Former student of Nikos Salingaros

Attendee, "New Science, New Architecture" Conference and Masterclass at The Prince's Foundation, London, September 2004

I am currently working for the French State, in a "Laboratory" which depends on the Public Works Ministry, near Paris. We are participating for three years on a European research project named "New Road Construction Concept" (NR2C) that is expected to finish at the end of 2007.

 

I have the responsibility to develop one particular task of this research project, and I have started, 3 years now, to re-use the pattern language of Christopher Alexander, to develop new patterns, essentially oriented on street design. Our work here has been carried out on two points :

1. creating a new "format" for the description of a pattern ;

2. making new patterns of streets, more detailed and numerous than those that are included in APL.

 

For point 1., the main transformation we have brought to the initial format of a pattern is to reinforce and to explicit the process-oriented point of view : we passed from the 3-fold "Alexandrian" scheme [Context - Problem - Solution] to a 5-fold one : [Context - Problem - Parti - Intentions - Configurations]. "Intentions" and "Parti" (French term meaning "formal concept" or "formation", which may be assimilated with the "formal cause" of Aristotle) were already implicitly present in Alexander's patterns, whereas the "Configurations" are analogue to Alexander's "Solution". So there is not a change in content, more in the way information is organized.

 

We describe the "Parti" as a process of transformation of an "initial form" which evolves into a "specific form", through a little number of steps. The initial and final form of the pattern, as well as the result of the intermediate steps of morphogenesis, are described in a way that is similar to Alexander's "centers" : by coherent and central pieces of space that interact with each other (we symbolize the centers by circles).  The actions of transformation are described by "verbs of action that have a spatial meaning" as : to turn, to cross, to cover, to open, to close, to overlap, etc.

 

So we have proceeded on passing from a problem-oriented formulation to a more form/process-oriented formulation of a pattern. We are also more explicit than Alexander about the scheme of organization of one pattern. But the philosophy remains the same : a pattern is meant to communicate the essence of an architectural idea. I attach some examples (in French for the moment...) so that you can have a look at it.

 

Rue_disponible_fiche_resume-3.pdf

Rue_mixte_fiche_resume-3.pdf

Rue_multi-fonctionnelle_fiche_resume-3.pdf

Rue_niveau_fiche_resume-4.pdf

Rue_rideaux_fiche_resume-3.pdf

Rues_principales.pdf

 

We are also fully using now the "wiki" tool, to organize the documentation and links between patterns, which makes the navigation much lighter.

 

For point 2, the fact of extending the stock of patterns on one special subject will be very crucial for the development of this tool : with a large number of patterns, each will appear more as a possibility than as a norm or something imposed by doctrine, which has been the main criticism that has been formulated against pattern languages. In focusing our work on the subject of streets, we are also, we believe, going into the direction of the analytical approaches, such as the one of Stephen Marshall for example as questions of classification appear.