(2022-04-10) Arango Taylor On Christopher Alexander

Jorge Arango interviews Dorian Taylor on Christopher Alexander. And I wrote something a while back when I was sort of thinking about like these 15 properties — there are 15 of them — from the perspective of information theory, like they kind of like cluster into three categories. And of course, I’m going to categorize the categories here. One of them is like, it generates information. The other one, it compresses. And then the other one kind of like de-noises, I think? (Nature of Order)

they don’t really deliver the goods (pattern language)

His testing criterion was this thing he called “the mirror of the self” test. The idea was you would hold up two objects and you would look at them and you’d be like, okay… you don’t ask, which one of these is better?

in many ways, he was quite radical in that I don’t think that his ultimate project was about designing those buildings per se. It was more about, like we said earlier, designing the thing that designs the thing

Alexander asked it himself in this keynote he gave; you can find it on YouTube. It’s called Patterns in Architecture.

think I’m calling myself a consulting designer. I think that was like what I landed on. Sort of borrowing from Sherlock Holmes — the recent one with Benedict Cumberbatch or whatever. I’ve been working with the web since I was a teenager. That was in the nineties. Front end, back end, side end… all the ends at once! And now it’s kind of moving into more governance-y stuff and more “the strategy-ry” — you know how it goes.

the design patterns in software are helpful, I guess

he’s like, “number one, I don’t know if you guys actually care about like making the environment good with this. I don’t know if you guys do… I think it’s a technical format that communicates useful ideas, but I don’t know if you guys are really trying to make things good.”

organizations are coming to the recognition of the need for this type of higher-level work. Somehow that design is not just about cranking out screens; there’s something that needs to happen at a more fundamental layer.

I read the documentation that nobody else has time for, and I can internalize that and I can re-represent that and [come] up with formal models of processes and structures within organizations to help people be able to converse about things

This notion of conceptual integrity might be a good segue for talking about the subject that brings us together today, which is the work of Christopher Alexander.

he abandoned the patterns themselves.

he really thought that there was a sort of objective criterion to that.

he was using phrases like, “creating life,” “make God appear in a field,” you know? He was saying sort of, quasi-mystical things, but he was almost serious about it. I mean, he was definitely serious about himself

It’s kind of like shared mental models, right? I mean, we’ve heard of the term “conceptual integrity” — that was a Fred Brooks thing, The Mythical Man-Month. You know, it is one of those books that everybody has, but nobody reads. Conceptual integrity is the sort of state of affairs of everybody working on a project kind of understands what it is and what’s going on, and they’ve got a shared unified mental model of just what the damn thing is. And they have like shared language to talk about it. It’s about an anatomy that people are referring to.

how did Alexander actually implement his process?

the first thing he did was he went and he became a general contractor. He started a company to be the general contractor because he couldn’t work with general contractors because they wouldn’t let him do his thing

There’s another book that like totally randomly, totally by chance, I bought at the same time that I bought A Pattern Language for the first time back in 2001. It was called Cognition in the Wild, by Ed Hutchins. And that is an ethnographic study. If you’ve never read it, oh my God! It’ll blow your mind. It is an ethnographic study of the navigation team of an aircraft carrier.

And so, anyway, one of the things that Hutchin says in that is he says, “a navigational chart is an analog computer.” And I’m like, “just like that!” The building sites of Alexander were analog computers. They were a computational environment to the extent that you use Claude Shannon’s definition: you say computation is revealing latent information using known… from known information,

what comes to my mind is Gall’s Law, this idea that a complex system that works evolves from a simple system that works, and you can’t really design a complex system from the get-go. And something like a building is a fairly complex system.

And it sounds like what Alexander was trying to do was to create the conditions for that complexity to emerge from conditions on the ground, as opposed to some kind of abstract imagined representation of what the ideal conditions might be

How do we make that, “this is a real thing to me” experience come to a client? I still would love to know.

It’s like, I want to talk about an information system-shaped hole in your organization, and we’re going to fix that hole.

Jorge: What I’m hearing here is, aspire for conceptual integrity. But not just internal conceptual integrity, but making a thing that is conceptually whole with the context and systems that it’s a part of and that it informs.

Yeah. And I mean, the extent to which that is a thing that like… you might want to try to read that as like, “oh, what you’re talking about is systems integration.” Because so much of what we do in software is all mediated one way or another by platforms and vendors

And they’re all proprietary. None of them talk to each other in any meaningful way. And each one of them is all about getting users locked in and then kind of like giving you whatever’s on their menu.

And so, like all of this software is verbs, but it all controls the data, which is nouns.

The Oregon Experiment, one of Alexander’s mid-career books, talks about this kind of replacement culture where you go and you do a ten or a hundred million dollar contract with one of these IT vendors. And they’re in there like a tick. And then, eventually, you just get more and more frustrated by… you know, you can’t do the something or other

Alexander, you know… the analog to that would be like going and building a giant building, and then it’s like, why? Because the big, big deal it makes the administrators look good

and Alexander was talking about like, “Hey! What about a couple hundred bucks to fix that fence? What about being able to put a rose bush down?” What about that?

And it’s like the granularity, the addressability of the money is so chunky that like, you can’t do it for less than a million dollars, sorry! And if it’s going to be a million dollars, it might as well be ten million

And this is why I’ve been paying more attention to the actual money aspect of it. And in the case of software, it’s kind of like… You know, I had a client long-term, a while back. Whenever they had some coin, they’d dial me up and be like, “oh, can we fix this little thing here?” And it’s like, a lot of their infrastructure was stuff that I wrote, but I wrote it piece by piece. We could imagine regimes where an organization might set aside fifty thousand

But there would be these sort of repair jobs on their internal infrastructure.

This is an adaptation thing. You know, computers are almost as old as television now, and we’re still treating them like, “ooh, mysterious technology thing.” And it’s like, no, no, no! Okay, we’re manipulating information

anytime you introduce computers it becomes an IT issue. And it’s really an issue of communication, memory, comprehension. You know, that was really my big thing, comprehension.

And for me, I was like, “well, how do you get a complex mental model into somebody’s brain?” Because you’ve got to feed it through and then you have to kind of assemble it inside their head. And I think, for me, like, one of the ways to do that was hypermedia. The ability to oscillate between storytelling and just the facts

It’s like some things are just complex and you’re just going to have to deal with the fact that they’re complex.

You’ve got the Minsky and McCarthy over on one thing that says like, “well, we want the computer to be this intelligent thing that just serves you. It’s kind of like a slave.” And you got the J.C.R. Licklider and Douglas Englebart camp who are saying, “well, we want the computer to be a dumb tool to make smart people smarter.” I’m in that camp, you know? I’d rather have a tool than a slave myself. I don’t need Siri, I just need a sharp knife.

People who are listening, if they can pick up one thing to read of Alexander’s or a way to get into his work so that they can think about integrating this into their own work — do you know of any resource?

Dorian: If they’re software people, I think that the best thing that they can probably do is watch that keynote from 1996, the one that’s called Patterns in Architecture.

I would start there because that is like an hour of your time whereas any book is going to be… I mean, you will blast through A Pattern Language if you get it, you know? But like Alexander himself said like 25 years ago, A Pattern Language is abandonware from his perspective

But I mean, you might go for Richard Gabriel’s Patterns of Software.