Warwick Rowell, Tuesday 18 March 2014
The first of a series which outlines the key ideas from Christopher Alexander’s many books.
Since 1964, Christopher Alexander, with many colleagues, has written a series of books about Patterns and Pattern Language. They range from a detailed computer based analysis of patterns in an Indian Village (1), to one of almost Buddhist philosophy that underlies our reactions to built space (2). In between there are some even more useful books that can both inspire and direct people on how to build their own homes and neighbourhoods and service centres, using patterns. (3-8). He has recently completed a four volume set, titled “The Nature of Order.” This article sets out the basic concept of a pattern, and their collection into a language.
A Pattern describes the solution to a problem in a particular context. There are patterns about what makes an entrance transition work, patterns about how to create rooms that are welcoming, patterns about how to orient a house, patterns about how to make a courtyard live, and so on. There are 253 in the major book, and many more under development, for example (hyperlink to Patterns for South Walls).
Many of the Patterns – the generic solutions to general problems – have a human, almost psychological tone to them. This is one of the major values of patterns. They are at once imperative – “If you want an entrance transition that works it will have all or most of these features: change of direction, change of view, change of level, change of texture, change of sound” – and subtle – “An entrance transition works when it provides people with the physical space and a change of visual cues to help them adjust to moving from one space to another, where different expectations and behaviours are needed.”
All pretty straightforward, when you think about it, I hear you say. Sure, once it is pointed out to you, so you do think about it! And here lies another value of patterns – many of them are “obvious”, but having them clearly stated helps us think about them, and so determine whether they really are appropriate to the situation we are trying to design. Another value is that they have a scale relative to each other; grouping patterns means we can make sure that we don’t “lose” one solution in addressing another, larger or smaller, problem.
Alexander says some of these patterns are quite tentative. He gives two asterisks to those he is absolutely sure about, one asterisk to those he feels quite confident about, and poses some as tentative, and needing further work. He also describes over and over again that there will be millions of different versions of each pattern depending on the taste and situation of the people using patterns to build or modify the structures that surround them. He goes as far as saying that each partilcular selection from the 253 patterns will be a separate language; one example he gives is the patterns for a front courtyard. He invites people to build their own patterns, and gives examples of patterns not yet developed – one is a pattern for a spa. Many criticisms of his work (mainly by architects) that Patterns are too simplistic and prescriptive are not sustainable.
Let’s use a new pattern we have developed as an example of some of the principles and values outlined above.
The problem is strong winds, the context is protecting houses and gardens and field crops on mainly level ground. The generic solution proposed is screens at spacings of 10 times their (eventual) height.
This statement can be expressed in diagrammatic form. It can also be expressed as an imperative; “On level ground, ensure there are screens spaced at 10 times their height”.
Note that we can verify this pattern empirically at a number of levels:
Does the context exist for us? Do we have structures and things we want to protect from the wind? If not, we don’t need this pattern now.
Does the problem exist for us? Are our winds strong enough to require protection?
And finally, Does the solution proposed actually work? Are there other patterns that could make it work better? The brief answer to this is yes – refer to (hyperlink here) for seven more patterns about windbreaks.
Note carefully that this pattern is saying nothing about whether we use trees, shade cloth, bunds or buildings. It is not saying anything about what species, what shape of tree, what other uses the trees or shrubs might fulfil, and so on. It is talking only about the nature of the solution required to solve the problem in the context given, and so providing a useful framework in which we can exercise our creativity, adapt our resources and, if necessary, make our own compromises.
Alexander talks about how some people use language loosely, and so their results are not heartfelt. He also talks about how a careful and dense use of patterns can bring a vibrancy and richness to our buildings, almost like poetry.
Alexander tested patterns in built spaces. He asked people how they felt in spaces with a particular pattern absent. They reported feeling uncomfortable, disjointed, ill at ease. He then included the pattern in the same space, and asked people how they felt then. They reported feeling comfortable, and congruent. To the 95% confidence level. This aspect of patterns – that people feel comfortable and at ease when the patterns are present – will help people who want to design their own infrastructure, homes, or just revamp a room. A Pattern Language can give them, perhaps for the first time, a means of talking about both their sense of uneasiness about a space, and a means of identifying the nature of the solution that will make that space more alive, and rewarding.
Alexander’s most important achievement was in defining the process of building a language that ordinary people could use to design and build their own living space.
He defined each pattern as firmly as possible using both practical experience and hard empirical analysis. He and his colleagues asked “How do you feel in this space?” (with a particular pattern present). Response: “good”. “How do you feel in this space?” (with the pattern absent). Response: “bad”. Over and over and over again, modifying the pattern (its words, its sketches, its construction) until the answers were as unequivocal as they could obtain. It is hard to ignore this sort of evidence. But as Dickson and Mann and Kelly and Hampden-Turner amongst others have pointed out, much of our pathology as individuals, as organisations and as societies, comes from not seeing the evidence that is clearly before our eyes. Until too late.
A pattern is a paradigmatic statement: “In this context, to solve this problem, any solution must look like this.” So it must, by definition, be a panacea.
Each pattern is NOT a rigid rule but a statement of what seems to be true for an overwhelming majority of people so far. Alexander grades his patterns based on the amount of empirical support they find in the professional and other literature, and the empirical experience of the patterns. He says that like any language it will grow, based on how much it is used, tested and criticised.
Part of his brilliance is taking us so far down the track that we can build on his work as we refine and test our language, from the wonderful grounding his work has given us. He has covered so much of the fundamental territory we can easily use it to start our own design work.
The next article will focus on one of the books below, and summarise its key concepts.
- Alexander, Christopher, Notes on the Synthesis of Form. Harvard University Press, Cambridge Mass., 1964.
- Alexander, Christopher, et al, The Timeless Way of Building. OxfordUniversity Press, New York, 1979.
- Alexander, Christopher, et al, A Pattern Language – Towns, Buildings, Construction. Oxford University Press, New York, 1977.
- Alexander, Christopher et al, The Oregon Experiment. Oxford University Press, New York, 1975.
- Alexander, Christopher, et al, The Production of Houses. OxfordUniversity Press, New York, 1985.
- Alexander, Christopher, et al, A New Theory of Urban Design. Oxford University Press, New York, 1987.
- Alexander, Christopher, et al, The Linz Cafe/Das Linz Cafe. Oxford University Press, New York, 1981.
- Alexander, Christopher, et al, A Pattern Language which generates Multi-Service Centers. Center for Environmental Structure, Berkeley, Calif., 1968
Best source: PatternLanguage.com