On simplicity #7 • Unfolding complexity

Something truly simple and beautiful requires a strong identity. To create that, we need to become tangled up in our creation as we evolve with it towards fittedness.

How can we gradually build up organized complexity in such a way that we keep it simple and intelligible? Both in the design process, where we as designers need to co-evolve with our creation towards intelligible complexity. But also as people enjoying such creations, fascinated by the gradually disclosing depths of them. Both suggest an intimate feedback loop of reciprocal realization instead of analytical planning grounded in universal rules.

Simplicity is an emergent property that is highly context-sensitive. We can’t methodically make something simple by following a universal set of rules, by following a grand master plan. There is no universal best next step we can take. There is no 10-step method or proven formula or cookie-cutter approach. The next best step will always depend on the context.

Each differentiation, i.e. [design] decision, is made in sequence and context. It is reworked right then and there until it is mistake-free, i.e., it takes into account all the connecting relationships. This must be done in sequence and context because the necessary information for a successful decision is not available prior to that step in the unfolding.

— Christopher Alexander · NoO.II.6 · Conclusion of the discussion on generated complexity

We must engage in a creative process of reciprocal realization, an iterative dance between us and our creation existing in its environment. We have to rely on a tight feedback loop in which we reciprocally disclose new information about each other, incrementally adapt to each other, and slowly evolve together towards better fittedness.

It is a process of unfolding, where everything is already there, just not fully expressed yet. It is our mission to discover and manifest the latent properties, step by step. It is not a prescriptive approach, but an inherently creative process that requires our full attention and engagement at every step of the way.

Of course, as I pointed out in the beginning, this is not as straightforward as we would like it to be, because our favorite highly optimized scalable mass-production approaches fail at this.

7. We may not always be able to find ways to make something more coherent and comprehensible. This is surprisingly difficult to do because it cannot be intentionally achieved systematically but has to be discovered through a creative process. And we usually don’t reserve that much time for such a process to properly unfold.

The most challenging part for us is not differentiating, adding more detail. The most challenging part for us is integrating, preserving what is already there, so that we don’t lose the coherence and identity we have already achieved.

Preserving existing structure

Don’t we love greenfield projects? Unencumbered by constraints and legacy we can dream up “clean” solutions and worry later about integrating them into what already exists. That independence also enables us to scale. Different teams can work on different parts independently, as long as we figure out a way to merge their work back together later.

To a certain extent this works. But it necessarily leads to complicated designs because we treat integration as a less important step of “making it work together” later. We design new parts in isolation, because isolation enables scale. Then, when it is time to integrate the new and now already considerably complex part into the larger whole, we like to minimize change to both the new part and the existing system, often utilizing copious amounts of “glue code” that binds them together.

Christopher Alexander suggests a different approach: Parts don’t make the whole. The whole makes the parts. New parts cannot be developed in isolation. They need to come out of and grow within the existing structure.

We have always assumed that the process of creation is a process which somehow inserts entirely new structure into the world… in the form of inventions, creations, and so on.

Living process teaches us that wholeness is always formed by a special process in which new structure emerges directly out of existing structure, in a way which preserves the old structure, and therefore makes the new whole harmonious.

Thus the process of making wholeness is not merely a process which forms centers or the field of centers in space… it is a process which gives special weight to the structure of things as they are.

— Christopher Alexander, NoO.II.12. Respect for what exists

It makes sense, if you consider Alexander’s primary domain: architecture. There is always some environment that already exists. If you are lucky, it is literally a green field, which supposedly makes it easy. But even then it is probably shaped in a certain way, and is embedded into an existing landscape with streets and neighboring lots and trees and views which all need to be accounted for in some way or another.

In software, preserving existing functionality — or legacy support, as we like to call it — is not what is meant by “preserving structure”. It is not enough to keep existing parts in place to preserve structure. Because adding new components will still change the structure of the whole, it will change what the product is, it will change its identity. Preserving existing structure means to preserve its identity. To protect that which made the first early adopters fall in love with our product.

We need to have respect for what is already there. Not in a technical sense, but in an experiential sense. The way our happy customers connect with our product, how it makes sense to them, how it fits into their context, how they enjoy interacting with it. That quality is not based on a feature list. It is more like a gestalt, an eidos, a whole, an identity that intuitively makes sense as long as it is intelligible and recognizable — identifiable as the unique thing they fell in love with in the first place. That is what intelligibility can provide and why it is so important.

If we just keep adding features, we risk changing its identity into something less sharply defined, something more generic — something that takes on a categorical identity, which makes it less special and more replaceable.

If we manage to preserve its identity, and extend its capability through differentiation with harmonious integration, we can further sharpen its definition and intensify its coherence — its identity becomes stronger. That, of course, is much harder to do.

We have a chance to build something people want to identify with, engage in reciprocal realization with, and enjoy interacting with. It becomes important to them, not in a “solves my problem” functional tool kind of way, but in a “I love being here” delightful environment kind of way.

Creating novelty out of the familiar

The enigma is that something new, unique, previously unseen — even innovative and astonishing — arises from the extent to which we are able to attend to what is there, and able to derive what is required from what is already there… and that all this, then, will lead to astonishing surprises.

In this context, creativity looks entirely fresh. The act of creation is not a willful process like the art of Michelangelo, which we evaluate according to its novelty. It is, instead, a process in which we most deeply express our reverence for what exists.

— Christopher Alexander · NoO.II.12 · Respect for what exists

This is pretty much completely against our usual tendencies.

We tend to optimize for universality — generic use cases that have the potential to attract more users without adapting too much to an individual user’s needs and wants. High differentiation and specialization is undesirable and needs to be avoided. Because the more adapted software is to specific use cases and environments, the smaller the addressable market. On top of that it is more likely that the amount of effort it takes to implement all this specific adaptation stands in no relation to the engagement or revenue added.

This is why over time most products lose their identities and turn into just another less exciting instance of their category. This is also why startups often begin with a unique product with a strong identity — they just haven’t had the resources yet to make it more universal and generic and had to make tough choices to implement just a few things that really mattered. They created something opinionated, almost by accident, but that established its identity.

A strong identity becomes easier for people to like who adore the particular choices through which this identity came into being. It also means it becomes more likely that people don’t like the choices made. It is polarizing. More people connect to it strongly, they become passionate, but not necessarily just in the positive. A strong identity yields passion. But passion can be both love and hate. The opposite of passion is indifference.

Do you want people to be passionate or indifferent towards your creation?

Growing intelligible complexity

The interplay between differentiation and integration — complexification — preserves existing structure and further develops the creation simultaneously. It is like slowly increasing the resolution of a picture while it is loading. At first everything is quite blurry, but it is all there already, although you can only make out a few big parts. With each iteration, the resolution increases and smaller and smaller structures become discernible. But never along the process does the image of what is being depicted change. It just becomes clearer to see.

Differentiation adds higher resolution details but keeps the existing structure in place. Integration makes sure that new details fit in their place, and that they form the right relationships with what already exists. Keeping the structure coherent, intelligible, harmonious, is key — even as complexity, in the form of a dense network of interdependent relationships, increases in density.

Of course, along the way we will make mistakes and learn, so we also need to be committed to cleaning out everything that turns out not to be essential.

It is important to grasp that each differentiation adds relationships and brings more interdependence among the centers. Of course, as a result of the many adaptations, and the growing centers and properties, the structure slowly becomes thick with relationships. It is getting denser and denser all the time. And it is vital, for success, that the process is able to keep on cramming in more and more relationships, so that the mistake-avoiding adaptations can continue to be generated.

This “cramming” of complexity brings with it a need to constantly clean out any non-functionalities and leave only the most simple possible geometry in place. It is simple structure that allows for maximum relationships (you need only think of the sphere whose simplicity allows for so many properties at once). […]

In short, then, to make room for more and more relationships, there is a cleaning process going on in parallel with the differentiation process. The process keeps cleaning itself out. Any garbage that accumulates has to be flushed out. The process is simplifying itself, getting rid of debris, and leaving itself, at each moment, with the cleanest and most spare structure possible. Only then, can the system be certain that there will always be room for more relationships, and only then will it truly be possible to keep injecting further transformations and maintain a coherent structure. It is this simplifying process […], which makes the beauty and majestic structure we think of as deep primitive art, or nature generating nature, at its best.

The [generated] structure seeks, above all, to avoid mistakes. To do it, it promotes an activity of structure-preserving transformations, to maintain coherence. In addition, to make the structure capable of containing the vast density of significant relationships which eventually builds up, the process is also cleaning the structure and simplifying itself continually, at the very same time that complexity is building up. To do this, the leveling and sharpening that is typical of a process trying to preserve relationships, and get rid of non-relationship stuff, so that a spruce, spare structure is being built, and we keep moving towards the fine, profound simplicity which is typical of the greatest art, and typical in nature. This is the only way a profound, well adapted structure can be built.

— Christopher Alexander · NoO.II.6 · Conclusion of the discussion on generated complexity

Every part unique

Notice how each part individually goes through a kind of development by itself. Each part gains its own identity, and its identity becomes stronger through more defined and detailed structure within it, as each part adapts more and more to its environment it is embedded in. Parts within parts within parts. Recursively.

Attention to detail matters. As each part differentiates further, adapts further, the network of relationships to its environment becomes denser, each part retains a strong identity — each part becomes more and more unique. In nature we can see this for instance in the petals of a flower:

The petals of a flower are not identical. They are similar, but each one is slightly different according to its position and history in the whole. When parts repeat we never have identical repetition. Instead we have repeated parts as centers which are changing and variable according to their position in the whole, as they repeat within the whole.

In nature, this follows directly from the fact that parts are induced by the whole and created by the whole. The whole is not created out of them. The flower is not made from petals. The petals are made from their role and position in the flower.

— Christopher Alexander · NoO.I.3 · Wholeness and the Theory of Centers

If something differentiates to respect its environment as much as possible, it has to end up becoming unique. Similarities can emerge that are so similar that they look the same to us at first glance. But if we pay attention to detail, we discover that they are not the same.

Nature does not mass produce. Everything in nature unfolds from the whole.

Life is unique. We could even say: Life is exactly that property of space in which each spot becomes unique according to its place in the larger scheme of things. So, if there were a spatial formula which would explain in detail how to make living space, it would fail, because, by virtue of being a formula, it could not succeed in treating each place as unique.

But there is a formula of living process. If you pay attention to the wholeness, intensify it, intensify it some more — gradually then it becomes unique. […]

Everything hinges on the understanding that every part must become unique when living processes are working. This is the key. It is a particular type of geometry which, though highly regular, has every part unique because it is true to its context and therefore to its essence. What we began to appreciate is that every repeated entity is different: that we have to look, carefully at every single case as fresh.

— Christopher Alexander · NoO.II.12 · Every part unique

We like to generalize. Because that enables scale. Looks like we have it backwards. This explains why we treat fabricated, mass-produced, identical objects categorically as replaceable.

At the same time, as we go through reciprocal realization with something (or someone), adapting to each other, mutually changing together, a stronger bond forms between us.

To become important to us, and to have the chance to become important to others, it has to become a strong identity, it has to become unique. And it becomes unique by absorbing parts of us, and we absorbing parts of it — we become tangled up in each other.

This is how we make meaning.

This point is so simple. There is nothing complicated about it. It seems rather astonishing that all this complexity boils down to something as simple as this. It is a little bit like the proverbial slap from a Zen teacher. The student thinks it is all complicated, deep, profound. He twists his head this way and that way, trying to “get” it. Then the teacher finally slaps him to wake him up, as if to say, “Stop thinking about such complicated matters: Just eat, just walk, just sleep”. Or (in the case of architecture), Just make it nice at every spot. […]

“I got there, but I don’t know how I got there. It is like reading a book, over and over, and then finally I have absorbed its meaning, but I do not know just how or when this happened.”

“I understand it now: but I do not know at what moment I understood it, or what happened. It is like an enlightenment. Suddenly it seems obvious, and the difficulties are gone.”

— Christopher Alexander · NoO.II.12 · Every part unique

Mirror of the Self is a fortnightly newsletter series investigating the connection between creators and their creations, trying to understand the process of crafting beautiful objects, products, and art.

Using recent works of cognitive scientist John Vervaeke and design theorist Christopher Alexander, we embark on a journey to find out what enables us to create meaningful things that inspire awe and wonder in the people that know, use, and love them.

If you are new to this series, start here: On simplicity…

The previous series starts here: 01 • A secular definition of sacredness.
Overview and synopsis of articles 01-13: Previously… — A Recap.
Overview and synopsis of articles 14-26: Previously… — recap #2.

Another great way to start is my presentation Finding Meaning in The Nature of Order.