How London Became a City
Professor Bill Hillier, University College London, and Bartlett School of Architecture
Why are we talking about London as a collection of villages, a city of villages, and what does this imply? I’m a Londoner and I’ve always been amazed that wherever you are in London you feel you are part of a local place and you feel like you are part of a big city. That’s not like Paris, for example, where you just feel like you’re in part of a big city. So how did London do this amazing thing, being sort of local and vast scale at the same time and what does it tell us about the future of Archway?
When we talk about villages I don’t think we really mean villages. Villages are small, they’re bounded and they’re far from each other. And I don’t want to live in one, I have to say. The fact that most of you live here suggest you may agree with that.
What we are trying to say is that when we talk about villages it means wherever you are in London you seem to be within walking distance of some small local centre and not very far from a much larger one. And both of these seem in some sense community focuses, maybe not everything we would want but places that do kind of acquire a kind of community identity as well as being a useful sources of local services.
Of course, these centres are sustainable as well. We now know a great proportion of the kind of trips we make when living in London are pedestrian trips, especially going to these centres. If we look down at London from above, there is a pattern of large scale centres in London within the M25. It’s a kind of mesh. I think there are, in fact 1200 or 1600 of them of descending size down to small runs of shops and a café. It is an astonishing thing that something as complex as this has arisen out of a very, very long historical process with very, very large numbers of people acting for different reasons rather than planning it.
Now what really interests me is how this type of thing arises - the self-organisation in cities. How did London grow in this peculiar mixture of the local and the large scale if you like? This is what gives London its peculiar character and makes it such a good place to live and probably makes it more sustainable than most other cities of this size.
We’ve tried to understand London in the past. People have tried to draw boundaries around areas. We go back to the Abercrombie plan in the early 1940s and try to imagine, if you like, that we can divide London up into bounded areas, liking housing estates. But London isn’t a housing estate. It’s continuous. You never know when you are on the border of one area and going into another.
What we really have to understand is not to try to set it up into bounded areas. Instead we want to try to understand what the process is that creates these patterned centres with different scales which seem so conveniently related to the ways in which we need to live. How does this pattern of centres which you see, how does that relate to the background of residential places - because most of London is residential? Most of any city is residential. In some ways, this kind of network, set into a residential background, makes the city what it is. And how do things relate to each other and to London has a whole.
Now these questions have always seemed to be rather mysterious. Cities grow by a gradual accumulation of buildings which makes a network of space which we call the street network, if you like. But somehow this network comes to life and creates a system of differentiated but connected spaces, the quiet roads where we live and busier bits are the different scales. It makes a network of centres and this is what we mean by a good city, I think.
For most of the twentieth century planners and architects tried to design a network of places but they had very mixed successes, as you can see. We can easily say what we don’t like about the 1960s development on the other side of the river. Can we say exactly why it fails? I think we can now. Because of what happened in the last 20 years in development we have, if you like, a much better understanding of materials.
But for much of the twentieth century it was people guessing about these issues and trying to do things which sounded good, had good words attached to them. That was because people didn’t understand what they liked about good cities and they were trying to create a process of self-organisation, living, making things happen, things that are changing all the time. But nothing happens all at once and nothing stays the same. It is a process that we have to understand and take care not to do the same as earlier designers again.
In the last two decades of the 20th century with high speed computers, our understanding all changed. We can now produce a network of every second of every street in the whole of London within the M25. It’s 285,000 street spaces and what we do with the computer is to model these and begin to experiment.
This is a kind of collection of theories which we call Space Syntax. You may have come across Space Syntax as a way of forecasting and movement in developments and this is one thing that it is useful for. But it’s actually much more than that. It’s a tool for understanding how to be self-organised and how we can use this knowledge in generic design, and I think this is what we need here. There’s a lot of trial of error, and trying to understand the dynamics, if you like, of what it is about this method that seems to come to life.
I think because we’ve made a large number of studies, in a large number of cities, we now understand both generic things which we like about how cities work, but because we live in London we’ve given a lot of attention of how this one works and what is so peculiar and so distinct about it.
So how is it that cities evolve and become what they are, various networks of differently scaled centres and sub-centres, set into a background of residential space, that is essentially what they are, and how does London acquire its peculiar version of this we call network of villages? With this understanding we can be much more precise about why Archway has huge potential, but it’s a near miss at the moment and it can be a heck of a lot better.
So how do cities evolve in this kind of way? By far the most important thing that has come to light as a result of using computers to study very complex things like cities is one very, very important discovery. In the past we’ve thought that to understand how people moved in cities we had to understand where you put the elements – the tube station, the shops. But what we know now is that the prime determinant of the movement though an area is actually the structure of the street itself.
It’s obvious if you think about it. When you’re crossing London there will be a certain route that will be available to you and if you imagine people going from everywhere to everywhere else it’s obvious that certain streets are going to get much more use than others. This is the kind of things we can now analyse. We can give this to a computer and what you can see is an analysis of the whole of London within the M25 colour coded for each segment of the street between junctions, worked on a very, very small scale, coloured from red to blue according to its potential for through traffic, for people going from everywhere to everywhere else in London.
This is doing something we couldn’t conceivably do without a computer. It’s imaging people moving everywhere and showing, if you like, the place in terms of the potential of every street scene for carrying movement.
If you think about it a movement involves two things: you decided where you are going to go - you select a destination - and then maybe select the space you are going to pass through on the way there.
If you are an ordinary human being most of your destinations will be not far away. We go to the local shop than you go to see your Aunt Hilda in Willesden. But some streets and different segments of streets are much more interceptable than others so much more available as destinations. And they are also different in the sense that much more movement passes through some streets than others. So what we can do is we can build a picture, if you like, of the movement potential of whole cities on this kind of scale in quite an effective way.
What happens when we check maps which are purely analysis of the street network, against actual movement we find the accuracy with which a computer predicts the movement is 65-85%. In other words, most of the pattern is determined in some way by the layout.
This is really beginning to give us a clue about what we need to understand. So let me just come down a scale and look at something a bit more close to home. What you find is that in the areas of greatest predicted movement are the centres like Hampstead, Belsize Park, South End Green, Belsize Village, Englands Lane, West Hampstead. It is not a perfect match but you do find this just by analysing the map and can begin to see what is happening, this time in a very micro way.
The first map covers trips of any length at all across London. The second takes only the trips up to 750 metres. And if we begin to understand the local patterns of movement rather than the larger scale we can begin to pick out and isolate these amazing patterns which are created by the structure space itself rather than anything else.
How does this come about? In fact the process by which cities self-organise is very, very simple. As we accumulate buildings we extend, if you like, the network of streets and this creates movement potential everywhere, and that grows. So, as the city evolves it creates different potentialities for people being in different places, if you like, potentialities to result in movement.
Now this has got two aspects to it. There’s what you call the local factor which generates the highly localised movement. Then there’s the position in the larger scheme of things and in Archway you’re on an intersection of very, very important routes. That’s not how the small scale centres tend to be. They tend to be much more sensitive to much smaller scale and if you have small blocks of buildings near where you are you’re actually accessible to a lot more people than if you have very large blocks. A very important element in the formation of a centre is this scaling of blocks.
What happens, of course, as the grid is producing movement in one place rather than another, land use alters in response and uses like retail go to those locations with more movement. It’s not that people move towards the attractions but that the attractions become located where the movement is going to go anyway. Once you’ve got shops there, then that attracts more movement and that in turn attracts more and more diverse uses. And by a kind of multiplier effect you see that the centres begin to develop in response to the scaling, if you like, of the local urban grid. You get a small one here, a bigger one there, and of course they compete with each other and relate to each other in terms of scaling and dynamics.
So what you can see if you like is a land use pattern which evolves naturally from the fact that some things seems to be where there’s going to be movement, where it’s busy. Other things, like where we’re going to live, for example, we prefer to be a little quieter. So there is a natural differentiation of this kind of pattern, built over a long time. Northwest London formed into this kind of marvellous pattern which created all these wonderful little local centres. And this is, of course, how London became a city of suburbs and villages.
Now we can use this as a kind of diagnostic and ask why is that High Street is on the twisty bit? Do people like the twisty spaces? Well no. It turns out that if you analyse this within a radius of 1,250 metres you get this kind of scale of centres. This analysis highlights the main part of Marylebone High Street and you begin to detect the centres of the scale, if you like, of Marylebone High Street. London is full of these marvellous little high streets.
Why is Conduit Street successful? It turns out that the much smaller scale is very much protected by the small scaling of the blocks here. This is a 250 metre catchment area centre. So we are beginning to see it’s just purely how the space was organised and how people moved in this area which gave rise to this wonderful little live centre here.
Why are all the best shops in the City of London in that wonderful little Bow Lane that is hardly wider than an alley? It’s because at 200 metres again it’s the most powerful space. It’s more powerful than Cheapside. Leadenhall Market again shows a very important principle. As centres grow they become more important and they go through what we call grid intensification. They can break up the grid to allow accessibility from all points in London.
Shopping tends to start in a linear pattern and then it grows into a kind of two dimensional pattern and when it does happen it’s marvellous. So St Christopher’s Place congregates behind Selfridges in London or the Carnaby Street area behind Regent Street. And this tool allows us to understand how things develop.
If you take the Edgware Road for example, from the North Circular Road to Oxford Street, you’ve got three High Streets essentially. They occur where the grid intensified. In between the activity fades away. In other words the large scale and small scale facts are beginning to interact with each other to show this.
What is happening with Archway? Well if you take my first map of London and look at it for trips of any length it prioritises the different streets. As you begin to reduce the scale of the map and look at the potential within 2 kilometres you can see that all kinds of interesting structures are beginning to appear. At 750 metres little things begin to appear like the little shopping centre further down Junction Road for example.
What this is telling us is that Archway has got the big stuff and you’ve got some of the small stuff but away from Archway you’ve got some very interesting small stuff, and very interesting small centres.
The central site with the tube station and the swimming pool should by now have become diversified into a grid intensified area, better connected into its apparent residential area and better linked to the large scale.
The kind of thing we should be looking at is this remarkable little route through the Girdlestone Estate. It is a twisted little route but remains powerful all the way through because you can actually walk through here from Dartmouth Park Hill to Archway Centre. That could be a marvellous connection and it opens up into a really nice green space which has real potential. It shouldn’t be a damned fool twisted thing at all and it’s one of the things you have to take into account if you want to animate this part of the space here.
So it begins to become clear that the process of Archway growing into a good place has largely been interrupted by planning and design, by making the mistakes we’ve made. What Archway gives to London as a landmark has been lost.
The future of London is about distributed nodes of work. It’s not about pumping more people into the centre. It is about finding new ways of moving round the City, because the transport routes into the City are already overloaded and the infrastructure is falling over. The capacity is for moving people sideways and Archway is one of those nodes which could benefit from that.