Schneier on Security
A blog covering security and security technology.
« The Death of Ephemeral Conversation |
| Lousy Home Security Installation »
October 19, 2006
Architecture and Security
You've seen them: those large concrete blocks in front of skyscrapers, monuments and government buildings, designed to protect against car and truck bombs. They sprang up like weeds in the months after 9/11, but the idea is much older. The prettier ones doubled as planters; the uglier ones just stood there.
Form follows function. From medieval castles to modern airports, security concerns have always influenced architecture. Castles appeared during the reign of King Stephen of England because they were the best way to defend the land and there wasn't a strong king to put any limits on castle-building. But castle design changed over the centuries in response to both innovations in warfare and politics, from motte-and-bailey to concentric design in the late medieval period to entirely decorative castles in the 19th century.
These changes were expensive. The problem is that architecture tends toward permanence, while security threats change much faster. Something that seemed a good idea when a building was designed might make little sense a century -- or even a decade -- later. But by then it's hard to undo those architectural decisions.
When Syracuse University built a new campus in the mid-1970s, the student protests of the late 1960s were fresh on everybody's mind. So the architects designed a college without the open greens of traditional college campuses. It's now 30 years later, but Syracuse University is stuck defending itself against an obsolete threat.
Similarly, hotel entries in Montreal were elevated above street level in the 1970s, in response to security worries about Quebecois separatists. Today the threat is gone, but those older hotels continue to be maddeningly difficult to navigate.
Also in the 1970s, the Israeli consulate in New York built a unique security system: a two-door vestibule that allowed guards to identify visitors and control building access. Now this kind of entryway is widespread, and buildings with it will remain unwelcoming long after the threat is gone.
The same thing can be seen in cyberspace as well. In his book, Code and Other Laws of Cyberspace, Lawrence Lessig describes how decisions about technological infrastructure -- the architecture of the internet -- become embedded and then impracticable to change. Whether it's technologies to prevent file copying, limit anonymity, record our digital habits for later investigation or reduce interoperability and strengthen monopoly positions, once technologies based on these security concerns become standard it will take decades to undo them.
It's dangerously shortsighted to make architectural decisions based on the threat of the moment without regard to the long-term consequences of those decisions.
Concrete building barriers are an exception: They're removable. They started appearing in Washington, D.C., in 1983, after the truck bombing of the Marines barracks in Beirut. After 9/11, they were a sort of bizarre status symbol: They proved your building was important enough to deserve protection. In New York City alone, more than 50 buildings were protected in this fashion.
Today, they're slowly coming down. Studies have found they impede traffic flow, turn into giant ashtrays and can pose a security risk by becoming flying shrapnel if exploded.
We should be thankful they can be removed, and did not end up as permanent aspects of our cities' architecture. We won't be so lucky with some of the design decisions we're seeing about internet architecture.
This essay originally appeared (my 29th column) in Wired.com.
EDITED TO ADD (11/3): Activism-restricting architecture at the University of Texas. And commentary from the Architectures of Control in Design Blog.
Posted on October 19, 2006 at 9:27 AM
• 43 Comments
To receive these entries once a month by e-mail, sign up for the Crypto-Gram Newsletter.
I was at the Library of Congress Monday night, walking around between the Capitol building, the Supreme Court building, and the Jefferson building of the LoC. These are all beautiful buildings, but the hideous barriers, quasi-temporary barriers, fences, and the like make this an ugly area, especially to pedestrian traffic.
"When Syracuse University built a new campus in the mid-1970s, the student protests of the late 1960s were fresh on everybody's mind." ...
"Similarly, hotel entries in Montreal were elevated above street level in the 1970s, in response to security worries about Quebecois separatists."
Combine the two and you get the Flashcube from Hell, a.k.a. Pierce Hall at the University of Chicago, which was built to be "riotproof" by elevating the residence part of the dorm on large concrete pillars and developing chokepoints in the entranceways so that rioting mobs couldn't force their way through.
Nowadays, of course, the main mobs are people trying to get to and from their dorm rooms on the way to class, and it's a screaming nuisance to have to go through in more-or-less single file. (Or at least it was the last time I tried to get in -- perhaps they've finally remodeled.)
They're not coming down, they're being replaced by permanent concrete and steel bollards, often with a decorative cap.
The install is "simple" -- dig a 4-6' deep ditch. Install a row of 10' long, 12" diameter HY steel tubes, and a bunch of rebar. Weld the tubes to the rebar.
Now, fill the trench and the tubes with high strength concrete. When cures, install a decorative cap.
What you get -- something you can't drive through, and something that will almost certainly outlast whatever it is protecting by several orders of magnitude.
If anything, security architecture is exploding. The Federal Reserve Bank in St. Louis has closed a street and rebuilt the entrances to make it impossible to reach the build with a vehicle. This isn't going away.
The Washington Monument landscaping project? A very cleverly designed way to turn grassland into a vehicle barrier. Only cost $15 million. You can still walk up to the monument, or even ride a bike, but you can't get a vehicle up there.
And so forth. Security architecture is exploding. The only thing that's happening in New York is they're replacing K-bar and concrete pots with more permanent solutions.
In 10,000 years, you'll easily be able to find the "important" buildings we had. They'll be surrounded by bollard fences. Never mind that the buildings will be long gone, those fences will survive.
This is a timely post; the World Bank and IMF has spent the summer transitioning from the eye-sore concrete barricades to (somewhat) more visually pleasing poles and bench installations. - To get an idea, you try searching for "IMF building" in Google images, you'll get some picture of the old building and a few of the new one as well. The Washington post did a brief article discussing the transition from barricades to the newer deterrents on Oct. 9, 2005 (by B. Forgey if you're interested)
After the IMF, the Bank followed suit and has installed poles and benches around the main building and is in the process of doing so for the various other buildings as well. To be sure, these new deterrents are much more permanent than the barricade predecessors, but they're also much less of an eye-sore.
Another thing to consider. The poles and benches seem to be primarily meant to deter would-be vehicle attacks on the building. But you may also consider them to be dual-purpose; traffic on the surrounding streets can be pretty heavy and pedestrians (especally Bank staff) are not always the most aware of their surroundings --to see them cross Pennsylvania Ave. with no regard for oncoming cars is depressing. So these security poles and benches might also serve to avoid any potential car-on-the-sidewalk accidents that probably happen from time to time.
The barricades were probably errected around 9/11 but I can imagine that they were also used in some way as deterrents to the masses of protestors who gather around the time of annual meetings. The poles and benches will certainly be much less effective than the barricades for "crowd-control"... so in a sense the Bank has traded one deterrent for another with the new installation.
Security architecture doesn't need to be ugly. The Atlanta Fed building has a beautiful lawn which surrounds the building, and is raised 4 or 5 feet from the surrounding street, with a granite restraining wall. It's a very effective protection against truck bombs.
And they only protect against one type of threat as if a determined attacker could not choose one of a million alternatives.
Ummm.........I'm not sure where you get that idea.......
The buildings on the main SU Quad come in two vintanges: 1960's and OLD.
Look at the white space surrounded by blue building shapes and criss-crossed with black lines: http://map.syr.edu/ That is the Quad you just said doesn't exist. Now it might seem that you could potentially shut off traffic to/from that area. I've seen it tried (big politicos visiting, etc)--and it doesn't work. Another decent view to illustate the smattered unplanned mess that is the North Campus there: http://www.syr.edu/syracwis/imagerep/maps/... Note how open that quad really is. Please do better research next time.
Don't think there were too many hotels built in Montreal in the 70s! Which one has an elevated entry?
it's funny how the security architecture of one generation is the bane of another. Syracuse built small and narrow to prevent riots; Hausmann in 19th century Paris built large and expansive boulevards to prevent barracades and riots. as for "security"... i live and work in midtown manhattan and have yet to see any of these bollards be removed. on 58th street between park and lex, there is an extremely annonymous building which has a line of them outside... there is a major subway station there and the difficulty of getting around these ugly, misbegotten, scarred, defaced and butt filled objects starts everyone's day off badly. they don't stop anyone from blowing a building up in any one of a million conventional ways or even from running an airplane into the building -- on purpose or not. they are -- like so many "security" measures from airport searches to signing in at building lobbies -- merely cosmetic. but what's even uglier is the big bollard they leave on the soul.
"Whether it's technologies to prevent file copying, limit anonymity, record our digital habits for later investigation or reduce interoperability and strengthen monopoly positions, once technologies based on these security concerns become standard it will take decades to undo them."
I'd argue that the same 'embedding of code' to control users into physical space - for political or economic reasons - is increasingly part of the mentality of many of those planning new systems and products as well. Using Lessig's term 'architectures of control', I've been trying to investigate and analyse examples in a few fields, from urban planning (anti-homeless benches) to DRM, and much in between - some are trivial, others much more serious.
Architectures of control in design - http://architectures.danlockton.co.uk
Finally, I now understand why the SUNY Buffalo North campus is so ugly - it was built around the same time as the Syracuse campus.
The Montreal Hyatt (which is a dreary and dreadful place to stay for other reasons as well) has one of those elevated entries.
The UK Ministry of Defence "Abbey Wood" site, next to my offices, were built in the early 1990's. when the IRA were the main threat. They have quite a nice design.
The site is split into car-side and buildings; all parking is as far away from the buildings (car bomb defence), especially the visitor section. you have to walk over a narrow footbridge to get in.
Between the buildings and the (no parking enforced by armed police) road is 'lake'. This stops suicide bomber raids without the ugliness of the concrete barriers.
What we effectively have is a modern variant of an old castle. The lake supplants the moat, but it and the narrow choke point/drawbridge.
When they build in paranoia from the outset it looks pretty.
Yeah I'm pretty puzzled about that comment about hotels in Montreal. The main thing people did in response to Quebec separatists was simply move away -- there wasn't much construction at all in this city for the next 20 years, let alone new hotels being built.
That said I can think of a few hotels on Mountain that have raised entrances.
The planners of the planters were shortsighted - they're too short. They needed to be as tall as the building itself to prevent airplanes flying in.
I second the comment by SUGrad about the Syracuse University campus. Whatever crowding exists seems due to one simple factor: trying to squeeze as many buildings as possible into a limited amount of university-owned property. In fact, when I was attending SU (1986-1990) there was a near-riot upon the opening of the Science and Technology Center, not to mention the 1987 basketball riots just off campus. What specific buildings/features are you referring to?
Erik Olsen wrote:
"They're not coming down, they're being replaced by permanent concrete and steel bollards, often with a decorative cap."
Exactly. In my small town, barriers went up around the Federal building right after the Oklahoma City bombing. They lasted about two years, which was how long it took building management to fund and complete an external landscaping project that included concrete walls about eight feet from the building, clad in faux slate and capped with polished granite.
The security architecture is permanent, and it's now quite a maze trying to find your way from the street into the post office. Folks with mobility issues have to walk or roll more than twice as far as they used to from parking to the main entry.
True, I was also thinking of the many "tourist rooms" establishments in downtown Montreal, but these are the 'second storey' hotels that used to be common in many North American cities. To spell things out, they cater to prostitutes and their clients, but they do get tourist traffic. The last really big hotel built in Montreal was the Chateau Champlain in 1967 - nothing bigger than that has been built since!
Thirding the SU comments. I'm puzzled by the "second campus in the 1970s" reference. I graduated in 1993 and am fairly familiar with when the different buildings were constructed. I don't know what significant construction took place in the 1970s. And having been there during the Gulf War, I can safely say that students always had a place to protest ... and play Frisbee, suntan for the two sunny weeks out of the year, etc.
Good examples of "riot-proof" housing occur at SUNY-Binghamton, a short ride from Syracuse. College in the Woods, a dorm community built after the "SUNY Riot" nickname got pinned on Binghamton, features concrete "quads" with steps breaking them into multiple levels to prevent charges; extremely steep, but very wide, stairs, to make it difficult to defend the central quad (these stairs also have a completely laughable "wheelchair ramp", at about a 20% grade, on one end); and a dining hall with lots of glass walls, amounting to what basically looks like a shooting gallery...
(Of course, this didn't prevent a sit-down strike from taking over the administration building when I was there...no one expected the students to demonstrate *there*, I guess.)
I'm seriously confused about the reference to Syracuse here. The Main, or Northern, Campus has an old large quadrangle and another very large open green area.
South Campus, which was built in the 1970s, is a residential campus with very, very ugly squat concrete buildings. However there is lots of space between them, although no real 'centre' to the campus, maybe that's it? But for protest everyone would head to the large main quad.
I'd love to know more, maybe there is something interesting about the South Campus architecture, but I ain't seen it in quite a few years here. Any references, Bruce?
I have to agree with J. Hale, Larry Z. James H., and SUgrad. I graduated in 1984, and was there in 1980 when the 50,000 seat carrier dome opened, only steps away from the main quad. The main SU campus is built on a hill, bordered by a park & residential neighborhoods to the east, a very large cemetery on the South, I81 on the West, and M street (retail, hotel & residential) on the North. Expansion options are limited to bulldozing former greek housing, or building on/within the existing footprint.
But that doesn't necessarily mean the campus feels cramped or crowded, even on football game days. There is still plenty of green space remaining. In fact, since the mid '80s, they've been adding green space, such as the removal of the last block of University ave between the student center and newhouse.
I'm curious, Bruce; what was your source for the SU comment?
Sixthing (seventhing) the SU comment. Have you actually BEEN to Syracuse University? The campus is very old, and was built in the 1870s surrounding a grassy Main Quad that still exists and is the scene of many gatherings and demonstrations (including one just last night!). There was no "new campus" built in the 1970s except for a remote residential area (which has lots and lots of green space). Now, the current administration seems to be trying to smack new buildings on every patch of green space they can find, but that's a different issue entirely.
University of Texas at Austin did something similar in the 1970's. I have a whole article about it somewhere... The west mall (next to the Union) used to be open and grassy. They paved it over with pebble-y pavement to make it painful for hippies to walk barefoot and installed giant planters to break up the space. They also installed those concrete walls along Guadalupe (the drag) to create a barrier between town and gown, and many other "improvements".
"University of Texas at Austin did something similar in the 1970's. I have a whole article about it somewhere."
I'd appreciate a copy of the article if you find it.
What about the W and the Intercontinental? Don't those class as big hotels?
Where did you get the reference for Syracuse? I had heard the the steps at schine for example were made shallow and long in order to prevent students from charging up or down them. I always thought this was a rumor. Can you point me to more information?
The steps at the Schine Student Center at SU were most likely made that way out of sheer stupidity. The student center was built in 1985, so I doubt student protests were on anyone's mind.
the new oklahoma federal center is a great example of a contemporary response to this very discussion - the implication that architecture can solve social issues is like using a screwdriver to hammer a nail... architecture can solve architectural issues - social solutions can solve social issues ... therien lies the dilemma . what is the architecture of democracy? solid and exclusive or transparent and accessible?
Yeah, your source regarding SU's architecture is definetly off-base. The campus has a HUGE greenspace in the middle of 10 or 12 very old buildings, some dating from the 1870's. In fact, it's the largest campus (in acreage) of any private university in the U.S. It would still be very easy to protest there and there indeed have been (lame) protests since the 70's. (When I was there in the 80's apartheid was the big issue). Most of the post-1950 construction has been in the periphery of this space and doesn't seem very inaccessible or "secure".
"When Syracuse University built a new campus in the mid-1970s, the student protests of the late 1960s were fresh on everybody's mind. So the architects designed a college without the open greens of traditional college campuses. It's now 30 years later, but Syracuse University is stuck defending itself against an obsolete threat."
1)Syracuse University did NOT build a new campus in the mid 1970's.
2)Syracuse University has, and has always had, since the 1870's, an open quarangle with lots of green space, very much a typical traditional university quadrangle.
3) Any buildings added on the main campus since the 1880's have always been added around the edges of the main quad, with significant amounts of green space between the buildings, and standard doorways and stairways.
Please check your facts before posting. Maybe, like another poster suggested, it was SUNY Buffalo you were thinking of?
The University of California at Irvine (UCI) was built in a manner designed to discourage riots and large congregation of students. It has a ring shape with no central gathering area. There is a large park in the middle, but it's full of hills, trees, and other obstacles which would prevent a large crowd from gathering. UCI was planned in the 60s, so the Berkeley riots was definitely on the designer's mind.
At the beginning of the autumn term 2006, I took my son back to his university accomodation- the renowned Brunnel in UK and was so appauled by a hostel design that have taken security issues too far. Windows, were very narrow and low such that visual contact with the residents and the environment was completely lost, fire exits were virtually non existent.
If a resident is trapped in the room, it is practically impossible to escape through the windows without first smashing the brick/concrete walls.
To the architect of those blocks of hostels, I'll recommend that he/she be de-registered whatever the brief was. A prison design could'nt have been much better than the hostels and even then the confinement of inmates and the spatial configuration lacks the social and functional dimentions that I will attribute to a student hostel.
Northwestern University has similar architecture for some of the buildings put up around those troubled times. In particular, the main library and the Rebecca Crown Center (main admin building).
Both buildings feautre multi-tower construction surrounding and protecting an elevated, central courtyard; vertical spearation of entraces from the rest of the interior; choke points; and don't forget the narrow, shielded windows looking like nothing more than arrow slits. You can see a couple of pictures here:
Every now and then, we hear the refrain that [insert 60s campus here] university was designed in that hideous way to quell student riots.
The problem with these claims is that they're nearly impossible to verify. A couple things tend to militate against them, though. First, if you look at _anything_ institutional built in the period, you'll note that the achitectural style was "prison". This is as true of university campuses as it is of anything else built in the period. Indeed, there are plenty of supposed luxury buildings of the period that look like they were lifted directly from the minds of a perverse Soviet architect. Further, the same "security concerns" claims are sometimes used to explain the silly layout of period public-housing construction, but it's clear from previous decades of architectural writing that the causal arrow, if it travels at all, travels the other way. By concentrating a lot of poor people in desolate buildings with no services around, governments created the ideal conditions for riots: lots of bored young men with nothing to do.
The second thing is that the buildings date from the _actual period_ where the unrest was going on. But university campuses and other such large projects take _years_ to get built, in bureaucratic wrangling alone (and never mind that actual construction used to take longer, dependent as it was on smaller machines). By the time projects are actually built, they usually look like previous architectural fashions. Certainly, there is no way that the SUNY Buffalo campus at Albany (which ought to be the canonical example of the prison/school line of architecture) could have been responding to "riots", because it was conceived in the mid-60s before such riots had really become a problem at suburban northern universities. (I conclude this partly on the basis of what I'm able to learn from http://library.albany.edu/speccoll/findaids/... Moreover, it opened in the early 70s, which means the buildings had to have been designed and the campus laid out in the period prior to the riots happening, or the building wouldn't have been done in time.
It could be, of course, that easy access by National Guard troops was really what was on the minds of the designers. But in the absence of an actual citation of someone claiming that's what s/he was doing, I'm inclined to be pretty sceptical. In fact, I'm more inclined to attribute the common design elements to the loss of anything like good architectural (that is, Vitruvian) sense by practitioners: it's easier to explain things by incompetence than by resort to conspiracy.
There's an interesting side note to all of this, though. The campuses that are the ones most supposed to be suitable to easy troop access, prevention of student gathering, &c. end up being terribly unsafe because there's never anybody around. So they make us less secure, not more, precisely because they eliminate all the things that make public spaces welcoming to humans and therefore nurturing of a healthy society. Even if those campuses weren't designed to be "defensible", we ought to take this negative lesson to heart before we slap down more concrete barriers everywhere.
Seconded on the UT Austin "west mall" construction. I suspect the planters were installed to prevent vehicles from Guadalupe ("the drag") driving onto the mall, which is widely said to be a "free speech zone", as if the First Amendment only applied in certain places. During the early 90s, there was an occasional "smoke out" by NORML, where people would smoke pot openly in some kind of civil disobedience. If I recall correctly, the city passed a law that made possession of small amount of marijuana punishable by a $1 fine, presumably so officers could, at their discretion, appear to charge people with a crime, while only giving a token penalty.
Actually, I heard another rumor that they were designed to prevent tanks from driving onto it, but I sort of doubt that.
I didn't have time to read all the comments, but obviously there is a lot of interest in the examples. i like to get some more exact data myself (drawings, photographs). If only because i am doing research on the influence of terrorism on architecture and the city and how we, consequently, behave in the city.
Of course a one on one relationship between terrorism/riot control and design can never be established beyond doubt. But i think these cases make clear there is, at least in hindsight, a link. Just as you can find a rough link (in history) between the layout of the defenses of the city, and the developments in weaponry.
Here's the article on how UT rebuilt the West Mall to discourage protests, gatherings, free speech, etc.
Quote from article:
"Meanwhile on the West Mall, the only method available to the administration to disperse gathering crowds was to turn on the sprinkler system, and this could be more fun than trouble if the protestors were prepared. During the summer of 1973, construction began on the West Mall "renovation" by Stokes Construction and James E. Keeler, Landscape Architect, to the tune of $280,000. It involved removing the existing grassy areas and replacing them with limestone planter boxes around the trees and a paved court and fountain in front of the Union. Eight trees were planted in the center, raised boxes. The paved area of the mall increased over 50%, and a later Student Senate resolution dubbed the project the Frank C. Erwin Memorial Highway."
BTW, there were castles before the reign of King Stephen of England (reigned from 1135 to 1154). The earliest known castle is in the Loire Valley in northern France at Doue-la-Fontaine dated 950 and it is were feudal society originated.
Interesting the two-door vestibule origin :)
"Lawrence Lessig describes how decisions about technological infrastructure become embedded and then impracticable to change. Whether it's technologies to prevent file copying, limit anonymity, [etc.], once technologies based on these security concerns become standard it will take decades to undo them."
It's worth noting that this cuts both ways. Most of the old, core internet protocols were and are quite naive, trusting, and insecure -- but it's just as hard to replace *them* with something more secure. ssh has gradually replaced telnet, and scp is replacing ftp, but there is still no meaningful replacement for smtp, and the problems caused by smtp's lack of any real authentication mechanism keep getting worse.
“God bless you, my son,” the priest said to Benny, making the sign of the cross, then he continued walking down the street, a little unsteadily, Howell thought. Surely there couldn’t be enough Catholics in Sutherland to warrant a full-time priest, and enough for him to worry about to get him looped this early in the day, he thought as he got back into the car.“What a night, indeed,” he said back to her, then fell asleep.“One should be plenty.”
Schneier.com is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc.