Environmental & Architectural
Phenomenology Newsletter
Grasping the Ineffable: From Patterns to Sequences
Jenny Quillien
Quillien teaches in the Laboratory of Anthropology, New Mexico University at Highlands, Santa Fe, New Mexico. She worked for six years with Christopher Alexander on his four-volume The Nature of Order and on the Pattern Language website (www.patternlanguage.com). An earlier version of this essay was presented as a paper for a special session on Alexander’s work held in October, 2006, at the annual meetings of the International Association for Environmental Philosophy in Philadelphia. jenny@jqsolutions.org. © 2007 Jenny Quillien.
Many faithful readers of Christopher Alexander’s A Pattern Language were disconcerted to discover that The Nature of Order was not its obvious sequel. A Pattern Language had afforded access to a well established and broadly pragmatic response to ordinary problems in the art of building. Now, nearly 30 years later, The Nature of Order confronts those readers with demanding material having few comforts of the user-friendly handbook style of the earlier work.
This article reviews one of the significant differences between A Pattern Language (APL) and The Nature of Order (NO) and explores the advantages of bridging the fundamental projects of these two works. The central theme reviewed here is that of patterns versus sequences.
Attributes
of Patterns
APL
is a compilation of architectural patterns
(honed solutions to recurring problems), combinatory rules, and techniques
for practical results. Key attributes of patterns include the following.
MINED
Working
patterns are, in Richard Gabriel’s expression, mined. Like diamonds,
patterns are the result of many years of process. We don’t make them—we
find them, polish them, use them, and value them. Developing patterns from
scratch and all in one go has proven to be extremely difficult.
MANAGEABLE CHUNKS OF
INFORMATION
Whether large scale (e.g.,
pattern #3—city country fingers) or small scale (e.g., pattern #200—open
shelves), each pattern is immediately graspable as a manageable chunk
of information.
BUILT & SOCIAL OVERLAP
Building patterns are obviously correlated to social patterns.
Consider an example such as the weaver in Libya (photograph below) who has
constructed his own place of work—pattern #80 (self-governing
workshop). Or consider patterns #133 (staircase as stage) or
#139 (farmhouse kitchen).
NESTED
HIERARCHIES OF SCALE
A hypertext structure supports
selecting and combining patterns of different levels of scales into a
coherent whole. Illustrated here is the simple example of #159 (light
on two sides of every room) calling for overlapping patterns at the
smaller scale (e.g., #192—windows overlooking life) and at the
larger scale (e.g., #106—positive outdoor space).
PROCESS & RESULT
Patterns are written as
mid-level abstractions and work as design constraints. Their concrete
guidelines do not unduly limit the builder. Each pattern offers both a
finished state (a verbal blueprint of the desired result) and a process
description (a guide for action). The guides for action are simple and
direct. More like origami, they are not difficult in the way that
following a blueprint is difficult. The instructions allow for an infinite
number of renditions.
INFINITE
CLUSTERS
Patterns are clusters of
geometry. A pattern language is a coherent subset of patterns and
combinatory rules that, like an individual pattern, can give birth to an
infinite number of variations.
Below are three renditions of a simple traditional house form found on the Caribbean island of Aruba. The pattern language for this house type has not yet been explicated but would include such patterns as #127, intimacy gradient (these modest homes all have a small front room for visits from the local priest and for other ‘formal’ occasions, while the more intimate spaces are located toward the rear).
The pattern language for these houses would also include patterns not found in APL. For example, behind the front room are cooking facilities and access to a pattern that might be called “winnowing breezeway.” The houses are placed so that their back give out onto a large working area, placed for maximum breezes.
COMPLEX OVERLAPS
Consider
the streetscape from Sarlat, France, below. The potential for the subtle
complex beauty of deeply organized spaces emerges from non-simple
juxtapositions and overlaps of individual patterns. No doubt, this is the
aspect of working with patterns that is the most difficult. It is also the
aspect that has been the least elucidated in the writing about patterns
and their implementation in the built world. It may also be one of the
reasons why so many well-intentioned attempts at using patterns have
resulted in rather mediocre spaces.
Attributes of Sequences
NO
is a ‘trail-blazing’ intellectual
work and a tour de force. Its goal is to understand the very nature
of order and through this understanding bring us closer to grasping the
ineffable life of beautifully built spaces. Creating profound spaces
occurs largely through generative sequences. Key attributes of sequences
include the following.
PROGRESSIVE
DIFFERENTIATION
An Alexander doodle,
below, provides a simple example [from NO,
vol. 2, p. 280]. A sequence is the mindful ordering of
decisions to be made. Decision n creates the context for decision
n+1, which, in turn, creates the context for n+2. The net
result of a building sequence is a progressive differentiation of space.
WHOLENESS AS GUIDE
& GOAL
The process is recursive. At
each decision point, we must grasp again the wholeness that exists to
determine which next move will best call forth the latent structure.
EACH STAGE
STRUCTURE-PRESERVING
The key is in getting the
sequence of decisions right. For example, starting with the wholeness of a
building site, understanding what to preserve and enhance will lead to
initial decisions about where not to build. This is very different,
say, from starting with engineering decisions about the most efficient
layout for water and sewer pipes.
Recognizing a structure-preserving differentiation is intuitive but usually involves: (a) minimum symmetry breaking; (b) connecting to the next smaller and larger scale; (c) enhancing existing centers or creating new centers; and (d) transformation through one or more of the 15 geometric properties Alexander identifies in volume 1 (centers, boundaries, echoes, deep interlock, good shape, not-separateness, simplicity, and so forth) .
CENTERS BOOTSTRAPPING CENTERS
Structures that have ‘unfolded’
have a much higher density of connections and are therefore more robust
than other structures. The whole and the part reinforce each other as
centers are added and strengthened. For example, bones, the overall shape
of which is often asymmetrical, are subtle, complex and very robust. A
hipbone carries much of our weight and allows us to stand, walk, and sit.
During growth, calcium is added according to the stresses placed upon the
structure.
UNFOLDED
FORMS
Morphogenesis and the mathematics of growth and structure preservation are
not fully understood, yet we do intuitively distinguish unfolded forms
from those that are not. Consider the two photographs, directly below, of
unfolded streetscapes in Amsterdam, Holland, and Lawrence, Kansas. Compare
them with the next two photos of streetscapes in
Rotterdam
and Aruba, where the buildings are not unfolded but are image-based and
template-assembled.
SHAPE AS THE
TRACE OF TIME
A plastic
flower, no matter how clever its design and fabrication, cannot match the
original. In a real daffodil, the form, subtle variations in color and
texture, the infinite differences between that daffodil and all others in
the same field are the consequences of growth and constant adaptation
between parts and whole over time. This fine-grained complexity can be
obtained in no other way. It is this particular quality that provides so
much pleasure and wonder.
Bridging the Two Works
It would be “preaching to
the choir” to argue the general importance of the built environment for
our individual well being, our communities, and the health of the Earth.
More to the point is the reaction of those who have now worked their way
through the four volumes of NO.
Those who have read both APL and NO, in ways that are difficult to define, sense that sequences are more fecund than patterns. When we encounter profound spaces, the concepts of unfolding and wholeness offer more insights into our experiences of ineffable comfort than do problem-solution protocols. It seems also possible that sequences might offer a deeper level of intervention when we strive to create profound spaces.
NO is unwieldy. Doing some absolutely necessary development work is an immediate first step to more fully adopting the advantages of sequences as real-world practice. Alexander himself would not disagree, having been involved some 30 years in the arduous task of constructing a logical chain of arguments as clearly as he could and looking to a new generation to carry the practical work forward.
What follows are some pragmatic proposals and questions to help develop the concepts sketched out in NO. Essentially, my suggestions are to take cues from APL and own up to a few hypocrisies.
A RETURN TO
ETHNOGRAPHIC METHODS
Much of APL’s richness is due to the ethnographic methods used
to define patterns. These same methods of mining, documenting, and
polishing could be used to compile a repertoire of sequences of
existing and successful building forms.
As with APL, the first repertoire might consist of small, manageable sequences that represent intuitive chunks of information relating to social patterns. Maintaining a level of mid-level abstraction and writing sequences as recipes for action guarantees the potential for infinite renditions.
Readers of APL enjoy a multi-leveled use and can dip in piecemeal for individual patterns or simply to spur their own creativity. They can play with different pattern combinations to create their own language. They can approach APL from a meta-level of nested, problem-solution protocols as the computer programming community did so successfully. A repertoire of sequences for existing built forms might aspire to that level of achievement.
FRONT & BACK STAGE
MAKING
Use of such sequences as
“recipes” need require no more understanding of underlying phenomena such
as geometric properties or implicate order than cookbook use requires
scientific understanding of chemical reactions of ingredients under heat.
Preparation of the sequences as sets of instructions will eventually require extensive knowledge and experimentation with the mathematics of growth, wholeness, force fields of centers, implicate order, the interactions of geometric properties, and so forth. This level of development work on sequences would surely bear fruit and take us beyond the useful but more mundane stage of pattern languages into the deeper realms of profound beauty, living structure, and dwelling in the sense that Heidegger used the term.
FACILITATORS TO
THE RESCUE
The
patterns compiled in
APL
and the vast majority of examples used in
NO
are gleaned from centuries of “architecture without architects” and from
cultures where those who built were those who dwelled.
In these cultures, knowledge of local patterns and the building skills of carpenters, roofers, masons, and other craftspeople were widespread. Small communities raised houses and barns with local materials and basic tools. This fusion is now largely lost, a fact constituting a major challenge. The average family today is clueless as to design and construction.
Although APL is still enthusiastically adopted by do-it-yourselfers and, although it is the basis for many charming remodeled homes, large-scale grassroots building based on pattern language has not taken place. Those few larger projects that were attempted all used intermediaries to steer the end-users and translate their desires into actions.
The ethos of Alexander’s material is “bottom up,” but 30 years of APL history show a reality of “top down” management and users who too readily defer to professionals. The New Urbanists, Alexander’s philosophical neighbors, have been more successful in carrying out large-scale projects. Andres Duany, a founder of the New Urbanist movement, is less hypocritical about the loss of the connection between building and dwelling. Yes, he and his colleagues run charrettes with end users and other stakeholders. But these charrettes are always planted with what Duany calls his “black shirts”—trained experts who manage the interface between lay people and professionals.
It may be that progress for Alexander’s approach will also require such intermediaries—practitioners, thinkers, and instructors trained in interviewing, consulting, designing, and building living structure through sequences. The New Urbanist movement is an important model and perhaps a potential partner.
EXPERIMENTATION
Experimentation with novel
sequences in both green-field and brown-site construction would also
provide insight. How do people, both professional builders and lay
audiences, actually experience the reality of creative building with
sequences?
After six years of working with Alexander on the text of NO, I personally felt the need to try my hand and took on a small project—remodeling a modest 1930s adobe house in New Mexico. In this hands-on experience, the concepts that turned out to be the most useful were not the obvious ones. Rather, they included the following:
Mistakes redefined as mis (leave out) take. In a sequence (n, n + 1, n + 2) a decision must take into account all the variables present at each stage. A mistake is to miss—to fail to take into account—a variable. For sure, I made mis-takes, but my skill and results improved with practice.
Working with step-by-step differentiation does reduce the number of variables to be contended with at any one decision point and allows for evaluative hindsight at frequent intervals. The current practice of blueprint-to-completion contracts with very costly stipulations for change is a guarantee for mediocrity.
Double mapping. For example, in a kitchen or office sequence, start by mapping activities in a detailed, personal way. Only then map out that sequence within the given space.
Stay qualitative as long as possible and do rough, large-size mock-ups to increase the odds of structure-preserving moves. For example, nobody can predict the effect of a color from a square-inch sample. Get butcher paper and small quantities of several possible colors. Brush on the paint and pin up the butcher paper. See what each color does to the whole. Most of us should avoid computer-aided-design programs, which in theory can be modified repeatedly but, in practice, tend to lock people into premature and poor decisions.
Levels of scale as a tool. Making sure each decision informs the next larger-scale and next smaller-scale decision.
Attitude. Every mis-take, large and small, came from being in a hurry to finish and thus inattentive. Quality came from a still point where the boundary between subject and object (me and the work) dissolved into calm reciprocity.
The simplified fundamental process. The simple question, “What is the next, most simple step to bring forth more life?” does work. Go with that. Don’t worry about the theory.
Realpolitik
& Plugging in
Someone
has suggested that if you want to learn something, go to where the
disagreements are. The disconnect between Alexander and the New Urbanists
is instructive.
Andreas Duany is fond of saying that his whole career has been founded on finding ways to take Alexander’s thinking and “plug” it into mainstream construction and urban renewal. Alexander’s stance is that current building practices will never lead to good buildings. “You can’t get there from here.” he says. In contrast, Duany says, “We have to get here from here. Here is where we are.”
It is fair to say that much of the success of New Urbanism is due to embracing the world as it is and, at least partially, coming to grips with current building practices and identifying leverage points for change: codes, zoning, design methods, financial regulations. Even if the results fall short of Alexander’s vision, taking the world as it is may be the only way to get beyond armchair philosophy and a rarely read four-volume opus on the library shelf.
There may also be other, less known avenues that would allow open-ended experimentation with the NO approach. We could make an analogy with the medical field where health is defined by absence of illness. Simple elimination of illness might give scope to healthier building.
Consider the work of Iraqi historian Besim Hakim, who has studied the old Muslim cities of the Mediterranean, where building codes were not, as ours are today, mechanical (i.e., standard setbacks of so many feet). Rather, based on more general laws of intention, these laws were really cultural and moral injunctions.
For example, a builder of a new house would have the intention of not interfering with the privacy or views of existing houses. Just how this is to be carried out could be creative and finely adapted to the specific context. Or, consider a building code that simply limited the kinds of building material to those already present, rather than imposing Disney-like formulas. Consider home financing that would favor families upgrading a current home rather than speculation and house “flipping.” Changing the financial context would naturally change the decisions.
Theories of
Fun
During
the years working with Alexander, I participated in early and limited
attempts to transcribe two patterns (entrance
transition and
one-room
cottage) into
sequences. Experiments asking volunteers to follow the sequence and
envisage an entrance or a cottage fell flat. People found following the
sequence to be frustrating, confining, counter-intuitive, and boring. No
real construction, nothing beyond the roughest sketch, was ever tried.
One source of help could come from a surprising corner. APL’s unexpected readership in the 1970s and 1980s was the computer community, whose programmers saw the meta-level applicability of the pattern-language approach to programming problems. Perhaps this same community will now take their turn as the provider of new approaches. Will Wright, author of computer games such as Sim City, readily acknowledges inspiration from APL. Work on the theories of fun by programmers such as Raph Koster, offer insights into why computer games are addictive, why APL is widely appealing, why early draft sequences were not, and how one might construct experiments with sequences engaging wide audiences.
The key to “fun” in computer games is the experience of discovery as players go through nested levels of partial revelations. Abductive computer modeling affords the player increasing apprehension of unity. The play is its own reward as the player generates new perceptions and resolves anomalies and finds order. The validation is through an increase of one’s scores.
If ethnographic methods and backstage work can produce a first generation of building-form sequences, theories of fun may guide their recasting into a second generation so that the user’s exploration of how to differentiate space is, in itself, rewarding. The third generation would be beta testing with a larger audience. Perhaps forms of extreme programming can be used for reiterative corrections–hastening the natural process that requires centuries for patterns and sequences to emerge naturally.
One caveat is in order: The best game programming is still based on closed menus. Tutorials on known cases would work. Open-ended exploration on novel sequences would not work.
Emerging
Forms and Methods
Often
criticized for being “stuck in the past,” Alexander’s standard reply is he
chooses old buildings as examples, not because they are old but because
they are better. Fair enough—up to a point. Understanding the sequences
behind deeply beautiful places from the past is not always nostalgia.
These efforts can inform new sequences, new forms, and supporting
technology for our own time and place.
But our society is unlikely to return to locally-based communities that remain coherent and stable. Ours is a time of globalization that ironically leads to both fragmentation and commodification. Strident narcissism is expressed in the built environment through the tidal wave of banal McMansions, aggressive stakeholders gambling on speculation, and professionals striving for signature buildings.
Indeed, we can argue that there has been a change of state in the relationship between the built environment and what Thomas de Zengotita calls in his recent Mediated, the overly “flattered self.” Institutionalized seduction has become the normative mode of thought and behavior. We expect to be seduced, only dimly aware of how vacuous it all is.
We can build on the strengths of A Pattern Language. We can train black shirts. We can use open-source shareware for assembling results from a wide pool of experimenters. We can build on the concepts presented in The Nature of Order to create beauty.
But there will have to be some change in awareness before we can learn how to dwell and, thus, how to build. The real conundrum is how we get ourselves into the future by finding what philosopher David Levin has called a “rediscovered primordial attunement.