SO THERE you are, waiting to cross the road when a car pulls up, a passenger
leans out and asks you the way to the nearest bank. There happen to be two
routes, a long drive down the main road with a single left turn and a much
shorter way that twists and turns through backstreets. Which do you tell
them?
The answer is the longer, simpler route. Try giving directions with too many
left and right-hand turns and the travellers are likely to end up lost. Perhaps
because our brains can cope with only so much complexity, humans prefer easy
routes over labyrinthine ones. It is the same desire for simplicity that
underpins the old music hall joke. “How do I get to the Old Kent Road from
here?” runs the gag. “Well if I were you,” comes the reply, “I wouldn’t start
from here.”
Our preference for straightforward journeys may seem obvious, but it’s
something that many architects and planners have missed. Not so Bill Hillier,
who first noticed it in the early 1980s while trying to fathom why some
buildings “work” as places that promote creativity and communication. But then
to Hillier, who is professor of architecture at University College London (UCL),
roads and corridors are one and the same: spaces. Roads are spaces that link
buildings; corridors are spaces that link desks with photocopiers, faxes and
coffee machines. “People use space intuitively, much like they use grammar,” he
says. And just like grammar, space has its own rules, which we use without
thinking when moving from one place to another.
Advertisement
Those rules, which stem from people’s preference for easy journeys, are the
basis of a theory that Hillier calls space syntax. The theory challenges
established design practice and is couched in difficult
terminology—factors that have tended to stop it spreading quickly. But
there is growing evidence of its effectiveness. Wielding Hillier’s ideas,
researchers at UCL’s Space Syntax Laboratory can explain, for example, how
cities grow and why shops thrive in particular areas. Space syntax even helps
archaeologists to deduce the organisation of ancient societies from the ruins of
their buildings (see “Ancient space”).
But the real power of the theory is that it provides a way to design spaces
that work. Architects have used space syntax to design offices that foster
creativity, revitalise urban wastelands, and even make houses less likely to
being burgled. Recent research also shows that space syntax is a more accurate
way of forecasting traffic than the costly surveys used today.
Hillier’s contention is not just that how spaces are connected determines how
much they will be used, but that this relationship is mathematically
predictable. “If you design something in a particular way it will influence
people’s movement, and you can predict when people will be in the same space,”
he says.
Hillier’s analysis is topological—it focuses on the way elements in a
network are linked. Going back to the lost travellers, it is easier to direct
them to some roads than to others. It depends on how the roads are connected.
Hillier brings out the accessibility of roads—or lack of it—in a
more formal way. He and his colleagues catalogue routes between each road and
every other road in a network by the number of turns that must be made to travel
between them. One way to depict this is to draw up a stylised line graph for
each road (see Diagram).
When these are combined using the lab’s computer
model, they reveal a hierarchy that reflects how accessible every road is from
all other roads.
One of the Space Syntax Laboratory’s case studies is the road network of
inner London. Oxford Street, which runs east to west through the city, is the
road with the best links to the rest of the network. “On average,” says
Hillier’s colleague Alan Penn, “you only need to turn left or right nine times
to reach anywhere in inner London from Oxford Street. On the other hand, if you
start from a back street in west London, you’ll have to turn twenty times or
more.” In the completed analysis of inner London, Oxford Street is at the top of
the accessibility hierarchy while back streets are at the bottom.
Two important consequences flow from these hierarchies. First, researchers
can forecast how much traffic will use a road from its position in the
hierarchy. Secondly, they can forecast which roads will be attractive locations
for offices and shops.
The conventional method of forecasting traffic is based on a loose analogy
with Newton’s law of gravity, in which the attraction between two bodies is
proportional to their masses and inversely proportional to the square of the
distance between them. In the traffic engineers’ version, the amount of travel
between any two zones within a wider area will be proportional to the number of
vehicles that start and end their journeys in each zone and inversely
proportional to the time it takes to travel between the zones. If the engineers
plan to build a fast new road, the model forecasts that the new road will
capture traffic from slower roads linking the zones.
To gather data for this model, engineers carry out a hugely expensive survey
of drivers, asking them where they have come from and where they are going to.
From this, they estimate how much traffic each zone generates. They add a few
fudge factors to make sure the model agrees with recorded data, factor in the
quicker journeys the new road will provide, and then run the calculation to
forecast levels of traffic at some time in the future. The technique is
notoriously unreliable in urban areas. “Conventional models are absolutely
dire,” says Phil Goodwin, professor of public transport policy at UCL and the
British government’s adviser on transport.
Penn and his colleagues found a strong correlation between a road’s position
in the hierarchy combined with its width, and the volume of traffic running
along it. In research published last year, they show that this method can
account for more than 80 per cent of the variation in traffic flows from street
to street. The results are impressive compared with conventional methods. An
official audit of the government’s forecasts in 1995 found that half of them
were wrong by more than 20 per cent.
Space syntax also helps to explain why traffic can vanish when a road closes.
Traffic engineers normally argue that if you close a road, you need to create a
relief road to take the displaced traffic. But a study carried out last year by
Goodwin challenged this wisdom. It looked at a number of roads that had been
closed and found that a substantial chunk of traffic simply evaporated. In
practice, drivers have a lot of flexibility: instead of driving to work, for
example, they may work at home one day a week. Penn says this is exactly the
sort of result you would expect from space syntax. If you sever a link in a
network, a number of roads will become more difficult to get to. They slip down
the hierarchy, so the volume of traffic on them will fall.
One of the beauties of the technique for modelling traffic flow is that it
works as well for pavements as for roads. From detailed surveys of pedestrians,
UCL researchers have found they can predict pedestrian movement from the street
pattern.
The logic behind these traffic patterns is crucial for understanding how
cities grow, and Hillier’s ideas provide a way to predict this process and
perhaps to avoid creating urban wastelands. According to space syntax, a road
that can be easily reached—on foot or by car—will tend to be busy.
This offers shopkeepers and office developers a golden opportunity. “Passing
trade makes all the difference to whether shops are let or not,” says Brian
Raggett of the Royal Town Planning Institute in London.
Penn quotes the estate agent’s adage: “There are only three things that
matter: location, location and location.” A prime location, says Penn, has a
multiplier effect. At the turn of the century, Oxford Street was already one of
the most accessible roads in London. In 1906, Gordon Selfridge, an American
shopkeeper, started building the department store that still bears his name. The
store rapidly became a destination in its own right, adding to the street’s
attraction. Other stores followed in its wake, attracting still more people to
the street. Today many of London’s department stores are still located on Oxford
Street.
Merchants of Venice
The accessibility of spaces also gives a clue to how car-free Venice has
developed. The city’s labyrinthine map of streets and alleys appears on
Hillier’s computer. As on other Space Syntax Laboratory maps, the hierarchy is
coloured: red for the most accessible streets and blue for the least accessible.
Glaring out in red is the Rialto bridge over the Grand Canal, which is where for
centuries the merchants of Venice have plied their trade.
Then there’s Berlin. After the Second World War, when East Germany built the
wall that cut the city in two, the old commercial district on Friedrichstrasse,
was cut off in the eastern zone. It could not be reached from the west and in
the east it was a backwater. Berlin grew two new office centres. In East Berlin,
the centre moved to the east. In West Berlin, a new commercial centre flourished
on Kufürstendamm.
Then, ten years ago, the wall came down. Conventional models of urban
development predicted that the stronger of the two centres—presumably
Kufürstendamm—would come to dominate the city and the other would
fade away. But Jake Desyllas of UCL uncovered a different outcome. Sure enough,
the eastern centre faded, but office rents also dropped in Kufürstendamm.
The smart money has moved back to Friedrichstrasse. Its accessible location in
the city with its reunified road network makes it the prime site for commercial
development—just as space syntax predicts.
Penn contrasts the way the multiplier effect has worked in Oxford Street with
the way its absence has created the wasteland of London’s South Bank. The South
Bank was designed after the Second World War to be the cultural centre of
London. It has concert halls, theatres, a museum, an art gallery and office
blocks. But despite all this, the area has never attracted many shops.
Among the reasons for this is that the planners segregated pedestrians and
vehicles—to cut accidents and prevent pedestrians holding up the traffic.
This left cars at ground level and condemned people to windswept walkways above,
and has kept the volume of people passing potential shopping sites too low to
make it worth building shops. “The area was used by residents, office workers,
cultural audiences and general visitors. But there was little mixing. They all
used it at different times and used different routes,” says Mike McCart,
commercial director of the South Bank Centre, which commissioned the Space
Syntax Laboratory to analyse pedestrian movement. “The result was places that
felt unsafe and hostile.”
The lab’s plan for redeveloping the site includes removing some walkways to
make the layout of the area easier to understand, and to allow more mixing
between pedestrians and vehicles. The changes are intended to increase passing
trade and kick-start retail development. Penn predicts that once one shop opens,
it will rapidly attract others.
The space syntax researchers have used the same techniques to redesign
offices—starting with advertising agencies. In these hives of creativity,
the most productive offices are those where the workers talk to each other.
Research from the Massachusetts Institute of Technology has shown that about 80
per cent of seemingly casual conversations are about work.
So, in one office, UCL researchers installed video cameras to monitor how
people used the spaces. They found that conversations happened largely by
chance. To increase the chances of these accidental encounters, they positioned
desks in an open-plan office so that most workers sat close to the main
through-route. The result was a large increase in conversations—which,
says Penn, helped to stimulate the creative spark. The lab also designed the
Coventry headquarters of Powergen, one of Britain’s electricity generators. The
rate of chance conversations in the building increased by 9 per cent. “It works
very well indeed,” says a company spokesman. “You bump into a lot of people
informally and you can sit down at special spare tables for a brief
پԲ.”
One of the most controversial predictions of space syntax is that crime can
be designed out of housing estates. In Britain, the 1970s saw a major effort to
pull down tower blocks and replace them with houses. To make the new estates
child-friendly, planners tried to exclude speeding cars and passing strangers by
designing them without through-routes for cars or pedestrians. But these designs
had an unexpected side effect. Space syntax research on housing estates in
England revealed a direct link between the layout of estates and burglaries.
Houses in cul-de-sacs are far more likely to be burgled than those on
through-routes.
Hillier’s computer plots the laboratory’s analysis of one such housing
estate. As normal, through-routes show up red, while alleyways and cul-de-sacs
are blue. Then the computer shows the sites of burglaries. Almost without
exception, the burglar got into the house from one of the blue spaces. “What
would you do if you were a burglar?” asks Hillier. “Break a window that’s
visible from a through road or choose one that’s not overlooked?”
Hillier’s solution—which flies in the face of established
thinking—is to avoid designs that create zones without fewer people or
less traffic. The threat of road accidents can be reduced by slowing traffic
with devices such as road humps. “We now know a lot about how to calm traffic,”
he says.
After twenty years, space syntax is starting to be widely used. “It’s taking
off,” says Hillier. It has a growing international following, and local councils
in Britain are increasingly asking the lab for help. If you want to steer urban
design, you have to understand how the system works before you intervene, says
Hillier. “It’s evidence-based design and planning, rather like you have
evidence-based medicine,” he says. The patients just happen to be made of stone
and concrete, rather than flesh and blood.
THE ruins of multistorey buildings in Chaco Canyon, New Mexico, have long
puzzled archaeologists. These once grand buildings are about a thousand years
old. But how were they used and by what sort of a civilisation? The theory of
space syntax is helping to answer some of the puzzles.
By studying how the rooms in the buildings are connected, it’s possible to
build a hierarchy showing which rooms were easily reached and which not. This
analysis can then be used to back up or contradict conclusions reached from
other reasoning. “We use space in the same way today,” says Wendy Bustard,
curator of the US National Park Service’s Chaco Canyon collection.
Archaeologists have found grinding stones in some of the rooms, which the
inhabitants used to make flour from corn. Conventional archaeology concentrates
on what rooms were used for. By analysing how these rooms connected to others,
Bustard deduced that the grinding stones were communal and shared by several
families. The inhabitants, it seems, had a highly organised social structure.
“This implies a level of social organisation above the nuclear family,” she
says.
What makes space syntax a useful tool, says Bustard, is that it is a
“completely different” approach to archaeology. Of course, it does have its
limitations. Because ruins are incomplete, archaeologists have to make
assumptions to plug the gaps. Consequently, interpretations are often hedged
around with ifs and buts. “That,” she says, “is true in any kind of
archaeological analysis.”
Ancient space
-
Further reading:
Space is the Machine
by Bill Hillier (Cambridge University Press, 1996). -
The original theory of space syntax is laid out in
The Social Logic of Space
by Bill Hillier and Juliette Hanson
(Cambridge University Press, 1984) -
More information about space syntax and its uses can be found at
www.bartlett.ucl.ac.uk/spacesyntax/
