A minimalistic definition of a complex adaptive system, acceptable by most people in the field of complexity theory, would be like: “Every complex adaptive system (CAS) has a history, and consists of adaptive heterogenous agents (elements) that have unpredictable interactions with each other, through some kind of communication, leading to an unpredictable emergence of the system as a whole.”
Examples of different CASs in the literature are, nature-made or man-made; ant colonies, bird flocking, human organizations, cities, the stock market, the brain, the immune system, the internet, and many, many more.
After this minimalistic definition, when adding more characteristics of a CAS, the opinions go apart. We will focus on some of the most important characteristics that is valid for organizations and their ways of working, here exemplified with some questions:
- Do all complex adaptive systems have a common purpose? Never?
- If the complex adaptive system has a common purpose, does all agents necessarily need to know about it?
- Can the system have a leader? Never?
- Can we visualize the history of the system? Sometimes? Never?
- Can there be a cause-and-effect chain in a complex adaptive system? Never?
The many different systems from different domains, stated as CASs in the literature mentioned above, as well as that the field of complexity theory is a rather new field, makes it understandable, that there are many different definitions. But, no matter, having different views about the definition of a CAS, is of course cumbersome and to a big extent prohibiting sound and common advances in the field. The different views also obscure if complexity theory is a discipline by itself, or transdisciplinary work together with other disciplines, or a result of transdisciplinary work, using science from already existing scientific disciplines. With transdisciplinary work also means to achieve a scientific basis on how to eliminate the transdisciplinary complexity to achieve a well-working (man-made) human system. Different views also lead to different vocabulary which in itself is problematic. A common vocabulary is always key, to be able to sort things out, especially when doing transdisciplinary work, i.e., since when synthesising actual existing science from different disciplines we must have a clear and agreed upon (novel) vocabulary for the resulting (applied) science on the transdisciplinary top-level, when (and if) needed. Let us now with some reasonings try to sort some characteristics out, valid for developing a way of working in an organization, by having a truly open mind, maybe we also need to change our thought patterns somewhat. Who knows.
All different CAS will not be covered in detail, but let us start the reasonings, and except from organizations that is the main focus in System Collaboration, also take a closer look into ant colonies for a comparison. Ant colonies are actually a very good comparison to organizations, with similarities and differences, which are bringing up and sorting out, several characteristics not having full agreement in the field.
We start with considering a common purpose, the overarching WHAT to do, valid for any organization. An ant colony as well as our immune system has a common purpose built-in to their DNA, that is unknowable to the agents in the respective system. A common definition of an organization is “a group of people who work together in an organized way for a shared purpose” (from Cambridge Dictionary), and in System Collaboration we use the definition “People that interact to solve activities with interdependencies for a common purpose”. This means that an organization on the other hand has a common purpose that is known by the employees, and where this common purpose also will be divided into the different parts of the organization. If we instead take the stock market as an example, it has agents working only for their individual purpose and not for a common purpose of the system. This means that a CAS can have a common purpose or not, or we need to differentiate these two types of systems as it seems, with a more specific vocabulary.
For an ant colony, the survival of the ant as a specie, is built into their DNA by simple nature-made rules, where no ant knows about this need of survival, neither the rules to achieve their survival, they only follow these rules. Without these rules, we would have chaos in the ant colony (or rather extinction already). But, to be able to follow these rules, the ants need to not eat each other (at least under normal conditions, most probably rules as well), which means respect for each other, as well as to be able to interact with each other by communication (most probably rules as well), i.e., having a common vocabulary, so to say. Respect and communication are therefore also necessary to avoid chaos. Note that all these rules are needed for the survival of the ants, and are the built-in HOW, leading to the achievement of the (hidden*) common purpose, the built-in WHAT.
With rules built-in to their DNA, this means that there is no leader needed**, to orchestrate the direction or actions for the ant colony’s survival, the WHAT, and HOW to reach there. This also means that the ants do not need any central planning or central architecture making, when building the structures of their ant colony. There cannot be a clearer decentralized way of working than this, meaning that all the ants’ interactions in for them a normal habitat, always will aggregate up to the common purpose. This is actually very close to the aggregational work in a production process at Toyota, that follows its process rules, the HOW, that will (to a very high degree) fulfil the WHAT per se, the common purpose of the process. This production process will in turn together with all other production processes, aggregate to the full common purpose of the plant, i.e., manufacturing the whole car. The adaptivity of the ants, needed due to differences in the environment and due to environmental changes (the latter may need evolutionary changes in the ants’ DNA), in combination with that the ants do not know anything about the existence of their own system, can only lead to an unpredictable emergence of the ant colony and its nest as a whole, in order to fulfil the common purpose, their survival. This is similar to the phenomena of bird flocking, where the birds in the flocking situation also follow built-in rules, that the birds are not aware of, as well as they do not know what it is meant for, giving unpredictable emergence of the flock.
A man-made organization is of course very different from the nature-made ant colony, but there are of course also similarities. For example, respect, a common vocabulary, a way of working (HOW) leading to a common purpose, the organizational purpose (WHAT), are all needed to avoid chaos in an organization, as well. But, the rules of our way of working in our organization, as well as the organizational purpose, are of course not built-in to our DNA. Rather we instead have cognitive limitations built-in to our DNA, which we cannot ignore. Our way of working in an organization, is a developed man-made construction, to be able to fulfil the organizational purpose. A bad way of working, where we for example are ignoring our cognitive limitations, will finally create chaos, even though we are fulfilling respect, a common vocabulary and having a direction from start. That means that both the ant colony and the organization have a way of working, but the former is built-in following rules, the latter is constructed, but need to avoid violating our cognitive limitations (the science), i.e., also need to follow mandatory “rules”.
The organizational purpose, the WHAT, is a necessity to have for any organization, since from this we get the context, which in turn is a necessity for making our way of working, the HOW. This means that we need to have a holistic view in all contexts, no matter context, since we always need to plan our work top-down, outgoing from the WHAT. This top-down approach is even more necessary in a complex and complicated context like product development. This because, here we also need to eliminate different kinds of complexity, i.e., gain more knowledge to be able to develop our products, which compared to ant colonies, gives us a range, with low possibilities for decentralization regarding novel products in product development, but high for production with low complexity.
We must have a humble approach to that every human way of working also need to follow “rules”, the science, as mentioned above. The science is our context dependent Organizational Principles, OPs, from different disciplines, like anthropology, logic and complexity theory. This means that every way of working (human (eco-) system per se) is an axiomatic system, where the needed science from many different disciplines (context dependent), must be fulfilled, for our system, our way of working, to be flourishing. Another way to put it, is that the transdisciplinary complexity needs to be eliminated (gain all the needed new/novel knowledge to be able to synthesize the parts), to achieve a well-functioning way of working (system). This means to at the same time, fulfil all the needed science from the different disciplines when making the systems design for the way of working, which is the same needed approach as when developing new/novel products. For the context of product development, we therefore also need to find ways to eliminate the transdisciplinary complexity in a completely novel product (also a system) or a modular platform (a more complex system to sort out). Which means two nested feedback loops to reduce two different kinds of transdisciplinary complexity, definitely a highly complex process. But, thanks to that we can follow our System Collaboration Deductions, the transdisciplinary complexity for our way of working is already eliminated, making it easy to make a solution for a flourishing way of working for product development, in any domain.
The non-negotiable presence of science when setting up an organization (any human (eco-) system) and its way of working, also means that if we do not fulfil the needed science, that will per se be our corresponding root causes, generating insolvable symptoms in our organization. These symptoms are mostly related to failing to eliminate transdisciplinary complicatedness or complexity, which is a must, in order to avoid violating our cognitive limitations or other science, when synthesizing the parts of our way of working. Many times, these symptoms are due to failing to make our tangible artifacts (compared to the intangible interactions needed when trying to make them) that we have planned for, or at the integration of them, to higher-level artifacts. All these artifacts are indirectly generated from smaller parts of the common purpose. Here are some concrete examples what the symptoms can be related to, when we are not fulfilling science in our way of working in order to reach the common (organizational) purpose; our people in organizational structures, our activities and their (inter)dependencies, our architectures of our products and the total architecture, and of course responsibility for the all of the structures and their parts, which in turn can lead to the symptom of excessive and unhealthy number of meetings, due to all symptoms. A reminder is that planning also regards, administration, competence, career, tools, educations, etc., and the activities that can be the result, which is normally the responsibility of the line hierarchy.
YES, I guess you could see this coming!
This means that we in organizations, when our way of working fails to fulfil all needed science in the different disciplines for a given context, will have cause-and-effect networks, with symptoms out-going from the non-fulfilled science (a.k.a. built-in root causes). We can therefore, when having problems (symptoms) in our organization, in hindsight back-track to the built-in root causes. We can use the problem picture analysis method SPPA – Systemic Problem Picture Analysis***, to easily find this network of built-in root causes and the symptoms they generate, which can be done at any time, since no igniting incident is needed. Do not forget that SPPA can differ between different kind of problems, not only the ones generated from built-in root causes. After that, with help of the System Collaboration Deductions, we can easily find a flourishing solution for our way of working, since all OPs need to be fulfilled, i.e., organizational adaptivity at its best. This flourishing solution is fully systemic, by fulfilling all the needed science for the context. Since symptoms are insolvable, this also directly tells us that if we in one way or another, are trying to handle symptoms**** here and there in the organization, it will only lead to sub-optimization, which rather is in the opposite direction to where the root causes are to be found, and can therefore never lead us to a systemic solution, only sub-optimization. An awesome property of this cause-and-effect network, is that it is more or less repeatable in a context, i.e., the same non-fulfilled science, will generate about the same symptoms, no matter domain for the given context. The evolving of artifacts according to our necessary top-down planning when we are fulfilling the science, or a cause-and-effect network of symptoms and root causes, when we are having unfulfilled science in our way of working, also clearly visualize the existence of a tangible history of our organization, in comparison to the intangible interactions. It is important to point out that the development of man-made artifacts in our organization, or network of symptoms and root causes if we are failing to achieve the artifacts, are not on the low granularity (intangible) level of our peoples’ interactions, that neither can, nor as any complex system’s emergence can, be predicted.
But for ant colonies, since the ants are following nature-made rules, when making all their interactions, this leads to that there cannot be any non-fulfilled science, since the science per se are built-in to their rules, i.e., otherwise, if we reflect on it, there will be no ant colonies. This in turn gives that there cannot be any built-in root causes, due to not following the science, in the HOW (the rules) of an ant colony, as there can be for an organizational way of working, when science is not fulfilled. The ants’ adaptivity and their decentralization, gives that they do not (need to/cannot) know anything about their system, which merely will emerge for fulfilling their survival (WHAT). As an observer, it is also impossible, due to the ants’ adaptivity, to conclude if there are intermediate artifacts that they are aiming for*****, or if they are failing to achieve them, since there is no plan to make a specific artifact and their adaptivity always make them successful, i.e., they can, under normal conditions, not fail their common purpose, due to their adaptivity. Due to no built-in root causes or no tangible artifacts, there can therefore under normal conditions not be any tangible cause-and-effect networks. If there anyway are non-observable artifacts that the ants are doing that we cannot see or understand that they are doing, they can still not come from built-in root causes. This also leads to that it will, for an ant colony under normal conditions in their right habitat, be very difficult to visualize the history of the system.
This results in that, having visible artifacts or not, the ant colony can only fail to fulfil its survival, due to environmental changes (or of course wrong habitat), which then will be the root cause, i.e., failure only due to external root causes, which will give us a history that we most probably can back-track. This is a huge difference to organizations, which of course can be extinct due to external factors, e.g., that no one wants our products any longer. But, by ignoring science, our OPs, about organizations (or human systems per se), when we are developing our way of working, we will get internal built-in root causes. This means that even though we humans are adaptive, we will in best case get poor efficiency, wrong effectiveness and bad quality as a result. In worst case, and with high risk, especially when our way of working is not eliminating the transdisciplinary complexity needed for product development of novel products, our adaptivity cannot help us much. In these cases, it often leads to a non-working product, or with severely reduced life-cycle for our product, and in the end of course, we are putting our organization at extremely high risk for extinction. These root causes are built-in, which means that we only can eliminate them by making a new systems design of our way of working that fulfils all the Organizational Principles valid for our context. To eliminate these built-in root causes must always be our top priority, never by focusing on that we humans are adaptive. Focusing too much on human adaptivity, easily leads to simplistic solutions when dealing with complexity and insolvable symptoms, like; continuous improvement, PDCA cycles, value streams, hypothesis without reflections, etc., which means that we instead let our adaptivity become our Achille’s heel.
With these reasonings regarding ant colonies and ways of working in organizations, we have shown that complex adaptive systems can be very different to each other, and thereby also answered the questions above, most probably giving valuable input to advances in the field of complexity theory.
*The common purpose of being able to survive, or to replicate, is valid for all species; for nature (in general)
**most probably it cannot even be one, since nature has already, found that solution for animal complex adaptive systems, to be non-working
***a traditional root cause analysis is from a production context, and often searching for single root causes to the problem/incidents that occurred, in order to solve the root causes. This is not appropriate for organizational problems, since all root causes need to be found, in order to be able to fulfil all science in the updated way of working.
****solve the symptoms directly (or after analysing more data) by putting them in a context, divide into smaller pieces to get more details/data, solve the ones having with lowest effort, bring many people with different backgrounds for the solution, collect more data about the symptoms for the solution, etc. Trying to solve symptoms generate more symptoms, which induces the generation of an enormous amount of more data/information, that will obscure our minds and in the end lead to a cognitive overload; an Achille’s heel, since we then think that we have reached the solution. But, no big or rich data, or narratives in the world can help us, since symptoms are always insolvable. Remember that, just by calling the symptom by another name, will not change the fact that it is still an insolvable symptom.
*****most probably, the absence of built-in root causes, makes it impossible for the ant colony to use artifacts, i.e., nature-made CAS, like ant colonies and bird flocking, must have adaptivity, which in turn will result in emergence of interactions, as well as the system itself, in order to fulfil the common purpose (survival or flocking).