In the second half of the last century in multiple domains of research the insight grew that a classical causal-mechanistic way of thinking cannot explain all natural phenomena – much of what we see exists without a final purpose, without predefined steps and unfolds through movements that happen both simultaneously and sequentially. This was the birth of complexity theory, that in de midsts of the 1990s first was recognized as a science in itself, with its own rules and scientific framework. Still, complexity theory is a field of research that is essentially interdisciplinary. Models from different domains of study complement each other, all highlighting different features.
In this chapter we explore three theoretical models on the dynamics of complex systems throughout time. One stems from synergetics and is often used in CTC, the other two stem from ecology and are applied to describe complex processes in multiple domains (Davis & Nikolic, 2014). One by one these theories will be projected on the displacement of stuff in order to discover whether they can adequately describe our observations. This lens generates a vocabulary to describe a variety of elements and events.
One of the basic theories on the process of self-organization stems from synergetics, an interdisciplinary science that from the mid-eighties researches the origination of patterns and structures in open systems. In the book The Science of Structure (1984), founder Hermann Haken, physicist and mathematician, proposes a model on self-organization defined by circular causality.
The process Haken describes starts bottom-up, as a local interaction between parts in a system out of its (thermal) equilibrium. A local fluctuation, which can be small, causes nearby parts to rearrange themselves and pattern starts to form. This global structure is represented in the figure by the top line and is called the order parameter. The individual parts gradually loose degrees of freedom as this overall structure forces them to rearrange, a process Haken terms enslavement (in this text referred to as structuring, because of its more neutral connotation). Without any disturbances this constant oscillation between top-down and bottom-up forces is a self-reinforcing process, a circular causation.
Different from thermodynamics, the complex systems this model describes are defined as open; they stay in a continuous interaction with their environment and exchange energy, material and information (Haken, 2012). The influences the environment poses on the system are included in the theory by the concept of control parameters. These are external factors, such as incoming light, a change in temperature or a material added to the system that has the ability to cause a systemic phase transition (Portugali, 2011). This changing condition thus alters the order parameters within the system itself until it finds a new mode of self-organization, a new balance.
The development of an order parameter, that in the process becomes more rigidly structured and gains an increasing ability to control, is the most characteristic process of self-organization. In a house, this can be found in the development of rooms or parts of rooms that often co-evolve with a certain activity; during the process more and more stuff is added to the place, which reduces the probability that other activities will be initiated there. This is especially visible just after moving, when both stuff and function quite quickly find their place. A clear example from the data set is the ‘office’ part of the room. Whereas the table was first relatively neutral, and would occasionally be used for dinners, model making or gift wrapping, it slowly grew specific. Throughout the year this ‘office’ has attracted a lot of stuff around it; first a tomado rack, a light and a magnet board, then a set of nightstands holding supplies, a paper organizer and a poster to block the light from the window, next a pegboard with office supplies and display for letters and booklets, a pile of A3 paper to pick from, speakers for music and most lately some boxes for desk organization. This structure is even so controlling that, even though there are chairs on the other side of the table, it would feel unnatural to change sides. The order parameter is in this case the identity of ‘desk’.
The stronger, the less likely to move the table around or use it for something else and the easier the desk attracts stuff. When tidying up the room and finding some item that belongs in the right category, such as an eraser, it is logically moved to the ‘office’, even if there is already an eraser and it would be more of use somewhere else.
Stuff systems are in constant interaction with their environment; the conditions of the surroundings – light, sound or silence, or temperature – can, just like in thermodynamic systems, exert pressure on the organization in the system. These control parameters can cause the self-organizing system to either adapt itself to overcome these changes or completely collapse. In other words, it changes towards a critical point, which, when reached, induces a rigorous transition.
A change of conditions is a frequent occurance in a house. When someone is reading in the garden, but after dinner it has become dark, the activity can initially not be continued. However, the stuff system might prove resilient enough to withstand this challenge, e.g. an outdoor light is taken from the shed and added to the cell. When the next day it turns out to be too cold to comfortably sit in the garden and attempts to warm oneself up prove not to be satisfactory, the activity is either discontinued or restarted somewhere else. Both the place and activity loose some of their stuff, thereby some of their structure, and are again more susceptible to bottom-up influences. The order parameter builds up anew.
Additionally, stuff systems are open systems, because of the constant exchange of individual parts with the rest of the world. New items enter the system and existing ones are thrown or given away.
The model of synergetics is able to give a complex description of the related Diderot effect, a phenomenon in consumption theory described by McCracken (1988) and named after Denis Diderot, who wrote about it in an essay (1769). After buying a luxury new robe, that did not fit the rest of his more ordinary collection of stuff, Diderot felt the urge to replace all his belongings with higher-quality substitutes.
“ My old robe was one with the other rags that surrounded me. A straw chair, a wooden table, a rug from Bergamo, a wood plank that held up a few books, a few smoky prints without frames, hung by its corners on that tapestry. Between these prints three or four suspended plasters formed, along with my old robe, the most harmonious indigence. All is now discordant. No more coordination, no more unity, no more beauty. “
– Denis Diderot.
From: Regrets for my Old Dressing Gown, or a warning to those who have more taste than fortune, 1769
The effect works in two ways (McCracken, 1988). Firstly it can constrain the consumer in what they buy, as only items that fit the overall style, aesthetic or level of quality of the whole are considered. Secondly, when a deviant item enters the collection, it can change the overall order so that every other item feels as if it should be replaced.
In the first case the existing parts fit so perfectly together that their order parameter is rigid enough to enslave all possibly newly bought parts. When these cannot obey to the existing order parameter, they do not enter the system at all.
In the second case, the newly obtained item is the start of a new order parameter and thus takes over the structure of what binds the elements together, starting a self-reinforcing effect of enslaving the other items because of their specificness in level of quality, or style, shape or color.
Another widely used conceptual model in complex system theory is that of the adaptive cycle (Holling, 1986) that highlights the dynamics of systems throughout time. This model, originally stemming from ecology, explains how complex systems not only self-organize into a solidified structure but also ‘experiment’ and innovate.
Any complex system continuously iterates through four phases, or ecosystem functions: the exploitation phase (r), the conservation phase (K), the release phase (Ω) and the reorganization phase (α). In general the trajectory alternates between longer periods of slow accumulation, the building up of a structure (from exploitation to conservation, r to K), and shorter periods that create opportunities for new innovations (from release to reorganization, Ω to α) (Gunderson & Holling, 2002). During the processes of accumulation generally stability increases and often material builds up, in the example of ecosystems in the form of, among other things, nutrients or biomass (Holling, 2001). As the system matures, it commonly becomes of more value for other systems of the same scale as it develops its possibilities and network. After a certain amount of time, however, it is unavoidable that resources will be depleted, together with the fact that the connectivity and therefore dependence of the system on its niche grows more and more rigid. In a moment of release the tight organization is lost and a more chaotic and free state of the particles make room for a new structure to form or, in the terms of Haken, makes room for a new order parameter to arise. In this phase of re-organization, or the early stage of self-organization, random coincidences can highly affect the path of the system. Within the chaotic surroundings it is easily possible to attract particles to follow a certain structure.
The previous paragraph described how a stuff cell in a garden survived the lack of light through the addition of a lamp, but collapsed when a drop in temperature could not be managed. The stuff cell was in its exploitation phase, as it already developed a structure with useful emergent properties (i.e. the possibility to comfortably sit and enjoy a book). When the first challenge occurred, the system could conserve itself through the addition of more stuff. As described by Gunderson and Holling (2002), the system is conserved for a longer time, through the accumulation of material.
In the second case, the system was too rigid to survive the changing conditions – thus reaching a critical point – and collapsed. It re-organized itself in another place, and both the garden and the system around the activity found themselves in the open and volatile phase that is typical in this new beginning. In this unstable state the system is open for suggestions, local interactions changing its path. When there happens to be a pen, a box of tea or a standing desk, for example, there is a chance it noticeably steers the course of the activity and thus changes the stuff cell, whereas in the initial situation actively obtaining those items would not have come to one’s mind.
In the introduction Holling’s term panarchy quickly passed in review. This is a substitute word to ‘hierarchy’ (which sounds too top-down and determined for the matter) given to describe the (vertical) nestedness of adaptive cycles of different scale levels within each other; think of a cell, a leaf, a branch, a tree and a forest. An exponential relationship (hence: Powers of Ten) between both the size of the system and the lifetime of its adaptive cycle seems to apply in many cases. The smaller systems are faster and the larger systems are slower. Fig 3.12 shows this concept on a logarithmic scale in both space and time (Gunderson & Holling 2002).
Two things distinguish the panarchical representation from the traditional hierarchical ones (Holling, 2001). Firstly the fact that the whole of it is intrinsically dynamic; it consists of widely varied adaptive cycles, which are sometimes rigid and domineering and other times open for innovation and renewal. Secondly because of connections between different scale levels, described with the verbs ‘revolt’ and ‘remember’.
Revolting, rebelling or upward causation of a smaller system happens when it enters its Ω-phase; the collapse can cascade to the larger and slower level and trigger a crisis. This is most likely to happen when the larger system is at the end of its K-phase, as it is vulnerable already.
Remembering, or downward causation, is when the structure of a larger and slower system defines the action of the smaller and faster system, particularly after its collapse, in the chaotic and unstable moment of re-organization. If the larger system is in its rigid and strongly structured K-phase, it is especially influential.
From the other side, smaller systems can revolt against a settled system, especially when this is in the unstable phase in which it is about to collapse. The Diderot effect as described in paragraph 3.1.4 is a clear example of this, as one small item can because of its specific style, aesthetic or color, force the entire system to find a new balance. This effect can also be seen in the configuration of furniture, especially when a living space is limited; the movement of one piece of furniture or the initiative of one small stuff cell in an unusual place triggers a cascade of displacements, possibly leading to a completely new room arrangement. In other words, one can say that the collapse of a system is brought about by a shortage of stuff or space at a specific point, which can be pushed by one single event – a straw that breaks the camel’s back.
Also in stuff systems, we can see that different scales of organization influence each other in the way described above.
Remembering occurs when a larger and slower system is in its highly developed K-phase, something that in a house can be clearly seen in activities that are paired with very specific stuff configurations. These actions, such as sleeping, showering, brushing teeth or, in the case described in paragraph 3.1.2, working in the home office, are often repetitive and can be referred to as highly scripted behavior (Abelson, 1981) in which the action (and therefore, the interaction with stuff) is intuitive, habitual and embodied.
When a more spontaneous activity, such as sorting out clothes, needs a large surface, the bed is an easy and logical place to partake. As long as the activity is performed, the stuff cell will maintain itself. However, at bedtime both the stuff cell around the bed and the activity of sleeping, that together form a larger and slower system, will ‘remember’ the place belongs to them; it is highly unlikely that this disturbance will cause the inhabitant to sleep somewhere else or not sleep at all. A similar event happens when in the midst of cooking, an activity that mostly consists of a series of smaller stuff cells in the kitchen, the table is used after putting smaller stuff cells (reading the newspaper, filling in crosswords) aside.
It seems that this planning behavior, which is associated with chronesthesia (Portugali, 2011) in stuff systems mostly occurs in higher scale levels. The placement of furniture is something of which various options are consciously considered while imagining possible futures. Stuff cells around this placed furniture, however, are more intuitive. The exact position of stuff on a table when working is not mapped out, but seems to be given by direct behavior.
In reflection we can say that complex system dynamics quite sufficiently describes the self-organization of stuff, yet under one condition. For order to emerge in a way defined by both bottom-up and top-down forces, a two-way interaction needs to present. This necessary ‘feedback-sensitivity’ is present when either the system is substantially larger and slower than the planning and controlling of individuals can reach (such as in a city), or when the interaction is so direct, ad-hoc and intuitive that long-term planning is not even considered. The latter is visible when people create the space they are immersed in by direct action, such as in the flux of (small-scale) stuff.
In the next chapter the basic understanding of stuff system dynamics though the singularly complex models given in this chapter will be extended into a doubly complex theory, in which the cognitive capabilities of humans are embedded.
Holling already projects this model on different domains, such as physical systems, meteorological systems, biological and ecological systems, and also political events or businesses. His conclusion is that although the model remains a sustainable explanation, the exercise to find exceptions does lead to different variants with for example an unusual rhythm or an oscillation between two phases (2001). Amongst them are human systems, that are defined by ‘foresight’.
The three elementary models on complex system dynamics in this chapter describe material and organic complex systems, in which the parts are not able to (directly) adapt their local rules through conscious decisions. Since stuff systems are human systems, the fact that a human being is conscious of what is happening around them has to be taken into account. The most important human cognitive ability in this context is that of chronesthesia, the cognitive ability of humans to be consciously aware of subjective time (Tulving, 1983). This enables to mentally travel back and forward in time, recalling and imagining events. More than an ability, chronesthesia is a default function of the brain and impossible to not do; people spend up to 50 percent of their awake hours thinking about events in the past or future (Raichle et al., 2001).
Awareness of the dynamics of a situation and the possibility to imagine possible futures can alter the model in a variety of ways, which is a study on its own. In general and as suggested by Holling (2001) people tend to both stabilize variety and exploit opportunity. Translated back to the model, this means that the exploitation and conservation phases are kept alive as long as possible, thus actively preventing a collapse through targeted action (fig 3.15). This is for instance done by increasing the resilience of the system to changing (climatic) conditions (i.e. by timely purchasing a fan or heating system) or averting rebellious parts to enter the house (i.e. not impulsively buying a bright red couch). Of course, when consciousness is high and the system is well understood, the dynamics can be steered even more dramatically by, for example, triggering a collapse, or directing the rise of an order parameter in the re-organization phase by giving forethoughtful hints.