In a house that is being used, stuff is constantly getting reordered, not by its properties, but in a dynamic relation to the other particles around. The items are moved through ad-hoc use – quick, intuitive and for direct purpose – and form bonds with each other, that are most of all functional. The toothpaste finds the toothbrush, the cushions find back the couch and stuff cells arise that contain all the functions needed for a certain activity. The system does not fall into chaos, but comes to life.

In this chapter the idea is introduced that stuff systems are complex systems with self-organizing abilities. Stuff systems differ from typical examples of complex systems because they can in principle be controlled  individually (as demonstrated in Overal Spullen). However, in daily life, there is an interaction between the human being questioning what to do next and the possibilities the stuff around them is offering. This circular causality of both the next action and the configuration of stuff that makes the next action possible results in an intuitive and impromptu interrelation between user and stuff. Although people might be able to specifically plan the placement of their belongings, they don’t. They are immersed inside the system themselves. 

2.1 Spontaneous order

The crystallization of stuff is a form of order. It is a restriction of the freedom of the parts, or, in other words, the maintenance of a low level of entropy. This phenomenon happens spontaneously, in the sense that no intervention for the sake of ‘ordering’ is consciously performed. Spontaneous order is since the second half of the previous century recognized in multiple domains of research, in which the classic causal-mechanistic way of thinking could not explain the behavior of certain material, organic and societal systems, and forms one of the central notions in complexity theory.

Spontaneous order, or self-organization is the formation of spatial, temporal and spatiotemporal structures arising from local interactions. It is triggered by random fluctuations and amplified by the self- reinforcement or positive feedback of formed structure. It is the most exemplary property of complex systems, systems recognizable because of their network-like structure.

Missing any one formalism to adequately capture all of their properties (Mikulecky, 2001), the global patterns of a complex system are explained by the cooperation and competition amongst the particles through time, that unfolds a structure with a highly dispersed and decentralized control ( Waldrop, 1993).

Fig 2.1: The process of self-organization, stylized representation.

2.2 Stuff as a doubly complex system

Three types of complex systems are distinguished, that differ in the ability of the parts to adapt the local rules they follow. Material complex systems, such as the pattern formation between two reacting materials or the weathering of rocks through wind and water, cannot change the rules of the game according to their specific situation. Organic complex systems, however, such as fur patterns, cell structures and animal architecture, do change overtime, through evolution. A natural selection of system variables and local rules gradually specifies itself to the genetic variation with the fenotype best to survive (here the complex viewpoint on how the leopard got its spots1). Human systems form the third category, that of doubly complex systems. Since humans have the ability to mentally travel in time, foresee certain outcomes and change our behavior accordingly, the process of adaptation in, for example, economical, social or political systems, is direct. When a global pattern is recognized, it can be fueled, impaired or something else can be ignited. The local rules change when decisions change (Stolk, 2015).

Another example of a doubly complex system is a city. Although cities are in essence large- scale artifacts, they showcase properties of complex systems, such as in the patterning of their urban fabric (i.e. especially visible in older cities that grew over time, a little less in the Chicago city grid). Their complexity is explained in three ways, as described by Portugali (2016). Firstly, while the city as an environment emerges out of the interactional activities of its agents, this environment itself influences (enslaves) the agents again (1). The artifacts, the buildings, highways and streets, are the media of interaction (2) – decisions are not made by a direct communication on how to build the city at a set moment in time, but all individual decisions do react on what is already built and planned by others. There is thus a two-way causality between the acting agents and the acted-upon environment, which, together with the factor of iteration through time shows swarm-like properties of global pattern forming and path-dependancy. As humans are complex beings themselves, the development of a city has a fast and active process of adaptation (3) which makes it doubly complex.

Fig 2.2: A city is a doubly complex system, because of the circular causality between acting people and the acted- upon city that is a medium of interaction itself.

Complex systems structure themselves throughout time, in a constant iteration of decisions; the current environment is the input of the next decision, the environment following that the input for the next. In cities this turntaking between the decision-making individual and the environment is very obvious. Firstly because the city has a time-span of centuries; it existed before the inhabitant was born, and will continue to develop even after all current inhabitants are gone. Secondly, it is a collective. There are legal procedures determining any building plans; builders need to get the approval of the municipality, review committees and many other scales of organization. To design a complete city on your own is simply not possible. This is different in stuff systems, that are slower, smaller and mostly individual.

But although our cognitive and physical abilities in fact give us power to completely regulate our direct surroundings, the circular causality through the immersion in the environment we act upon, operates also here.

Although the dilemma in Complexity Theory of Cities why a city that is essentially an artifact still displays complex behavior could in stuff systems be even sharper, as there is the possibility to individually control all agents, also the solution to this dilemma applies. The continuously changing configuration of stuff around us is the media of interaction itself: in the same way, we act upon a system we are right inside of. Stuff influences us as much as we influence stuff.

Fig 2.3: Although it is in fact possible to completely control our direct surroundings, the same circular causality as seen in cities exists also in stuff systems.

The doubly complex theory of Portugali will therefore be taken as the starting point for this research, as a requirement for complexity, that is explored at a smaller scale. This translation is possible as our relation with stuff is so direct, ad-hoc and intuitive.

2.3 A state of self-organization

When considering stuff systems as complex, a third state can be added to the spectrum, one of self-organization, giving a triangle of possible stuff states with five additional processes. The processes numbered 6, 7, 8 and 9, in other words, all that do not lead to order through classification, can happen spontaneously, albeit under particular circumstances. Self-organization happens when one is engaging in an activity, in which the emergent properties of different belongings together will be needed. This can arise both from an ordered and a chaotic state, whose differences will be explored in chapter IV. The process of self-organization turning into disorder happens when the activity is ended; the relationship that bound the individual particles is lost and the system looses its emergent values. Things fall down or are, in the hassle of other activities, set aside to where they are not in the way; the typical free and random particle movement that is part of the chaotic state. Self-organization to order is the non-spontaneous process of tidying up directly after an activity.

Fig 2.4: The identification of five additional processes between the states of order, chaos and self-organization.

Fig 2.5: Five additional processes and their equivalents in a house.

* the words ‘planned’ and ‘unplanned’ are too limited, chapter IV dives further into the difference.

2.4 Stuff explained

With this framework of three possible stuff states, order, disorder and self-organization, and the nine different transitional processes in between them, we can describe observed stuff configurations in a more accurate way. In the picture in fig. 2.6 that is taken of my desk, an ordered state at the left side is clearly distinguishable (fig. 2.7). The pencils and markers are sorted on their type and color, to obtain the advantages of ordering; it saves space, it gives a clear overview of what is available and it is easy to specifically pick a desired part. On the desktop, closest to the chair, we see an example of self-organization (fig. 2.8). The laptop (which is closed for the picture, but open during the activity), notebooks, pen, cup and light are organized around the activity and all support it; some more literally, like the pen, notebook and laptop, some in a more general role, like the cup of tea that supports the host. All artifacts in the stuff cell are part of a network in which their functions complement each other, in a dense cloud-like structure dimensioned around the (ergonomic) characteristics of the user and the activity.

Fig 2.6: The desk on January 18th 2017, 00:11

The combination of the plate, knife, cup and chocolate sprinkles in the back used to be a form of self-organization around the activity ‘breakfast’ but was left and not cleaned up. The items lost their functions and were one- by-one moved to either a self-organizing centre where they had value again (e.g. re- using the cup), or where they stand less in the way. In this case, they are moved to the back of the desk (S-O to chaos, fig. 2.9), on top of some loose things that were already traveling around. These now lie on the bottom of the pile (chaos to chaos, fig. 2.10).

2.5 Morphology

This illustration might suggest that every item in a configuration can be ‘labeled’ as being in an ordered, chaotic or self-organized state, which is not the case. Instead, the states are a concentration of a certain type of organizational pattern – i.e. a global pattern (order), a global pattern emerged from local interactions (self-organization) or no pattern at all (chaos) – that constantly overlap. The states depend on each other; the sorted office supplies on the left (fig 2.7), for example, are something the activity itself (fig 2.8) constantly draws from and needs to keep its resilience. A ‘stuff cell’ is therefore more than what here is called a state of self-organization alone. It does have a nucleus, but no cell boundaries.

The clearest way to distinguish between the three states is therefore by their morphologies, that are all of a significantly different kind. Chaos is soup-like, as it consists of loose particles in a uniform mix. Order is grid- like, possible to split at any point without changing its arrangement. The pattern emerging from self-organization is quite different in nature; it is more of a crystal. The pattern is inseparable; whole and uniform throughout its scales (Kwinter, 1994 & 2001). Or as Schrödinger put it in his book What is Life:

“The difference in structure is of the same kind as that between an ordinary wallpaper in which the same pattern is repeated again and again in regular periodicity and a masterpiece of embroidery, say a Raphael tapestry, which shows no dull repetition, but an elaborate, coherent, meaningful design traced by the great master.“

What is Life? Ch. 1, p. 2

Fig 2.11. Order, the grid-like or the wall-paper.

Fig 2.12. Chaos, the soup-like. Ultimate uniformity.

Fig 2.13. Self-organization, the crystal-like or the tapestry.


Ordered items are sorted by their properties, which means that piles of ordered stuff are piles of the same kind of stuff; a bookshelf full of books, a pencil case full of pencils and a kitchen cupboard full of plates. Those similar items have a similar form. Another property is that ordered collections are arranged in a space-saving way, thus follow a grid with a space-filling pattern. Box-formed items give a rectangular grid, but this, of course, does not have to be the case.

Fig 2.14: Cups with handles finding their optimal pattern given by their shape, gravity and minimizing their footprint.

Fig 2.15: Wine bottles in a space-filling pattern.

The ordering of stuff is something which requires energy input, as it gives no emergent properties (except for being stored away as space-efficiently as possible). Still, the pattern that arises, can be seen as self-organizing. The rules, space-fillingness + gravity, are so decisive, that there is no conscious choice made by the person ordering; there is only one best option which is already given. It is the local rules (the side of an object that fits to the side of another) that create form. The pattern comes forth from a material (singular) complexity, in which material finds its optimal form, given the forces acting on it.

Other self-organized grids are defined by form, space-fillingness, gravity and one or more additional rules. This is, for example, the possibility to have an overview (resulting in a sightline from all items to the observer), the possibility to dry (resulting in enough space for air flow and for water dripping off ) or the possibility to easily reach all items without disturbing the others (resulting in a pattern with no inaccessible center).

Fig 2.16: Rules: Space-fillingness + gravity + possibility to be reached by the water from the spray arms + air flow.

Fig 2.17: Pattern of drying dishes too large for the rack. Rules: Space-fillingness + gravity + water dripping off.

How material finds its optimal form can be an inspiration for designing the architecture supporting it (in this case the shelf, clothes line, wardrobe rod etc). This can be seen as a way of parametric design, in which the parameters of the local items create form, instead of the expressionist designer. In this sense, parametric design is the ultimate form of ‘listening to the material’.

Fig 2.18: Support that makes a space-saving pattern possible, that avoids the wine glasses collecting dust inside.

Fig 2.19: A more optimal wine rack design, in which more space is saved.

This idea can be of inspiration when designing for ordered patterns, such as for the storage spaces in a kitchen or attic of a house.