3
Steering Complex Adaptive Systems: Managing Weak Signals

In the previous chapters, we highlighted the need to look at organizations as complex, evolving systems, i.e. discussed how they adapt in multiple, often self-organizing ways that can be steered but not controlled. In this chapter, we will return to the concept of the complex adaptive system (CAS), understanding that adaption includes the twofold ability to adapt on its own and to bring about desired changes through governance in response to external or internal challenges. A synthesis of these two objectives leads to the concept of guided self-organization.

The goal of this chapter is to understand what is happening in the organization and what explains its apparent behavior in order to guide us in both the design and management of such entitites (including new firms and complex projects). We will often use the term “steer” rather than “manage” in order to draw attention to the fact that, as with captaining a boat, proper governance involves considering the behaviors of both the boat and the sea – understanding that both are critical to undertaking a successful voyage. Steerage cannot be a fully deterministic activity. Consequently, it is a more innovative use of the term management.

The classification “adaptive” corresponds to an intrinsic quality of the CAS organizational system. A crucial aspect of adaptability is the emergence of the phenomenon that we will call weak signals and the capacity within the structure to identify, read and interpret them. In a cognitive description of the system, we are dealing with atypical information, in the sense that it does not match what individual mental models, or organizational routines, are used for recognizing and processing. In a linguistic metaphor, we could speak of foreign words that our ordinary syntax is not prepared to articulate.

For the organization, the weak signal is information that is off the radar, outside the normal parameters of relevant information or easy interpretation – hence the proximity of this concept with that of “noise”. The concept of noise is used in many contexts to similar effect – in electronic systems, it is unwanted signals or disturbance; in the theory of information, this unwanted data clutters up input quality. The art of the observer or system steerer lies in spotting this information as potentially relevant and in distinguishing it from global systemic noise. Therefore, how can we detect such exotic signals? And how can we build a language capable of interpreting them in a meaningful and shareable way? This is the steerer’s role, and this is what sets a real leader apart.

A person who knows how to manage business in a stable state does not need to be an entrepreneur in the noblest sense of the word. Yet stability is not the common state for a complex system such as an organization for any period of time – a state of dynamic flux comes into play, and indeed an organization is a system of “multiple states of flux”, meaning that there are always some areas which are stable, some bedding down change and some in a state of dynamism or change at any one time. In order to cope with such differentiation and potential ambiguity, those who lead and manage successfully in such an environment are “complex thinkers”, a competence allowing them to deal with this state of messy yet normal emergence.

This enables them to be visionary, yet visions and mental representations that break away from the usual schemata cannot be pulled from a magician’s hat, relying just as much on accumulated experience and the synthesis of knowledge and experience (into what we call wisdom) as on the visionary individual’s imagination. An essential aspect of knowledge likely to feed such strategic intuition – visions that are pertinent to the system’s future – is this category of weak signals that emerge in the organization, most commonly at the interface between the organization and its environment.

3.1. Navigating the ocean of signals

Weak signals are atypical forms of information and thus ones that risk going unperceived; however, they must also be distinguished from rare, insignificant events. The art of steering lies in recognizing them as signs that are harbingers of change, and once they have been recognized, we must know how to wait for the right moment to act on them – or more precisely, with them. They may be internal or external: an unexpected change in the boat’s behavior or an indication of a coming change in the weather conditions. They are emergent properties in the global self-organized system, and they must be processed simultaneously with the other, more typical dynamics of the system, like the inertia that we will deal with later in this chapter.

3.1.1. Understanding the nature of the ocean

It is clear that the system we are observing here is halfway between order and disorder or, in Henri Atlan’s words, “between the crystal and the smoke” (Atlan 1979). We can also use the expression chaordic, i.e. a mix of order and chaos.

For such a system’s management, the name of the game is knowing how to subtly steer the organization by playing on the interconnections and seeking new stability conditions through a policy that is more motivational than authoritarian. Knowing how to recognize and act on the weak signals requires managers to demonstrate the abilities to observe, identify and strategically adapt. Part of what is considered by the observer to be a weak signal is in fact already a form of systemic adaptation, often pre-emptive in nature. The goal is to formulate strategic propositions that take this emerging information into account, undergoing organizational change by dynamically steering it instead of trying to get it to adhere to locked-in strategic plans that will, by default, be wrong. This is what distinguishes deliberate, planned strategy from emergent strategy in the traditional managerial literature. The chaordic organization floats between order and chaos, and its strategic management must navigate between the two risks of planning (utopic and inefficient) and business behavior that goes with the flow through simple reactions to events (which is not strategic).

3.1.2. Observing the ocean

Once we begin to understand how things are connected in a complex system, it can become overwhelming as we see increasingly more connections – the team is a system but connects with the department, the department with the section, the section with the company and the company with the market – just like a Russian matryoshka doll, there is one inside the other. Complexity shows us how each influences the behavior of another – hence the term co-evolving complex systems. Just as the political or economic environment impacts an organization, so the organization can impact the environment, especially if it is large, and creates a significant positive or negative impact on the system. One is not independent of the other. Yet we cannot work at every level of the system. Our scope would be impossibly large. Therefore, how do we know where to draw the boundary and concentrate our gaze?

3.1.2.1. The system dynamics lens

The best way to consider this is to remember that “the question bounds the system”. This means that at any one time, we look at a particular aspect of how that system is operating, whether to improve something, deal with an issue, report on operational progress, design a new strategy and so on, the part of the system to look at is what is relevant to the question being asked. This stops people from being overwhelmed by not knowing what to leave out!

In our metaphor of sailing the sea, the sailor knows that the boat is not distinct from the ocean, but is instead part of an interactive system formed by both. The quest of adapting during the dynamic voyage includes the consideration of many things – the “internal system” includes your supplies, the capabilities of the boat, time constraints and the experience and skill of the crew. External conditions include weather conditions (both forecasted and emergent), tide, wind, current and wave size and type. The skill of the successful sailor lies in identifying both weak and strong signals from various parts of the system – reading the interconnected system and keeping their radar up to pick up changes in any parts of the system, then deciding on how relevant they are (both identifying and understanding the signals).

In terms of a company’s long-term strategy, this co-evolution of internal and external conditions gives us another way to look at a major strategic choice: adapting to the environment and/or building an environmental niche. Co-evolution means adapting its skills to the environment and also adapting its environment to its skills.

Apart from evolutionary change as discussed above, there are also events that better match revolutionary change in complex systems, when external events have a large impact and change the internal dynamics of the system – in this case, the organization. The general theory of adaptive cycles explains resilience in terms of a system’s slow accumulation of resources and increasing connectedness, which lead to decreasing resilience over time. Thus, in order to adapt, these cycles are interposed with stages of crisis, transformation and renewal (Holling et al. 2002; van Eijnatten 2004). The interdisciplinary theory of complex adaptive systems offers a number of models that we can use to understand this process. Prigogine’s model of dissipative systems is used when describing convection, cyclones, hurricanes or chemical bifurcation, which then change the state of the system. Panarchic systems (Holling et al. 2002; van Eijnatten 2004) add a further level of detail, which is the different time frames of evolving hierarchical systems with multiple interrelated elements that sit within each other and act upon each other in an organization, creating the possibility for a fast jump loop (revolt, innovation) that allows small-scale local change to quickly affect the much larger evolutionary path. Panarchic loops exist in all natural settings – a forest is such a loop, with a cycle of new trees requiring at times the death of old forest. In the living world, as in organizations, we also observe systems that reorganize at a higher level of complexity following external shocks throughout their lives (Henri Atlan). In this situation, there is a direct connection between environmental shocks and the internal mechanisms of self-organization.

The butterfly effect is a term sometimes used for a small input having a large effect, and we shall explore this further into this chapter. Of relevance here, General Stanley McChrystal (Team of Teams) writes that this term is commonly misused in popular culture, having become synonymous with “leverage” – the idea of a small thing that has a big impact, with the implication that, like a lever, it can be manipulated to a desired end. This misses the point of Lorenz’s insight. The reality is that small things in a complex system may have no effect or a massive effect, and it is virtually impossible to know which will turn out to be the case.

A specificity of complex systems is the existence of nonlinear connections. In linear systems, cause leads to effect in a way that can be planned and traced, and even though the path from one to another may be complicated, the known unknowns can be defined and built into a forward plan. In complex systems as discussed earlier, the path to the outcome cannot be predefined as it emerges along the way, only becoming clear when tracing back the journey. Gupta and Anish (2011) describe organizations evocatively as “webs of non-linear feedback loops that connect to other people and other organizations by webs of non-linear feedback loops”. This is a great metaphor for explaining why it is so difficult to trace how a process or idea grows, travels and changes over time.

Another thing that is useful to apply to organizations in dynamic environments is that cause and effect are proportional to each other in linear systems or processes, whereas in nonlinear cases, they are not, making it less easy to predict the impact of activities and the reactions that occur at both the individual and organizational levels. The explanation for this in system’s terms is a little technical, but we are all familiar with examples of organizational change or political unrest where a large action has caused very little impact, and also where a small action has had an impact way out of proportion, often called the butterfly effect.

The technical reason is linked to our previous discussion on the stability of organizations that at any given time, some parts of an organization are stable and other parts are changing. For example, one department may be undergoing a restructure, an IT system change or new operational equipment. This is called multiple states of equilibrium, and for this situation, there are always multiple solutions – one reason why strategy is more difficult and long-term plans will always be wrong! In linear systems, there is one stable state (equilibrium) and hence linear processes can be optimized; on the contrary, complex processes and systems, such as organizations and economies, cannot – something many theorists, consultants and leaders fail to understand, with negative consequences.

Thus, organizations are by nature paradoxical, travelling between drivers for stability and control and the opposite ones of decentralization and adaption. This is accentuated by organizations being a human system, as people require fluctuating levels of security and stability, excitement and innovation. Leading such a dynamic system has been likened to being the conductor of an orchestra who knows the whole music score and thus coordinates each instrument accordingly, or Meadows’ lovely description of “dancing with the system”. It is a critical skill because if the organization is too driven towards stability, then it will fail, as it cannot adjust and change as required, but if it is pulled into chaos, then it disintegrates, so a successful CAS lies at the border of stability and instability – on the edge of chaos.

3.1.2.2. The economist’s lens

The paradox is that chaos could bring more stability in the long-run as long as it can be brought back into a level of order. We can draw a parallel here with the concept of crisis in economics. Jacques Attali speaks of “order from noise” to characterize the crisis, which is undergoing systemic evolution through qualitative leaps. The expression order from noise comes from the theory of systems (Heinz von Foerster), and noise must be understood as random perturbation (or deviation). Such perturbations can help a self-organizing system to find more stable states in its “fitness landscape”. Unexpected signals contradict old logic and lead to a new form of organization through a systemic crisis. Attali (1976) explains that crisis is the installation site of order and not the exacerbation of disorder within organizations. The crisis is not a breakdown, but a process of resorbing imbalances accumulated throughout the previous period of “normal” operations. It is therefore system metamorphosis.

In the latter approach, order from noise, the weak signals may be local micro-crises that pre-empt necessary global reorganization, but that an inattentive observer would not pick up. At the international level, for example, Attali reminds us that large worldwide crises generally start with one country, maybe even a simple, local event, before spreading across the planet. Shrewd analysts can anticipate global metamorphosis if they decipher a sign that is a harbinger of global crisis in a weak signal, if they perceive a serious contradiction within the system as a whole.

To return to our concept of complex adaptive systems (CAS), we can reinterpret them in the framework of order from noise theory by stating that as part of the adaptation process, weak signals are created through the system reacting to changing conditions, mainly on the periphery of the system (in this case, where the organization interacts more directly with the environment, such as clients, vendors and partners). They are often mistaken for noise (irrelevant data, unacceptable interference) as they are not predictable. They are unexpected and often invisible to most actors whose perception is still being shaped by the strong signals of stated goals, rules, measurement of what is seen as relevant data and so on. They are structuring interference because they will bring about systemic evolution, which will take place according to the system’s self-organizing logic. Indeed, the system maintains its identity through metamorphosis: the butterfly’s genes are the same as the caterpillar’s. The ability to adapt manifests itself not by having seen the direction of the future in strategic plans, but by being able to pick up such weak signals, to observe and understand their importance and to move to act on them in a timely way in this dynamic situation. Put another way, it is the ability to reorganize according to a new “grammar” imposed by the appearance of some new “words”. This adaptation is the manifestation of system resilience, which is not inertia; quite the opposite, it is proof of the ability to change.

3.1.3. Taking a course

Since we are talking about human systems here, it is important to introduce the dimensions of choice and intention into our analysis, as we are reminded by Mitleton-Kelly (2003) in particular. The innovation process, for example, is a major case of human CAS at both the micro- and macroeconomic levels, characterized by the intention and will of the person Schumpeter calls the entrepreneur (see Chapter 4 of this work for other definitions of the entrepreneur). This project leader, the actions of whom may change an entire aspect of the world in the end, is a strategist. To further his vision, he makes use of information, skills, financial means and tactics to convince and lead partners. This human dimension is essential in the process through which a new order will emerge by adopting new ideas, objects, methods, attitudes and so on.

3.1.3.1. The necessary autonomy of individuals

In a system involving humans (alongside physical and virtual artifacts), there is one essential point to consider: all people seek to maintain a position of internal control over their thoughts and actions. If they are placed in a situation where they have no choice, this will create anxiety, which, in turn, not only triggers health issues, but also creates forms of disengagement, non-cooperation and even marked aversion towards the organization. The organization will not only lose potential creativity, but also suffer in its ongoing activities. A bad manager ignores the intrinsic motivations of other actors.

Good leaders, on the contrary, know that in order to ensure “intelligent” operations in the organization, they must create an understanding of the organization’s intentions and goals shared by all of its members (Kerr 2014). It is a powerful concept to manage the organization like a thinking system made up of individual human intelligence, which, if aligned, can create amazing outcomes shaped and driven by succeeding in a shared goal and intention. Creating a collective intelligence is part of the goals that governance must set for itself. When the vision is clear and shared in regard to what the organization is there to achieve, it is possible to develop the desired abilities to learn, adapt, innovate and be resilient.

3.1.3.2. The role of human interactions

Shared intention is created through interaction and connection. These are therefore central in the operations of the CAS. They predominantly take place through informal networks of individuals, which are based on relationships. This flow of horizontal exchanges is distinguished from the formal information and communication networks that are orchestrated by the hierarchy (Wheatley 2006; Seijts et al. 2010), and in an adaptive organization, both work well in terms of allowing people to gain the information they want when they want it, in the most useful way and at the right time.

Informal communication networks are essential in the management of weak signals, as often the information is not in a form that can be formally presented to those measuring what is “valid” at that time.

Liant (2007) analyzes intelligent CAS, capable of operating in environments rich in knowledge and requiring rapid reactions and learning. These organizations are also characterized by knowledge of how to co-evolve with their environment, hence a “smart survival” evolutionary model. For this, they must know how to renounce the bureaucratic fantasy of total control and allow the system to exist at the margins of self-organization (although with enough structure and rules that there is clarity around things such as purpose, values and operational learning). Complex thinkers understand that any attempt to force through vertical control will also lead to forms of self-organization, however these may be opposite to the desired direction, and in the less severe cases, will result in behaviors such as white anting (tricky sabotage), obfuscation, creatively getting around the rules and shadow systems.

One result of a state of allowing guided self-organization is that it brings out unexpected ideas, often through a greater ability to read weak signals. The art of governing these intelligent CAS is summarized beautifully by Lichtenstein and Plowman (2009, p. 628): “Emergence can lead systems and leaders can foster emergence.” For this, the design and implementation method of the organization’s strategy cannot be linear; for beyond the linear informational categories of known knowns and known unknowns (classical strategic planning works on such variables and tries to quantify a degree of potential uncertainty), there is the unknown unknowns category that exists in complex systems, created by the process of emergence. This state of “we do not know what we do not know” cannot be brought back into a predictable, linear category, but instead should be listened to in order to learn what is changing, what novel conditions are emerging, and thus how the CAS may self-organize around them.

3.1.3.3. The maturation preceding emergence

Weak signals potentially have the ability to spread throughout the entire organization and to become “common knowledge”, but initially they are hardly detectable to anyone other than those close to the source of such signals. They can only manifest themselves through complex interactive processes that are not apparent as at this early stage of emergent change as they remain informal. By the time the organization’s management is able to detect these new signals, recognize their significance and act on them, they have already been identified and somehow captured by a select number of people who know they somehow matter. They have an underground life in various types of groups.

Such groups do not have to be in a specific unit or from a particular area. Sometimes there is a shared neural network of knowledge (Kerr 2017) that leads to a shared understanding of what to look for in the changing landscape. This typically occurs at the intersections and edges of the organization where people interact with each other and the outside world in different ways. Thus, it is vital to allow such “boundary riders” to have a voice mechanism through which to communicate. In adaptive companies, these people are valued – they are the canaries in the coal mine and have invaluable insight.

Another group of people who are relevant here can be understood via the theory of communities of knowledge described particularly by Cohendet et al. (2006). Firms – particularly large ones – benefit from the cognitive contributions of informal groups built around a particular field of knowledge, practice or theory. These members of the organization voluntarily take part without official reference to the hierarchy in the construction and the exchange of a repertoire of resources in terms of information, knowledge and know-how surrounding a given profession or function; they construct a shared identity and behavioral norms particular to them. These communities greatly contribute to the maturation of new ideas. In a description of the firm as a cognitive system that is permanently learning, Amin and Cohendet (2012) define it as a platform of communities.

3.1.4. Navigating in symbiosis

Communities of knowledge are often organized around professions: computer programmers, machine repairmen, designers, scientists and so on, which can exist across different organizational silos or sections. Like people with a passion for an activity that is their hobby, these communities often spread through the “firm’s frontiers”, for people have exchanges among themselves with no particular respect for organizational perimeters and demonstrate greater solidarity between themselves than with their respective hierarchies.

This cross-connection is a rich source of adaptivity as it allows diverse ideas and information to be combined across hierarchical structures, and adaptive companies not only allow, but also encourage such cross-collaboration by rewarding the activities and outcomes that support it. However, in many of the past economic and hierarchical management models, this type of activity is not seen to meet the interest of firms because they do not recognize or reward the benefit they gain from the common construction of new ideas, practices and competencies.

From an economic viewpoint, the concept of creativity founded on externalities of knowledge was described for the first time by Alfred Marshall in the early 20th Century in his theory of industrial districts. Let us recall that for Marshall (1920), the ability of companies to innovate in the first industrial revolution is explained by external effects that belong within precise territories (like Manchester). The field of externalities of knowledge constituted by the first communities of entrepreneurs and workers was a creative factor shared by the entire sector-specific ecosystem. The history of Silicon Valley is also that of a Marshallian district, but in the age of globalization, as shown by Saxenian and Hsu (2001), the interpenetration of communities, which has led to an exceptional surge in major innovations, has spread to both sides of the Pacific, even though the location in California has played a key catalyzing role.

Such shared sources of value remain generally hidden, in the form of many weak signals, until the formal emergence of ideas. Explicit formulation is certainly an important stage for organizations’ management, but the underground creative factors were already at work, both produced by and impacting the system in as yet unrecognized shifts. Historically, numerous large enterprises have recognized the importance of the communities of knowledge phenomenon – under a variety of names and noting a range of facets of the phenomenon such as communities of practice for Xerox technicians (as defined by Wenger 2000), learning groups at Hewlett-Packard and knowledge networks at IBM. These reflections have upset the conception of managing change in an organization (knowledge management). If management by communities is practiced, as suggested by Amin and Cohendet (2004), then there is a certain degree of refusal to introduce creativity through a hierarchical approach (management by design) with its emphasis on quantitative reward and measurement. Instead, a learning infrastructure is created that motivates and lends value to the communities’ social dynamics. In this framework, work is done through intrinsic motivations (encouragement, symbolic compensation, nudges, etc.) rather than principally through extrinsic ones (hierarchical control, premiums, etc.), as we mentioned in Chapter 2.

The informal communities that exist in a CAS is one of the features that allow them to function as ecosystems (Attour and Burger-Helmchen 2014), maturing over the long-term by not only producing creative systemic modifications in their own ecosystem, but also co-evolving with the system within which they exist. Mittleton Kelly calls such organizations complex, co-evolving systems (CCS). Think of the image of underground roots shared by mushrooms (spreading synergies with other living creatures such as trees). The emergence phenomenon in an organization must often be understood as a self-organizing process within a larger system. In this highly interconnected world, all emergent behaviors are rewarded (whether positively or negatively) and interconnected via feedback loops, although not necessarily in a formal, official manner. The participants’ intrinsic motivations suffice, generally, to maintain motion. Seijts et al. (2010) recall the small world of actors, sometimes physically distant from one another, who communicate thanks to a global society that is becoming increasingly interconnected. This is the paradox of “globalization”, which causes certain local events to have much larger repercussions in a short time frame.

The smart leader’s art lies in detecting these weak signals and partially hidden mechanisms to surf on the unpredictable ocean of emergent information instead of trying to impose plans elaborated over a long period of time to try and control the system. Complex systems are nonlinear, with multiple feedback loops, high levels of interconnectedness and a range of both dynamic and static states across the system, all of which produce emergent behavior. This makes linear thinking at best oversimplistic and inaccurate, and at worst dangerous with regard to strategy and governance.

3.2. Managing interdependences and dancing with the system

We have already pointed out that communities of knowledge go beyond the limits of organizations. Furthermore, the complexity of the economic interactions has extended yet further with globalization. The current overlapping of companies’ value chains and their attempts to impose hierarchical organization (capitalistic ties, subcontracting chains, etc.) create a context where peripheral events can easily trigger a series of impacts spreading throughout the global structure and climbing as far as multinationals’ headquarters. This is the meaning of the small world proposed by Seijts et al. (2010). Here, the notion of a weak signal is described as the spreading of ideas and new representations that start at the periphery and not at the center. The signals are perceived as “weak” not only because governance did not plan them, but also because they are difficult to recognize by those who see change as a centrally controlled process, rather than as a possible result of peripheral events (sometimes happening at geographically and culturally very distant locations). Moreover, through various instances of translation on the journey through organizational channels, the initial message often arrives at the center in a very different form, especially where there is a lack of adequate measures and methods of identification and capture. All of this makes governance a delicate task, as the weak signals are both significant (they have already initiated change) and ambiguous. Interestingly, these signals are not ambiguous for those closest to their origin and hence these people recognize that something is changing. However, these people rarely have the agency to share their suspicions, and as there is not yet clarity as to what a subtle shift may mean, in a hierarchical organization that is commonly driven by quantifiable data, such intuitive insight is not seen as valid input.

Here we require a more precise analysis. What are weak signals and why are they weak? This concept is linked to the issue of sensitivity to initial conditions. The structure of the system and the general rules of the organization strongly condition the impact of such novel information.

Weak signals are small shifts that allow the new and novel ways in which a system may be adapting to be identified. They are weak because current practice rewards stability, and it contains strong connections and feedback loops to maintain that stable state, including processes and infrastructure that reinforces current methods, even in the face of mounting evidence that change is necessary. The weak signals, made up of what Aaltonen and Sanders calls perking information, have not been strengthened or reinforced in any way, either in individual thought or action or in the organization as a whole.

Perking information is made up of the small ripples that are taking shape just below the surface in emerging conditions and changes, but are not yet visible. They are especially difficult to see in emergent rather than planned change, as they are a consequence of local interaction among people at a local level, and are not as easily identifiable or monitored as in planned change. Yet they are the norm in a CAS. As local interaction is based on local principles, rules and beliefs rather than via a plan imposed by a CEO or a director, this reinforces the need for the leader to have created a shared mental model of purpose, goals and values to align and guide rather than attempt to control activities at many levels.

The signals are also weak because at an individual or whole organization level, more effort is commonly put to maintaining present practice than to acting on evidence and shift, and this can be explained in CS terms. Complex systems have a natural resistance to absorbing and applying information and forces that show a need to alter the present path, as it requires the system to make more effort than it puts into maintaining the status quo. This can be a biological CAS trying to maintain homeostasis, the equilibrium of technical CAS or the entrenched patterns of behavior in human systems. The feedback loops that reward behavior already in the system are well established and on automatic pilot, so any shift requires energy to pattern break, taking apart the current ways of thinking and doing and building new ways in their place.

3.2.1. The transmission of signals as a creative process: the example of composite materials

When we speak of signals that spread in organizations and between them, such signals are not concepts such as standard signal (in electronics, in Shannon’s sense). We can now see them as something more complex, emergent and qualitative. Weak signals are forms of knowledge that have evolved from one micro-context to another, like chains of translation from one language to another.

The adaptation of a message through translation is always an operation that modifies the meaning. Traduttore, traditore, as the Italians say. In reality, there is no “treason”, but depending on the case, simplification, enrichment, extension to a broader semantic field or, on the contrary, a more specialized one, occurs. If there is spread, it is not identical, but according to a process of adoption-adaptation at each stage of transmission, further altered by the local micro-environment.

We find the same idea in the theory of innovation diffusion (Rogers 2003; Mercuri-Chapuis and de Bovis-Vlahovic 2016). The first works by technological historians built models of homogeneous innovation diffusion in a sector and/or country using simple models like a logistical S-curve. However, if models like this can be used to handle cases like the progressive mechanization of agriculture, in many recent technological fields, we can observe more of a creativity process throughout the diffusion. At each state, the adopter adapts the technological idea to his or her problem, which causes the idea to evolve. Finally, innovation is created through the very fact of its diffusion. This is what can typically be observed with the emergence of technical composite materials in the 1970s and the 1980s: “The developmental progression of this technology is very characteristic of the progressivity with which technologies are generally created using a chain of technical solutions” (Zuscovitch and Arrous 1984, p. 185).

In the technology diffusion chain of technical composite materials, each new adoption of the general principle was also a chance for the adaptation of solutions tested in a new situation. Creativity not only precedes diffusion but also manifests itself throughout the chain. These innovations that can be classified as incremental contribute to the reinforcement of the technological paradigm as well. It is only after an entire phase of diffusion with multiple adaptations that we can say: “the principle of composite materials is a major innovation”.

From an economic viewpoint, another characteristic of innovation through the accumulation of adaptations is that it places the dialogue between the providers and customers of technical solutions at the center of enterprise strategy. CM are functional materials in the sense that they are designed to meet a demand for precise user functions. Based on this fact, interaction between organizations is essential. The subject of innovation is not an individual economic actor, but a complex inter-organizational system. The conclusion we can arrive at upon analysis of the spread of CM in the concerned sectors is that “competition plays more on the ability of firms to manage customer-provider networks and industrial partnerships” (Willinger 1989, p. 65). Current discussions on the concept of open innovation cannot help but revisit this idea – but based more strongly on major technological developments of the moment, notably digital and the Internet.

In complex systems terms, CM development was driven by the emergence of novel thought, which occurs when actors in the system are interconnected and self-organizing enough that information can flow as required along human networks, allowing them to apply diverse views and knowledge, learning and building layers of individual knowledge into the bricolage of a new concept. Open innovation incorporates many of these elements, but both outbound and inbound innovations need absorptive capacity. Cohen and Levinthal define this as “the ability to recognize, assimilate and apply relevant new external information” and they consider that the organization needs prior related knowledge to assimilate and use new knowledge. They call those who enable this to occur knowledge catalysts, who can bridge the organization and the external environment. An adaptive organization with minimal structure and the capacity for cross-connection also fosters such capacity.

Our conclusion in terms of signal transmission in complex systems is that the process cannot be planned or modeled linearly (expressed with diffusion curves for an unchanged object), but through creative interactions that modify, and sometimes amplify, the signal while spreading it.

3.2.2. The nonlinear changes at the source of evolution

A cybernetic metaphor is often used to speak of the cognitive evolutions observed in interconnection networks by suggesting nonlinear transformation operations. Gupta and Anish speak of “networks of nonlinear feedback loops” that connect people with one another and organizations with others of their type (Gupta and Anish 2011). The result is that the nonlinear properties produce qualitative modifications that cannot be predicted as they emerge over time, complicating the job of a manager who needs to maintain order and forecast results accurately. While the emergent path of change can be understood by looking back and seeing how the developments emerged, it is not possible to predict the path at the beginning of an emergent process. Thus, long-term strategic plans that describe the organization’s journey through space and time are illusory.

Another characteristic of complex adaptive systems that makes strategy and long-term plans inaccurate is that in such environments, cause and effect are not proportional to each other, whereas in linear systems, they are (a concept that shaped outdated “rational” management frameworks that saw organizations as both linear and controllable). A phenomenon that illustrates this lack of proportionality, and which is irritating for Cartesian minds, is the butterfly effect, where a small action in one place can have a disproportionate impact, even modifying the state of the whole system after a certain number of interactions. It also works the opposite way, with a large action having very little impact, again with little ability to predict beforehand. This is due to a phenomenon mentioned earlier, which describes the organization as containing multiple states of equilibrium (or flux). One of the consequences of the butterfly effect, which includes stability in some parts and metamorphoses in others, is that multiple states create the possibility for multiple solutions, rendering strategy and long-term plans precarious. Nor is it ever possible to expect a single, complete equilibrium in the system, as would be the case in a linear model.

Some managers or consultants still have a tendency to describe the system as linear, which is inadequate to say the least, even though they hope to do this “as a first approximation”. This is not necessarily a sign of incompetence but of linear thinking, and a lack of understanding that while identifying what the parts of a system do, be it a car or an organization, it is the understanding of the connections between these parts that show how the system works. For some, it is a way of affirming their power by gaining hold through a simple and thus understandable and communicable schema. We find this temptation in the political world, for better or worse.

3.2.2.1. The human characteristic

The butterfly effect is a metaphor initially used to describe chaotic systems in atmospheric physics. Human complex systems also illustrate this coexistence of stable and evolving parts within the system. Indeed, individuals across the system have varying needs and preferences concerning stability or instability – a key element for a firm’s innovation capabilities. Stability is preferred where security takes precedence. Innovation satisfies the need for adventure and excitement that is absolutely essential in the life of other individuals. The governance of organizations is complicated further by the need to take this supplementary complexity of variable preferences into account, often made even more so by hierarchical position. To describe the manager’s role, Meadows (2015) uses the metaphor of the orchestra conductor who “dances with the system”.

The metaphor of the dancing conductor is interesting to express the idea that managing a complex system formed by a number of creative interacting individuals involves finding the right balance, avoiding two major pitfalls:

  • – excessive stability, which would restrict and even prevent system adaptation;
  • – an excessive push for change, which would render the system chaotic through a lack of capacity to embed change and learn.

Good managers are able to think about complexity. This does not stop them from occasionally using linear discourse that, for the same reasons of political rhetoric mentioned above, can have a cohesion effect of creating a clear, simple vision around shared goals that should be able to be expressed simply. However, they must not allow their own rhetoric to lure them into a false sense of simplicity or premature success. This is minimized in complex thinkers who recognize and are comfortable with the emergent state that complexity creates, giving them a different view of how to handle common management challenges:

  • – The relative inefficiency of chains of command and classical motivational schemes is recognized, and instead such managers utilize and build in “soft” motivational tools like nudges. Dancing with the orchestra can then be interpreted as the preference for methods reinforcing intrinsic motivations through listening to the individual instruments and knowing the score.
  • – The ambiguity of multiple options and signals, which makes it difficult to control outcomes and define processes. This reinforces the importance of knowing how to interpret weak signals.

Vis-à-vis weak signals, managers in adaptive organizations not only gain the ability to use “radar” to watch and listen, but also understand the importance of creating organizational permanent radar coverage to capture and interpret such signals. A key attribute here is to not only resist the temptation to refute disconfirming information, but also actively look for it. Indeed, some messages that do not respect the linguistic rules of the organization’s discourse may be important to consider. They may be the ones that will structure the common language of the future.

3.2.2.2. Virtual stability

One of the fundamental aspects of complex systems that we have not yet described is their virtual stability. Richard A. Voorhees provides the following definition: “The ability of a system to gain in flexibility and become more maneuverable while maintaining its self-control to remain in a state that is normally unstable” (Voorhees 2008, p. 133).

These are normal states in nature, and they go so far as to constitute the norm in human systems. A cyclist’s virtual stability allows him or her to operate a technical object that is rather unstable to be on. A surfer, who seems to be a prodigy for common mortals, has developed a system of reflexes that permanently compensate for the countless destabilizing mechanisms of his or her paradoxical situation (staying upright on the sea with a slippery board). Managing an innovative enterprise and steering global finances are other examples of exercising virtual stability. As for weak signals, virtual stability involves taking potentially revolutionary phenomena into account while feigning that the system is working as usual.

Virtual stability has a price: the system must have enough resources to assume a high level of permanent reactivity, or in other words, it needs some slack built into the system – when the state of instability causes the cyclist to wobble, there must be enough time to correct the steering while traversing the road. Indeed, if this person is going too fast, or holding on too tight, he or she will crash! Ashby’s famous first law of cybernetics, that of requisite variety, states that the variety of possibilities for controlling a system must respond to the variety of outside disturbances. For Ashby, it is not necessary to provide more flexibility than is necessary to deal with all possible eventualities, or in the words of requisite variety – to control the spectrum of variety in terms of ordinary environmental fluctuations. However, some additional resources could be put aside for highly unlikely, but high-risk events with a very strong impact (we could call this the Fukushima clause).

Maintaining virtual stability requires finding a balance between expending too much and not enough energy to correct the small alterations currently in the system. The extreme situations that must be avoided are:

  • – a too high level of control (monitoring), which could monopolize attention on irrelevant signals and particularly limit the possibilities for innovation;
  • – a too low level of control, which could create instability through a lack of synchronization with external fluctuations.

We can draw the following lesson from the theory of virtual stability: “life is not about stability, it is about managing instability (so as to produce the illusion of stability)” (Voorhees 2008, p. 137).

This is a good intellectual model for the ideal “adaptive” leader who understands emergence, interacts with it, knows how to identify weak signals and can act on them.

3.2.2.3. Organizational inertia

The opposite of creative motion is inertia. This does not always have to be understood in a negative sense. The system’s resilience requires both the ability to adapt and the inertia to stay on course. However, an organization’s management constantly faces pernicious inertia that runs counter to its ability to adapt.

One example, somewhat paradoxical, is that of the optimism bias created by initial strategic success. Individuals, like organizations, are often blocked by hanging onto old ways of thinking about how things should be done due to them having worked in the past. This can occur to the extent that they do not recognize that the context has evolved and these ideas are now less pertinent. Fixed thinking minimizes the ability to see that tomorrow will probably be different from today. The cognitive bias caused by initial success prevents atypical signals from being perceived, those emergent peripheral weak signals of perking information that are important to recognize. What has not been understood in the case of an organization that remains overly proud of its initial success is that this only represents an example of a possible attractor for the system – to use the terminology of chaos theory, revisited by Irene Sanders. The system can perfectly swing towards another attractor, and it is best if the management is aware of this fact.

Another source of inertia lies in the games of internal actors, with coalitions that lead to rigid political equilibriums. Game theory has taught us that a game’s equilibrium may not be optimal – even in the very limited sense of economic theory (Pareto-optimal). In fact, in complexity theory, optimization is not the goal, as when one part of a system is optimized, it de-optimizes the rest and skews the capacity for system-level adaptation. Instead, maximization is sought through intelligently combining and aligning sub-goals while ensuring that one part of the system does not succeed at the expense of another. Inertia in games theory presents a paradoxical but stable systemic situation, resulting from the actors’ behavior, such that everyone would be better off changing states to move towards a new balance, although no one is interested in moving on their own (Umbhauer 2016). In this so-called prisoner’s dilemma, we can clearly see the eminent role played by informed managers if they manage to force movement through their hierarchical weight.

It shows us the need for strong leadership, supported by minimal rules and structure against which people can steer. Self-managed complex systems still have such things – they are not laissez-faire, disorganized mobs with no structure or goals, but instead, one of their key features is a clear, shared goal that is only achievable through each part of the system working with others to do so. As organizations, they are actively shaped from both the top down and the bottom up, with the leader clarifying and maintaining both the organizational purpose and the rules around the values used to achieve such purpose, thus setting the tone and creating trust (or not, in some cases). They then tend and nudge in order to ensure that things go in the right direction but they do not get in the way of innovative ways to get there. Only when people go against the clear, value-based boundaries will they step in and enforce their power. Thus, people in such a system have a clear idea of where they are going and what the boundaries are and can innovate within this space. With too much structure, or too little, however, inertia will occur for different reasons.

3.2.2.4. Chaordic operations of evolving systems

Regardless of its own cause, inertia is a cause of organizational collapse due to a lack of flexibility. It can even become quite schizophrenic where the dominant discourse is at odds with the reality of a system that has already begun to change. Interestingly, in these circumstances, it is common for the only ones who do not recognize this to be the leaders still pushing the dominant discourse!

“Chaordic” systems, presented at the onset of this chapter, provide a model for how organizations either avoid or fall into this kind of crisis. At certain moments in its history, the system moves from stability to instability along the path of change before stabilizing again in its new state. This is a necessary course for adaptation, provided that it does not remain in this state of ambiguity for too long. As emphasized by van Eijnatten (2004), complex systems have discontinuous, evolving trajectories, and the order/chaos duality manifests itself particularly in the phases of regime change where the system’s ambiguity is at its maximum.

In the phases of regular growth, the system is in a state of relative stability and modifications are incremental (see the textbox and figure below). Approaching the end of growth, the system seems more unstable. We can speak of bifurcation in the sense of system theory or a catastrophe in René Thom’s morphogenesis models (1979).

image

Figure 3.1. The discontinuous growth of a chaordic system (source: van Eijnatten (2004, p. 431)). For a color version of the figure, see www.iste.co.uk/heraud/creative.zip

In the critical phases, the organization is in “the eye of chaos” with multiple possible paths. When the system moves to such a strongly nonlinear regime, it is also particularly sensitive to exogenous shocks. This can be a crisis situation in the etymological sense of the term: in Ancient Greek, “crisis” is the moment in the evolution of a disease where the patient is lingering between life and death. If equilibrium is reached by adopting new ways of thinking and doing, the organization will survive the crisis and grow through a qualitative leap towards a level of greater complexity. A new cycle begins.

3.3. Surfing on the wave

How can we manage to forget old dominant signals or move them to the background, especially in a period of crisis, to keep our ears open for new ones? Within an organization, it is not enough to change the boss’s way of seeing things; everyone’s mental representations need to be shifted, old practices changed, loyalty to old methods and processes questioned and the new behaviors that herald the way forward in the transition phase must be accepted.

The most successful way to reinforce changes in behavior involves initially tackling individuals’ mental representations rather than dictating new actions to be taken. Unlike robots, humans must be mentally prepared for change. Contemporary psychology and experimental economics have taught us the importance of the phenomenon known as cognitive dissonance: a discomfort that occurs when a person’s actions do not align with his/her beliefs. There is an internal drive that pushes the individual to change in a way that aligns the two again. Such alignment to doing something new is consistently more successful when people change the way they think rather than what they do – once they accept the need for the change, they change their actions accordingly. In a similar vein, the bias towards the status quo (a relative preference to that which does not change) is discarded more quickly if people know why there is change. They can then be more creative and positive about the question of how to bring the change about as they have an understanding of the underlying purpose.

3.3.1. Preparing the actors means first listening to them

The dominant signals are constantly reinforced by learning loops and infrastructures, which are set up to correspond to current methods. Even when faced with evidence of change, these routines produce powerful inertia against organizational adaptation. However, the weak signals are already at work. As described above, perking information (Aaltonen and Sanders 2006) are facts and representations that are just below the surface, often created by local interactions between agents. These barely observable phenomena, like wrinkles on the surface of ordinary operations, bear witness to the existence of another barely perceived reality.

What management – truly wishing to observe weak signals to evaluate them can – do is establish a system that trusts and enables actors in a position to reveal them, knowing that most organizational system’s rules rarely encourage this. The first step in preparing actors is thus giving them the right to express representations that do not reinforce the standard model, and even go against the mechanisms for aligning attitudes – mechanisms that, we must not forget, lie at the heart of the very concept of an organization.

Artigiani (2015) showed that the largest successes enjoyed during historical battles came from leaders who did not use overly formal and rigid plans of action (like the Royal Navy’s permanent fighting instructions), but instead, allowed plans enabling dynamic changes in battle. This was possible because those who constructed the plans had spent many hours together building a set of alternative action plans based on common goals and values, and this shared knowledge of strategic alternatives allowed for easily broadcast changes in precise action. Building this type of flexible strategic skill involves largely renouncing fixed, comprehensive plans, as these are not adaptable when the environment changes faster than the strategy. It is much more a matter of building systems for reading the environment, observing local responses and understanding each other’s agreed signals. This means knowing how to decentralize responsibilities, accepting the right to err, and above all else – the primary point we wish to emphasize – allowing on-site collectives to construct their mental representations of goals and tactical means based on what they see and not what the comprehensive plan foresaw.

3.3.2. Choosing the right methods to design a strategy

We saw above that preparing agents for change begins with the freedom to form their own representations based on local information at their disposal. These local interactions allow them not only to gather and synthesize local information (the “acquisition radar” to find weak signals), but also to share mental representations and the values that will guide their action. This organizational method is more effective in complex situations than giving on-site teams formal missions. The objective is to establish a form of guided self-organization.

In practical terms, there are minimal tools available to organize the governance of organizations in a way that maintains ambiguity and identifies a future direction in a non-orderly environment. Aaltonen and Sanders (2006) proposed a classification of prospective methods based on categorizing them in a way compatible with the philosophy of complex adaptive systems. This is not the place to go into details on the nearly 30 methods found by the authors, but it is useful to look at some examples illustrating the typology proposed:

  • – The first category of methods corresponds to the engineers’ approach and largely attempts to remove ambiguity. These apply to systems designed ex ante and managed using straightforward rules: science and technology roadmapping, genius forecasting, relevance trees and so on. These are useful for linear transactional problems, but they contribute little to complex issues or the emergence of new ideas within the organization and in dialogue with it.
  • – The second category of methods leaves room for heuristics by accepting ambiguity. The tool typically found in this category is the Delphi method. Let us recall that this method involves consulting the system actors during a number of stages and returning information aggregated at each stage to gather not only their information but also their future visions and their judgment of others’ opinions, thus constructing a common representation. If this method does not manage to bring the actors together towards a common vision, it will at least have the interest of examining a set of contrasting visions from all sides and allowing the strategic debate to be organized on clear bases familiar to all concerned. Although Aaltonen and Sanders do not consider this tool to be typical of complex adaptive systems, we believe that it can help large organizations evolve in this direction. It is commonly implemented by an external designer and allows for some bottom-up control depending on how much the feedback is accepted and acted upon.
  • – The third category applies to an emergent system – through interactions between agents. Models exist to attempt to define rules for the future system. This is the domain of mathematic complexity (structural analysis, interactive scenarios, etc.). We believe that the drawback of this approach is not co-constructing the future system with the agents.
  • – The fourth category also applies to emergent systems, but working through heuristics. This is the domain of social complexity. It is a matter of participatory methods. Only in this last category can we find processes authorizing the emergence of truly new forms during operations by identifying future directions in an environment with no preconceived order.

The overview of known methods put forth by Aaltonen and Sanders shows that at the end of the day, there are only a few methods fully adapted to managing complex adaptive systems. Instead, the most common methods present a linear bias by imposing a consultation or calculation model. Ethics of complexity, on the contrary, would suggest seeking and preserving a wide variety of options for the future (without seeking perfection or working forever) by opening spaces where every actor can participate in a loyal and transparent spirit in order to collectively construct the future.

3.3.3. Choosing a good steerer

We can start with the principle that in order to steer a complex system, it is best to have a leader with the qualities of a complex thinker. How can we summarize these qualities? This chapter has strongly insisted upon the ability to detect weak signals – insofar as this atypical information is ambiguous and thus does not fit into quantitative analytical categories currently used by the organization. It is born of diverse input and diversity is critical for a healthy system. It is a condition of the system’s adaptability (as defined by Loasby 2000). In many cases, weak signals are even harbingers of the system’s future operations if its trajectory is recognized and supported, thus pushing it towards a new attractor (beyond the crisis) and creating a new path. From the standpoint of the organization’s structure and communication flow, it is best to reflect on whether there are current systems that can help the manager detect and evaluate these signals. However, if the head of the organization is neither capable nor inclined towards this kind of listening, it is an undeniable fact that they will not be acted upon! In any case, something will happen, but maybe not what the boss wants and what is in the interest of the organization.

The last part of this chapter poses the question of the ideal qualities of a complex adaptive system’s manager. Some essential dimensions are explored: at an individual level, training and career profile is relevant, but also complex cognitive capacity and behavioral characteristics related to the cultural context in which the individual and his or her organization evolve. We will make some digressions concerning language as both a metaphor in which we think of management and imagine the future and as a fundamental context of managerial behavior. Linguistic context formats our understanding of the world and the relationships between human beings, and perhaps the habit of knowing a number of languages prepares managers to detect and interpret weak signals.

3.3.3.1. Nonlinear thought as the foundation of an elaborate form of resilience

Successful leaders’ training should influence both their attitude and ability to deal with complex questions. Training that is strongly oriented towards classical problem-solving techniques risks creating a cognitive bias by reinforcing an analytical, quantitative approach only. Many professions (engineer, financial officer, lawyer, economist) feel comfortable with well-expressed problems for which an optimal solution can be found, and this often reflects their training and what is rewarded as an outcome. This skill is well adapted to simple and complicated situations, but not complex ones. CAS demand nonlinear thinkers, and for all but the most linear thinkers, regular exposure to complex problem solving; the agency to act in order to put their solutions into practice and learn from the outcomes increases the capacity for cognitive complexity (Kerr 2013).

Unfortunately, even where such exposure is distributed throughout the organization and thus increases adaptive skill throughout, it is rare for agency to also be widely distributed, so much of this learning is only done at the top. Indeed, the greater distribution of appropriate permission to act is a key characteristic of truly adaptive organizations. They not only task people with complex problem-solving and give permission to act (at an appropriate level of authority), but their infrastructure and processes also reinforce such engagement throughout the organization. This is rare in hierarchical organizations as it is seen as “messy” – difficult to control, measure and direct – yet we see from our Artigiani example that it allows for nimble adaptative response at all levels of the organization in the face of change.

The nonlinearity of systems creates emergent outcomes, occasionally translated as brutal changes in the operational regime. As described previously, the person steering the organization must be able to deal with flux, allow systemic mutation and not seek to shut it down, control it or optimize current practice. He or she must not be thrown off by the ambiguity of information in transition and the inability to make clear choices in highly uncertain situations.

To characterize the qualities of successfully managing a complex system, let us revisit the analysis from the previous chapter concerning system resilience. The steerer must think of the organization’s long-term future, shaping the path for the system’s metamorphosis. The old concept of ecosystem resilience introduced by Holling (1973) describes their ability to remain in an original state subject to disturbances when a number of attractors are possible, something more akin to homeostasis. The literature moved over time (see Folke et al. 2010) to include the possibility of integrating resilience, adaptability and transformability, particularly for complex socio-ecological systems, i.e. by introducing human parameters – implying intentionality and thus requiring long-term strategies. In this broadened framework, where actors take the initiative and risk instigating profound transformation, the leader’s visionary abilities involve understanding the pull of inertia to remain the same, and encouraging people to “step into the next adjacent”, thus, over time, allowing the system to change to another possible state. In this case, the primary quality of managers is not prudence, but a capacity to inspire trust in people in order for them to take such a step into the unknown. They have an ability to clearly frame their goals and ideas, and they take people with them through to the end by convincing them that the future will be different but safe. They believe in their vision and thus are able to convince others. The ability to inspire and convince is an essential skill of leadership.

Understood in its usual sense, resilience is sometimes used as a synonym for force of character. Resilient people know how to overcome life’s challenges and more specifically, they engage with the potential risk of failure when they undertake individual or collective projects. Not letting themselves be thrown off balance, learning through their failures: these are a priori the characteristics of the resilient individual. However, complex systems theory adds to the concept of resilience with regard to steerage. One who steers well knows how to stay on track through the tempest, even when it means changing course. We have already seen that when they are faced with multiple attractors during a crisis period, systems are in a very strong state of instability and ambiguity. It is thus not simply a matter of sticking to a plan but also of adopting an open attitude vis-à-vis regime changes. As such, obstinacy is not a successful quality of leaders if not presented with the pursuit of bold ideas founded on visionary analyses (which may be revised). Indeed, interpreted as a bias towards the status quo, such a mindset is the opposite of the flexibility that is required. The resilience of complex adaptive systems allows them to remain viable precisely by adapting and transforming.

3.3.3.2. Knowledge angels

The central message of this chapter is that managers must know how to decode weak signals. We have specified the meaning of this expression by explaining that their “weakness” is not necessarily an intrinsic characteristic of the signal, but rather its reception by the system in its current form. As such, a highly sought after quality in those who steer is knowing how to decipher snippets of information in the form of a new and unusual code. For a sailor, deciphering fragments of nautical charts in new waters is such a challenge. The more experienced the sailor, the more able they will be, as they will have built a robust body of knowledge that their brain can call on to compare and contrast patterns, and thus make informed predictions.

We have also discussed the adaptive value of people at different levels within the organization being able to tackle complex problems, identify weak signals as and when they recognize them and have mechanisms to feed them into tacit and/or explicit knowledge flow. As noted, such signals are often picked up at the intersections and borders of the organization, as this is where the system interacts with an external environment.

This leads us to considering the external environment as another source of such radar, as those outside the organization can compare and contrast what is occurring within and without. This may be the new employee who brings in different experiences and mindsets, although they often tend to be discouraged from such observations and encouraged to “adapt”. Consultants are another source, and they are indeed a mixed bag. Many have cookie cutter approaches and either little capability or too little time to observe, dance with and understand the system. However, there are other external sources that are of value in deciphering weak signals and assisting adaptation.

We hope to illustrate this theory of deciphering with an example observed in a particular sector and a particular type of consultant who contributes to innovation at enterprises. The notion of KIBS (knowledge-based business services) was set forth by Ian Miles in the 1990s to characterize business service providers who impart elements of knowledge and lend advice regarding engineering, management, economic analysis, finances, law and so on (see Miles 2005). Muller, Zenker, and Héraud (2015) observed an international sample of companies in this B-to-B sector to detect the particularly creative individuals within them who contributed to each business’s success – as well as to the success of clients who benefited from this advice to undertake and manage projects and innovate. The essential function of these people, known as knowledge angels, was knowing how to contribute the right information, idea or knowledge at the right time to bring about an innovative development or a change in the strategic course. Such people also tend to have enthusiastic and creative temperaments, as well as having developed their activity in various contexts – successively or simultaneously. These are people who are well aware of how to decode weak strategic signals because these signals are only “weak” in the current context and with their current client, but not elsewhere. We note that knowledge angels are similar to the concept of knowledge catalysts described previously in the section on absorptive capacity.

Just like the experienced sailor described above, the message that goes unperceived in the enterprise they are evaluating is sometimes detectable by consultants who have seen similar signals elsewhere and thus have a cognitive map to refer to. They can also recognize the similarities across different environments, such as other clients, other organizations like basic research centers or various creative communities.

Here are the typical qualities of knowledge angels (KA) detected in a survey:

  • – ambitious character and will to fulfill themselves through projects;
  • – playful character;
  • – a wide variety of interest;
  • – ability and enjoyment in taking on multiple tasks;
  • – imaginative and potentially visionary character.

Because the survey was carried out in various corporate consulting sectors and in multiple countries, we can generalize the usual professional profile of KAs from the sample:

  • – they occupy a relatively high hierarchical position but are not necessarily the head of the organization, because they only moderately enjoy exercising power; they actually prefer influence to hierarchical power, and their personal satisfaction is connected to project management, not general management;
  • – they lend a great deal of importance to trust as a form of relationship. Teamwork (in small teams) is their preferred framework to act in;
  • – they evolve or have evolved in various professional and intellectual contexts, and their comparative individual advantage is being idea ferrymen.

We have also managed to observe differences between national (and therefore cultural) contexts for the perception KAs have of their role:

  • – in France, they tend to define themselves first and foremost as creative “idea givers” on behalf of their clients;
  • – in Germany, they see themselves more often as intermediaries on the idea market, “knowledge brokers”;
  • – in Canada, they consider themselves to be “business pushers”, emphasizing the entrepreneurial dimension of creativity;
  • – In Spain, they mention their role as a “facilitator”, in the sense of an intermediary between actors, institutions, etc.

The few contacts made in Asia seem to indicate a certain difficulty in aligning the notion of KAs with their cultural and institutional reality. In China, respondents saw themselves more as solution providers, which does not completely respond to the idea of creativity (providing breakthrough responses). The ideological framework remains relatively linear: for each problem, there must be one solution a priori in a known repertoire.

In Japan, there is a different problem. Faced with the idea of KAs and the examples already seen, the reaction of Japanese respondents is as follows: the model is seen as very interesting, but unfortunately impossible to apply. KAs will be poorly received because they contradict the Japanese concept of collective identity. Although a sense of belonging to multiple communities or organizations is tolerated, even encouraged in the West, it is considered bad form in traditional Japan, where individual identity is meant to be strongly shaped by the group. This has not stopped Japan from becoming a very innovative country, but the processes and methods of innovating are not the same as the ones observed in the other model countries. The concept of breakthrough thinking and radical innovations are difficult to accept, particularly when brought about by individuals, who should at least bring about advances in an anonymous way.

Although they are limited to the specific domain of the enterprise (KIBS), works by Emmanuel Muller and his partners to bring about the concept of knowledge angel are interesting to consider when considering individuals capable of steering complex inter-organizational projects. In particular, we can stop to conclude that the “steerer” of change is not always the head of the organization. Among the companies interviewed that belong to the KIBS category, only a portion of these mentioned their boss as the breakthrough/innovative thinker.

This means that there are many situations as described throughout this chapter, where the manager is better advised to listen to what a subordinate worker or partner has to say, as this person is filling the role as a visionary. Better still, they build infrastructure that not only allows but also encourages such information flow. In doing so, the boss demonstrates a form of meta-competence – wisdom.

3.3.3.3. Navigating between languages

We have used the metaphor of language to speak about the concept of weak signals, and we shall now take linguistic skills literally to illustrate the skills necessary to steer a complex system and render it adaptive in the long-term. Interestingly, many leaders are polyglots, so what can this skill contribute beyond its obvious concrete utility? Indeed, languages are simply useful operational skills in management, just like technology, law, accounting and so on. However, the mastery or simply the learning of foreign languages has cognitive similarities to an essential meta-competence for managing CAS.

Speaking a number of languages (two is already a good deal) means being able to interpret the world and express one’s ideas in multiple cognitive contexts. This consequently increases the intellectual agility, considering multiple points of view. The habit of getting by even in languages that speakers have not perfectly mastered like their native language is an asset insofar as the person is psychologically prepared for ambiguous situations typical of complexity. Polyglots tolerate fleeting uncertainty in information better a priori. They will tend not to suffer blockages as often, not to wait to understand everything before taking a somewhat new course. As such, they are patient and know how to take limited risks by trusting progressive learning of meaning.

Why? Let us turn to the brain for one aspect of the answer. If we look at the cognitive process of learning a language, there are many interesting aspects regarding the way in which it increases cognitive complexity. First, it makes the brain work harder, thus increasing the strength of our willpower to not just focus on a quick answer or habitual response. The brain has to actively suppress the dominant language with each word, so it gets better at working instead of being on autopilot. This “fitter” brain means the person can bring to bear more cognitive power and apply more thinking power to a problem. The reason why this is important is that our brains are both efficient and lazy, and they will not work if we are not excited or compelled enough to engage willpower to make such a cognitive effort. However, willpower is like a muscle – the more it is used, the stronger it gets, and the better we get at bringing more of our thought processes (cognitive resources) to bear a problem. This is not only true in the case of adults who are trying to deal with complex issues, but is shown in all sorts of situations, such as studies of first-grade children who are better at math when they take the bus than if they are driven to and from school. The reason is they have to use willpower to be where they need to be at the time required and work out the details necessary to catch the bus. That same capacity allows them to apply more cognitive effort to work out the mathematical problem and solve it.

Language acquisition is also the gift that keeps on giving, for each time we have to translate, the brain has to work. As such, learning one or multiple languages means building new real estate to hold the information (we increase cognitive density or “reserve”), increased perceptual judgment and more use of executive function, all reasons why language acquisition in adults is a recommended method for minimizing dementia and has been measured to protect against symptoms for 4–5 years.

Another key advantage is that it is not only the effortful memory of learning words and symbols that occurs when learning a language. It is also the embrace of a new culture and a different way of seeing the world, as the words in a particular language represent how we frame it. This is also why people learn faster when stationed in the country where the language is spoken – it is not just that there is more opportunity to practice, but they are immersed in this novel culture of different behaviors, rules, food, relationships, esthetic preferences (art, music, dress, beauty) and so on. All of these things change the way our brains are programmed, expanding these areas of our own awareness and memory.

Embracing the culture and the language creates a more flexible concept of reality instead of the often quite rigid one we have built when only exposed to one language and one way of life. Such exposure and the added conceptual richness that comes with learning a language that captures new ways of seeing the world are obvious elements that build the polyglot’s capacity for immersive adaption and indeed enjoyment. Now let us turn to the role of language itself.

The French language is pervaded by Cartesian culture, which wants every statement to be as clear as possible. The Japanese language does not make the same demands. It is common for sentences not to have subjects – each time the subject is logically superfluous because it can be deduced from the context. In some cases, this can pose problems; however, this is culturally expected. In French, the speaker’s minimum courtesy is to be precise; in Japanese, the responsibility for clearing up ambiguity lies much more on the listener, hence an occasionally visible form of panic when a Japanese man or woman is not sure to have understood instructions correctly, whereas speakers of French or English do not hesitate to ask for clarification.

We can also analyze the question of the subject and intentionality through languages. This is important with regard to complex systems, where matters of subject, causality and responsibility pose problems.

Let us imagine that the pilot of a ship asks a sailor to describe the state of their environment (the sea) at sunrise to him:

  • – The Japanese sailor could respond [天気はとても良い] (Tenki ha totemo yoi). The approximate meaning is, “regarding (は) weather (天気), completely (とても) beautiful (良い)”. Renowned French grammarian Vaugelas (one of the first members of the Académie Française, † 1650) would not have appreciated our translation! He would have chosen to add at least a subject and a verb to the sentence. Most of European languages require the same. Nothing like in Japanese. The presence of a verb is not systematic (even a function as essential to the verb in European languages as expressing tense can be ensured in Japanese with an adjective). Since the subject is not necessary, either, we see that the subject–verb pair, so fundamental to our understanding of the world in the West, is much less structuring at the other end of the world. Logically, this cannot be neutral in the way people think about systems.
  • – The English sailor would say, “The weather is very good”, which respects the European discipline of having a subject and a verb. It is clear and pragmatic. Let us simply note that, logically, “the” is useless, because the question of definite and indefinite is moot here, yet the language demands the use of an article. The German sailor would say exactly the same thing: “Das Wetter ist sehr gut”.
  • – The French sailor’s statement is more surprising: “il fait très beau temps”. Literally, we are suggesting that someone “makes” the weather. In fact, there is a purely formal subject and a verb that clearly does not express an intentional action in this case. We are satisfied thinking that this is an idiomatic expression and that this does not change anything at the end of the day, but the question remains: how did the French language, throughout its evolution, end up with this formulation? Why not simply say, “The weather is very good” (which is, perfectly possible, but less common)? It is clear that language expresses and reinforces cognitive biases and that French, as well as English or German culture, through the ages, has pushed mental representations towards values of subjectivity and action. Speaking in terms of subject and action is maybe not the ideal mental representation of systemic phenomena.

As the preceding linguistic digressions have shown, learning to speak a number of languages broadens people’s horizons by making them aware of their own representational biases. For example, we will ask the question of knowing whether the mechanisms observed in an environment are more causal and relatively linear or, on the contrary, more systemic. What is true individually for an adaptive leader who speaks multiple languages is also true for the adaptive organization. Intentionally building a culture that contains and rewards diversity of thought and approach, and a tolerance of ambiguity and difference, is a key element of innovation and adaptation. Looking at the role of language in increasing such characteristics was indeed a digression, but it shows the complex nature and advantages of recruiting members with diverse cultural and/or linguistic backgrounds, or at least with exposure to different settings and ideas, to better carry out the function of seeking weak signals. We know that innovation is a bricolage of ideas that already exist, layered upon each other in new and novel ways, so the greater the diversity of thought and experience, the more creative and innovative scope the organization has.

3.4. Conclusion

The notion of a complex adaptive system has led us to question the classical approach of rationality and to consider the role and skills required of those who lead and manage complex organizations in a different way. One of our conclusions can be formulated as follows: a successful leader and manager cannot control but can only steer a system that is largely self-organizing. This does not take away from the role which is still vital in orienting and clarifying where people are going and why, and tending them on the journey of adaptation, but it changes the approach and the journey.

An essential task for those who steer is knowing how to correctly identify the challenges facing the organization. To tackle this in an adaptive way, instead of building a hierarchical, vertical management structure, they must establish an internal environment capable of intelligently evolving along with the external environment. Their objective will be to establish strategic intelligence distributed throughout the system, capable of perceiving and welcoming weak signals. These signals will be often be hard to identify, confirm and capture with methods and goals of the current system. The more static and quantitative the norms around such things as the organization’s identity and its reference language, the more difficult it will be to escape embedded cognitive routines and to both identify and interpret weak signals. Yet if this rich information source is not only recognized, but encouraged and built in, then it does not endanger the organization, but quite the opposite.

This type of dynamic flexibility is also related to the building of resilience, whereas some organizations unfortunately build an ability to resist change. Organizational flexibility becomes crucial in periods of flux, change and crisis. These episodes, which are necessary for long-term survival, are highly unforeseeable situations that are often uncomfortable at the time due to the natural anxiety that change creates in people. During such periods, organizations on the whole have not built a cultural, cognitive system with adequate language that gives meaning to many of these situations. Most have also not yet embraced the challenge of building a structure and system that enables adequate self-organization, within the boundaries of simple structure and rules to steer against and imbed learning, while still allowing sufficient cross-connectivity for creative curiosity.

When we speak here about flexibility, this is not the static, planned contingent flexibility of having alternative plans for situations known in advance, but a more dynamic characteristic, which allows the system to reinvent itself while being tended adequately.

Steering a complex adaptive system requires both a great deal of ambition, insofar as leaders will occasionally bear the responsibility for major organizational changes, and a great deal of self-awareness and even modesty through the recognition of how much knowledge will at times be lacking when making decisions. When this is the case, complex thinking managers know how to embrace, harness and rely on the collective intelligence in their organization – which implies that they have cultivated this beforehand. Together, the organization and its management will know how to keep their eyes peeled for weak signals that emanate from the system itself and/or its environment.