Despite Christopher Alexander’s notable application of complexity theory in design during the 60's and 70's, the two fields have mysteriously grown apart.
Thank you for putting this into writing! My internal voice was constantly screaming "YES!" while going through it. I think one reason why any pattern language has not yet materialized in the physical realm is of course a matter of designer arrogance, but also an issue of coordination. The medium of A Pattern Language (just like the medium of Gödel, Escher, Bach) is not fit for purpose. I had the same feeling when I first read A Pattern Language after graduating architecture (and never hearing a single utterance of Alexander.)
The concept of "eyes on the street" is most often perceived as one of Jane Jacobs' anecdotal observations of life in New York’s Greenwich Village. We relate Jacobs to social "bottom-up processes" but not to problems of "organized complexity" — despite her dealing with both topics in the chapter "The Kind of Problem a City Is" from her 1993 book, Death and Life of Great American Cities. Her work seems qualitative in nature, but she is coarse-graining her experience in patterns, which we struggle to quantitatively describe more than 60 years after her writing.
I believe there’s a case to be made that both Christopher Alexander, Jane Jacobs and other “qualitative” urban thinkers are echoing a deeper order, which we are not yet able to comprehend and might never be able to decode. I suspect however that there are fantastic opportunities in merging desirable urban patterns, with various levels of community control. This could transform communities from lethargic participants in tokenistic public consultations, to proactive agents of change – which access the deeper order of cities by excelling at community-driven local interventions. A distributed but well-coordinated exercise in urban acupuncture which might allow cities to access that higher order that Alexander hints at without the need to fully comprehend it. An organized and coordinated distribution of local governance might have a much higher effect on cities than current cartesian attempts of imposed order and legibility.
It was clear that the book enhanced my wholeness, it was also clear how one could individually put it into practice; what was unclear and I would dare say impossible at the time was a matter of coordination - a murmuration or an ant hill is a complex work of simple interactions between equal individuals. An ant, a bird, a fish, needs to respond in real time to coordinate with it peers. A decade after I first read A Pattern Language, I had the same feeling of "wholeness enhancement" when learning about DAOs.
I think the idea of an interdependent network of DAOs, all working and governing different scales/aspects but bound by rules / regs / conditions that are curated by higher hierarchy DAOs (in this context higher hierarchy meaning more members) is the future of urban governance, planning, architecture and design. A high level DAO for example might set the FAR for a certain neighborhood (development density) and then the neighborhood DAO might decide how that FAR is distributed. And than a lower level plot/building DAO might decide on the architecture and design (based upon constrains set by the neighborhood-level DAO). More about this specifically https://thinkthinkthink.substack.com/p/thefutureofgovernance
I am actively starting to work into this space and would love to set an hour and riff if that's interesting to you. Really enjoyed the article! Keep them coming.
But I do think that there are two qualifications that need to be made to this article. Firstly, Alexander was addressing complex issues, but not from a CAS perspective (and he couldn't really be expected to) but he follows Cybernetics and others in trying to identify and critically codify patterns that can be used. My view is that still has utility but a CAS based approach is different. Secondly I and others make a difference between computational complexity (Santa Fe are the exemplars of this) and anthro-complexity (the study of complexity in human systems) which used but is not constrained by computational model. I'd argue against the value of the DIKW model and the systems dynamics models you demonstrated for CAS in design not to mention concern about behavioural economics - but all that would be an interesting discussion. In general I think its a mistake to homogenise systems dynamics, pattern languages and CAS.
I was just this morning writing to a friend that natural structures structure themselves. We recognize natural structuring structures (clouds, mountains, trees, ...) from culturally structured structures (stools, houses, cars, ....) . Nature doesn't do design, she designs. Only human beings do design.
In nature form structures (follows) structure and structure forms form. Following what one can call "Law of the Least Work" (least action, least resistance, ...). Everything is in the only way it can ever be.
Nature "designs" herself. Any sign (cloud, mountain, tree, ... ) signifies it/her/him/them-selves. That's why they're all unique and are having the same "structure". The latter may come as a surprise, as I'm using the word structure in a slightly different way.
The structure - the ways it has structured itself - of a natural thing or creature accounts for its usefulness. Like the ways on a map show (you) the structure of the city streets. (Reversing Korzybski's "the map is not the territory").
The Law of the Least Work "makes" a projectile follow parabola, and we use it to explain why a projectile makes a parabola. It is as-if the projectile "knows" what path to follow, naturally. Nature "knows", without explicit knowledge, how to structure herself.
These - in our idea of design - complex results are actually also the simplest. They cannot be different, because of the Law of the Least Work (I know, it look like a fallacy) . Any other (more complex) shape would mean "more" work. The interesting thing about the "laws" of nature - all these laws are human laws -, is that they maintain themselves. There's no "least work" police, advocate, judge or even punishment. I could call it "Occam's Reaping Machine" or better, "Occam's Harvester" :-).
Our designs follow rules prescribed by us. In our concept of design design should be "fit for function". Things have an intended purpose, one assumes. We tend to think, for instance, that the beaks of birds have a function. To eat seeds, insects, build a nest, ... . Or that the function of one's hearth is to pump blood. But birds have beaks; you've got a hearth. One wouldn't live without them. They were designed by themselves, without "purpose" or function, because they "work".
Warning! Complexity ahead! Animals make sound to recognise each other and to signal situations to each other. Human being expanded this into language. As we've learned to use language to instruct (long story, no time for it now) to make constructions - called artefacts, because they're arti-ficial and not nature-ficial - we tend to think in terms of principles, rules, laws with a purpose, with de-sign.
One can only have a theory on design, when having a language. But the structure of our language (artificial: order, rules, linear, static, ..) doesn't accommodate for the structure of nature (natureficial: out of control, chaotic, dynamic, ....)
(Because we also still use these sounds to recognise each other (in Dutch the word for "belonging" is "behoren" or "being heard"), language also induces a "double bind": we tell each other rules and laws and what to do to belong.
Things have become "complex", because living lives is not simple any-more. And thanks to the printing press and the computer, we're able to complicate complexity. I like to say, "we used to have language and now language has us". )
Simply put: complexity theory in design is an oxymoron. Complexity is natural, practical, universal and "undesigned". Our designs and our artefacts, paradoxically, are too simple, that's what makes "complexity". They don't work for themselves. (I always say that AI in cars will only work when it becomes afraid for scratches in the paintwork).
Alexander's writing are heuristics, not actual rules or principles. He's been predated 2000 years by Vitruvius. Simple architectural "rules of thumb".
I want to stress, that I don't see these issues as problems. We cannot "solve" these. They're paradoxes and will resolve themselves. Paradoxes are "at work" and as one tends to exclude paradoxes from thinking, it will take some time to adjust our thinking. Thanks for reading this.
Thank you for putting this into writing! My internal voice was constantly screaming "YES!" while going through it. I think one reason why any pattern language has not yet materialized in the physical realm is of course a matter of designer arrogance, but also an issue of coordination. The medium of A Pattern Language (just like the medium of Gödel, Escher, Bach) is not fit for purpose. I had the same feeling when I first read A Pattern Language after graduating architecture (and never hearing a single utterance of Alexander.)
The concept of "eyes on the street" is most often perceived as one of Jane Jacobs' anecdotal observations of life in New York’s Greenwich Village. We relate Jacobs to social "bottom-up processes" but not to problems of "organized complexity" — despite her dealing with both topics in the chapter "The Kind of Problem a City Is" from her 1993 book, Death and Life of Great American Cities. Her work seems qualitative in nature, but she is coarse-graining her experience in patterns, which we struggle to quantitatively describe more than 60 years after her writing.
I believe there’s a case to be made that both Christopher Alexander, Jane Jacobs and other “qualitative” urban thinkers are echoing a deeper order, which we are not yet able to comprehend and might never be able to decode. I suspect however that there are fantastic opportunities in merging desirable urban patterns, with various levels of community control. This could transform communities from lethargic participants in tokenistic public consultations, to proactive agents of change – which access the deeper order of cities by excelling at community-driven local interventions. A distributed but well-coordinated exercise in urban acupuncture which might allow cities to access that higher order that Alexander hints at without the need to fully comprehend it. An organized and coordinated distribution of local governance might have a much higher effect on cities than current cartesian attempts of imposed order and legibility.
It was clear that the book enhanced my wholeness, it was also clear how one could individually put it into practice; what was unclear and I would dare say impossible at the time was a matter of coordination - a murmuration or an ant hill is a complex work of simple interactions between equal individuals. An ant, a bird, a fish, needs to respond in real time to coordinate with it peers. A decade after I first read A Pattern Language, I had the same feeling of "wholeness enhancement" when learning about DAOs.
I think the idea of an interdependent network of DAOs, all working and governing different scales/aspects but bound by rules / regs / conditions that are curated by higher hierarchy DAOs (in this context higher hierarchy meaning more members) is the future of urban governance, planning, architecture and design. A high level DAO for example might set the FAR for a certain neighborhood (development density) and then the neighborhood DAO might decide how that FAR is distributed. And than a lower level plot/building DAO might decide on the architecture and design (based upon constrains set by the neighborhood-level DAO). More about this specifically https://thinkthinkthink.substack.com/p/thefutureofgovernance
I am actively starting to work into this space and would love to set an hour and riff if that's interesting to you. Really enjoyed the article! Keep them coming.
I'd agree that design needs complex adaptive systems theory and we've been working on that for 4/5 years and starting to shift new methods out - some linked to the EU Field Guide on complexity. https://publications.jrc.ec.europa.eu/repository/handle/JRC123629
But I do think that there are two qualifications that need to be made to this article. Firstly, Alexander was addressing complex issues, but not from a CAS perspective (and he couldn't really be expected to) but he follows Cybernetics and others in trying to identify and critically codify patterns that can be used. My view is that still has utility but a CAS based approach is different. Secondly I and others make a difference between computational complexity (Santa Fe are the exemplars of this) and anthro-complexity (the study of complexity in human systems) which used but is not constrained by computational model. I'd argue against the value of the DIKW model and the systems dynamics models you demonstrated for CAS in design not to mention concern about behavioural economics - but all that would be an interesting discussion. In general I think its a mistake to homogenise systems dynamics, pattern languages and CAS.
I was just this morning writing to a friend that natural structures structure themselves. We recognize natural structuring structures (clouds, mountains, trees, ...) from culturally structured structures (stools, houses, cars, ....) . Nature doesn't do design, she designs. Only human beings do design.
In nature form structures (follows) structure and structure forms form. Following what one can call "Law of the Least Work" (least action, least resistance, ...). Everything is in the only way it can ever be.
Nature "designs" herself. Any sign (cloud, mountain, tree, ... ) signifies it/her/him/them-selves. That's why they're all unique and are having the same "structure". The latter may come as a surprise, as I'm using the word structure in a slightly different way.
The structure - the ways it has structured itself - of a natural thing or creature accounts for its usefulness. Like the ways on a map show (you) the structure of the city streets. (Reversing Korzybski's "the map is not the territory").
The Law of the Least Work "makes" a projectile follow parabola, and we use it to explain why a projectile makes a parabola. It is as-if the projectile "knows" what path to follow, naturally. Nature "knows", without explicit knowledge, how to structure herself.
These - in our idea of design - complex results are actually also the simplest. They cannot be different, because of the Law of the Least Work (I know, it look like a fallacy) . Any other (more complex) shape would mean "more" work. The interesting thing about the "laws" of nature - all these laws are human laws -, is that they maintain themselves. There's no "least work" police, advocate, judge or even punishment. I could call it "Occam's Reaping Machine" or better, "Occam's Harvester" :-).
Our designs follow rules prescribed by us. In our concept of design design should be "fit for function". Things have an intended purpose, one assumes. We tend to think, for instance, that the beaks of birds have a function. To eat seeds, insects, build a nest, ... . Or that the function of one's hearth is to pump blood. But birds have beaks; you've got a hearth. One wouldn't live without them. They were designed by themselves, without "purpose" or function, because they "work".
Warning! Complexity ahead! Animals make sound to recognise each other and to signal situations to each other. Human being expanded this into language. As we've learned to use language to instruct (long story, no time for it now) to make constructions - called artefacts, because they're arti-ficial and not nature-ficial - we tend to think in terms of principles, rules, laws with a purpose, with de-sign.
One can only have a theory on design, when having a language. But the structure of our language (artificial: order, rules, linear, static, ..) doesn't accommodate for the structure of nature (natureficial: out of control, chaotic, dynamic, ....)
(Because we also still use these sounds to recognise each other (in Dutch the word for "belonging" is "behoren" or "being heard"), language also induces a "double bind": we tell each other rules and laws and what to do to belong.
Things have become "complex", because living lives is not simple any-more. And thanks to the printing press and the computer, we're able to complicate complexity. I like to say, "we used to have language and now language has us". )
Simply put: complexity theory in design is an oxymoron. Complexity is natural, practical, universal and "undesigned". Our designs and our artefacts, paradoxically, are too simple, that's what makes "complexity". They don't work for themselves. (I always say that AI in cars will only work when it becomes afraid for scratches in the paintwork).
Alexander's writing are heuristics, not actual rules or principles. He's been predated 2000 years by Vitruvius. Simple architectural "rules of thumb".
I want to stress, that I don't see these issues as problems. We cannot "solve" these. They're paradoxes and will resolve themselves. Paradoxes are "at work" and as one tends to exclude paradoxes from thinking, it will take some time to adjust our thinking. Thanks for reading this.