A Pattern Language
Christopher Alexander, Sara Ishikawa, and Murray Silverstein,
with Max Jacobson, Ingrid Fiksdahl-King, and Shlomo Angel, 1977
A Pattern Language presents 253 "patterns," or tips for ordering
the built environment. And by "the built environment" I mean the
entire built environment: the first pattern is about how the
globe should be divided up into countries (the authors suggest that 1000
countries is just about right) and from there the book gets smaller and
smaller in scale until it's talking about where to place beds within
bedrooms. So far I have read the first section, "Towns," which
shockingly enough is about city planning. (Though I do plan to continue,
after 94 patterns and 460 pages, I'm taking a break.)
(By the way, what is with the number 253?)
While I read this primarily because I'm very interested in city planning,
my understanding is that many people who aren't especially interested in
the built environment have become big fans of this book because of its
structure. The structure is indeed very interesting! It immediately
reminded me of a couple of things. One is go strategy. Consider the
Black to play. Now, there is a rule of thumb in go that says that a
two-space extension along the third line is proper. This would suggest
that Black should play at the circled point. But there is another
rule of thumb that says to play away from thickness! Playing at the
circled point would essentially be a matter of pitching a stone right
at the white wall. Figuring out how to determine which of two opposing
rules takes precedence, or how to compromise between them, is a huge
part of go strategy.
The analogy to go strategy, incidentally, was how I came to understand
why these tips are called "patterns." In go, the terminology is obvious:
you actually see the same arrangements of stones over and over, and you
start to recognize those patterns and react accordingly. "Okay, so when
I see three-fourths of a bamboo joint, I should usually play the fourth
point myself." "Aha, that arrangement of stones calls for a shoulder
hit right here!" And so forth.
The other thing A Pattern Language immediately reminded me of
was standardized test tutoring. Again, there are an awful lot of rules
to convey, and the trick for the student is to learn not just the rules
themselves but also which ones to apply to a given problem and how to
resolve apparent conflicts between them. For instance, when trying the
answer choices in a math problem, it's a good idea to start with the
middle value, since you can narrow things down — but it's also a
good idea to start with the easiest value to work with, since you can
test it faster.
Needless to say, it was a revelation to see an approach to one of my
favorite subjects, city planning, take shape in a manner that resonated
with my occupation and one of my hobbies. It's easy to see why the
format has been so influential. Structure is so important in
non-fiction — the whole point of which, I would argue, is to
organize information and thereby aid the reader in understanding
it. I've mentioned this before: rambling books such as
Salt may be full of interesting data, but without a framework
for it, not much tends to stick, at least for me.
And it occurred to me that the very fact that this idea —
"non-fiction must have a clear structure" — has appeared in several
of my earlier Calendar articles makes it a "pattern" like those in this
book. Then I realized that, in fact, there must be many such patterns in
the evaluative "language" I have unconsciously developed over the past
seven-plus years. So I trawled through my archives looking for places
where I had the same reaction to different works. And then I
wrote them up. (Aside from the 28 on that
page, there were others that came up only once but which I recognized
as reactions I'd had to books I'd read or movies I'd seen before I
started doing these writeups. So the pattern page may grow over time.)
But the actual subject of this fricking book — or the "Towns"
section, rather, since that's what I've read — deserves a few
words. What I like most about the authors' approach to city planning
is their recognition that we often want opposite things at the same
time. For instance, I would like to be able to walk out my door and
have all my favorite shops right there within a block or two. And I
want to live surrounded by nature with no one else around for miles.
I recognize that these goals are not easily reconciliable. But
thank you, authors, for not insisting that I therefore simply
drop one of them!
I grew up in a place where, to get to the nearest building that wasn't
a house, you had to walk a long, long way — if "walk" is even
the word, since the steep hills meant that "climb" was more appropriate.
And even if you had a car, where would you go? Unless you left town,
you were limited to shopping centers with slightly different combinations
of chain stores. And in the US, at least, this is more the rule than
In 2001 I moved to New York City. There I could walk to a subway
station, board a train, and go visit some of the world's greatest
museums, or have a meal so good that I would remember it years later,
or see a show or a movie that wasn't playing anywhere else on earth.
Endless opportunities for fun and personal enrichment were just a
Metrocard swipe away.
But given the choice, I'd take the first one. As this site documented
at the time, it felt like the city of New York was actively trying to
kill me. Being stuck in the middle of all those people, all those
buildings, all that everything crammed into such a tight
space... I can't believe I lasted a year there. I often hear people
talk about how to deal with the problem of the suburbs — the
problem being that they're ecologically unsustainable and that they're
intrinsically soul-destroying — and often their solution is to
just gather everyone up into a bunch of New Yorks. To me that's
a cure every bit as bad as the disease.
So I was thrilled to see that the main thesis of the "Towns" section
of A Pattern Language is that the goal of city planning is to
conjure up the magic of the city, to concentrate the population enough
to bring about the cultural opportunities that only arise in cities,
without the madness of the city that tends to accompany it. Though
they recognize that for a society to function people must live
communally to a certain extent, they don't lose sight of the fact
that people need their own space. Furthermore, they acknowledge
that people are different and that they therefore must have the
opportunity to position themselves at different points on the continuum
between places in the middle of the action and places with peace and
So a lot of the "Towns" section of A Pattern Language is about
different types of clustering. The idea is to stave off both total
centralization, where everything ends up in one massive downtown,
and total decentralization, in which the urban area is a sort of
homogenous gruel with nowhere people really want to go. And the
clusters are intended to be different, so you can find the one that
speaks to you.
One aspect that the authors don't address as much as they probably
should is economics. Perhaps this is because they were writing in
the 1970s, before the economies of Reagan and Clinton and the
Bushes cratered the middle class and left us with a vast gap between
rich and poor — and consequently polarized neighborhoods into,
on the one hand, increasingly interchangeable affluent enclaves, and
on the other hand, slums. I read an
article recently about the way Seattle neighborhoods have become
less diverse: "Ballard was old Scandinavians. Fremont was hippies.
Capitol Hill was gay. Kent was where whites of modest means moved to
escape Seattle school busing. Bellevue was the same for the rich.
Today, you can make a joke about Ballard but it's a bunch of wealthy
people who work in the information industry. You make a joke about
Wallingford and it's a bunch of wealthy people who work in the
information industry. Fremont? That would be a bunch of wealthy people
who work in the information industry." I suppose it beats the
alternative of them all turning into dilapidated nightmares but it's
still kind of depressing.
The closest thing I've found to an ideal city is Berkeley, with its
combination of great eateries, attractions (the university being #1
for me), climate, and scale — while there's not as much
great stuff as in New York, I actually feel comfortable while making
use of Berkeley's great stuff and not constantly assaulted. The authors
of A Pattern Language seem to agree! Berkeley is the example
they turn to again and again to illustrate their principles. So why
don't I live there? Mainly because the polarization of wealth and the
increasing disappearance of middle-class neighborhoods that has
resulted from it mean that the options around here boil down to
too pricey, too scary, or too far. I went with door number three and
took a place twenty miles away.
Also I'm not a huge fan of Berkeley's housing stock. But that's a
discussion for after I read the "Buildings" section.
Return to the Calendar page!