner (though of course there are examples of cities being designed this way) a city tends to naturally “organize” itself into districts of concentration. Similarly, the overall organization of ant colonies naturally results from the simple, genetically coded behavior of individual ants. This is “emergent” organization, where the system “structures itself” without centralized planning. In systems sustained by repeated adaptation… the “homeostasis” of individual sub-systems collectively causes the organization of the super-system into a “sustainable” structure
- People in relationships typically arrive at “ways of doing things” ways of dividing up tasks, and ways of handling situations without much conscious decision making, as a result of individual responses to one another and the external environment.It’s a common experience in couples therapy for partners to realize they have developed a sophisticated and organized “way of doing things” together as a couple (often with parts they would rather were organized differently) without ever talking about it out loud, or intending to organize it that way.
- Network Structure: Complex systems are characteristically structured as “networks” composed of the nodes of individual components or subsystems, and links between those nodes representing “functional relationships” where the nodes connected to one another influence one another’s states. In this way, “signals” (changes in the states of nodes) can “move” through the network, as the state of each node influences the states of the nodes to which it is linked.
- I’ll expand on this idea in a moment, but in a relationship, the “nodes” of the network represent the behaviors of the participants, with behaviors influencing one another’s states through the ongoing process of stimulus and response.
- Fractal Structure: Similar to the idea of systems nesting inside of systems, complex systems tend to display similar structures at multiple levels of analysis. This is characteristically the result of “recursive” processes, closely associated with the “feedback loops” discussed within general systems theory, or the cycle of action and perception discussed in relation to predictive processing.
- Often the overall dynamics of a relationship can be “seen” at every level of analysis. We might see a behavioral process of “pursuit and withdrawal” emerge within the micro-interactions of eye-contact, as well as the high-level dynamics of emotional proximity over the course of a long-term relationship.
- Adaptation: The feature of being reshaped, or reshaping oneself based on contact with an external environment over time, typically self-reorganizing in order to better confront future perturbations or better “fit” the environment. The examples of learning and evolution discussed in the function post are primary examples. In that discussion, though, the focus was on the process of partners adapting to one another’s behavior, which is the granular process from which the collective relationship’s “adaptive behavior” emerges. Adaptation is the result of interactions of conflict and cooperation between components of the system, as well as interactions between the whole system and the external environment.
- The “mutual adaptation” of partners behavior in relation to one another is kind of the whole deal, but it does bring up the question of what effects different approaches to conflict might have on the qualities of the overall system.
- Healing and Growth: Technically called autopoiesis, the ability of (particularly living) complex systems to “create their own parts” or to replace missing parts. You could think of this as a higher-order version of homeostasis, in representing systems which can “return to equilibrium” even when part of the system is absent, by “growing” or “healing” the missing piece.
- When a relationship begins, there is a whole world of interactive behavior to be “grown” in the space between partners. As discussed previously, there is a general tendency for me to subtly shape my partner’s behavior into a kind of behavior I am used to dealing with, such that in a sense, I may relationally “regrow” missing parts or “roles” from elsewhere in my life. Similarly, when circumstances impose a substantial change to a relationship, that relationship often “find ways” to replace the functions that had been served by the now-impossible behavior.
- A “Mind of Their Own: Complex systems characteristically appear to have a “mind of their own.” They behave as if they want things, have intentions, or are motivated by pursuit of goals. This brings up an interesting question of whether to interpret this tendency as evidence that human minds are themselves complex systems, or just to note that we tend to feel a sense of identification or at least animation when we observe the collective behavior of such systems.
- Relationships have a mind of their own. ‘Nuff said.
To summarize a bit: the informational process we call a “human mind” may be understood as a “complex system” made out of behavior, which may display all those characteristics of complex systems we’ve discussed so far. Though not enough to draw conclusions from, it’s at least an interesting point of comparison to note that in computer science, a neural network is a piece of software representing a model constructed of nodes and functional relationships, which itself functions to “learn” as a result of that structure. The fact that neural networks get their name from their resemblance to physical brains might simply prompt us to wonder whether the “software” of mind might be usefully understood as structured in a similar way to the “hardware” on which it’s running.
So, even within an individual mind, behaviors may be arranged in a “fractal network” of relationships with one another. In a relationship between two individual minds, our behaviors prompt and shape one another’s behavior alongside the process of our individual mind’s behavior prompting and shaping other behavior within itself. The two-mind system may then represent a sort of superordinate “mind” organized at a level “above” the minds that constitute it.
Mental Diversity
Closely related to the concept of “complexity” itself, the “diversity” within a system refers to the range of different kinds of components or nodes within that system, and the level of diversity has significant implications for how that system functions and behaves.
Among the advantages of increasing diversity are the characteristics of “resilience” and “flexibility” within a system. To use the ecosystem example, if there are a broad range of similar but not identical kinds of organisms, which fill similar “functions” within that ecosystem, if one of those species goes extinct, it is relatively easy for a similar species to “fill in the gap” left by the extinct one, and for the overall ecosystem to recover from “external threats” like pollution or natural disasters. When considering an individual human mind, ”behavioral diversity” would represent having a broad, flexible, and specialized range of behaviors to fit a broad range of potential situations, rather than applying a single kind of behavior to all situations. Having a diverse range of behaviors to “choose from” confers the flexibility and the freedom to move through a broad range of situations, functioning adaptively in each.
Within the context of a relationship, “behavioral complexity,” might confer something like the advantage of a complex economy—in which it is generally relatively easy to make a living in a wide range of different occupations–to find a career that’s a good fit for your unique personality. In a couple, or a family, behavioral diversity might mean that individual members have the space and support to manifest a relatively large range of ways of being rather than experiencing pressure to conform to a narrow or rigid set of interaction patterns. In other words, a more complex relationship might be one in which it’s easy to be yourself.
So how does a complex system become diverse? Though this may start to sound a little mystical as we reach the limits of my mathematical comprehension, somehow the cycle of mutual adaptation among interacting subsystems within a complex system is thought to reliably produce increasing diversity and complexity in organization as it is repeated over time—under the right conditions, i.e. absent external, human, catastrophic, divine, or “rational” intervention. As far as we know, the ecosystem on Earth evolved towards increasing complexity and biodiversity up to the point in history where humans became dominant, and are now threatening the collapse of the whole global organism. This is an important theme across a range of complex systems: that the dominance of one subsystem is the way the system loses complexity—or in other words, dies.
Modeling Complexity
As discussed in relation to predictive processing, human beings—along with all other organisms—function to “minimize uncertainty” in order to survive, while at the same time the complexity of the systems in which we live make our environments inherently uncertain and unpredictable. Humans often prefer the order of simplicity—monocultures in industrialized agriculture, or the homogenization of appearance and behavior common in military organizations—because for our rational minds, complex phenomena are just plain confusing compared to more linear ones—where events have simple and proportional causes and effects, and firm conclusions can be drawn from prior premises.
We are inevitably faced with the task of functioning within complex systems, where our tools of rational prediction and control tend to be particularly ill-suited. The natural environment, human society, verbal language, market economies, and the internet all represent exactly those kinds of systems that traditional methods of scientific/rational inquiry have the most difficulty predicting. Of course, as discussed in relation to emotion, that “conscious” process is not our only tool for predicting and perceiving our environment.
When science seeks to understand the mechanisms and behavior of complex systems, such as global weather patterns—it often turns to simulations and models rather than equations or a particular predictive formula. These simulations are typically attempts to recreate the behavior of the systems they model by representing components of the system and the simple ways they interact with one another—such that the network of representations behaves similarly to the collective behavior of the system being modeled. We might guess that the mechanism of prediction used by our subconscious, “animal minds” might work this way, on an implicit level. There’s good reason to suspect that the mechanisms of intuition, gut feeling, or felt sense, might be best understood as predictive systems adapted specifically for the task of predicting complex phenomena, and as such may have distinct advantages in the approach to relationships when compared to rational analysis.
Somewhere in the middle ground between the “conscious, logical prediction” and the “subconscious-intuitive” sense, stands the particularly human predictive tool of imagining. Though much of the rest of the animal kingdom acts on implicit models of their environment, “imagining how my partner will respond if I do a particular thing” may also be thought of as a prediction based on a simulation rather than a rational conclusion.
Relatedly. I originally discussed the concept of intention as an “explanation” for behavior contrasting with the “functional” one. In the context of complexity, we might now wonder whether the nearly universal (among humans) experience of perceiving intention might also be understood as a predictive technology particularly suited for application to complex systems in particular. When I look at an ant colony and “see” it trying to do something, much like when I see my partner trying to start a fight, that might just be my subconscious-prediction-system telling me—in very concise language—what it thinks the complex system of colony or partner is going to do.
Managing Complexity
As people in relationships, we are often faced with the choice between “logic” and “intuition” in guiding our behavior. We want our relationships to be certain and predictable, it’s just less work to make decisions when the outcomes are known in advance. We naturally want stability and dependability, reliable access to things we want and reliable safety from the things we don’t. There are also definite advantages to simple, orderly systems—they can be spectacularly productive (in the short term), efficient, and “easy to understand.” In other words, our preference for ease and the maximization of short-term productivity makes it very tempting to try to reduce the complexity of our interpersonal environments, even if that means losing out on the many benefits of complexity and diversity. It’s so much less complicated when one person is right and one wrong, when you settle on a “right answer,” when you just follow a rule. Thinking of relationships in particular, this commonly manifests in the choice to “hide” or “suppress” parts of oneself or one’s partnerin order to keep the relationship stable, peaceful, and predictable.
Referring back to the concept of differentiation covered in the parenting post, human relationships—in their more complex forms—may naturally represent “growth machines” in their tendency to increase the behavioral complexity of their participants’ minds as those participants ongoingly adapt to one another. When partners successfully resist the temptation to simplify the relationship—balancing authenticity and autonomy with sustained connection, and consistently confronting conflicts without either compromising themselves or separating—the relationship that emerges between them will naturally become more complex, diverse, and resilient over time.
To return to our evolutionary ecosystem metaphor yet again, we might wonder what factors preserve biodiversity in ecosystems in the networked “food web” One observation is that predators in particular function to preserve diversity by limiting the dominance of any one species of herbivore, which in turn limit the dominance of any one species of plant. Though it’s a very loose analogy, the metaphorical predators in a relational ecosystem are the set of behaviors we generally call “interpersonal conflict” which—when managed effectively—functions to limit the stultifying dominance of any one behavioral subsystem of the relationship.
Complex systems are generally impossible to “control,” in the traditional top-down sense, at least without killing—or weakening—them in the process. When we bulldoze the natural ecosystem and replace it with concentrated masses of those particular organisms we find useful, we’re left with something very fragile and dependent on outside intervention in order to sustain. On the other hand, complex systems can often be steered or at least affected, by the behavior of their participants.
One general observation is that in trying to shape complex systems in a particular direction, it tends to be more effective to apply a relatively large number of little “nudges”rather than attempting a system-wide, top-down overhaul. In other words, rather than traditional “planning” and “design,” instead of trying to predict the distant future, or conforming our relationships to some pre-existing template of “health,”—our ability to respond adaptively to the current circumstancescarries the most promise for moving our relationships in a healthy direction in the long term. So rather than trying to “cut out” the bad parts of our relationships and “replace” them with something more desirable, it may be far more feasible to instead focus our efforts on “redirecting” the existing behavioral dynamics.
Illustrating this principle, take this delightful example provided in a 2003 paper by Cynthia Kurtz and Dave Snowden:
A group of West Point graduates were asked to manage the playtime of a kindergarten as a final year assignment. The cruel thing is that they were given time to prepare. They planned; they rationally identified objectives; they determined backup and response plans. They then tried to “order” children’s play based on rational design principles, and, in consequence, achieved chaos. They then observed what teachers do. Experienced teachers allow a degree of freedom at the start of the session, then intervene to stabilize desirable patterns and destabilize undesirable ones; and, when they are very clever, they seed the space so that the patterns they want are more likely to emerge.
As I alluded to before, the mechanism of intuition may be particularly applicable to decision-making within complex environments. If your felt sense of what I’m likely to do, and what you should do in response is a predictive model shaped by experience as described by reinforcement learning or predictive processing, then it is theoretically a far better basis for effective decision-making in complex environment of me than is a “rational” calculation of what you consciously believe I’m going to do. Of course, this makes it doubly problematic when the predictive model implicit in your intuition is systematically biased in a particular direction.
If I know I want our relationship to be more affectionate, for example, rather than trying to impose a rule, or convincing my partner of the strategic advantages of choosing to be nice instead of mean to me from now on, I may be better served by consistently seeking opportunities to be affectionate, each of which have the potential to start self-reinforcing cycles of affectionate interaction.
Sources
Gregory Bateson (1972) Steps to an Ecology of Mind: collected essays in anthropology, psychiatry, evolution, and epistemology
Urie Bronfenbrenner (1979) The Ecology of Human Development: experiments in nature and design
Fritjof Capra & Pier Luisi (2014) The Systems View of Life: a unifying vision
John Holland (1996) Hidden Order: how adaptation builds complexity
John Holland (2012) Signals and Boundaries: building blocks for complex adaptive systems
Robin Kimmerer (2013) Braiding Sweetgrass: indigenous wisdom, scientific knowledge, and the teachings of plants
Humberto Maturana & Francisco Varela (1987) The Tree of Knowledge: the biological roots of human understanding
Melanie Mitchell (2009) Complexity: a guided tour
Scott Page (2011) Diversity and Complexity
James Scott (2020) Seeing Like A State: how certain schemes to improve the human condition have failed.
Ralph Stacey (1992) Managing The Unknowable: strategic boundaries between order and chaos
William Stolzenburg (2008) Where The Wild Things Were: Life, death and ecological wreckage in a land of vanishing predators
M. Mitchell Waldrop (1992) Complexity: The emerging science at the edge of order and chaos
