The Nature of Order: The Process Of Creating Life
How do beautiful creations come into being? Nature can make an infinite number of human faces, each one unique, each one beautiful. The same is true for daffodils, streams, and stars. But man-made creations - especially the towns and buildings of the 20th century - have only occasionally been really good, and more often mediocre.
What this means is not only that these buildings have the familiar ugliness of so much modern design. Most often, they also do not work. Being conceptual - not based on step-by-step application of practical reality - they do not fit practical human needs; they do not fit gradually and softly into the harmony of the world which envelops them. They stand, awk-ward, unrelated to all that is around them. And this is to be expected, simply because they have not been created by an unfolding process. Inevitably, they do not work. (Pg. 128)
What has caused the new tradition of structure-destroying forms of this era, are mainly the machine-like processes of planning, conceiving, budgeting, developing, construction contracting, construction labor, and so forth. The architects who fully accepted the modern machine have hardly been more than pawns in the game which is much larger than they are. (Pg. 128)
Yikes!
Engaged in the process, we architects covered up its defects, made it seem palatable, made it seem - even if only marginally- acceptable. (Pg. 128)
It is unfortunate that many of us - instead of standing against this production process, and trying to create a workable and life-creating process - instead chose to stand with the money and the jobs, becoming spokesmen and propagandists for the problem. (Pg. 129)
Alexander talks about the rise of the hyper-image. Photographs becoming more important and more real than the thing they depict. The emergence of the hyper image and its reward by capital sets the agenda and the culture for destructive process and outcomes.
At each moment you ask yourself, What is the most important thing I have to do next (pg 129)
On starting the process with an image in your mind:
Gradually, slowly, step by step, every part of the preformed “image” will dissolve, because reality - and common sense - will lead to something different. (Pg. 130)
To build from and image is to ignore the intersection with reality.
Alexander recognizes a structural problem as the core of the issue;
Many of these mass buildings which shaped the 2oth-century world were almost innocent. Some were made by idealists in the housing department of socialist coun tries. Others were made by developers trying to make a living, and trying to provide housing for a capitalist context. The two look almost the same. (Pg. 135)
They turn out dead because the systems of creation make no room for life.
if we do one thing at a time, and if what we do is wholesome and sound, then whatever comes next will work. We do not have to tie it down ahead of time for fear of some imaginary potential catastrophe of “design.” Instead, we just go step by step, doing what is required as well as we are able, with confidence that the next thing, too, will work out somehow when its time comes, but that it need not be worked out now. (Pg. 174)
In beginning to talk about process, and how we can have a good process to create life, Alexander talks explicitly about software in a way that until now he has talked explicitly about architecture and geometry.
Computers and computer networks, and software, are all enormously complex. It would be natural to expect, therefore, that we must have a theory of complexity, that we have an effective and sensible way of thinking about the best way to create complexity. Faced with the need, growing every day, to create successful complex structures all around us, one would expect that we have at least asked ourselves how, in general, a complex structure may become well-formed. (Pg. 180)
Summarizing with the law:
ALL the well-ordered complex systems we know in the world, all those anyway that we view as highly sucessful, are GENERATED struc tures, not fabricated structures. (Pg. 180)
In software, we pretend like we know this and we make lip service to it in our standards and culture, but I think that we ultimately play more to the hyper-image and deformations it places on our society.
Than he goes back to geometry, which for intents and purposes we must treat as topology.
When creating a complex structure there are many, many opportunities for mistakes. When making all the possible decisions quickly, that is compressed in time, it is far more likely to make many mistakes. Spending more time on the decisions, or perhaps, making the decisions one at a time so each next decision has the benefit of all the information of the previous decisions, means that far fewer mistakes are made. Not only this, but compressing the number of decision makers also creates a higher likelihood of mistakes. Expanding the agency across many people allows for a spread of perspective, and therefor a wider view on what could one agent could myopically destroy.
Because there are so many things you can get wrong, to assume that you will get any amount of them right in one sitting is simply Hubris.
Alexander talks both of maintaining control and letting go of control. One must make decisions and thus be an agent in the process of creating a system, but if the system is to be good one must respond to the state of things and not impose and image or idea into it, and thus work in a way without agency. Or rather, work with agency but not ego.
a typical house nowadays has about 2000 man-hours of labor in it. (Pg. 192)
That’s interesting! By my own calculations that would be about 500 blobs of work, about 2 years of full time work for one person, which would bill at a total of about $500,000. Of course, we have the benefit of deeply cheap materials, whereas Alexander has elsewhere suggested that construction in the modern era is about 50/50 labor to material, making a typical house, if built by software engineers, about a million dollars to build.
Thus the importance of time, is not merely that you have a chance of tinkering and adapting. It also allows you to get each next layer of structure from the previously established layers of structure. Complex, generated structure cannot be arrived at in any other way. (Pg. 195)
Alexander holds that the cost of all these mistakes can be calculated. This calculation starts with number of decisions over time. A system therefor has a mistake-density based on its size and mistake density. Each individual mistake can be said to be costly in some way (and this not quite the same as the concept of a bug. A mistake can be software working as intended, bug free). With a house, the pitch is that good house with a low mistake density can command a higher sale price than a similar house with a high mistake density. Alexander estimates an order of magnitude in dollar amount between a mistake-free house and a mistake-full one.
The key to complex adaptation in a generated structure lies in the concept of differentia-tion. This is a process of dividing and differentiating a whole to get the parts, rather than adding parts together to get a whole. (Pg. 197)
Let us consider the case of software. A typical computer program contains tens of thou sands of lines of code; others in daily use contain a million lines, even as many as ten million lines of code. In recent years some attention has been given to their theory of design, and some improvements have been made in contemporary ways of thinking about their design. Computer scientists have told me proudly that they consider computer programs the most complex objects designed by human beings. Yet, to date, there is little recognition of the following commonsense point: If indeed the programs are so complex, then it is likely that they, too, will be potentially subject to hundreds of thousands, perhaps millions of egregious mistakes of adaptation. Here I am not only talking about “bugs”- failures which stop a program from running altogether. I am talking about mistakes of adaptation, ways in which the program fails to do what it is supposed to do, fails to meet the needs of the people who use it, or is more awkward, more annoying, less useful, than it is supposed to be. If the analysis given in this chapter is correct, then it is fair to say that truly successful programs can only be generated, and that the way forward in the next decades, towards programs with highly adapted human performance, will be through programs which are generated through unfolding, in some fashion comparable to what I have described for buildings. (Pg. 197-198)
Blowing my miiiiiind. Apparently he told a bunch of computer science people this they said they do it already in iterative evolutionary adaptation. This made Alexander sad:
I realized that I had failed, in my lecture, to emphasize the real essence of all generated structures. The real essence lies in the structure-preserving transformations which move the structure forward through time, and which are primarily responsible for the sacess of the generating process. The needed transformations are not merely trial-and-error steps, or some neat way of continually checking and making things better. (Pg. 198)
Small step by step evolutionary ratchets are not enough by themselves. We need the structure preserving transformations to move the whole structure forward.
What I mean by a generated structure is just such a thing. It is a mistake-free structure which is beautiful and coherent, and which has become beauriful and coherent because of the impact of successive applications of the fifteen transformations which create, bring out, and generate its beauty and its elegance of structure. (Pg. 199)
Each differentiation, i.e. decision, is made in sequence and in 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 in context because the necessary information for a successful decision is not available prior to that step in the unfolding. (Pg. 201)
The system must be kept clean, pruned, and simple.
The transformations called siMPLICITY, INNER-CALM, and THE-VOID have as a direct function the task of keeping a structure clean of useless debris and open to the possibility of further useful differentiations. … In short, then, to make room for more and more relationships, there is a cleaning process going on in parallel with the differentiating pro-cess. 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, together with the fifteen transformations, which makes the beauty and majestic structure we think of as deep primitive art, or nature generating nature, at its best. (Pg. 201)
You must clean out the debris that you put there in the first place. That’s fine! That’s the point!
Living systems are not built by adding and aggregating things together, but rather by dividing and differentiating what is already present. Space becomes divided into inside and outside, walls become divided into windows and doors, doors are differentiated from each other, all grows from this.
Differentiating acts are formal, undertaken with an eye to the functional.
The rules of the game - the way architects, builders, planners, painters, gardeners, and ordinary citizens play their roles and undertake their actions - that is what controls the form of the world. And the rules of the game needed are precisely those generative processes introduced in chapter 6. The question is, “Are the present rules of the game likely to help people generate a living structure in the larger whole?” It is this question that confronts us. (Pg. 215)
And here is the magic; the fundamental differentiating process:
- At any given moment in a process, we have a certain partially evolved state of a structure. This state is described by the whole-ness: the system of centers, and their relative nesting and degrees of life.
- We pay attention as profoundly as possible to this WhOLenEss - its global, large-scale order; both actual and latent.
- We try to identify the sense in which this structure is weakest as a whole, weakest in its coherence as a whole, most deeply lacking in feeling.
- We look for the latent centers in the whole. These are not those centers which are robust and exist strongly already; rather, they are centers which are dimly present in a weak form, but which seem to us to contribute to or cause the current absence of life in the whole.
- We then choose one of these latent centers to work on. It may be a large center, or middle-sized, or small.
- We use one or more of the fifteen struc-ture-preserving transformations, singly or in combination, to differentiate and strengthen the structure in its wholeness.
- As a result of the differentiation which occurs, new centers are born. The extent of the fifteen properties which accompany creation of new centers will also take place.
- In particular we shall have increased the strength of certain larger centers; we shall also have increased the strength of parallel cen-ters; and we shall also have increased the strength of smaller centers. As a whole, the structure will now, as a result of this differen-tiation, be stronger and have more coherence and definition as a living structure.
- We test to make sure that this is actually so, and that the presumed increase of life has actually taken place.
- We also test that what we have done is the simplest differentiation possible, to accomplish this goal in respect of the center that is under development.
- When complete, we go back to the beginning of the cycle, and apply the same process again.
Pg. 216
I define a living process as any chain of differentiating steps, each of which carries out the center-intensifying process by means ofthe fifteen transformations, applying them, it-eratively, to the whole. (Pg. 217)
Step by step adaptation
Gradual progress toward living structure.
Alexander provides the example of 30 coins, to represent system state. A successful system has all variables at heads. Two approaches exist, flip all the coins and see if they are all heads. If not, flip them all again. Repeat.
Or: flip one coin. If it’s heads, keep it. If it’s not, flip it. Once one coin is heads move on to the next coin.
One of these ways is faster to arrive at an end state.
But what about flipping all coins, than just refilling tails? Where does that metaphor fit in? Feedback becomes important.
Think of the Sudoko puzzle - 81 variables, each with one correct answer and 8 incorrect answers. Some of the variables are given, providing some information for the rest. Writing a number in the square is akin to building some part of the real world.
In Alexander’s first example, solving without feedback, we randomly guess every spot at once. If the puzzle isn’t correctly solved, we randomly generate them all again.
In the second example, we start with the empty variable that has the most information, randomly generate until it’s correct, then move on to the next.
Rough models - intentionally rough - hell works through designs. The roughness makes them not precious, and keeps bodges and adjustments from being distracting.
It must lack a fixed, predetermined end-state. This is necessary because adaptation itself means nothing if changes cannot be made in response to the process of ad-aptation. By definition, such changes cannot be foreseen. (Pg. 240)
The Whole
Each step is always helping to enhance the whole.
Ego and greed, or pressure and stress from systemic forces, can force us to focus on the local, and forget or ignore the larger impacts and implications.
The larger context of the whole stretches out from the whole of the thing that is the the thing we are making. This whole has latent centers that are suggestions, and step by step we strengthen the centers with an eye to strengthening the whole, and the whole that stretches out beyond that.
In the early stage we must concentrate, of course, on broad structure, on the emergent structure of the whole. A difficulty is that we do not have a good notation for the emergence of form during this carly process. The notation that architects traditionally use is a language of drawing or computer representation. Using this notation, we try to get the glimpse of an idea, and then put this down as a simple sketch, or as a simple model. But such a sketch always in dudes too much information, too early, so that the sketch (or computer drawing) is invariably over-specific. Sketches and computer drawings are seductive and may be interesting. But if only 20% of the information in a sketch is based on real decisions (Pg. 256)
Instead of detailed sketches and comps, Alexander uses “morphological ripples”. A fuzzy suggestion of form that feels right and good but has yet to be determined definitely. More ripples come and interact, and we try and capture as much as them as we can in a written spec.
AT EACH STEP DECIDE ONLY WHAT YOU KNOW WITH CERTAINTY (pg. 257)
The content of each decision should be small, and only finalized when your fairly certain it is correct. The scale of the decisions will start quite large and gradually fro finer and finer.
Similar to our Sodoku example above, most of the open variables will be bad choices for the next thing to decide. We need to be very critical of what ideas jump into our head for the next thing to do, as the odds are it is the wrong thing to do.
Always making centers
So that every center is shaped by the next step in the differentiation
Alexander talks about how spending time and money on projects can and should be a healing process for the environment:
We need to look for a place which can benefit from the expenditure of the capital (pg. 270)
Hold the rough idea of what you are making lightly in your minds eye, let it guide your work and let your work guide your idea. This is conceptual labor, and being aware of the conceptual labor you are doing.
All the centers, from the biggest to the smallest must mesh together, work together, always be improving the whole and learning and adapting to the ones that came before. There must be no space between them, they must all melt together. This is how the proper effect is achieved.
The Sequence of Unfolding
Generative sequences are the key to the success of living process whenever complex structure is being formed.
Rather, a good artist knows that an emerging work will come out well to the extent that the artist does things in the right order. This order is not usually written down, nor is it pre-fixed. But many artists are aware, while making a new work, that what is first laid down, what is done next, what third, what fourth, that controls the quality of work that comes out. (Pg. 300)
In process, fix the bigger structures earlier.
A chant in Samoa is the process and instructions for making canoes;
The Samoan canoe chant is an example of a generative sequence. It generates canoes. The generative sequence is essentially the same each time that it is used, but each canoe that is generated is unique. (Pg. 301)
More notes and examples on path dependency, after talking about he sequence of placing a house and garden on a lot - which one you place first makes a huge difference. Alexander says place the garden first.
And as you can imagine, if the sequence - the order of steps - is significant when dealing with only two operations, how much more significant it will be when dealing with ten, or twenty, or fifty steps. There the chance of getting the sequence right by accident, is extremely small. (Pg. 303)
A sequence works, or does not work, according to the order of the steps. (Pg. 305)
Figuring out the path dependency of your process is itself a difficult task:
By doing experiments on test cases, it is possible to winnow out, and ultimately to eliminate the bad sequences, thus gradually finding one’s way to the few see quences which have the property that no such backtracking occurs as the project unfolds. (Pg. 307)
Generative sequences allow for the operationalizing of conceptual labor processes
The crux of every design process lies in finding the generative sequence for that design, and making sure that sequence is the right one for the job. (Pg. 317)
Do one thing at a time, one decision at a time. Envision the entire project as a whole, and each step you take improved the whole, as an entirety.
A student of Alexander’s sums up:
“… Just do the most important thing. Finish it. Then do the most important thing. Finish it. And so on.” (Pg. 318)
Every part unique
A process so finely tuned to circumstance that every part becomes unique.
Living structure though arrangement of modular identical pieces is not possible in principle. What you need to make it work is repetition and symmetry, similar but not the same. Each moment is adapted to its own place and informed by the preciseness of its own context. This theory is held out with photographs of atomic structures - even atoms are not identical. This undercuts a huge philosophical underlining of modular design systems in the 20th century.
Even simpler than the student above:
Just make it nice at every spot. (pg. 337)
What that means at every spot is different.
Patterns
Generic rules for making centers, or, making life enjoyable.
To this point we haven’t talked about defining “what it is we’re trying to do”. How do we define and describe the function of the project?
The patterns that define the function of a project are centers that reflect the human culture that sites the project. Definition of the patterns must primarily be sourced from the human culture they extend from.
Alenxader makes an interesting point about how human culture and expectation creates a physical salience in a space, and actively adjusts the wholeness in a room - a wholeness which is itself geometric. Therefor human culture has pressure and physical presence on geometric structure. He follows this with the example of objects in proximity to each other being able to change each other through their proximity. I accept this as it seems fundamentally accurate.
What we see as the physics of the chair, its mathematical structure, then appears subtle and vaincrable it is not constant and fixed. It is changing, subtly, all the time, as influences from the world around enter the picture, and modify the relative salience of different centers in and around the chair, thus forming, reforming and altering the wholeness which was “ the chair.” (Pg. 348)
Alexander argues that the physical wholeness of the world is actually a different place when inhabited by indigenous peoples or colonial peoples. Not just the perception of the place, but the place itself is physically changed by the presence of indigenous participation in the place.
The tricky work of defining patterns for the generation of centers;
It is a question of rearranging the fune-tions, redefining the nature and meaning of the way the thing works - shich saddenly opens the door to a new comfortable life, which will actually work. (Pg. 357)
In this way, understanding what a thing is and how it works is the key point,
…the “design” is less important than the choice of just this center. (Pg. 360)
Going to the essence in a way that matters is often simple, rough, and cheap. Building to an image to create an impression - the opposite.
The essence of the point, is to find - or create - a set of centers which, together, will generate a complete and coherent object of the type we are looking for. We keep trying out our rudimentary list to see what sort of whole this list of centers will generate. We then use intuition and feeling to judge the deficiencies in the whole which is created, to make us aware of more centers that still need to be created. (Pg. 363)
Deep Feeling
THE AIM OF EVERY LIVING PROCESS IS, AT EACH STEP, TO INCREASE THE DEEP FEELING OF THE WHOLE
This chapter is beautiful, and largely full of examples of how to let the deep feeling of liveliness guide your work when it comes to making of decisions. In short, work through the options quickly and lightly, waiting and looking for the stirring of a deep feeling that builds upon and makes the entirety of the project more poetic and rich. Than do that thing.
This is very, very hard.
Emergence of Formal Geometry
Appearance, finally, of coherent form.
There’s a moment in a project where the time comes to impose order upon what has been grown naturally from the process so far. In architecture this moment is when it comes to determining the structural underpinning of the building, where the columns and load bearing walls go.
Alexander imposes an aperiodic grid upon the form as it has grown so far, reacting to all the decisions made up till this point. This grid creates the buildings middle-scale structure and order, and is an important part of building the living whole.
He calls this imposition of structural order a “brutal” process (in the French sense of “raw” rather than the English sense of “violent”).
Thus, the elaboration of form which occurs, is a raw harmony made of the simplest elements, in a telling and geometrical order, which is balanced and gauged to convey raw feeling. It re-peats. It is often brutally simple in its regulari-ties. But each mass is calculated to oppose, or complement, or complete, the harmony of the other elements. (Pg. 428)
Alexander also notes that any living process will demonstrate this “brutal” imposition of order:
I believe - but am not in a position to affirm with certainty- that a similar “brutal” and purely geometric process always occurs somewhere in other kinds of unfolding that generate living order. I have had conversations with colleagues who are trying to extend what I have presented in this book. They tell me that at least something similar is going on in poetry, in dance, in the formation of social structure, in planning, even in the creation of healthy family relationships. (Pg. 429)
This is Alexander’s way of seeing what Daghlian calls “Conceptual Labor”. Conceptual Labor is the work we do the understand the work that we we think we are trying to do while we are doing it. There comes a point in that recursive process where the conceptual labor resolves, flattening our understanding of a context and a problem back to the realm of traditional labor. Like Alexander, Daghlian notes that this is really hard. I think that here Alexander is describing the shift that occurs as the large-scale conceptual labor of the whole building resolves, forming the middle-scale structure, and leading to the next phase of the fine details.
This is the pivot point from “what are we trying to do” to “how are we going to do it”.
Form Language and Style
HOW CAN HUMAN BEINGS IMPLEMENT A GEOMETRICAL DIFFERENTIATING PROCESS SUCCESSFULLY?
Regarding design and style,
It is not merely an individual problem, a natural personal concern of each individual artist. Rather, it is a fundamental issue that belongs square in the middle of the analysis of living structure as a product of society. (Pg. 432)
How we build our sense of this from our society,
Somehow, we learn, over years, the ingredients that make a building good. Somehow, this knowledge is kept by us, occasionally transmitted to others, but mainly kept by us, in our minds, as a store of knowledge, which we may draw upon … the combinatory system we grow in our own minds, the form language we use to speak the words that come out as buildings (pg. 423)
The form language supersedes the pattern language, a unification of function, form, style, and process. The form language is manifested in the physical act of creating a thing in a way that performs a function.
tools which directly give people the language of form (pg. 434)
The form language is made of small, modest, simple steps that help bring about an unfolded wholeness. They are not patterns to plug and play, but actions to take and methods of creating differentiation and transforming structure.
A critique of the programmers approach to the pattern language;
they have been undertaken, so far, within the context of the value-free, postmodern attitude, which almost by definition is not able to deal with problems of life and tries instead to study its subiect within a purely neutral conception. (Pg. 443)
Wrapping up the book is a chapter on appliquéing living processes to broader society, with the eye to making society a system that outputs a living world rather than a destroyed one. Alexander places a lot of the blame on Taylor, but quite surprisingly shows that Taylor had a good sense of the importance of living process:
“Now, in the best of the ordinary types of management, the managers recognize frankly that the workmen who are under them possess a mass of traditional knowledge, most of which is not within the possession of the management. The most experienced managers frankly place before their workmen the problem of doing the work in the best and most economical way. They recognize the task before them as that of inducing each workman to use his best endeavors, his hardest work, all his traditional knowledge, his skill, his ingenuity, and his good will, in a word, his initiative, so as to yield the largest possible return to his employer.” - Frederick Fucking Taylor, Apparently
Then he destroyed it! On purpose! What the hell man!
Contra Taylor, on the interaction between design and engineering;
I argued that the questions I wanted the soils engineer to ask, would depend on the preliminary design I had in mind, which in turn depended on his first rough findings, and that I needed to keep the question and answer process going, dynamically, between myself and the engineer, while the work went forward. (Pg 517-27)
And with a client refusing to bring engineers in until after drawings are done:
This made genuine and eftective cost-squeezing impossible during the early design process - the very phase of work when it had to be done in order to be effective. (Pg: 517)
And a great summary of conceptual labor:
Just to decide where to put the houses, one had to clear some trees, some underbrush, grade some land. And in doing those elementary first steps, to see where to put things, one was already making de-cisions, merely in the act of finding out how and where to make the decisions. (Pg. 518)
And regarding floating as much agency downward in a system as you possibly can;
We see here, a principle of local autonomous adaptation, needed for successful adaptation of the larger system. It requires a level of knowledge and expertise in the hands of people of society at large. (Pg 521)
As for how to create, share, and evolve small patterns than have the potential to make the world more living:
The natural answer to this question for the aist century is, of course: The Internet. For the first time in human history, there is a public, world-wide, communal memory that is - in principle - accessible by every-one, and instantly accessible. (pg 542)
Of course this is Alexander’s California Optimism showing through a bit, but I think the principles are sound. My aim is to take these processes and goals and apply them to making the Internet, and thereby putting them on the Internet by demonstrating their use and function within the Internet.
Regarding his own published sequences at patternlanguage.com,
The sequences are relatively simple in char-acter. And they are operational. That means they give practical instructions which allow people to create or generate versions of things unique to a given circumstance, in a natural and unfolding way. And they are simple enough, so that judging their effectiveness, and seeing how to make improvements is relatively easy. (Pg 542)