A Matrix About Change: Adaptive Capacity, Transformation, & Dissolution

Part 2 of a discussion of the talk I gave at Red Hat "Risk, Vulnerability, and the Precarity of Identity." This post focuses on bringing together the theory in post 1 to discuss change.

A Matrix About Change: Adaptive Capacity, Transformation, & Dissolution

Developing new Domains

This is the second post discussing a talk I gave in December at Red Hat OpenShift Commons titled "Risk, Vulnerability, and the Precarity of Identity." The first post discusses the theory foundational to the talk and these articles while this post ties together the theory with my own experience to create domains for thinking through hazards, vulnerabilities, and responses.  In the following, theory introduced in the last post including Maturana's Domains of Structural Determinism, and the understanding of identity from Di Paolo, Buhrmann, & Barandiaran, (2017) are braided together with ideas of change to produce a matrix. The matrix is intended to be used for thinking about the hazards faced by an organization, its vulnerability to them, and how it might respond.

Domain of Robustness

The Domain of Robustness remains largely the same as Maturana's Domain of Perturbations though it has been widened to include an absence of change and placed within the organizational context to indicate changes in operations. This domain is where all events that might lead to brief variances in inflows or transformative processes but not a change in outflows belong (Anderies, Folke, Walker, & Ostrom, 2013). In this domain, organizations are able to withstand variances and continue reproducing themselves as themselves in the same way even if a slight departure from the norm is needed. Events in this domain might take the form of a change in where inflows come from, a brief interruption in inflows that is buffered by inventory, or a momentary breakdown in the technologies underlying transformative processes corrected by redundancies. Robustness and the closely related idea of engineering resilience that also belongs in this domain are concerned with maintaining or quickly returning to a certain system state (Gunderson and Holling, 2006). This domain is about stability.

Change, Flows, Openness, & Delimiting

Part of maintaining stability is reducing the amount of inputs that can cause variances by delimiting the system from its environment through the creation of a boundary with controls that determines what enters and exits (Di Paolo, et al., 2017). To achieve complete robustness, nothing would be allowed to enter the system. In this ideal state, everything is seen as a potential threat to the reproduction of the system's identity and can therefore not be allowed in. However, it is impractical as without certain inflows the identity of the system cannot be reproduced and would ultimately be lost as resources ran out. Inversely, an ideal state of identity reproduction is total openness where every inflow in the environment can be used to reproduce identity. In this ideal state there is no boundary. In their book, Di Paolo, et al. (2017) indicate that in this scenario of total openness all flows are used in the reproduction of identity. Cilliers (2005) departs from this understanding and writes that if a system is moving from flow to flow there is no identity and in fact no difference between the system and the environment - the system is moving too fast to have an identity. Di Paolo et al. (2017) suggest there is a middle ground where systems exercise agency and select some flows and reject others in the reproduction of their identity. Of course and as discussed earlier, what flows are desirable versus not desirable is a matter of identity while at the same time what identity is possible is contingent upon the availability of flows (there is a bidirectional relationship between flows and identity). Leafy greens are a desirable inflow to a giraffe that eats a herbivore diet while the survival of the giraffe is entirely dependent on the continued inflow of leafy greens.

An expanded version of Maturana's domains of structural determinism

Domain of Adaptive Capacity

As it is presented here, the Domain of Adaptive Capacity is broadly consistent with Maturana's Domain of Structural Changes in that it considers events that may lead to change within the constraints of an identity. In other words, "organizationhood" remains the same but the way it is realized changes. This domain is moved into intentionally or unintentionally as the need arises to enact change while keeping identity constant (which could possibly be an error and lead to a forced move into the Domain of Transformation or Dissolution later). In this domain, a certain degree of change is possible before things no longer remain the same and the boundary into the Domain of Transformation is crossed. This boundary may resemble a liminal space found in Cynefin or otherwise understood as "thick" as it was referred to in the first post. Whatever the language, there is a space between the domains where a system is not fully in either one that may be advantageous in turbulent environments.  

The focus of this domain is all the events that might trigger the need to "manage resilience." Adaptive capacity is understood in Walker, Holling, Carpenter, and Kinzig  (2004), as the ability of actors in a system to manage its resilience. Resilience is defined here as the "capacity of a system to absorb disturbance and reorganize while undergoing change so as to still retain essentially the same function, structure, identity, and feedbacks—in other words, stay in the same basin of attraction" (Walker, Holling, Carpenter, Kinzig, 2004, par.21). This is similar to Gunderson and Holling's (2002) definition of resilience as the "magnitude of disturbance that can be experienced without the system flipping into another state or stability domain" (p.50).

In the article by Walker, Holing, Carpenter, and Kinzig (2004), they provide four critical aspects of resilience:

Latitude: the maximum amount the system can be changed before losing its ability to recover; basically the width of the basin of attraction. Wide basins mean a greater number of system states can be experienced without crossing a threshold (L, Fig. 1).
Resistance: the ease or difficulty of changing the system; related to the topology of the basin—deep basins of attraction (R, or more accurately, higher ratios of R:L) indicate that greater forces or perturbations are required to change the current state of the system away from the attractor.
Precariousness: the current trajectory of the system, and how close it currently is to a limit or “threshold” which, if breached, makes recovery difficult or impossible (Pr).
Panarchy: how the above three attributes are influenced by the states and dynamics of the (sub)systems at scales above and below the scale of interest (Pa; par 23).

Systems travel through a three-dimensional state space defined by the "variables that constitute the system" (Walker, Holling, Carpenter, & Kinzig, 2004, par. 11). To explain state space, Walker, Holling, Carpenter, and Kinzig (2004) use the example of a rangeland system, and write that all the possible combinations of grass, shrubs, and livestock are the system's state space and the state of these variables at any time define the system. They continue and explain that each of the varying distributions of grass, shrubs, and livestock, are different basins of attraction found across a stability landscape (Walker, Holling, Carpenter, & Kinzig, 2004). Basins of attraction, also known as stability domains, vary in size and depth and may be harder or easier to move out of accordingly.

Resilience, then, at least by the definitions given here, is concerned with a system staying within a basin of attraction. Unlike the notion of engineering resilience concerned with returning to a point within a basin, the idea of ecosystem resilience used by the sources cited here allows for systems to move around within basins of attraction as needed. In the related literature, the stability landscape of a system is constituted by basins of attraction characterized by different combinations of the same variables in different distributions (Holling, 1973; Walker, Holling, Carpenter, & Kinzig, 2004). As a system changes domains featuring different combinations of grass, shrubs, and livestock, it maintains the same identity (e.g., a rangeland) even though the variables have changed.

The Domain of Adaptive Capacity introduced here is concerned with changes in the values of variables across basins of attraction as well as changes in variables as new basins are created. The constraint to the introduction of new variables is the continued reproduction of the same system identity. This idea finds its roots in Maturana's work where structural changes are only constrained by the "organizationhood" (identity) of the system. The core question of this domain is, "How much can things change and remain the same?" The identity of the system can be "stretched" only so far before it begins to change and there is a transition from the Domain of Adaptive Capacity to the Domain of Transformation. Like with other domains, the boundary between the Domain of Adaptive Capacity and the Domain of Transformation is "thick" where a change might not fully enter the adjoining domain but become close to doing so. From this position, both retreating and advancing are equally viable options and may be an ideal state if there is turbulence and a system can be held there.  

In the Domain of Adaptive Capacity belong all events that could trigger the need to manage resilience by widening or deepening desirable basins of attraction, reducing the size of undesirable ones, eliminating undesirable ones, and moving the system further within a desirable basin or closer to the boundary of an undesirable one (discussed in the table below; Walker, Holling, Carpenter, & Kinzig, 2004)

Some adaptive capacity dynamics and methods for achieving them. 

Not mentioned in the above is the adaptive capacity dynamic of creating desirable basins of attraction (Walker, Holling, Carpenter, & Kinzig, 2004). The dynamic of moving to newly created basins of attraction involves changing the variables that constitute the system, but not the identity of the system itself. For example, imagine a bridge company that builds bridges out of stone when the stone basin of attraction starts shrinking due to less expensive alternatives. Recognizing this, the bridge company creates a new basin of attraction where they replace the variables of stone inflows, the transformative processes of masonry, and the outflow of stone bridges, with the inflow of iron, the transformative processes of iron working, and the outflow of iron bridges while continuing to reproduce their identity as a bridge company. Here, agency is exercised to change the inflows as well as the processes that transform them into outflows. Gaining the transformative skills to build bridges out of iron on top of those already possessed is an example of multi-skilling that will allow the company to move back into the stone basin of attraction as the market changes (Trist, 1993). The scenario is different if the identity the bridge company is wider and deeper than being a company that build bridges and instead includes the detail of being a bridge company that builds bridges exclusively out of stone. If this understanding of themselves is what they are holding on to and seeking to reproduce, they may stay in a steadily shrinking basin of attraction. From this undesirable domain it is possible the bridge company could enter the Domain of Dissolution where their identity and how it is produced both disintegrate.  

Domain of Transformation

In the Domain of Transformation belong all events that might trigger a change in identity and as a consequence how it is produced - its "organizationhood" is changed as well as how it is produced. This domain can be entered purposefully or as a consequence of external forces (Folke, et al., 2010). It is important to remember here, as in other domains, Maturana's notion of structural determinism. Structural determinism suggests that anything that happens within a system is a product of its structure and not a direct result of external events. In any of the domains, events trigger but do not directly determine change within the system. In the case of this domain, the events serve as triggers for change in the state space of the system through the introduction of new state variables but do not determine them.  

At times societies or groups may find themselves trapped in an undesirable basin that is becoming so wide, and so deep, that movement to a new basin or sufficient reconfiguration of the existing basin becomes extremely difficult. At some point, it may prove necessary to configure an entirely new stability landscape—one defined by new state variables, or the old state variables supplemented by new ones. For instance, in the rangeland case—defined originally by the amount of grass, shrubs, and cattle—a new stability landscape could be created by introducing new ways for earning a living, such as ecotourism, based on wildlife and rivers (Walker, Holling, Carpenter, & Kinzig, 2004, par.18).

The core difference between the Domain of Transformation and the Domain of Adaptive Capacity is a change in the identity of the system that enables the introduction of new state variables. In the Domain of Adaptive Capacity, new state variables can be introduced but only to the degree that they reproduce the existing identity. In the Domain of Transformation, a change in identity allows for the introduction of new state variables (inflows, outflows, transformative processes) beyond what would have not yet been conceivable in the Domain of Adaptive Capacity where identity is held constant. The Domain of Transformation is where the bridge company transforms into a railway car company and the herbivore giraffe turns into a meat eating tiger. By changing identity, the stability landscape opens up and new basins of attraction can be created.

Domain of Dissolution

The Domain of Dissolution combines Maturana's early notion of the Domain of Destructive Interactions with his later proposition that events in this domain can cause destruction while subsequently leading to the appearance of a new type of system (Maturana & Verden-Zöller, 2008). Events belonging to the Domain of Transformation may briefly move a system into and back out of the Domain of Dissolution as one identity is dissolved and a new one takes its place. During this foray into Dissolution the system losses coherence and its "organizationhood" only temporarily as it moves from one identity to the next (this dynamic may be short lived if successful). If this transformation is not successful, the system may move further into the Domain of Dissolution after suffering the destructive interaction of a failed transformation. In this case, coherence is lost and what was formerly the system becomes difficult to discern (Waldrop, 1992). There may also be deeper moves into the Domain of Dissolution when a system unexpectedly loses its identity and does not immediately possess the means to create a new one and develop the necessary flows and transformative processes to reproduce it. The pathway out of the Domain of Dissolution is to move to the left into the Domain of Transformation. Moving upward into the Domain of Adaptive Capacity is impossible as there is no identity, as is the Domain of Robustness as there is nothing to keep stable. It is likely the case that the longer the amount of time spent in the Domain of Dissolution and the further the system is moved into it the harder it is to get out. Like in any of the other domains, "further in" refers to how far the system has traveled from the domain's axes.

I have Either said too much or Entirely too Little

When I sat down to turn a talk into a lecture I was unaware of the difficult task set out before me. I tweeted that writing about things that you have only ever spoken about is challenging, much less putting it in a coherent form.

In the above, I have provided an account of Maturana's Domains of Structural Determinism expanded upon by ideas of identity and its reproduction by Di Paolo et al. (2017). Tied together, these understandings provided a starting point for a new conversation around change focusing on robustness, adaptive capacity, transformation, and dissolution.

The following discussion focused on identity and its dynamics at the level of an entire organization. Of course, everything said about organizational identity and change is also applicable to teams, departments, divisions, and even individual projects. For this theory to be applicable, all that is needed is an identity, an understanding of what a component of an organization is and does, and inflows of resources that are rendered into outflows by transformative processes. These multiple identities then emerge into the identity we discern at the level of an organization.

As an organization is analyzed with a finer degree of granularity, the various interconnected organizational elements can be placed in the domains to which they belong relative to the hazard events they are facing. This type of map can give a more detailed understanding of where the organization is at composite level as well as make threats and hazard events explicit. Of course, this matrix can also be used at the very coarse level of the organization's function - what events might threaten this function and what domain are they in? In the image provided below, a grocery store is used as an example of events threatening overall function as well as at one department level. Note that in all domains, polycausal narratives are constructed that account for the event as well as how the event became an issue internally. This is in alignment with Maturana's structurally determinism.

Mapping a grocery store 

If this were to be used, it might be applied as a thought experiment for prospectively thinking about what threats are out there, how a system might be vulnerable to them and in what way, and how it might respond. For example, would a breakdown in a grocery store's primary supply chain place them directly in the Domain of Dissolution or would it only require a change in inflows in the Domain of Adaptive Capacity? It may also be useful in considering what an organization can handle as well as how it might recover from events that outstrip the other domains and move it into the Domain of Dissolution. There may be a bias toward keeping organizations in the Domain of Robustness and for parts of organizations this is certainly the case, but certain projects and departments may thrive better within the Domain of Adaptive Capacity where they can manipulate their basin of attraction as well as their position in it, and create new basins of attraction to meet new needs.


In the above I have focused on identity at length as I think it can produce a detailed understanding of "who we are" as well as "what we are facing" and "how we might deal with the various threats." But I am also wondering about a less detailed version of the matrix introduced over the last two posts. As a former wildland firefighter, I wonder to what degree a tool like this could be used spread across the hood of a truck used to think through an operational period: "If we consider this operation, what are we robust enough to not directly be bothered by, what events might lead to the need to manage resilience and move basins of attraction (for example different tactics), what might trigger a change in strategy (transformation) and when do we need to abandon this operation (dissolution?).

More on this later.


Anderies, J. M., Folke, C., Walker, B., & Ostrom,  E. (2013). Aligning key concepts for global change policy: Robustness,  resilience, and sustainability. Ecology and Society, 18(2).  doi:dx.doi.org/10.5751/

Cilliers, P. (2006). On the importance of a certain  slowness. Emergence: Complexity and Organization, 8(3), 105-112.

Di Paolo, E., Buhrmann, T., & Barandiaran, X. E.  (2017). Sensorimotor Life: An Enactive Proposal. Oxford, UK: Oxford.

Folke, C., Carpenter, S. R., Walker, B., Scheffer,  M., Chapin, T., & Rockström, J. (2010). Resilience thinking: Integrating  resilience, adaptability and transformability. Ecology and Society, 15(4).  Retrieved from www.ecologyandsociety.org/vol15/iss4/art20/

Greenberg, R., & Bertsch, B. (2021). Glossary. In  R. Greenberg, & B. Bertsch (Eds.), Cynefin: Weaving sense-making into  the fabric of our world (pp. 26-30). Singapore: Cognitive Edge Pte Ltd.

Holling, C. S. (1973). Resilience and stability of  ecological systems. Annual Review of Ecology and Systematics, 1-23.  doi:doi.org/10.1146/annurev.es.04.110173.000245

Holling, C. S., & Gunderson, L. H. (2002).  Panarchy: Understanding transformations in human and natural systems.  Washington D.C.: Island Press.

Korsgaard, C. M. (2009). Agency, identity, and integrity. New York: Oxford University.

Maturana, H. R. (1983). What is to see? Archivo de  Biología y Medicina Experimentales, 16(3-4), 255-269.

Maturana, H. R. (1988). Reality: The search for  objectivity or the quest for a compelling argument. The Irish Journal of  Psychology,, 9(1), 25-82. doi:10.1080/03033910.1988.10557705

Maturana, H. R., & Poerksen, B. (2011). From  being to doing: The origins of the biology of cognition (2nd ed.). (W. K.  Koeck, & A. R. Koeck, Trans.) Kaunas, Lithuania: Carl-Auer.

Maturana, H. R., & Varela, F. J. (1987). The  tree of knowledge. Boston, Massachusetts : New Science Library.

Maturana, H. R., & Verden-Zöller, G. (2008). The  origin of humanness in the biology of love. (B. Pille, Ed.) Exeter, Devon,  UK: Imprint Academic.

Thompson, E. (2007). Mind in life: Biology,  phenomenology, and the sciences of mind. Cambridge, MA: The Belknap Press  of Harvard University press.

Trist, E. L., & Bamforth, K. W. (1951). Some  social and psychological consequences of the longwall method of coal-getting:  An examination of the psychological situation and defences of a work group in  relation to the social structure and technological content of the work system.  Human Relations, 4(1), 3-38. Retrieved from  doi.org/10.1177/001872675100400101

Trist, E. (1993). Introduction to Volume II. In E. Trist, H. Murrary, & B. Trist (Eds.), The social engagement of the social science: A Tavistock anthology (Vol. 11, pp. 36-60). Philadelphia: University of Pennsylvania.

Trist, E., Gurth, H., Murray, H., & Pollock, A.  (1993). Alternative work organizations: An exact comparison. In E. Trist, H.  Murray, & B. Trist (Eds.), The social engagement of the social science:  A Tavistock anthology. University of Pennsylvania Press.

Waldrop, M. M. (1992). The emerging science at the edge of order and chaos. New York, NY: Touchstone.

Walker, B. H., Gunderson, A. P., Kinzig, C., Folke,  S. R., Carpenter, S., & Schultz, L. (2006). A handful of heuristics and  some propositions for understanding resilience in social-ecological systems. Ecology  and Society, 11(13). Retrieved from www.ecologyandsociety.org/vol11/iss1/art13/