Pattern Language and Interactive Design.
Nikos A. Salingaros
Department of Mathematics, University of Texas at San Antonio, One UTSA Circle, San Antonio, TX 78249, USA.
Chapter 9 of Principles of Urban Structure, Techne Press, Amsterdam, Holland (2005). Originally published in Poiesis Architecture (Toulouse) No. 15 (2003). Earlier version entitled "How the Pattern Language Defines a Geometry for Urban Interfaces" published in the Proceedings of the International Seminar: "Design With the Community", University of Rome III, (2000), pages 15-21.
Using the work of Christopher Alexander, I present a method that uses "patterns" to aid in urban design. The method allows inhabitants of a particular neighborhood to participate in planning their own environment. Most important, this approach replaces the current practice of large-scale intervention by planning authorities, which is usually insensitive to local needs, and frequently results in unlivable environments that are resented for generations afterwards. The practical method is supported by a significant body of theory from mathematics, neurobiology, and evolutionary psychology. Nevertheless, user participation in urban design has not generally led to satisfactory results. This is explained in terms of anti-patterns, or "memes", that predispose the users to certain fixed images.
- Two contradictory models for urban interfaces
- The design method in practice
- Preserving what is most valuable
- Emotional dimension of design
- How patterns are displaced by anti-patterns
- Anti-patterns that destroy urban interfaces
Attempts in the past to enlist the participation of inhabitants in planning and designing their own environment have not been entirely successful. There is at present no theoretical basis for collective design, and there also exist severe impediments to any participatory process. Collaborative design is not as well established in architecture as it is in some other fields, such as research and computer software development, where the concept of "collaboratories" promotes interactions between the involved parties. Complex software cannot be written without two separate collaborations: (I) between individuals on the software team that have to deal with different pieces of code which are too complex for one individual to handle; and (II) constant feedback between the software team and the end user.
Some architects approach the design of structures with community participation, and serve as teachers, guiding a collective discussion towards a design goal. In each of these cases, the process changes according to the project and the participants. In mainstream architecture, at least, there has been no clear prescription for participatory design. With the publication of Christopher Alexander et al.'s A Pattern Language 1, a new and comprehensive technique was made available, and was subsequently applied by Alexander and others in completing successful projects. It is only because of a lack of published material that this method has not become more widely known. This situation is soon going to change, with the publication of Alexander's book The Nature of Order 2 .
Here, I am going to discuss a process of interactive design using Alexandrine patterns. My brief outline is no substitute for Alexander's description of his method, which should be consulted by every serious practitioner. What I am going to focus instead is on a fundamentally problem: why any method of participatory design is likely to fail because of ingrained anti-patterns in the participants' subconscious. In The Structure of Pattern Languages 3 , I describe how patterns combine in building urban boundaries (simply as one possible example of their application). After outlining an interactive design method here, I will spend some time pointing out the pitfalls that can make it ineffective, and which must therefore be avoided once identified.
A "pattern" is, quite simply, a recurring solution (here for a problem in architecture or urban design, but generally for any social or technical problem). Regardless of how the solution was originally obtained, as soon as human beings identify a pattern and communicate it either orally or graphically, this confers an enormous advantage to the group. The ability to communicate patterns eventually builds an arsenal of reusable design knowledge. Thus, the language of patterns is closely linked to culture and tradition 3 . Each pattern represents a rule governing one working piece of a complex system. A pattern language allows patterns on a smaller scale to combine and support patterns on a larger scale. Large-scale patterns themselves are necessary because they contain more information than the smaller-scale patterns on which they rely, thus showing emergent properties.
I am not offering my own or Alexander's personal preference, or an ideal theoretical model. I am outlining what I suspect to be the design process as it has occurred to create living cities the world over for several millennia. Finally, I will attempt to explain the failure of participatory design from a novel point of view. I will review supporting results that explain the theory in terms of the competition between patterns and anti-patterns. (Anti-patterns can in many instances be treated like mind-viruses, or "memes", a concept taken from the literature of biological evolution). Without this understanding, there is no point to introducing a participatory design method at all, because of the resistance from the modernist architectural tradition.
Two contradictory models for urban interfaces.
The Structure of Pattern Languages 3 derives a definite (fractal) geometry for urban interfaces, such as is found both in traditional cities, and unplanned human settlements and favelas in the third world. The twentieth century saw a deliberate reversal of traditional design rules so as to allow novel forms to be produced. People have argued at great length for or against this transition, but the discussion has always taken place purely on a stylistic level. Urban boundaries influence the activity matrix of people in cities. A boundary's geometry is determined by fundamental processes, and if this geometry is tampered with, it inhibits the everyday actions that contribute to make a successful city.
I believe that the imposition of a simplistic geometry on city form, by suppressing more traditional patterns, has curtailed or eliminated the traditional functions of a city that make it alive. Contemporary design philosophy gets rid of connective interfaces altogether. A deeper problem arises out of the modernist desire to visually "purify" areas by eliminating complex structure, subdivisions, and connections. Consolidation of functions by concentrating them geometrically eliminates the complex mixing that characterizes a traditional city. In our times, urbanists intentionally de-couple urban elements by separating them spatially.
The twentieth century has invented urban boundaries that are very bad interfaces. Many of these are made possible by technological developments that were unavailable in traditional cities. Urbanists have created a new set of rules that are used to define urban interfaces. The interfaces of twentieth-century cities prevent most of the activities that occurred there in previous times. People have been taught by schools, critics, and magazines to prefer smooth, unnatural edges and boundaries over ones with fractal scaling that resemble natural structures. Architecture schools instill a precise, sleek image of the world into our culture. Any participatory design process therefore, is bound to get hung up by the two contradictory views of the built environment.
For this reason, I emphasize the need for two objectives in participatory design: (I) education of the users by reintroducing patterns that are timeless; (II) careful avoidance of modernist images that are bound to work against the patterns. Otherwise, the contradictory forces will destroy the coherence of the final product. Unless these two points are clearly distinguished, then the design team is going to get irreconcilable demands from the users, who, being under the influence of the architectural media, will associate economic success with the most disastrous examples of the built environment in the twentieth century.
The design method in practice.
One first holds preliminary discussions with residents of any region to establish the major, i.e., most significant patterns for them. This could be by consensus or via a majority vote, but only after some discussion to "mine" for patterns hidden in the residents' subconscious. For example, in designing the Eishin School Campus in Tokyo, Alexander found to his surprise that most teachers and students interviewed put a lake as their top priority. He then built an artificial lake, and this feature is what makes the school most extraordinary. An insensitive architect and contractor would have dismissed this wish as outrageous, and actually, some interviewees were initially afraid to express their vision of the lake, because they feared that it was too crazy to even mention.
Once the residents' deepest wishes are noted, then the architects have to come up with patterns to express those desires written in pattern format, as in A Pattern Language 1 . If such high-priority patterns are not already documented in the pattern language, they have to be researched by looking at the most beautiful examples in existence around the world, and identifying what makes those examples so successful. It could be geometry, position, connectivity, or a relationship to other patterns. In the meantime, the architects should have picked out a dozen documented Alexandrine patterns that are most relevant to the job at hand. These are going to be presented to the residents in pictorial form.
The next step is to hold an educational session with the residents, and to teach them the twelve most important Alexandrine patterns that apply to their project. To these are going to be added any additional patterns that come from the residents themselves. The rest of the work involves combining all the patterns together in many different ways, and judging the end results. Here, enormous work is saved if one can do the simulations on a computer, and view the results graphically. Unfortunately, most existing software gives a very sterile picture of designed environments, which is totally useless for judging the life that a particular solution will have when built.
For practical reasons, the completion of the project will be in the hands of the planner/architect, with one or two residents closely participating until the end. These residents may be individuals who take a more intense interest in urban design, and can be expected to represent the entire community. It is not suggested that any more collective group decisions be made, as different members of the group will certainly diverge in details and preferences, and the project will then be delayed indefinitely. The input, and even veto power, occurs in the major patterns in the beginning, which is precisely where the most damage can be done by insensitive planning.
Another point is that, as the design process progresses, many more patterns will have to be brought into play as things get built on different scales. The project becomes more complex because of the increasing number of possible pattern combinations, and quick decisions will have to be made. Many of these will be driven by competing technical, practical, and legal considerations. As long as the scientific principles of patterns and their combinatorics are respected, however, one can choose equally well from among an infinity of different choices, all of which will lead to a successful result.
The great power of this method is that it is recursive on smaller scales. That is, once the largest scale is decided upon, individual regions on the next smaller scales may be tackled. One works with about five patterns to solve a specific design problem. The number five comes from a "chunk" of disparate pieces of information that the mind can juggle simultaneously 3 . Again, one can choose from among the original Alexandrine patterns, or if required, a new pattern is derived for the occasion.
Preserving what is most valuable.
In urban restructuring, it is essential to preserve what has most life in the existing environment. Because of the infection by mind-viruses discussed later, it is almost always the case that urban renewal destroys the few objects and places that have life in any neighborhood. It is therefore necessary to identify these by common consent, and to put a priority on preserving them so that they are saved intact in any new plan. Examples of failing to do this abound around the world. In one instance, the Indonesian Government rebuilt an entire village that was destroyed in an earthquake. The planners ignored the traditional sacred places in the old village, with nothing to take their place in the new village. The result was a loss of culture and identity, with a severe disorientation and cultural breakdown.
It is the people themselves, infected with anti-patterns from the media and education, that usually want to eliminate their most sacred urban places. Those are incorrectly identified with the past, and with a way of life that most residents are ashamed of and wish to erase. People are frequently seduced by empty images of prosperity that they have ingested from magazines, and imagine that if only their environment would look like empty North American suburbs, they too can aspire to a high standard of living. Almost universally, when people acquire the money to alter their environment, they invariably destroy what is most beautiful in it. This is most striking in our times of cultural disorientation.
For example, in a neighborhood, a corner with a tree and an old low wall might provide a meeting place for teenagers. This corner with its tree might not fit into a rigid rectangular plan of a new proposed rebuilding, and would normally be eliminated without the least resistance. I propose reversing the priority, and putting social needs first, so that any new buildings need to respect and hence save the corner and tree, precisely because it supports a social pattern for this community. Consequently, the whole urban renewal plan has to bend to accommodate that tree and its corner. This node should influence the shape of new surrounding structures rather than the other way around. Unless a planner understands the necessity for such an action, any intervention and rebuilding will probably destroy the neighborhood's life. The process requires some preliminary work to reveal which physical structures, however trivial, tie in with social patterns in this community.
The computer scientist Thomas Erickson 4 has reached the same conclusion. He reviews the work of the urbanist Randolph Hester 5 in revitalizing the town of Manteo, North Carolina. Although not originally expressed in those terms, Hester and his team "mined" for socio-urban patterns important to that community before rebuilding anything. These had to be observed directly, since no-one talked about them. Once this "sacred structure" was identified, all construction was aimed at reinforcing rather than destroying it. What is important here is that the town's sacred structure was not composed of any building or urban construction that would be classed as "important" according to current architectural and urban criteria. Architecturally, the town's "sacred structure" was a network of negligible and insignificant places, buildings, and bits of built form -- this nevertheless provided a matrix for the life of the town. By following this plan, the town was extremely successful in regenerating itself 4,5 .
Emotional dimension of design.
What characterizes a socio-urban pattern is an activity combined with a place that gives emotional pleasure to human beings. Patterns are thus fundamentally based on emotions. Even patterns that have to do with an efficient process involving some function have to be selected from among all the alternative possibilities that make the actor feel less comfortable. Here we come face-to-face with a basic incompatibility between patterns and "functional" design, as expressed by both modernist and postmodernist architects and urbanists.
In his projects, Alexander experienced again and again the emotional dimension of design when trying to "mine" for patterns. For example, in preparing for the Eishin School outside Tokyo, he asked the teachers and staff to imagine the most beautiful environment to teach in. At first they thought it was a joke, but then became very emotional when they related visions of walking along a lake between classes, and concluded that encouraging such a dream was cruel because it was incompatible with contemporary fortress-like concrete schools in an urban setting. At another occasion, in designing a new community at Chikusadai, Nagoya, also in Japan, he asked families to actually draw the plan of their house. While they were doing so, many of them cried out of emotion, since they had expected to be given a standard fixed, or at best a modular design.
There is simply no way to decide on geometry -- whether the shape of a building, the shape of a path, or the relationship between two structures -- without using emotions. The human brain is the world's most sophisticated and powerful computer, and it alone is capable of making the astronomical number of parallel computations needed to decide the best position for placing an object. Forget the simplistic method of alignment to an artificial rectangular grid: it is the easy way out for architects who wish to avoid difficult decisions. In designing the Eishin campus, Alexander had people holding flags on poles move around the site so as to experience the best position for the planned buildings. In the words of the client, the headmaster Mr. Hosoi: "We could feel the actual buildings ... standing there".
Modernism eliminated emotions from design, thus depriving the individual user from any say in the process. How can a place be loved without emotions? The modernists made machines out of houses and tried to make machines out of people, but it didn't work. You cannot reverse human evolution (not in one generation, anyway). We react to built form emotionally: it is either nourishing, neutral, or hostile. In the latter case, it poses a threat to our sensibilities, so we naturally wish to destroy it before it destroys us.
How patterns are displaced by anti-patterns.
The word "meme" denotes a transferable unit of information that propagates itself by going from human mind to human mind 6 . In direct analogy with the spread of a virus, a meme is an idea, a description of something, or a visual image or snippet of music. Once released into the collective human mind pool, it is picked up by someone, and then passed on to the population at large by imitation. The success of a meme depends on its efficiency to replicate, and has nothing to do with its beneficial or harmful aspects to its hosts: human minds. An unsuccessful meme simply does not spread, and may die off, or linger on marginally. A successful meme infects the population in an exponential fashion, and could just as easily be replaced by a more virulent competing meme.
Human language -- both spoken words, and the pictorial language of images -- quite possibly drove the brain to expand fourfold in order to accommodate the increase in information input. The problem is that destructive memes also use the replicating process intrinsic in the human mind to propagate themselves throughout a population. We apparently have no defenses against virulent memes, and cannot distinguish between them and benign memes. A frightening picture emerges of human beings being manipulated by inanimate pieces of information that, like viruses, care only for increasing their number at the expense of their human hosts. Major human catastrophes can be attributed to, or are certainly helped by, a destructive meme that spreads to the population and drives it to do what it does unquestioningly.
A meme is more like a simple visual image rather than a reasoned description of how something is made. Successful memes are very easy to remember. A collection of simple memes could pretend to form a language, which could itself be perfectly consistent internally; nevertheless, they cannot coexist with a pattern language that respects complexity. The best example comes from government. Fascism and totalitarianism clean up the messiness of human society by displacing our most deeply-held patterns of human values. They have an undeniable appeal, however, otherwise they would not take over the collective mind of nations every few decades. Each time that happens, people again fool themselves into believing the demagogues who tell them that life's complexities can be drastically simplified.
Anti-patterns that destroy urban interfaces.
Most of the known architectural and urban anti-patterns were created by Charles-Edouard Jeanneret (Le Corbusier). Characteristic of all viruses, there is no completeness in the sense that we have an organism that metabolizes and interacts with others in an ecosystem. What we have is a nonliving informational code, or meme, whose sole purpose is to reproduce itself. For this reason, a mind-virus is given as a simple image, and not as a formula or solution to a problem. I have noted below some of the most destructive urban anti-patterns. These have infected the minds of people alive today, and work to displace patterns from the collective subconscious. This is the reason why it is extremely difficult to reintroduce Alexandrine patterns back into today's society.
- ABSOLUTE RECTANGULAR GRID
- SEGREGATION OF FUNCTIONS
- SHEER CONTINUOUS WALLS AT STREET LEVEL
- BUILDING SETBACKS
- EMPHASIS ON THE LARGE SCALE
- SEPARATED BUILDINGS
- VERTICAL STACKING
- GEOMETRY OF STRAIGHT LINES
- NON-INTERACTING UNITS
- UNNATURAL MATERIALS
- SUPPRESSION OF GEOMETRICAL PATTERNS
- ELIMINATION OF THE HUMAN SCALE
This list underlines my point. There is no scientific support for any of these twelve anti-patterns, despite the false claims made by Le Corbusier, and repeated later by his apologists. Scientific investigation of human interactions proves that these twelve anti-patterns prevent the normal activity in a city that drives people to inhabit urban regions in the first place. Anti-patterns become so deeply embedded into a culture, however, that any questioning of them threatens many people's essential being. Those persons are certainly unwilling to admit that they have allowed themselves to be infected with mind-viruses. Their mind is their self, and so they will defend their prejudices as forcefully as they will defend their life 6 .
Taken as a set of working rules, the above dozen anti-patterns have been used in a method of urban design to build cities throughout the world. They combine well together, and support each other. They have a consistency which is mistaken for adaptivity 3 . Because of this consistency, they give a result that is standard and easily identifiable: it is the modernist anti-city that treats human beings as emotionless machines. In a recent essay 7 , Michael Mehaffy and I argue that the application of modernist urban anti-patterns around the world, by erasing the traditional urban fabric, is in part responsible for the rage the non-industrialized world feels against the industrialized nations.
I have offered some ideas on how to handle interactive design with local communities, based on the theories and practical experience of Christopher Alexander. People starting an actual project will have to go to Alexander's writings for more details. Here, the description was meant as no more than a proof that participatory design is possible; suggestions on how to proceed with it; and why it is very likely to fail unless some deep-seated problems are addressed in the very beginning of the process. Design today is hampered by anti-patterns (or mind-viruses, otherwise called "memes") that impose unnatural forms and shapes on the environment. These will prevent the application of design patterns, which offer the only way to create a living environment.
Alexander, C., Ishikawa, S., Silverstein, M., Jacobson, M., Fiksdahl-King, I. and Angel, S. (1977) A Pattern Language (Oxford University Press, New York).
Alexander, Christopher (2001) The Nature of Order (Oxford University Press, New York).
Salingaros, Nikos A. (2000) "The Structure of Pattern Languages", Architectural Research Quarterly, volume 4, pages 149-161. Chapter 8 of Principles of Urban Structure, Techne Press, Amsterdam, Holland (2005).
4. Erickson, Thomas (2000) "Lingua Francae for Design: Sacred Places and Pattern Languages", in: Proceedings of the ACM Conference on Designing Interactive Systems 2000 (ACM Press, New York) pages 357-368. [Brooklyn, New York, August 17-19, 2000]
5. Hester, Randolph T. (1993) "Sacred Structures and Everyday Life: A Return to Manteo, North Carolina", in: Dwelling, Seeing, and Designing, Edited by: David Seamon (State University of New York Press, New York) pages 271-297.
6. Salingaros, Nikos A. and Mikiten, Terry M. (2002) "Darwinian Processes and Memes in Architecture: A Memetic Theory of Modernism", Journal of Memetics -- Evolutionary Models of Information Transmission, volume 6, approximately 15 pages <jom-emit.cfpm.org/2002/vol6>. Reprinted in: DATUTOP Journal of Architectural Theory, volume 23 (2002), pages 117-139.
7. Mehaffy, Michael W. and Salingaros, Nikos A. (2002). "The End of the Modern World", PLANetizen <www.planetizen.com>, January, approximately 4 pages. Reprinted by Open Democracy <www.opendemocracy.net> , February 2002.
Salingaros Home Page Contributions to Architecture