adaptive morphogenesis

“In a structure which is differentiated, the structure will not, in general, be made by small piecemeal acts happening in random order. Rather, each step creates the context for the next step in the whole, and allows the process as a whole to lay down, next, what has to be laid down next in order for an orderly unfolding to occur.”
— Christopher Alexander, The Nature of Order

Underlying his specific interest in architecture, throughout his career Christopher Alexander was always searching for a more intelligent kind of technology — more adaptive, more integrative of the wisdom gained over centuries, more enduring and sustainable. For him, technology was nothing other than the original Greek idea of techne + logos, knowledge of making. At heart, what are we making, from a human point of view, and how do we ourselves know and understand its impacts? What are the “rules of the game,” so to speak, and how do we change those rules to create a more intelligent and wiser human future?

Like Jacobs, Alexander found reasons to be hopeful about the answers, particularly those beginning to emerge from the “new sciences” of complexity. It seems many of the answers are indeed to be found in nature, including human nature. Some surprisingly modern lessons are also to be found in the rich “collective intelligence” of human history and its traditional knowledge — a point we will come back to.

Hence Alexander believed his work had to deal more directly with the problem of process. What is the process by which a pattern language is actually used effectively to create form? What are the steps one must go through? Whereas pattern languages were about the structure of things, offering a kind of library of combinable fragments of that structure, perhaps his newer work must be about the process of creating that structure. The library this time might be of recombinable fragments of steps, rather like the steps in a recipe, that tell you how to get from one stage of form to the next.

Alexander’s insight came again from the observation of traditional cultures and their remarkable adaptive sophistication and success, as it so often did throughout his career. He observed that building traditions guided individuals in specific steps of building, and in how those steps should respond to their context. Often very sophisticated ancient codes functioned to do this. Often more direct linguistic concepts and “rules of thumb” guided individuals and groups, and the guidance was refined and handed down in tradition.

But in the technology of the last several centuries, this delicate contextual structure was swept away, replaced by a more powerful but at the same time cruder, much less contextual system. This system was more dependent on abstractions, and more likely to ignore or even to destroy the contextual structure — often with powerful results, but also often with disastrous unintended consequences.

As we saw in the last chapter, Alexander took his odyssey beyond human traditions, to ask basic questions about the processes of growth in nature. He made a simple, even obvious observation: nature regularly and almost effortlessly, it seems, creates a vast range of successful living forms, from astonishingly simple ingredients. These structures are exquisitely well-adapted, beautiful, sustainable. What are the processes it uses to do this? And what can we learn from those processes for our own human applications?

To answer that question, he drew on insights from many fields, including embryology, physics and others. He came to one central conclusion: nature does not use a “plan” or “blueprint” of some final end result. Rather, as we saw in the last chapter, nature acts to transform an existing whole into a new whole, and it does this repeatedly, over many cycles of evolution.

In doing so, it preserves the structure of the earlier whole, but it often amplifies, articulates and deepens it in some important way. We can see that process of transformation very clearly in the biological patterns of evolution. Alexander noted, intriguingly, that we can also see it in our own built history — in the structure-preserving transformations of the Piazza San Marco in Venice over 1,000 years, for example, where at every step, the whole was maintained. At no point was the piazza entirely bulldozed and rebuilt according to some architect’s bold new vision. It was rather a continuous evolution, with human plans playing a disciplined role within what could be seen as a kind of “dance of the centuries.”

But the steps of such a “dance” can appear deceptively simple and humble — much as a mere 26 letters cluster into words, sentences and soliloquies can create the complex beauty of Shakespeare. When presented with the 26 letters alone we might wonder how we could possibly create something so rich from such modest parts; but that is just what Shakespeare did.

So, too, in the process of creating form, as we see all over in nature, the steps can seem exceedingly simple and modest. But the key is in how they combine, how they multiply in repetition — much like the way two colors of putty will mix surprisingly quickly after just a few repeated folds, or the way a marvelous animal shape can result from just a few relatively simple steps of folding paper in Origami. There is an exponentially multiplying interaction between the parts, which manifests over repeated steps.

It turns out that this is very much how forms develop in embryology, through a very similar kind of “unfolding” process. This occurs not only in the DNA and RNA molecules, but also in the protein structures that they then form, that subsequently bend, fold and interact, and form various products, including tissues. These tissues then divide, fold, differentiate and articulate into new structures. In addition to the simple parts — just four molecules in the genetic code — all of this rich complexity comes from relatively simple steps too: combine, divide, fold, merge, and so on.

This “complexity out of simplicity” is a key to understanding the processes that create richly articulated, differentiated, living structure. It is at the heart of what biologists call “adaptive morphogenesis” — underlying the creation of thriving, stable ecosystems.

This was a major revelation for Alexander. It was not lost on him that age-old human processes share some aspects of this structure. He observed the way traditional craftspeople took relatively simple steps to gradually weave stunningly beautiful patterns in carpets, or the way traditional city-builders took small steps to position their houses and the spaces around them, gradually building up a marvelous urban structure with exquisite traits.

Our “modern” methods, he noted, are based on a very different, radical approach: creating templates and “blueprints” and "expressions" ahead of time, which can be thought of as little fully-developed models of reality. They produce powerful economies of scale because they allow for standardized repetition. But they also tend to impose rigid artificial aspects on the reality, instead of adapting to it to the very fine degree that nature requires.

Nature too uses templates, on occasion, or something like them. One might think of DNA as a kind of blueprint. But nature is much more subtle than current human technology: there is no little model of a finger encapsulated in the DNA molecule. Nature actually uses a strategy that is at once far simpler than that, and far more complex and sophisticated in its output. For every finger produced is a marvel of uniqueness, sophistication and complexity. The human version of a template methodology, though it has been enormously effective in conventional technology, is a far cruder and less elegant device. The implications of that are significant.

Perhaps an even more sophisticated, more “modern” approach, would re-integrate these other powerful processes into human methods — including the powerful if often unconscious processes of human tradition. Perhaps nothing less would be required to create the kind of well-adapted, sustainable, balanced structure that nature had done, and that was beginning to look like an essential requirement for a prosperous human future.

Alexander came to see that even his pattern language was guilty of the “template” limitation. If people used the language to come up with a design, planned in advance, without a careful generative process for adapting the form, then the form simply wouldn’t have that living quality that is needed, and that was achieved by previous generations across so many cultures. The reform of our unsustainable modern processes of morphogenesis was still incomplete.

Alexander asked himself, what were the methods that people could use to apply these kinds of processes to acts of building (and other form-making) in a modern age? What insights would they be built upon, and how would they function?

Here I will outline several of the key concepts of this work.

Centers

One needs a useful diagrammatic model of the structure of things that is undergoing a pattern of growth — an analytic understanding of the essence of what is going on geometrically. For Alexander, that model is a system of centers. Every form can be understood as a system of centers in some relation to one another — one inside another, one forming part of a boundary around another, and so on.

A system of centers can have a hierarchical relationship, a networked or semi-lattice relationship, or some combination. It can have all of the kinds of relationship that Alexander and other theorists described, in critiquing the limitations of modern design methods.

Thus the model offers a powerful way of understanding the structure of form, and the transformations it undergoes — and also its limitations. There are echoes of the “process” philosopher Alfred North Whitehead’s “actual entities”, and echoes too with theoretical physics. There is also an echo in the work of Herbert Simon and "The Architecture of Complexity". But there are also unique insights that have gone on to prove themselves useful foundations for the rest of the work.

Fifteen Properties

Alexander then made an interesting observation: in spite of the endless variety of configurations in which centers can be found in nature, he found that one can distil them down to just 15 different classes of organization, or geometric properties. Every form of structure that he was able to observe could be grouped into one of these classes. This scheme of classes turns out to be very useful in analysis and, Alexander believes, in aiding as a design tool. I will outline the 15 classes here briefly.

Alexander observed and reported the 15 properties phenomenologically, but then began to seek clues to their underlying formation and arrangement. He came to understand that they arose naturally as a result of the natural transformations in the processes of morphogenesis. That is, the process of structural development leads to these classes of order, through the workings of the transformations.

Structure-Preserving Transformations

We discussed the central notion of transformation earlier. The process of a “structure-preserving transformation” is capable of generating the different properties of natural morphology, as the result of its morphogenesis or generation of form. To see how this is so, we can go back to the famous milk-drop photographs by Harold Edgerton that I mentioned in the last chapter, and we find that even this simple process generates remarkable geometric characteristics.

On the next page is the famous series again, showing the drop of milk hitting a thin layer of milk covering a hard surface. Note the initial sphere, a simple center in the middle with no articulated centers around it. It strikes the surface, and its symmetry is immediately broken. The result is not chaos, but a new kind of organization. (This “symmetry-breaking” is a very important idea in physics.)

Specifically, notice how the displaced milk rises up and forms a ring, a boundary around the original drop. The ring expands, and as it does so it too becomes unstable — the equilibrium between gravity and velocity is exceeded — and its symmetry is broken as well. But again, the result is not chaos but the articulation of new structures — arm-like appendages, and at their ends, like exclamation points, new “baby” spheres.

The overall structure exhibits the features of organization, not unlike an organism. There is a unity of form without the overall composition. There is a hierarchy between the center, the ring, the appendages and the small spheres. There are interconnections between the different arms, which can be seen to slightly perturb each other. There is an irregularity too, which is intricate, varied and unique — not exactly like any other milk drop. It is not a perfectly repeated pattern, but it is a well-organized one all the same.

We can begin to see, in this simple example, how Alexander’s “15 properties” can be generated spontaneously through such a process. One can see levels of scale, boundaries, alternating repetition, and so on.

What is significant is that this structure has arisen as a direct result of the steps in the transformation. The structure of the original sphere transformed into the ring, which transformed into the appendages, and finally to the little spheres. It did not simply disappear, to be replaced with a new structure inserted into the environment, assembled from parts according to a template. (That is the exceptional, limited and extreme method that is unique to contemporary human beings — producing a fundamentally limited set of geometric properties, from this perspective.)

As I noted previously, it is a remarkable fact that the structure at each stage is a complete whole — first the sphere, then the splash pattern, then the rings, then the droplets and so on. It is not a combination of separate and independent parts. Instead, at each stage, the whole is transformed and extended in some important way. This is why Alexander later used the term “wholeness-extending transformations.”

Nature of course is full of far more complex examples than the milk drop — perhaps nowhere more than in embryology.

Unfolding

What is interesting for our purposes is that the DNA in some way guides the process, like a recipe, with instructions for the various steps. A similar process can be observed in human processes, including cooking recipes and medical procedures. The steps are very simple, but their effect becomes sequentially more profound, resulting in a delicious meal, or a process of complex healing.

I mentioned previously that similar process can be observed in the function of traditional urban codes. Relatively simple rules guide builders through various steps of construction, specifying contextual responses, such as position of windows relative to previously built windows, and so on. The result is an emergent contextual form. Alexander has found that traditional building processes used a similar kind of stepwise guidance, or “rules of thumb.”

In all cases, the evolution proceeds to “unfold” more and more complex, powerful, beautiful forms, not merely as a matter of “trial and error,” but as a compounding characteristic of the process itself. It appears that there are capacities built into the nature of space itself, or space-time, that allow these structures to be formed as the result of their adaptations.

The lesson for human designers and planners is a simple and one. We need to return to an understanding of these basic, powerful processes of adaptive morphogenesis, in responding to the context of our designs. We need to temper our use of abstractions (including abstract ideas of art) and place primary emphasis on the human being, the human context, and the natural world, including the world of human nature.

These living places and activities are the seeds of the renaissance of cities.