the nature of order

Pages from Christopher Alexander's book The Nature of Order, showing an x-ray scattering pattern from a beryllium atom, and a pattern of self-organizing crystal formation. Alexander wanted to find out what we can learn from natural processes and their resulting geometric structures.

“In the past century, architecture has always been a minor science — if it has been a science at all. Present day architects who want to be scientific, try to incorporate the ideas of physics, psychology, anthropology in their work, in the hope of keeping in tune with the “scientific” times. I believe we are on the threshold of a new era, when this relation between architecture and the physical sciences may be reversed — when the proper understanding of the deep questions of space, as they are embodied in architecture, will play a revolutionary role in the way we see the world and will do for the world view of the 21st and 22nd centuries, what physics did for the 19th and 20th...”
— Christopher Alexander, “The interaction of science and architecture,” www.katarxis3.com

For Christopher Alexander and his colleagues, the popularity of A Pattern Language was certainly remarkable, and gratifying — and in particular its evident success with non-architects seeking to design and build for themselves. (Its perennial bestseller status suggests it is still a major influence among this group.) But Alexander and his colleagues were disturbed to find that many of the designers inspired by the book produced crude work that lacked the simple dignity of older vernacular buildings. Clearly they had not succeeded in replacing the robust traditional pattern languages of vernacular building with an equivalent new technology. What was missing from the methodology?

Alexander came to believe that he had not sufficiently dealt with the detailed problem of geometrical form. Returning to the problem of the relation of parts to wholes, he asked, what is it about the particular geometries of the built environment that we find most satisfying — not just on the basis of one narrow problem-solving criterion or another, but more broadly, on the level of a whole system? What characteristics do they have, and what detailed processes actually created them, in nature, and in traditional societies? What can we learn from them to improve our own methods and results?

Answering this question, and documenting the ideas for his readers, was the task that would occupy him for the next 25 years, culminating in his magnum opus The Nature of Order, subtitled “An Essay on the Art of Building and the Nature of the Universe.”

Taking his cue from nature, Alexander studied the processes of morphogenesis in biology and other natural phenomena, and the characteristic geometries that resulted. Working phenomenologically rather than reductively, he grouped them into a series of categories, eventually distilling them down to 15 “properties”. They range from familiar ones like “boundaries” and “alternating repetition” to more esoteric-sounding ones like “not-separateness”. Even so, as always, the structural logic of even the esoteric-sounding ones is rigorous.

In addition to these geometric properties, Alexander also looked at the processes that shape them. He made an intriguing observation: each of the 15 properties has a corresponding kind of step-wise transformation that gives rise to it — and this goes for human acts of creation as well. These transformations do not create structure from scratch, but instead preserve some aspect of the previously existing structure. Hence Alexander referred to these as “structure-preserving transformations.”

The compounding of results through such a sequence of stepwise transformations turns out to have some surprising characteristics, as mathematicians like Stephen Wolfram have shown. What seems to be a straightforward linear process of steps turns out to produce, through the multiplication of pattern, astonishing complexity. One can think of the surprising results of just a few artful Origami folds, or the way two cooking ingredients can mix initially very crudely, but in just a few steps of kneading, can take on new properties. Compounding of the interrelationships creates some surprisingly transformative results.

This was a crucial finding. Our contemporary methods of technology have relied upon much more atomic processes of standardization, replication and combination. In architecture, they had their analogue in the technical act of “composition” — the assembly of parts to form pleasing wholes. This has been a powerful and productive strategy for humans in the modern era. But it is generally not the way that nature works to produce its bounty of successful, sustainable, beautiful structures. Rather, natural systems rely at least as much upon such stepwise processes of differentiation, transformation and articulation.

In natural processes, the whole characteristics of the system do not arise anew from the mere aggregation of parts. Rather, there is a wholeness at some level that is always present, at every step. It is merely transforming as the steps go forward, and the structure unfolds, often into a more elaborate configuration.

In Book One of The Nature of Order, Alexander refers to a famous series of photographs of a milk drop by the photographer-scientist Harold Edgerton. One can readily see that this very ordinary, simple process results in a remarkably elaborate structure with a deep interrelation of all the parts — a deep wholeness of the structure. Moreover, this is not occurring through only one logical step, but it is occurring continuously through a complex set of sequential transformations in time.

Figure II.3.1. A series of Harold Edgerton’s famous milk drop photographs.

Note that it is never true that at any stage the parts “went together” in any conventional sense. The parts did not “make” the whole. Rather, the parts were differentiated out of the process, and changed their structure and their relationship to one another, as the whole went through each transformation. In fact, it would be more accurate to say that the whole made the part! There was always a whole structure, at every step: the milk drop, and then the splattering form, then the deformed ring, then the articulated bumps, then the little arm-like appendages, then the emerging smaller spheres, then the cell-like regions at the bottom, and so on.

It now seems apparent that the process of morphogenesis in nature works very much like this, with structures responding to natural physical forces of this kind — articulating, extruding, mirroring, and carrying out similar “structure-preserving transformations.” Nor are such processes strictly confined to biological systems: indeed, it appears that many other natural phenomena undergo similar processes.

A second, closely related point is that at any point, each region or center is in some way closely fitted to all the others, and to the whole. They are always in some sense part of a deeper order, and the act of creation is not to bring together an aggregation of atomic parts, but to orchestrate a transformation of parts that already existed within a whole. (For this reason, Alexander later referred to “structure-preserving transformations” as “wholeness-extending transformations.”)

Put differently, a local action can contribute most to the global structure when its already existing, embedded relationship to that global structure is acknowledged and strengthened.

For Alexander, it followed that it is impossible to talk about the structure of a built environment and its effect upon the human user, without facing squarely the human users themselves, and the qualitative nature of their experience as an a priori condition that must be accommodated. One could pretend that qualitative experience didn’t matter, and only quantitative analyses and "facts" mattered. But it was increasingly clear that this omission was precisely the source of much of the current grief, and the reason that progress had begun to grind to a halt. In fields that had taken this "positivist" approach. Meanwhile, progress continued in fields like neuroscience, artificial intelligence and other subjects of complexity, precisely because these fields had recognized the necessity of facing the phenomenon of subjectivity, and the subjective experience of value.

As Alexander noted, value is the unwelcome guest at the party, simply because it is so hard to define in quantitative terms. It is the ultimate holistic, emergent phenomenon. And yet, Alexander noted, value is a sharable phenomenon, and a discussable one. There are straightforward scientific methodologies that can indeed reliably find large areas of shared value. Moreover, these shared areas turn out to have a surprisingly definable relation to structure itself.

When it comes to living organisms, and apparently, when it comes to the built environment, value is rooted in the structure of things.

For some design theorists and critics, this was treacherous, alarming stuff — the makings of totalitarian design, or an enforced blandness of “average tastes”. But for Alexander, there was no alternative but to face it squarely. While some things are variable, not everything about value is entirely relative or “subjective”. Indeed, there seems to be something more like a continuum, from the personal to the universal. Our job as scientists is to tease this apart carefully, using rigorous scientific methods.

In this respect, Alexander saw no problem whatever applying the rigor of science to subjective and qualitative phenomena. Indeed, he saw them as necessary allies in confronting the current challenge of the built environment.

A biological perspective may help to put a more comprehensible perspective on the situation. Clearly there are matters of individual taste and preference in the built environment, but equally clearly, matters of shared valuation. (To illustrate the point with one extreme example, car exhaust is likely to be universally regarded as an undesirable feature of built environments, whereas, say, fresh air is generally regarded as desirable.)

There are structures within the built environment that affect human health and well-being, and for Alexander, it is the business of built environment professionals, not unlike doctors, to diagnose and prescribe more healthful and more desirable conditions. There is certainly art to it; but there is equally science, to be applied to the professional care of the well-being of others.

While some architects see Alexander’s work as idiosyncratic, many in other fields see familiar insights and ready opportunities for collaborative development. As noted, there are strong correspondences with the insights of much complexity science, notably in the understanding of complex generative processes, adaptive algorithms, cellular automata and the like. There are resonances with biology, and the processes of morphogenesis. There are resonances with studies of wholeness in other disciplines, as for example in quantum physics. There are echoes of the insights of neuroscience and the study of artificial life, in the study of qualitative phenomena and the surprising relation of living systems to non-living ones.

As Wiki inventor Ward Cunningham has noted, many of the problems Alexander has explored are indeed inter-disciplinary in nature. Hence the interest of him and other computer scientists in this work. For example, the topic of generativity is a very important one in computer programming, and Alexander has explored it very directly, and in a way that Cunningham and others have clearly found useful.

Yet these inter-disciplinary topics have not been much explored by very many architects, except perhaps as devices to generate expressive ideas. This may be, again, because the emphasis in architectural design is on the abstract expressive qualities of the work, as opposed to its adaptive problem-solving power.

Alexander’s work also shares strong correspondences with other reformers in the planning and urban design fields. For example, we already noted Jane Jacobs’ extensive writings on the topic of “organized complexity” and the importance of its insights for the design of cities. She also spoke of the generative capacity of cities, and the responsibility of designers to facilitate this capacity. She warned about the practice of sweeping away context and composing from scratch, and she also argued that the best cities continuously generated their qualities of life from their existing ingredients, however humble.

More recently, members of the New Urbanism movement (which we will discuss in more detail later) have campaigned for greater mixed use and diversity in development, and more use of existing contextual patterns. A number of prominent academics and practitioners within the New Urbanism tradition have also explored Alexander’s ideas.

All of these collaborators have shown a willingness to work in an “open-source” format with Alexander and other colleagues, and to find new insights that will benefit architecture as well as other fields. At the same time, many of those from other disciplines have also expressed some bemused perplexity at the apparent isolation of many architects, and at Alexander’s “prophet without honor” status in his own ranks.

Still, The Nature of Order was a treatise, not a manual of applied technology in the style of A Pattern Language. While it gave a number of methods for design using its insights, it left a number of topics for further development, which I will consider in more detail in the next two chapters. Among these we could include:

1. Generative codes. How does one operationalize “structure-preserving transformations?” What are the steps that one can take to incorporate qualitative and diagnostic procedures within a stepwise methodology of this kind?

2. Collaborative development. What can we learn from the success of those in the software community for greater progress? How can we incorporate more “open- source” methods into the work ahead? Given that Alexander has been at work for over a half-century, how will this work be taken forward?

3. The remaining opportunities with pattern languages. As we have seen, the software community has also achieved much more success with pattern languages as a robust technology. What can we learn from them? What can we learn from Alexander’s own later work, which was meant to address the perceived weaknesses of pattern languages? What opportunities remain to return to pattern languages, and infuse them with useful new capabilities in architecture?

4. “Massive process difficulties.” Alexander notes that there are major constraints on the kinds of processes he describes, built into the current institutional system of design and building, and into the current flow of money and debt. As he notes, this problem needs much more work, as I will discuss further in the last chapter of this section.