Dissertation proposal! Artificial Intelligence, Architectural Intelligence: The Computer in Architecture, 1960–80

UPDATE: My dissertation proposal.

I’ve completed my dissertation proposal! My dissertation is tentatively titled “Artificial Intelligence, Architectural Intelligence: The Computer in Architecture, 1960–80.” At noon, I defend it. Wish me luck! Here is the abstract:

With the advent of the information age, architects in the 1960s and 70s found themselves contending with more complex design problems than they had in the past. In response to these changes, the architectural profession began to turn to computers and computer- related sciences including cybernetics and artificial intelligence (AI), and to ways to solve and represent problems using the computer. The computational shift promoted design process over formal object, moved the architect out of a central role in the design process, and generated architectural solutions beyond the capabilities of machine or architect alone. This dissertation will examine three architects, Christopher Alexander (b. 1936), Nicholas Negroponte (b. 1943) and Cedric Price (1934–2003) and the influence of, and their collaborations with, key figures in cybernetics and artificial intelligence. The period from 1960 to 1980 is significant because it marks the introduction of computing paradigms to architecture and the beginning of the mainstream of computers in architectural practice. Throughout, this dissertation will develop the notion of generative systems in architecture; that is, systems that incorporate models of intelligence, interact with and respond to both designer and end user, and adapt and evolve over time.

Writing a dissertation proposal has more to do with writing a brief, a pitch, or a grant application, and less to do with writing the actual dissertation. That was the hard part: I kept sitting down and attempting to write the whole thing. It was thanks to the help of my friend Janet Vertesi one afternoon in Venice, with two plates of truffle french fries and a glass of rosé, that I finally got my head around the fact that I needed to write the argument for the project, not the project itself.

Defense the proposal marks the final hoop before finally starting the dissertation and my work for the next two years. I’m delighted to begin.

weeknote 06

Enough with the snow. I’m in Los Angeles, or more precisely, Venice (and I missed the third snowstorm in 10 days in New Jersey). I will be shifting my time to be here more than not in the next several months, an audition for whether I might fully move here later this year. I’ve been running on the beach boardwalk in the mornings, something I ordinarily do later in the day. In the afternoons, I write. I’m pondering adding yoga to the mix since I have enough energy for it during the day. The sunlight here is beautiful and my freckles are out for springtime.

My dissertation is focusing ever more on generative systems. I’m working on the dissertation proposal and this week, I’m writing about what constitutes a generative system. Rather than turning out formal prose, I’m just writing between 1000-2000 words, written quickly. It feels lighter this way and it captures my insights better.
I’ve been doing a close reading of Nicholas Negroponte’s The Architecture Machine (1970), J.C.R. Licklider’s “Man-Machine Symbiosis” (1960) and Warren McCulloch’s (ready? this is long) “Toward some circuitry of ethical robots or an observational science of the genesis of social evaluation in the mind-like behavior of artifacts” (1956). In this case, I’m using some of the methods I followed when I worked on a paper about Adolf Behne’s work in the 1920s and the notion of the apparatus… I suppose that this isn’t too different, since Negroponte is all about generative apparatuses.
So what is a generative system? Here’s a broad list of attributes I’ve gathered so far.
  • Intelligence
  • Contextual (context-sensitive, context-appropriate)
  • Adaptive and adaptable
  • Bridges dissimilarities
  • Evolutionary
  • Symbiotic
  • Unfolds over time
  • Has disposition and agency
  • Appetitive  (it absorbs from the environment around it — a word that comes from the McCulloch piece)
  • Capable of learning
  • Social
  • Communicates in (somewhat) natural language
  • Self-organizing
I need to group these and boil them down: these come from the work of a few figures in architecture and information theory, cybernetics and AI. The funny thing is, as much as I will apply these attributes to architecture, they apply to a certain attitude of systems in general. (I suspect that we should build systems today to strive for more of these attributes.) I’ll take the initial framework and bounce it against the work of the people in my case studies: Christopher Alexander, Cedric Price and Nicholas Negroponte. It’s nice that I’ll get to take on the most exciting aspects of my master’s thesis research on Cedric Price.
Next week, I’m aiming to start pouring content into the actual proposal with the plan to finish it at the beginning of March. There are other things that may compete with that, in reality, but I’m trying to keep enough structure and momentum going so that it carries me forward.

weeknote 05

It’s the second snowstorm in a week and right now, it’s the strange moment where I can feel the pressure change and sense the rest of the front that’s about to hit. They’ve closed Princeton today–in fact, they’ve closed most of the East Coast, from the sound of it, which also means that Enrique’s dissertation proposal won’t happen till next week and the Richard Sennet-Eyal Weizman-Teddy Cruz lecture will be rescheduled for tomorrow. During Saturday’s snowstorm, I baked bread, made coq au vin for hours in the slow cooker, started sewing a dress, and dyed my hair. Today will be geared more toward work and toward my own dissertation proposal.

First, I passed my general exam! It was a two-hour, closed-door critique of my work by Beatriz Colomina (head of the PhD program), Christine Boyer (my advisor), Ed EigenSpyros Papapetros and Brigid Doherty. It was very positive. I found it fascinating to see how the commitee drew links and connections through the body of work I had presented. It was apparent to them that I had a method and a clear set of interests (though the method is not as clear to me as it is to them–I inhabit it). They actually said that they enjoyed the papers — that’s the word they used. The committee thought my work needed to be theorized better and that media theory seemed to be useful (I have a meeting with the fabulous Tom Levin tomorrow to discuss). They thought my research paper — the one that will undergirds my dissertation — was the weakest, but I knew that: I had gone through 11 drafts of it and it was out of control. But that indicates to me that I have found the right topic for a doctoral dissertation. Overall, I got useful feedback that I can apply to both the larger scale of my work and the smaller scale.
The whole process of the general exam, from selecting papers to revising and expanding, to editing and presenting, really boosted my confidence — something I did not expect. It made me realize that I have a genuine body of work rich with research questions. It gave me a chance to see the common threads passing through the work: the things that tease me intellectually and won’t let me go. I now have all of these ideas of things I look forward to working on in a career, not just a dissertation.
I’m also ready to get back to writing, not just for my own work but out in the world. Maybe it’s time to start writing a column somewhere? We’ll see.
I’m grappling with two different directions on my dissertation proposal. On one hand, I can write about the introduction of the computer to architecture. There are three themes: methodology, representation, and generation. But really, it’s the generative systems that I think are really interesting. I have an idea about how architectural computing becomes computing architecture, how it on one hand ends up as ubiquitous computing, and on the other, as spatial metaphors for computing. There are reasons to do both: one is a straightforward dissertation; the other really ties together my big questions but might be harder to convince an architectural committee. I’m helped by the fact that much of these things happened within MIT’s architecture school, where the Architecture Machine Group existed and the Media Lab still resides (even if they don’t cross over at all with the history/theory/criticism part of the school). Talking with my advisor, Christine Boyer, will help: she listens well, she was at MIT in the period that I’m researching, and she’s done a good job of steering me the right way.
I keep coming back to haptic and physical engagement with space. Nicholas Negroponte & Richard Bolt’s 1977 Spatial Data Management System is really interesting in that it gave rise to the desktop metaphor, but what really intrigues me is the importance of “motor-memory reinforcement” — the notion that by physically putting something somewhere, or by going somewhere, it reinforces memory. They give the example of Simonides, the Greek poet famous for his ability to memorize long oratory. Negroponte explained in 1986, “His secret was to tie each successive part of a to-be-remembered poem or speech to a specific locale within the mental floor plan of either an actual or imagined temple. For each successive subsection of the talk to be given, the orator would mentally walk from place to place within the temple, rehearsing the appropriate material before some specific piece of statuary.” (Stewart Brand, The Media Lab, 1987, 138). This points out what eBook readers get wrong: the physical, haptic engagement of reading. It also points out a key question of what “future of reading” projects miss out on: the physics of authorship.
There’s so much possibility in that idea! It’s not about creating a metaphor, or a bookshelf on a device — that’s done and usually, done poorly (the iPad is no exception). It’s also not a gestural mode of interaction with a device — but what would happen if we created things that help us learn by our own movements? I’m going to work more on that in the coming days and share my thoughts about it.
I’m going to light some candles, invoke some hygge, and watch the snow fall… and write. I’ll let you know where this puts me next week.

weeknote 02 and 03

Greetings from Mexico City! I’m on vacation, a tagalong for a commercial that Motion Theory is shooting. I keep having dreams about editing and rewriting, but I have nothing to do until February 3, when I have my oral defense. I think this is what they call vacation.
First: the Microsoft Social Computing Symposium was outstanding. This was my fourth and it was by far my favorite–maybe because it was in New York and not Redmond, WA, maybe because it focused on cities, maybe because I got to see some of my favorite people. I gave a 20 minute talk on the introduction of computing to urbanism and urban planning (see below)… the whole thing was great. I’d have more to say about it, but I came back to Princeton and put myself on total lockdown in Princeton for 8 straight days, pulling together the last of the revisions for my papers. I left my apartment maybe once a day, if that. The final day, I wrote and edited for 28 of 30 hours (1 1/2 of those I spent asleep, sort of). Writing nonstop like that is nearly impossible — it requires so much concentration, especially for academic writing. But somehow I got it done. It’s not perfect, but it will do, I hope. I’m mightily thankful for the help that I received in feedback and editing and layout: if not for that, it never would have come together.
What’s really hard about the way our generals work is needing to keep six separate topics in mind, moving from one to the next. No sooner was I finished with something on France in the 1880s than I had to move onto Pakistan in 2007, and back again to the 1960s in the US.
Here’s what I handed in.
  • Artificial intelligence and architecture: the introduction of the computer to the field of architecture (with Christopher Alexander and his interest in AI and cybernetics as a case study), 1960-75. I wrote a tidy version of this paper in May 2009. Then, I blew the whole thing up into a much bigger framework about how the computer affected architectural practice. I’ve written scores of pages that didn’t get included: the draft at one point was 50 pages long (what I handed in was 36 pages). After 11 different drafts, I whittled it down to a couple of key ideas. I convinced myself that it was okay–I would be writing a dissertation on the topic and I could reuse what I wrote and then deleted.
  • Paris & communication networks: The Hôtel des Postes in Paris, 1884 and the Parisian pneumatic tube network, 1866-1900 (something many people know I’m interested in, thanks to last year’s eTech Ignite talk). These two papers are parts of the same topic: urban-level communication in France in the late 19th century. When I first wrote the paper about Julien Guadet’s central post office in Paris in 2008, my central argument was that it functioned like a big computer atop a tangible network. That argument proved thin, so when I rewrote the paper, I instead focused on what made it a modern building and what made Guadet a modern architect — namely, the way that it served as a physical mechanism to organize and control bureaucratic processes. The pneumatic post paper, too, looked at how technology had shifted the relationship of space and time to the human body, goods, and the communication of information. I had originally thought I’d do a dissertation on 19th century communication networks but was talked out of it by the entire PhD committee. (I was blue about that, but now it’s fine: they were right.) The majority of my research for these projects involved French language engineering publications.
  • Levittown, PA and its mass-produced landscape (1950s). Levittown, the famous, mass produced suburb, also mass-produced its gardens. Most bizarrely, Levitt patriarch Abraham Levitt wrote a column on gardening for the Levittown newspaper. Why? The way to maintain the value of the investment the Levitts had made in the suburb was not through the house but through the value of the landscape. The homeowners (most of whom had been apartment-dwellers and were completely unfamiliar with houses and gardening) needed to be taught to tend their gardens.
  • Apparatuses in architecture: a close reading of two 1920s works by Adolf Behne, a German architecture and art critic. For this paper, I analyzed the way that Behne used the word “apparatus” (Gerät) and the notion of defensiveness — as objects develop their own disposition. In many ways, I think Behne presaged the holistic approach to design that software finds so popular (and architecture, well, doesn’t). My research was all in German; the most painful part was reading poorly photocopied Frakturschrift (old-fashioned German writing).
  • Contingent communication: how communication jumps from network modes, using Pakistan’s 2007 state of emergency as a case study. I looked at cable television, satellite uplinks, and FM radio. (People who are holding crisis camps for Haiti might want to consider non-Internet media as a way of establishing communication networks — especially radio.) The idea for this paper came from a question Usman Haque asked me during my eTech presentation on India and mobile phone sharing, although what I wrote had nothing to do with it.
I’m looking forward to being able to talk next week about location scouting and casting and shooting a commercial: not my work, but someone else’s. This week’s location scouting not only introduced me to rooftops, kitchens and backyards, but also canine sociology between well-socialized and not socialized. My favorite: a golden retriever named Archie who chased oranges and carried a toy steering wheel in his mouth.