Understanding Cities - Part 1

We've talked about cities on many occasions on this podcast. We've had Jeff Weiss talk about cities and we've had Luis Picncourt talk about cities as well. In this episode and the one that follows, this is a two-paradur. We're going to dig much more into the Malamathics of cities and how we model cities. To help us do that, we are joined by Michael Barley, professor of planning at University College London. Part one is going to be all about how we've talked about cities in the past and it's also going to be toolkit of complexity concepts that you need to better understand cities and frankly that you need to understand complex systems in general. Many of these you'll heard about before you've heard about emergence, you've heard about fractals, you've heard about positive feedback. Then in part two we're going to jump into the various models they have of cities and how we can use them to better make decisions for the cities of the future. So what I forderadoo, let's go to Michael. This is simplifying complexity. A podcast where we explore the underlying principles of complex systems. Systems that seem to defy our rational view of the world, like economies, ecologies, or even you or me. I'm forensic engineer Sean Brady and I'll be your host. Michael Barley, welcome on the show. Hi Sean, nice to be here. Can we start with how we use to think about cities before we had a complexity based approach to it? Well in fact if you go back far enough, if you go back before the industry revolution for example then most people thought about cities and so far as people thought about cities at all, people thought about them as visual objects in some sense. We can see the city, our immediate experiences of everything out there in that sense and that was the dominant way and people who were skilled in visualization, architects, artists and so on really developed ideas about cities. And the ideas were fairly ordinary, common sense basically, the idea that you know the city should be as beautiful as possible, it should be efficient and so on. So these were the ways we thought about it and then gradually as the industrial revolution sort of began and wore on, then essentially the ideas from the early industrial revolution, the ideas of mechanics if you like, the ideas that we could explain how things worked really, physical things basically with push and pull forces etc. These began to be translated into the social sciences from the physical sciences to some extent and indeed the social sciences and the empties on the built environment. So what emerged really in the early to middle 20th century that's about a hundred years or so ago, what ideas about what are called the systems approach and the idea was that we could see a city as an organized system and we could explain it really from the top down and really the so-called systems approach really dominated in the mid 20th century. Of course a lot of other things going on that we'll talk about such as the development of computers and so on. But generally speaking, prior to complexity theory it was the idea that we could see the city as a machine basically in that sense. A negative feedback was a huge part of that model, is that a fair statement? Yeah, very fair in the sense that the internal combustion engine, the idea that you could keep the engine working within certain limits etc. Negative feedback was the way of thinking about those limits in that sense and the idea would be that we'd keep the system on course, we'd keep it behaving as it was designed to behave basically and if it departed from that this would lead to a negative feedback which would lead to a correction in the course trajectory. A good deal of where we see modern mechanics, contemporary mechanics today is really all about negative feedback basically when we drive a car, you know, any kind of vehicular movement or motion in that sense really depends on negative feedback and that was the notion, the idea that if the system and in the case of a city, if the city was moving off course it was beginning to gentrify or segregate in undesirable ways then the view was that these trends should be corrected in some way and the plan basically the idea of the city plan was really a mechanism for correcting a system that was moving off course in this sense that the trajectory was departing from what was desired really in that way. So can you give me an example Michael of some of this negative feedback in practice that's a planner at that stage would have done to try to bring the system by equilibrium? Well the obvious example is traffic to some extent. Traffic has its own sort of equilibrium level and generally speaking traffic can really go off course you can get extreme congestion which actually clogs and the various mechanisms which are put in place by traffic engineers which actually control the level of congestion in that sense. So traffic lights themselves really where they're imposed are elements that enable you to feedback the volume. Indeed there are increasingly a variety of mechanisms in the built environment that has sensed the amount of traffic and make corrections basically we now have big control centers for traffic where you've got a mixture of sensors in the environment telling you how much traffic there is and whether there is congestion and so on cameras etc. Most big cities have control centers of this kind which is a mixture of mechanical devices to control traffic in some sense and a variety of human responses because everything can't be automated in that sense. So that's a typical example of negative feedback in that sense. And what changed the thinking before we get into some of these concepts? For what fundamentally took that view away from that purely mechanical view or purely predictive view? It was largely due to the fact that when you look at the complexity of cities, when you look at how complex they are I should say how complicated in that way we could get into an argument about what complexity means and what complicated means so on but in other words as we began to scrutinize cities more and this was really from the mid 20th century on. It became apparent that relatively tractable ideas about negative feedback and so on and the organization of the city into different systems and subsystems and so on. These basic ideas were not really up to the task of explaining what was happening in particular. When you look at a city it's changing all the time it's never in equilibrium. It's always changing and we as individuals are making changes all the time to the city and the idea that there was a grand controller in the sky if you like a hidden hand I suppose in some sense sort of controlling everything was really a figment of our own imaginations in a sense and really in a way cities actually evolved from the bottom up not the top down which the systems approach implied that you organized things from the top down in a hierarchy and so on and it was quite clear that cities didn't actually work like that. In other words there were lots of things happening in cities which were unanticipated largely because they merged really from the bottom up in that sense. So this thinking began to permeate people who were thinking about how we think about cities basically but it wasn't just academics who were thinking about it it was people in practice. People found that the rudimentary tools that were being developed to deal with things like negative feedback and so on to deal with things such as population growth urban sprawl etc. These tools were really very simple minded and we need to move beyond these tools and a new approach was needed. So in other words lots of things that were put in place on the assumption that we could correct and steer the city and produce a more livable and sustainable environment just weren't working basically in that sense. You can see it everywhere today this basic problem the housing affordability crisis how do we deal with migration, immigration etc. large waves of population moving which we have no control over. A whole range of things like that they're changing the nature of the city. So it was this kind of context that really led to people thinking about the fact that cities like other social systems were really complex systems really in that sense and that the basic ideas that we had were misplaced. I mean in some senses negative feedback really turned into the idea of positive feedback basically that in other words people are doing things from the bottom up in cities that are actually changing those cities in that sense and putting them on a different kind of trajectory really which is perhaps an unanticipated one. So that was really the basic idea that was just switched really from it's a shorthand the description really from negative feedback towards positive feedback in that sense but it's bigger than that as we'll see basically in terms of these podcasts for example that the whole idea of complexity is bigger than positive feedback let's say. Fascinating just hearing you talk about a Michael because we had Brian Arter in the show and he told an almost similar story from the perspective of economics that it's an equilibrium system and the negative feedback is there and everyone always push you back to a equilibrium and but it was essentially positive feedback and built and built and built and pushed you forward away from equilibrium into a new way of thinking about things and fascinating to see the same thing here. So that's positive feedback. What about emergence or the idea of emergence come from a city perspective. In one sense, emergencies contained in what we actually see in cities are difficult to explain. So, for example, segregation of different activities in cities, different land uses, different income groups and so on. Cegregation, in some sense, is a phenomenon that is emergent and a good example of segregation would be gentrification. The gentrification is based on the idea that cities go through different cycles of development and in the 19th century, for example, there was very rapid growth in Victorian, in Western cities, Victorian housing, basically, in that sense. And then basically, as people got wealthier and wealth spread out, basically, then those houses were vacated and they were turned into multiple dwellings and so on in that sense. So, they basically degraded. And then, of course, in the mid-20th century, many of those, and you see this, of course, quite classically in Australia, in places like Melbourne and so on, those older houses basically became gentrified. And it is process of growth and decline in the housing stock and gentrification and redevelopment and so on. Nobody planned it in that sense. It emerges, really, from the bottom up. Individuals begin to realise there are gains to be made if you want to live in a nice Victorian terrace and so on. You probably need to renovate it, basically, this was the thinking and this was what was actually happening. And really, what emerged, as it were, were newly segregated districts. And if you look at the city over time, you see a continual change in the character of different land uses in the sense from the centre. The central cities, in lots of central parts of cities, in lots of big cities, world cities, are very valuable, basically now, in terms of the cost and price of housing. And there is a slow move, really, of poorer groups to the edge of cities, whereas 100 years ago, the richer groups were moving to the edge of cities to get more space and so on. So, there is this sort of towing and throwing in terms of the dynamics of cities in that particular context. So, emergence, really, relates to phenomenon that we don't really anticipate. We can always understand it after it's happened in hindsight, but we don't really anticipate it in that sense. And there are lots of things in that sense. I mean, if we look at immigration in cities at the present time, we look at Britain with immigration is about 4 million people have moved into Britain in the last 10 years, basically. That's quite a big change in numbers of people. Were those people are going? We don't really have a kind of clear view, basically. They are changing the nature of cities in that sense. And not only are they checking that the immigration is changing in the nature of cities, it's changing people who are already living in cities. So, they're quite complicated sort of repercussions really in that sense. And that's true. It's not just immigration and population, it's true of industries, it's true of all sorts of things really in that sense. So, what about path dependence? Now, path dependence. Yeah. If we talk about emergence in the sense of the idea of something emerging, then it will follow a trajectory, we would say, which is a path really through time in that sense. And one of the other features I should say about complexity theory is that it's brought time onto the agenda. You remember I said a few moments ago that the systems approach thought of cities as being in some kind of equilibrium. Well, of course, once this concept was really turned on its head, we began to realise that cities are never in equilibrium, basically. Then we need to chart the trajectory of change in that sense. And the trajectory of change is often referred to the historical path. Path dependence means that if you get a change in some characteristic of the city, the change itself will lead to other changes. In other words, when we look at the city and how it's changed, we'll see that it's dependent upon, if you like, historical accident in that sense. If an accident takes place, basically, in a sense, it establishes a new path in the city. And a good deal of when we look at cities, we can't really explain, because we don't know the genesis of the original change or the original accident does as actually happened in that sense. So what we're really saying is, there's an element of randomness that we can't explain. You can explain everything if you have very, very detailed data on it and scrutinise it. I mean, there are limits to our explanation of course. But basically, the explanation itself is contained in this notion of path dependence, how an accident sets off a particular movement or change in the city, basically, that really depends on the original accident or the original change. So the dependence is based on this idea that history matters really in a sense. Have you got a good example? A weird example. I always enjoy weird examples of this sort of thing. I think a good deal of the history of how cities have grown and spread. It contains this idea of path dependence. I mean, I think in the 19th century, when the industrial city began, the industrial city was clearly where people were flooding in from the rural hinterland, changing their occupations to agriculture to industry in that sense. A lot of the growth of cities then was unanticipated basically in a sense. And then of course, when cities reached a certain level, people began to rather than be attracted to the city. So actually, working the city and then move out to the edges. So you've got the emergence of suburban sprawl. These general waves of phenomena were sort of unanticipated. We can now explain them, but at the time they were unanticipated. So, for example, we don't know what the future city is going to look like. We can speculate, basically, in that sense. I don't know what London is going to look like in a hundred years time, for example. I mean, if you look at a hundred years ago, then the structure of the city is fairly similar to what it is today. A lot more movement, basically, it's a sort of massive melting pot in terms of movement now compared to what it was a hundred years ago. But the way it spreads out will it continue to spread out. We might reach a situation with the emergence of new information technologies, for example, where a large, a large and larger proportion of people are working at distances away from a notional centre of work in that context. There's increasingly people might behave in that way. So, the original ideas of the city attracting industry and attracting wealth, the agglomeration economy, so-called, for the city, may change quite dramatically in the next hundred years in that sense. Certainly, we know that the growth of cities is everybody we're believing in cities within the next hundred years and so on. And the total world population will probably be, well, it looks like it's sort of becoming stable in some sense. But these are simply speculations. If you look at those sort of speculations that people have made over the last hundred years, they've always been wrong, basically, in that sense. So, the sobering issue is that we don't know, and that's why cities are complex. So, then going on to a daily that I think is really tricky, particularly a via layer person, and one of those areas that if you understand it well, it becomes so self-evident, but fractals. And, I mean, obviously, we're going to have to talk about self-similarity and get on to power laws and scaling, which our listeners will have heard a little bit before. But let's start with fractals. When we look at nature, we don't see the kind of regularity that we see in mathematics or geometry, in other words, that nature is geometrical, but it's not geometrical in the sense of, we would say, Euclidean geometry. You know, the world is not composed of objects based on straight lines, basically, in that sense. That's what we're saying. There is a degree of regularity everywhere. If you look at a razor blade underneath a magnifying glass, basically, then you can see almost a serrated edge. If you scaled it up, that serrated edge would look as though you'd never think if you scaled it up that you'd be able to actually shave with that particular edge. You would cut you in this sense. So, in other words, the idea that the world is regular in a geometrical sense is quite problematic, but it is regular in another sense and in our X sense. And the fact that an object is irregular in one sense, but also has a degree of order, the irregularity is ordered in some way, is essentially the idea of a fractal. It's only when we begin to measure that irregularity, basically, that we introduce the idea of fractals, basically, in that sense. And essentially, any object which has scale of this kind, where you can scale it and you can see the same structure emerging at different scales. So, for example, at the top of the hierarchy, we've got a center, which it depends on, let's say, four or five big district centers in the city, basically, an element of hierarchy. But from the district centers, we've got little centers or smaller centers that depend on those district centers. In that sense, we have this kind of hierarchy and the organization, how many centers we have at these different scales, really is reflected in the fact that we call this kind of object a fractal, basically, in that sense. So a fractal, the best example of a fractal is a tree. And the hierarchy I've just sketched is a tree like structure. But if you look at the branches of a tree, that is the very best example of a fractal you can get. You have a thick trunk, the trunk then branches into two, three or whatever, and minor trunk. basically and so on and the tree reaches out. In fact there's another element in fact in this which is the energy that's being passed to the branches really from the stem of the tree from the bottom of the tree. And in some sense the reason why the tree looks outwards into the atmosphere around it basically it's searching for food, it's searching for light basically the tree. And you can see the same thing in terms of the roots of a tree looking for nutrient in the soil and so on beneath it. So a good example of a fractal is a tree and the critical issue in terms of a fractal is we can measure it in particular ways. It does a new sort of algebra or geometry that really relates to how we measure the dependence of these branches, this hierarchy branching on the whole thing. That's in essence a fractal basically. So it's fundamentally it's a way of looking at a network structure as you say like a tree but putting the mat behind it so that it's got relationships between the different sizes of the branches and all those bits together and then when you have that mat presumably you can start then applying it to fractal like network type structures in real life like a city. Well I explain is both the network structure and the sizes. If we look at the network itself we think of a tree as a network basically. If you look at the network essentially we have some big branches and we have a lot of little branches basically and the relationship between the number of branches at these different scales is often follows a relatively simple law which is a power law for example. Again the very best example of a power law is to look at the number of cities. If you look at the number of cities saying Britain for example then we have one very large city London we have about 203 million plus cities like Manchester, Birmingham, Liverpool and so on and then the next level down we've got like maybe 20 cities which are the size of Sheffield and places like that and then all the way down the hierarchy. So we have a regular number of cities at different levels of the hierarchy and we can explain that using what is called a power law and the power law the term power really relates to the structure of the equation that describes this relationship it's not specific to fractals. A power is just a value in an equation that scales it to what we see in that sense but in other words the power law itself the nature of the equation the fact that it operates at these different levels is best explained using the idea of the power law basically in that sense. The power law explains this relationship between the number of the bigger and the smaller and all this branching and where does networks come into this and I mean I understand that you can account of one way or the other at this is it fractals close networks that give you the power law is that an over simplistic way of saying that or is it? If you take the network and you look at the elements of the network okay if you think of a situation where we might have three or four different town centers within a small region and we look at the links in those town centers to the neighborhood centers around each individual town center so we can think of these three or four different town centers we can think of little town centers each surrounding them in that sense so we can think of basically a series of networks in that particular context and as the network grows in extent then you get more and more smaller little networks the twigs if you like of the tree the twigs basically and at every stage from the entire network the center of the network to the edge of the network of the twigs basically you can actually count the number of elements and accounting the number of elements simply gives you a relationship called the power law now you can do that not just for networks but you can do it for entire cities you can do it for lots of things that depend on what we call constrained growth in that sense so for example if we open ourselves up and look at our vain structure or look at our long structure basically the lung is the excellent example it looks like a tree basically because it's actually pumping blood and air to the rest of the body it's a kind of set of networks that really lead to how we maintain ourselves in that sense and indeed the networks in cities are exactly that the ways of actually maintaining the city they're the energy where the energy pulses out the fact that we drive our cars every day and where we drive in the song really relates to how the city actually works how we actually produce things how we live in song in this context so fractals in this sense simply represents the well first of all the fractal represents the fact that you do have this what is called self similarity that's an important concept the idea is that the network looks the same at whatever scale you're looking at right so if you look at the ramification of the networks big cities basically and you scale in and you look at a a local bit of the network the local bit of the network has the same generic structure as the entire big network itself and at every level of the hierarchy you've got networks which reflect themselves spatially they're self similar in other words if you take the little network and scale it up simply scale it up to the big network then it looks similar it's not identical but it's similar in that sense so everywhere we look you have these networks which is similar say in the human body an excellent example is if you have a very strong light and you pick up your hand and you shine a strong light through your hand you can see the the network of how blood is distributed to the fingers and all of that kind indeed the very notion of fingers in some sense and different the way our limbs are organized is a kind of generic fractal in a sense that it's not quite a fractal basically because we don't have sort of finger we don't have hands on the end of fingers and so on in that context but there is an element basically of fractality and the whole process of growth both of the human of the city of the society any kind of growth phenomenon has got this idea of self similarity is basically using this module to actually reach out and to spread energy in the most efficient way in that sense and what we see in terms of cities is that we see the same shapes at different scales another example would be to take a small town of 50,000 people and to see whether you could actually scale it and map it into a big city of let's say five million basically well of course all one saying here is can you actually scale it in such a way that it looks similar in some sense you'd have to add things because a small city's doesn't have the number of components a big city does basically but that's essentially it scaling one thing into another across the spatial scales I mean said that the whole all of this all of these ideas related to temporal scales scales over time I'm not going to get into that because there's been not much done in cities on temporal scaling what is it I know you don't want to get into it but what is it well temporal scaling is the fact that a good example is in terms of the stock market if you look at the movement of stocks over over let's say a week then they replicate in some sense the same movement of stocks over over a day or over an hour basically in that sense in other words you have the same pattern the patterns are often look very random basically in that sense which is in the nature of the stock market but you can have random phenomena which is also fractal so you've got the same kind of thing appearing in time so we'll end part one there join us in part two we put all this together to show how we model cities thanks for listening to simplifying complexity when we look at the key concepts of complexity science with expert minds from across the world concepts like emergence self-organization adaptation networks scaling tipping points and much more this podcast was produced by Brady Heward and Wevelyn Creative to make sure you don't miss an episode be sure to subscribe to or follow the show in your podcast app I'm Sean Brady and I'll see you in our next episode [BLANK_AUDIO]