UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The synoptic and dynamic paradigms of city planning : re-interpreting planning methods through Newtonian… Heap, Nicholas Ian 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1997-0303.pdf [ 10.84MB ]
Metadata
JSON: 831-1.0087744.json
JSON-LD: 831-1.0087744-ld.json
RDF/XML (Pretty): 831-1.0087744-rdf.xml
RDF/JSON: 831-1.0087744-rdf.json
Turtle: 831-1.0087744-turtle.txt
N-Triples: 831-1.0087744-rdf-ntriples.txt
Original Record: 831-1.0087744-source.json
Full Text
831-1.0087744-fulltext.txt
Citation
831-1.0087744.ris

Full Text

T H E S Y N O P T I C A N D D Y N A M I C P A R A D I G M S O F C I T Y P L A N N I N G : RE-INTERPRETING PLANNING METHODS THROUGH NEWTONIAN PHYSICS AND CHAOS THEORY by Nicholas Ian Heap BA, The University of British Columbia, 1988 A THESIS SUBMITTED I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F ARTS ( P L A N N I N G ) in T H E F A C U L T Y O F G R A D U A T E STUDIES (School of Community and Regional Planning) We accept this thesis as conforming to the required standard J T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A April 1997 © Nicholas Ian Heap, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Denartment of c-<='—— ^ The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT The goal of city planning is the ordering of the city. Modern city planners have adopted a particular paradigm of order, originally developed during the Industrial Revolution when the profession of city planning came into existence. This "synoptic" planning method replaced traditional views of order with a new world-view which stressed a 'scientific' understanding of natural ordering processes. Because the natural ordering processes described by Newtonian physics were the only ones known to Victorian science, city planners have subsequently understood scientifically-valid 'order' to be limited to that produced by "Newtonian systems". The characteristics of Newtonian systems are examined in the thesis, and are related to specific aims and assumptions of synoptic planning as revealed in examples of theory and practice. Since the late 1950's, many practical and theoretical shortcomings of the synoptic planning method have become apparent. While there have been many attempts to reform the practice of synoptic planning, its fundamental reliance upon the axioms of linear dynamical systems has barely been acknowledged, let alone questioned. As a result, none of the reforms suggested in planning methods to date have managed to resolve the profession's current crisis of faith. However, recent scientific discoveries have been made regarding a second type of natural ordering processes, popularly termed "chaos", and referred to as "Lorenzian systems" within the thesis. Given that Lorenzian systems are order-creating processes, and that city planning seeks to promote order, this thesis argues that in addition to "synoptic" planning, there could additionally be a method of "dynamic" planning based upon the characteristics of Lorenzian systems. Consequently, the characteristics of Lorenzian systems are also explored, and their axioms are extrapolated to create a hypothetical i i method of "dynamic" planning. Independent precedents in planning theory and practice which accord with the aims and assumptions derived for this hypothetical method are employed to demonstrate the plausibility of dynamic planning. Because dynamic planning may well prove similarly ineffective in important areas of city planning, the thesis concludes that dynamic planning should be seen as a useful adjunct to, but not a replacement for, synoptic planning. in TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES vi LIST OF FIGURES vii PREFACE viii DEDICATION xi EPIGRAPH xii I N T R O D U C T I O N l 0 . 1 . A I M S A N D ASSUMPTIONS OF THE THESIS 1 0 . 2 . C H A O S BY A N O T H E R N A M E 5 0 . 3 . PARADIGMS A N D T R U T H 8 0 . 4 . O U T L I N E O F T H E T H E S I S 1 0 C H A P T E R O N E : O R D E R A N D P L A N N I N G 13 1.1 . W H A T I S O R D E R ? 1 3 1.2. N O R M A T I V E THEORIES OF C m ' ORDER 1 5 1.2.1. The cosmic city 16 1.2.2. The mechanical city 18 1.2.3. The organic city. 21 1.3. SCIENTIFIC O R D E R 2 6 C H A P T E R T W O : N E W T O N I A N A N D L O R E N Z I A N S Y S T E M S 3 0 2 . 1 . T H E ORDERS OF THE UNIVERSE 3 0 2 . 2 . N E W T O N I A N SYSTEMS.. 3 1 2.2.1. Deterministic 35 2.2.2. Linear or parabolic. 36 2.2.3. Size of effect proportional to perturbation 38 2.2.4. Constant or periodic equilibrium 38 2.2.5. Complexity of phenomena is proportional to complexity of system 41 2.2.6. Graphic representation: geometric 42 2.2.7. Accurate long-term prediction and control possible 43 2 . 3 . LORENZIAN SYSTEMS 4 5 2.3.1. Deterministic 49 2.3.2. Size of effect dis-proportional to size of perturbation 49 2.3.3. Non-linear 50 2.3.4. Equilibrium constantly changes. 52. 2.3.5. Graphic representation: fractal. 53 2.3.6. Complexity of phenomena dis-proportional to complexity of system 56 2.3.7. Accurate long-term prediction and control impossible 56 2 . 4 . C O N C L U S I O N 5 9 iv C H A P T E R T H R E E : F R O M CRISIS M A N A G E M E N T T O I N T E L L E C T U A L CRISIS 62 3 . 1 . A BRIEF HISTORY OF M O D E R N CITY PLANNING 6 2 3.1.1. The origins of modern city planning. 63 3.1.2. Planning becomes a profession 65 3.1.3. The ascendancy of synoptic planning 71 3 . 2 . T H E CRISIS OF FAITH IN SYNOPTIC PLANNING 7 5 3.2.1. The synoptic diaspora 8 0 3 . 3 . P L A N N I N G THEORY A N D C H A O S THEORY 8 3 C H A P T E R F O U R : N E W T O N I A N S Y S T E M S A N D S Y N O P T I C P L A N N I N G 89 4 . 1 . INTRODUCTION 8 9 4 . 2 . T H E A I M S A N D ASSUMPTIONS OF SYNOPTIC PLANNING 9 1 4.2.1. Planning policies determine city development 91 4.2.2. Specific future conditions of city can be determined 92 4.2.3. Newtonian order results if 'random' urban phenomena are controlled. 95 4.2.4. Complex by-laws are required to control urban complexity 98 4.2.5. Large-scale processes have greatest influence over city development. 102 4.2.6. Equilibrium is constant 105 4.2.7. Geometry is a sign of order 109 4 . 3 . C O N C L U S I O N 1 1 4 C H A P T E R F I V E : L O R E N Z I A N S Y S T E M S A N D D Y N A M I C P L A N N I N G 116 5 . 1 . INTRODUCTION 1 1 6 5.1.1. Christopher Alexander 117 5.1.2. Dynamic Planning and its Precursors 119 5 . 2 . T H E A I M S A N D ASSUMPTIONS OF D Y N A M I C PLANNING 1 2 0 5.2.1. Planning policies shape city's development 121 5.2.2. All scales of processes have influence over city development 121 5.2.3. Future conditions of city can be bounded. 125 5.2.4. Lorenzian order naturally produced by urban dynamical system 130 5.2.5. Simple by-law structure appropriate to stimulate urban complexity. 132 5.2.6. Equilibrium is constantly shifting. 136 5.2.7. Fractals are a sign of order; randomness need not indicate disorder. 140 5 . 3 . C O N C L U S I O N 1 4 5 C H A P T E R S I X : O R D E R S A N D P L A N N I N G 147 6 . 1 . L O O K I N G B A C K W A R D S 1 4 7 6 . 2 . T H E ORDERS O F T H E CITY 1 4 9 6.2.1. A tale of two orders 151 6 . 3 . A FINAL N O T E 1 5 3 BIBLIOGRAPHY 1 5 5 v LIST OF TABLES Table 1: Characteristics Of Newtonian Systems 35 Table 2: Characteristics Of Lorenzian Systems 48 Table 3: Comparison Of Newtonian System Characteristics Wi th Lorenzian System Characteristics 59 Table 4: Comparison of Newtonian System Characteristics with Synoptic Planning Aims and Assumptions 90 Table 5: Comparison of Lorenzian System Characteristics with Dynamic Planning Aims and Assumptions 120 Table 6: Comparison of Synoptic Planning Aims and Assumptions with Dynamic Planning Aims and Assumptions 145 vi LIST OF FIGURES Figure 1: Taxonomic Diagram Of Planning Methods 3 Figure 2: Taxonomic Diagram Of Dynamical Systems 7 Figure 3: Idealized And Actual Glissades 32 Figure 4: Results Of Newtonian System Experiments: Time (T) Over Left-Right Position O n Slope 36 Figure 5: Results Of Newtonian System Experiment 1: Velocity (M/Sec) And Acceleration (M/Sec 2 ) Over Time 37 Figure 6: Results Of Newtonian Experiment 3: "Attractor" Of Snowboard-And-Glissade System 40 Figure 7: Idealized And Actual Mogu l Fields 46 Figure 8: Sample Trajectories Of Snowboards In Idealized M o g u l Field 47 Figure 9: Results Of Lorenzian System Experiments: "Strange Attractor" Of Snowboard-And-Mogul-Field System 51 Figure 10: The Mandelbrot Set 55 Figure 11: Aerial View Of Siena, Italy 112 v i i PREFACE This thesis is the product of several books which have profoundly influenced me. I read T. S. Kuhn's Structure of Scientific Revolutions (1970) as a university freshman, Jane Jacobs' Death and Life of Great American Cities (1961) following my graduation with a BA in History, and James Gleick's Chaos: Making a new science (1987) shortly before embarking on a tour of Europe and Asia. Influenced as I have been by these works, and lacking (as yet) any substantial experience as a practising planner, it is fair to say that this is a work of enthusiasm and theory rather than one of experience and observation. It is also a work in progress - the ideas in this thesis have changed, and have (hopefully) improved as I have gone along. There is no reason to believe that these ideas cannot be further developed and refined. The potent ideas contained in these works came together for me while I was ' in the field' as a traveller. I toured the cities of Europe with a planning student's inclinations, and specifically included some of the regions' more notable planning showpieces in my travels -the London Docklands, Edinburgh's Georgian New Town, La Defence in Paris, and the dykes and "sea works" of the Netherlands. It didn't take long before I was struck by the remarkable lifelessness of most of these developments. M y destinations were greeted with incomprehension by my fellow backpackers. " W h y would you want to see something as boring as that?", I was asked. At other times, my travels took me to medieval villages in central France and fishing towns in the Cyclades of Greece. Tourists, myself included, would come from literally halfway around the world just to walk down the streets of these towns. Even drowned under the hordes of visitors, these places retained a comfortable human scale, and exuded a sense of comfort, ease and identity. And yet these places were 'unplanned'. How could vi i i something that functioned better as a human environment than the most meticulously master-planned developments not have been designed? Surely, these thousands upon thousands of traditional settlements could not al l have been the result of fortuitous accidents? It was during my trip through Europe that I became convinced that the beautiful villages of France and Greece were indeed ordered, but structured by chaotic processes (here termed "Lorenzian processes" in honour of their discoverer), rather than Newtonian ones. I also began to draw parallels between the observations of Jane Jacobs on city form and city process with the nature of Lorenzian systems. I began work on this thesis in the fall of 1994 with a search for literature on the subject of Chaos Theory and planning. A thorough review of relevant journals, dissertations and published works turned up only three works with a direct bearing on the subject, although a significant literature on applications of Chaos Theory exists in related disciplines like urban geography, economics, ecology and business administration. I reviewed Chaos Theory proper to develop lists of the differing axioms which underlie Newtonian and Lorenzian systems. I was then able to draw specific connections between the axioms of Newtonian systems and the aims and assumptions of synoptic planning practice. Work ing from first principles, I outlined a parallel methodology of planning, derived from the different axioms underlying Lorenzian order. Convinced that planning theorists such as Camillo Sitte, Lewis Mumford , Jane Jacobs and Christopher Alexander had long-ago recognised Lorenzian order, but had lacked the terminology of the new science to express their ideas, I re-read their works, and discovered the ideas of Jacobs and Alexander to be in remarkable accord with the experimentally-derived methodology of dynamic planning. M y initial findings were organised in a paper presented at the "Order and Chaos" interdisciplinary conference in Corner Brook, Newfoundland in M a y 1996. Following this, I substantially revised and refined my conclusions. In addition, I conducted substantial research into the history of planning in order to substantiate the links between the axioms of Newtonian systems and the aims and assumptions of synoptic planning practice. M a n y people have been instrumental in enabling me to complete this thesis. I should like to thank my first advisor Dr . Peter Boothroyd, for his cheery radicalism and encouragement, and my second advisor Doug Aberley for his astute and fair-minded criticism in the face of my own less-than-objective rhetoric. I am very grateful to my fellow student Robert Barrs, with whom I have worked through many of the issues and arguments contained in this thesis, to Dr. Roland Stull of the Department of Geography at UBC, who offered his time and expertise on non-linear dynamics, and to my colleagues in the School's computer lab for their helpful feedback. Dr . Wi l l i am Rees suggested that I use the term "dynamic planning" within this thesis, a far superior choice to my own initial proposal. Michael Coyne, Lois Boone, Louise McGi l l i s and all the others involved in the "Order and Chaos" conference deserve my thanks not just for helping me complete this thesis, but also for organising a tremendously interesting, friendly and enjoyable event. A l l of these individuals have provided me with sound advice, but their best recommendations did not always make it into this work - the flaws and opinions of this work remain my handiwork alone. Finally, I would like to thank my many fellow students at the UBC School of Community and Regional Planning, my numerous co-housemates, and my one-and-only partner Sharon Paterson for reminding me that there is more to life than working on a Master's Thesis. Nicholas Ian Heap M a r c h , 1997 DEDICATED TO JANE JACOBS who unlike so many of us observes and then theorizes instead of the other way around. x i M E N O F D I S O R D E R N E E D E X P E C T N O I N D U L G E N C E F R O M M E . - "Baron" Georges-Eugene Haussmann xn INTRODUCTION 0.1. Aims and Assumptions of the Thesis This thesis is intended to aid in the rebuilding of current city planning theory: it is an attempt to introduce new principles to planning theory and practice 1 . Nevertheless, while it is meant constructively, it is possible that some city planners may find this work as uncomfortable a read as the works of two generations ago which precipitated the discipline's long-standing crisis of faith. To be sure, this thesis is polemical in nature, and argues in no uncertain terms that the fundamental assumptions about 'order' which have long informed city planning are only partially valid at best. By mistaking one long-understood variety of natural order for the whole, the profession has been led into and kept within a state of intellectual crisis. If we are to find our way out again, it is crucial that we examine these assumptions, and broaden our perceptions of 'order' so that they may more effectively correspond to the processes and patterns of the wor ld in which planners practice their profession. Consequently, much of this thesis illuminates the history of planning theory and practice in such a manner as to highlight the limited scope of these long-unexamined assumptions. The thesis notes the inability of these assumptions to properly address certain types of phenomena and processes within cities, and the failure within planning practice to achieve the goals implicit within these assumptions. Moreover, these core assumptions are common to virtually al l planners across the wide spectrum of current practice. This is true even now, after two generations of schism, protest and reform spawned by the crisis of orthodox planning in the late 1950s. Rather than emphasize the differences which 1 With apologies to Jane Jacobs (Jacobs 1961, 3). 1 separate these myriad schools of practice, this thesis emphasizes their common intellectual heritage in the Newtonian conception of natural order. Consequently, the old "comprehensive" and "new town" orthodoxies have been deliberately conflated with the contemporary "radical" and "ecological" denominations; throughout the thesis, this diaspora of practices w i l l be regarded only as variants of a single "synoptic" planning method. That said, this work does not question the necessity of the profession of city planning, nor the continuing need for synoptic planning in particular. What this thesis attacks is the assumption that one can plan scientifically onlyby means of the synoptic method. It is synoptic planning's monopoly over the ordering of cities — even in areas where it has proved ineffective or actively harmful — which is under attack, not synoptic planning itself. Nevertheless, one can only break up a monopoly by taking issue with the monopolist. Furthermore, the grouping together of virtually every modern planning practice under the common banner of "synoptic planning" should not be interpreted as a statement that there are not significant differences between them. These differences were tremendously significant to figures like the organicist urban critic Lewis Mumford . Indeed, Mumford spent half a century engaged in his own polemics against just those "comprehensive" planners which this thesis has bundled h im together with: it would no doubt surprise (and enrage) Mumford that his writings are being used to illustrate the world-view of his ideological opponents as well as himself. Whi le Mumford's spokesmanship for the "comprehensive" and the "radical" planners alike may be a clear case of lumping the lamb in with the lion (or perhaps vice versa), we should also take note that taxonomists are guilty of precisely the same thing when describing the biological kinship of lifeforms with one another. Biologists do not dispute the fact that lions and lambs are different creatures, or that they are locked in eternal (one-2 sided) struggle with each other. Yet when one views the great Tree of Life, the lion and the lamb (and the human) all end up branching off the same small twig of the Class Mammalia: when seen in the large picture, their similarities far outweigh their differences. In like fashion, the overall free of Planning Practice is displayed below in Figure 1: Figure 1: Taxonomic Diagram of Planning Methods (traditional) chaotic organic mechanical cosmic "synoptic Lorenzian systems1 Newtonian' systems A ''unplanned' human settlements Lewis Mumford , and the vast majority of other planning theorists have concentrated their energies upon the real differences between the organic and mechanical normative theories of planning, a focus corresponding to the line labelled " A " in Figure 1. Our own task in this thesis is to examine the differences between the synoptic and dynamic methods of planning, a level of analysis corresponding to the line labelled " B " . It is not that the persepctive on planning theory available at Level " B " is inherently superior to any other. It is, however, the most effective vantage point for examining the particular issues we shall investigate in this thesis. 3 While this thesis is implicitly critical of previous planning theorists and practitioners, it must be admitted that we are now able to view the matter in a way that was simply unavailable to those who came before. While Mumford in particular comes in for some measure of criticism, Chaos Theory was only developed when he was 6 8 years old. No one should be called ignorant for failing to anticipate intellectual breakthroughs 2. Finally, planners are not the only ones who view cities — and the wor ld — through the synoptic paradigm: it is a way of seeing common to politicians and businessmen, engineers and much of the educated public. As Jane Jacobs noted in The Death and Life of Great American Cities. Bankers... have got their theories from the same intellectual sources as the planners. Bankers and government administrative officials ... do not invent planning theories.... They are enlightened nowadays and pick up their ideas from ... theoretical city plann[ers] 0acobs 1 9 6 1 , 1 2 ) . City planners are not the only ones to hold the synoptic paradigm, and they are certainly not the only ones currently unacquainted with Chaos Theory and Lorenzian systems. But if society is going to accept a second way of viewing the order of cities, then it is city planners, the acknowledged experts of this field, who must lead the way. The effect of politics, economic systems and power relations in determining the actions of planners is not addressed in this thesis. Clearly, these factors are important. Political and economic pressures rather than abstract conceptions of order are usually the immediate cause of a particular planning action, and it is these factors which have largely shaped the entire economic and political system in which we l ive 3 . Nevertheless, T. S. Kuhn's theory of "paradigms" posits that our perception of the physical wor ld about us is largely determined by the theoretical world-view we employ to make sense of it, and 2 This is, however, exactly what Jane Jacobs managed to do in The Death and Life of Great American Cities (1961). 3 For more analysis of these factors, see (amongst others) The Theory of Social and Economic Organization (Weber 1966) and Urbanization and Urban Planning in Capitalist Society (Dear and Scott 1981). 4 suggests that the world-views we hold may have a profound effect upon the way we justify and carry out our actions (Kuhn 1970, 1 -9). While a shift or broadening of intellectual paradigms could not be expected to alter the motives oi planners, politicians and plutocrats, the way they think of concepts like "order", it might well affect the way in which such people perceived and acted upon their motives. Even if this thesis succeeds in its goal of presenting a second scientific valid framework for perceiving and creating urban order, "dynamic" planning w i l l remain a set of ideas rather than a program for action simply because the crucial political factor remains unaddressed. That does not mean the ideas explored in this essay are not worth pursuing — for now — on a theoretical plane alone. 0.2. Chaos by Another Name Readers familiar with physics and Chaos Theory wi l l notice that this thesis employs the words "Newtonian" and "Lorenzian" instead of the more familiar terms of "linear", "non-linear" and "chaos". For several reasons, the terminology currently employed by dynamicists is awkward for the purposes of a city planning thesis. Amongst dynamicists (the scientists who study the behaviour and properties of deterministic systems), "chaos" is now the established term for referring to deterministic, but unpredictable processes which give rise to particular forms of order (Lorenz 1993, 20-21). Unfortunately, amongst city planners, "chaos" has long been used as a term of particular opprobrium, who reserve it for particularly horrid examples of civic disorder, anarchy and dysfunction. Anyone hoping to discuss the order of "chaotic" systems with planners has lost their argument as 5 soon as they open their mouth. If planners are to be convinced that there is a new and viable way to perceive order in the city, "chaos" simply has to be give a different name 4 . There is a second practical reason for employing a different terminology from the "linear" and "non-linear" processes referred to by dynamicists. As defined by dynamicists, the term "linear" has a very specific meaning, referring exclusively to straight lines. The term "non-linear" in turn refers to anything that is not a straight line. This latter category is, not surprisingly, broad-ranging, including everything from purely predicable geometric (or parabolic) progressions, to unpredictable "chaotic" functions. These definitions of "linear" and "non-linear" make it impossible for us to simply contrast linear with Lorenzian systems when discussing synoptic and dynamic planning. Awkwardly , both linear and 'parabolic' 5 Lorenzian systems share the same underlying characteristics, including predictability and controllability. Moreover, both of these types of systems have been familiar to scientists from before Newton's t ime 6 . Because it is the characteristics common to both of these types of functions which have shaped the aims and assumptions of synoptic planning, these parabolic Lorenzian systems w i l l be examined together with Newtonian systems for the purposes of this thesis 7. Consequently, for the purposes of this thesis Newtonian systems— named after the scientist who codified the laws describing their behaviour — w i l l refer both to linear dynamical systems, and to non-linear dynamical systems characterized by geometrical functions. Conversely, while "chaotic" systems are all non-linear, not al l non-linear dynamical systems are "chaotic": as mentioned above, they form only a subset of the larger category of 4 As far as city planning is concerned, "chaos" by virtually any other name smells far sweeter. 5 Such as the phenomena produced by a compound interest rate, or a falling object in a vacuum. 6 Galileo characterized them in his inclined plane experiments during the early 1600s (Bueche 1981, 16-19). 7 It is worth noting that "curvilinear" patterns produced by such systems can be reduced to straight line functions either through the use of logarithmic graph paper, or through the use of second-order functions: thus while a graph showing the behavior of a falling object is curvilinear in terms of velocity, it is linear, or constant, in terms of acceleration. 6 "non-linear" dynamical systems. Because it is the particular and peculiar properties of "chaotic" systems that we w i l l be examining in this thesis, we cannot simply refer to "non-linear dynamical systems" if we wish to avoid using the term "chaos". For the purposes of this thesis Lorenzian systems— named for the scientist who first investigated these types of processes, in 1963 — wi l l refer to those systems which exhibit "sensitive dependence" in most instances 8. Figure 2 below shows the relation between the different dynamical systems, and the names by which they wi l l be called within this thesis. Figure 2: Taxonomic Diagram of Dynamical Systems A l l systems dynamical (deterministic) random, stochastic (non-deterministic) linear non-linear linear geometric exponential bifurcations "l imited chaos" chaos "full chaos" random stochastic Referred to in this thesis as: not addressed Lorenzian '-/systems, Several other points in relation to chaotic systems should also be noted. As implied above, "Lorenzian systems" have been defined so as to include only those systems 8 This is Lorenz's own definition for the type of chaotic system he terms "full chaos" (Lorenz 1993, 207) 7 exhibiting "full chaos" (Lorenz 1993, 207). While the transition to chaos constitutes a fascinating part of Chaos Theory in its' own right, the subject is tangential to the aim of providing planning theory with a second paradigm of natural order, clearly distinguished from that of Newtonian systems. Consequently, for the purposes of the thesis, we shall ignore "limited chaos" systems (Lorenz 1993, 207). This thesis w i l l be based on the behaviour of Newtonian and Lorenzian systems within dissipative systems rather than non-entropic "Hamiltonian" systems, as the latter type would appear to have little relevance to cities. Finally, this thesis w i l l not address the concepts associated with "complexity". While many of the concepts arising out of "complexity" seem as applicable to cities and city planning as those from Chaos Theory, attempting to introduce two new sets of fundamental ideas is simply not feasible within a single Master's thesis. 0.3. Paradigms and Truth This thesis utilizes the concept of "paradigms" as presented in T. S. Kuhn's Structure of Scientific Revolutions (1970), and seeks to explain the actions and motivations of planners by examining the belief structures through which these professionals make sense of the world. Kuhn himself took great pains to distinguish between scientific conceptions of reality and Reality itself, and went so far as to argue that the movement from the Earth-centred Aristotelian paradigm to the heliocentric Copernican paradigm did not constitute progress towards Truth, but simply the shift to a different system better able to answer particular questions satisfactorily (Kuhn 1970, 154-173). In like fashion, this thesis does not argue that cities themselves are truly Lorenzian (or Newtonian) systems. As we shall see later in the thesis, an increasing amount of research is attempting to l ink the patterns and behaviour of cities to Lorenzian systems. Nevertheless, substantial proof of this is 8 neither available nor particularly necessary for the purposes of this thesis. Instead, this thesis argues that the paradigm of Lorenzian systems has the same potential to provide a theoretical foundation for the ordering of the city as the Newtonian paradigm has done. Indeed, adoption of the new paradigm seems to hold the potential of resolving many of the problems which currently bedevil practitioners within the old paradigm. In her 1991 business management dissertation incorporating Chaos Theory, Brenda Zimmerman adopted a similar approach: By viewing the strategic process through a lens with a markedly different language and conceptual base, it is believed that new insights into strategic management may be found. ... What is attractive about chaos theory in this regard is that the propositions it makes form a coherent pattern. M a n y of the axioms w i l l not be new to the field: the newness comes from the l inking together of these concepts into a theoretical frame which make the idiosyncrasies non-idiosyncratic. ... By looking at the things company does through a standard and a chaos theory lens, the process appears in different lights (Zimmerman 1991, 4-5). Zimmerman was also clear that Chaos Theory held the potential to enhance, rather than replace, current modes of thought: In the past few years there has been gradual acceptance of the idea that different modes of explanation , or metaphors, w i l l highlight different aspects of a phenomenon.... No one metaphor seems capable of capturing al l the complexities of organizational life (Zimmerman 1991, 11). As shall be made clear in the conclusion to this thesis, this is the same spirit in which this present thesis is intended. Ever since the 1950s, the shortcomings of synoptic planning have been evident. Even today, just about any point within a synoptically planned city w i l l fare poorly compared with the vitality, variety and human scale of a busy Medieval market town, a Greek fishing village, or an Arab medina. Nevertheless, it is foolishness to simply condemn the method outright — a moment's reflection in the heart of a metropolitan city about the conditions which would prevail in the absence of arterial roads, sewers, fresh water and power should convince anyone that synoptic planning has been responsible for 9 an incalculable amount of good. What we need is a choice of methods, so that the strengths of one w i l l compensate for the weaknesses of the other. What we need, this thesis concludes, is both dynamic and synoptic planning 9 . 0.4. Outline of the Thesis Because this thesis is concerned with paradigms of order held by city planners, the first chapter of this thesis begins with a brief exploration of what is meant by the term "order". We then,examine the three basic "normative theories" of civic order as defined by Kevin Lynch; the "cosmic", the "mechanical" and the "organic". In the course of illustrating these three metaphors of city planning, the thesis argues that the "mechanical" and the "organic" normative theories share a common origin in Newtonian science, and therefore comprise two aspects of a single "mechanical-organic" world-view. The chapter concludes by noting that science is no longer limited to Newtonian physics, and the study of Newtonian dynamical systems, but has recently recognized the existence of Lorenzian systems as well . Consequently, it would appear that a new "normative theory" is possible, based on this newly discovered type of natural ordering process. The second chapter of this thesis is a brief introduction to Newtonian and Lorenzian systems. The disparate natures of these dynamical systems are revealed by isolating the defining characteristics of both. The third chapter of the thesis returns to the subject of city planning proper. The history of the profession of city planning is explored, and the manner in which the 9 This conclusion, perhaps, violates the letter of Kuhn's own insistence that paradigms are mutually exclusive. Because shifting between one and the other requires a gestalt switch in which the same object will appear different, it is impossible to operate in two paradigms at once. True science, Kuhn argues, cannot be done unless all the scientists share a single, common paradigm. But while this thesis is centrally concerned with importing the scientific paradigms of Newtonian and Lorenzian systems (and the concept of "paradigms" themselves) into planning theory, city planning itself is not a science so much as an art. 10 "mechanical-organic" paradigm of city ordering gave rise to the "synoptic" method of city planning is highlighted. The theoretical and practical critique of synoptic planning in the late 1950s and early 1960s is then documented. Seemingly unable to abandon the impracticalities and shortcomings of synoptic planning practice without also rejecting its 'scientific' method and rationality, the thesis argues that city planning has languished ever since in a "crisis of faith". Although the recognition of Lorenzian systems as a second scientifically valid form of order process offers a way out of this conundrum, a review of extant literature shows that very little has been written to date relating Chaos Theory to city planning. The fourth chapter builds on the second and third chapter by l inking the specific aims and assumptions of synoptic planning to specific characteristics of Newtonian systems. Examples of each of the aims and assumptions of synoptic planning " i n action" are provided by quotes from planning literature, or by examples taken from planning practice. In parallel fashion to chapter four, the fifth chapter derives the aims and assumptions for a hypothetical method of dynamic planning from the characteristics of Lorenzian systems. In order to provide some indication of the practicality of this method, these aims and assumptions are illustrated by sympathetic statements within planning literature, and by precedents within planning practice and experimentation. While a surprising range of sources have "anticipated" elements of dynamic planning, most of these examples are found in the work of Jane Jacobs and Christopher Alexander. The thesis concludes with the argument that a new method of ordering the city, founded upon the scientific understanding of Lorenzian systems, would be a valuable adjunct to synoptic planning, rather than an effective replacement for it. Instead, dynamic planning offers practitioners an alternative course of action when faced with the wel l -documented failure of synoptic planning methods in certain situations. 11 Certainly, the reader w i l l realize (perhaps more than the author) that the scope of this thesis is immense. Perhaps it would have been more practical to have limited the thesis to one particular part of the task outlined above — the laying of the theoretical foundation, the erection of a basic ideological framework, or perhaps, the construction of a few concrete examples with which to furnish this new edifice. To judge by the currently underdeveloped state of this literature in this area, it would seem as though we are practically on our own in this endeavour, and would have to manufacture the whole structure by ourselves. In reality, however, this is very far from the case. The work of this thesis is not so much to scratch-build from raw materials as it is to assemble a number of pre-fabricated components together into a coherent structure. We shall make good use of the solid body of work on Chaos Theory which has come out of the physical sciences, well-crafted analyses of planning theory, and the sturdy insights of urban critics like Jane Jacobs and Christopher Alexander. It is only because the main components are already complete that assembling this thesis has been possible at al l . There can be no doubt that this work remains simplistic, rough-hewn, and lacking in detail. Nevertheless, this thesis has been submitted for defense in the hope that its basic structure w i l l prove sufficiently sturdy to remain standing upright, even against storms of criticism. 12 CHAPTER ONE: ORDER AND PLANNING 1.1. What is Order? If we are to talk of the 'order' of cities in this thesis, and moreover, of differing conceptions of order, it is crucial to get a clear understanding of what we actually mean by the term 'order'. Surprisingly, 'order' is a remarkably loose concept, despite our constant employment of the term as though its properties were self-evident. For our current purposes the relevant definitions given by the New Shorter Oxford English Dictionary iov the word "order" denote a "sequence" or "arrangement". Thus the noun "order" is defined variously as a/an: • ... [arrangement of things in which one thing follows another; sequence in space or time; succession of acts or events. Also, the way in which this occurs. • ... method by which things act or events take place in the world, society, etc.; a natural, moral, spiritual, or social system in which things proceed according to definite laws • Formal, regular or methodical arrangement of the position of the things in any area or group. Also more widely, the condition in which everything has its proper place and function • General state or condition; spec, normal, healthy, or efficient condition. ... (NSOED 1993, 2 0 1 6 ) 1 0 . A closer look at these common-sense definitions shows how fuzzy and subjective the concept of "order" actually is. They are, for instance, remarkably vague — how are we to distinguish that which is in a "general state or condition" from something which isn't? More importantly, these definitions are subjective rather than absolute, for they imply that 10 It is curious how closely, and literally, the word "order" is also bound up with notions of control, power and regimentation. Next to the definitions above in the New Shorter Oxford English Dictionary art others which refer to "rank" and "class", and those which talk of "the action of putting or keeping in order; regulation, control" and "an authoritative direction, an injunction, a mandate: a command, an instruction" (NSOED 1993,2016). In this regard, the ordered layout of monasteries — the order of orders — seems particularly fitting (Kostof 1991, 168). 13 things can only be "ordered" if the observer perceives "definite laws" in process, believes things are in their "proper place" or evaluates conditions as being "normal" and "healthy". It seems that we must look a little farther than the dictionary if we are to obtain an objectively valid definition of "order". Taking the loose and subjective nature of the concept into account, philosopher Peter Caws manages to come up with just such a definition in his paper "Order and Value in the Sciences". According to Caws, "order" is simply: [an] arrangement with respect to which / / would matter if it were otherwise (Caws 1968, 108). [Caws' italics] Under this definition it becomes clear that we can spend all day shifting the pebbles on a beach without affecting its "order" one wh i t 1 1 . The arrangement is random, and it doesn't matter if a bucketful of stones is moved from one end of the site to the other. Clearly, we could not say the same of pebbles in a carefully manicured Zen garden, nor of pebbles mounted in a display case at a mineralogical exhibit. As Caws notes, this understanding of "order" transforms the issue: The question about order, then, is not why things should be disposed in a certain way, but why it should matter that they are disposed in this way. And that transforms the question into a question about value. To say that something is ordered is to say that it embodies a value; to say that it would matter if it were otherwise is to say that it ought to be as it is. (And to say that order is preferable to disorder is tautological.) ... If something embodies value we do not wish to see it changed — it is properly ordered... Similarly, if something fails to embody a value we feel it ought to embody,... we have an impression of disorder and seek to introduce changes until the situation is corrected (Caws 1968, 106-107). It is now clear why the issue of "order" in the city has such relevance for planners. If we accept the nub of the rad ica l 1 2 planners' critique in the 1960s and 1970s — that planners inevitably hold certain values, and base their decisions upon them (Hudson 1979, 11 Unless we take the point of view of the sand fleas whose homes have been disturbed. 12 As defined in Hudson (1979). 14 390) — we should hardly find it surprising that planners have historically placed great importance upon 'ordering' the city. This then, is the fundamental premise upon which the rest of this thesis is built: the process of city planning, at least from the point of view of planners themselves, is inseparable from the process of maintaining and furthering "order" in the structure and functioning of the city. While it is not an end in itself, the "order" of a city is the discernible measure of how well or how poorly the city is informed by the values of the planners, and those of the larger society of whose behalf the planner acts. By ordering the city, the planner brings it closer to a perfect expression of a particular set of values. When the city slips into disorder, it increasingly embodies foreign, ignored or antithetical values. 1.2. Normative Theories of Ci ty Order To judge by Caws' definition about "arrangements" above, there would appear to be potential for a vast number of differing conceptions of "order" regarding the city. Fortunately for us, this does not appear to be the case: several works on comparative city form are basically agreed that despite the huge diversity of actual city forms throughout the wor ld and wor ld history, the types of civic order which planners have striven to impose can readily be classified into a few basic categories. Amongst these surveys of urban form, Kevin Lynch's typology of urban orders within his 1981 work Good City Form is particularly notable. 1 3 Lynch refers to "three normative theories" regarding the proper 13 Another classification scheme is developed by Lewis Mumford in his article "Theories and Ideals of Planning" in the 1974 Encyclopedia of Urban Planning. Mumford derives a somewhat inflated six-category typology of city planning in the Western World since the Reformation. Although his scheme is congruent with Lynch's own framework, Mumford's six categories are collapsed into three pairs of related "ideals". While his scholarship is somewhat compromised by tub-thumping for his own ideal of "organic planning", Mumford like Lynch specifically relates the values and intentions of the planners with the form of city order or patterning they created (Mumford 1974, 985-996). Note should also be taken of Spiro Kostof's beautifully illustrated work, The City Shaped (1991). While it is a valuable and wide-ranging work, Kostof is more interested in pattern than order per se, making his relatively 15 ordering of cities: the "cosmic city", the "mechanical city" and the "organic city". Each of Lynch's "group of theories... focuses on some comprehensive metaphor of what a city is and how it works" (Lynch 1981, 73). Clearly, Lynch's categories are broad and fundamental in nature but, even so, it is perhaps surprising at first that the world's diversity could be profitably shoehorned into just three categories. Nevertheless, accepting the existence of deep commonalties in the ordering principles of city-builders throughout human history becomes easier when we consider Caws' argument above. Orders are expressions of values, and the desired orders of human cities reflect the common stress that myriad societies have placed upon the values of constancy, timelessness and conformance to the laws of nature or, conversely, upon the values of progress, growth, and the efficient exploitation of the laws of nature. We shall examine the characteristics and logic of each form of city order in turn, bearing in mind Caws' distinction between the order or arrangement and the values such an order embodies. 1.2.1. The cosmic city According to Lynch, the "cosmic" normative theory of city order "asserts that the form of any permanent settlement should be a magical model of the universe and the gods" (Lynch 1981, 73). Such cities are planned as diagrams or working models of what their builders understand (or intend) the order of things to be: It is a means of l inking human beings to those vast forces and a way of stabilizing the order and harmony of the cosmos. Human life is thereby given a secure and permanent place; the universe continues its proper, sacred motions. The gods are upheld, chaos is kept off, and, not incidentally, the structure of human power — of kings and priests and nobility — is maintained (Lynch 1981, 73). relaxed framework insufficient for our present needs. In The Elusive C/(y (1986) Jonathan Barnett presents a four-fold classification scheme - "monumental", "garden city/suburb", "modern", and "megastructure" - but limits his scope to planned cities of the Western World since the Middle Ages. 16 The visual language of order employed in these different cosmic cities is remarkably similar, util izing the precision of geometry, the control of symmetry and exploitation of height and distance, all intended to make observers properly aware of the grand scheme of the universe, and their own powerless place within it (Lynch 1981, 79). There are many examples of such "cosmic" cities, particularly from the ancient civilizations of Asia: Lynch cites Beijing, Kyoto, Seoul and Madura i as examples. Lynch adds that Western civilization has also had a long tradition of cosmic city-building, highlighting the "Baroque model of ...interconnected... diverging and converging axes" because of its intended 'cosmic' function as "an expression and an instrument of power and order" (Lynch 1981, 75). He additionally discerns modern-day expressions of "cosmic" order in the capital city, the monumental axis, and the corporate skyscraper (Lynch 1981, 75). Whi le Lynch does a masterful job of summarizing the nature and characteristics of the order in such cities, he does not proceed to reveal the values that this order embodies except by inference. Nevertheless, we can reach firm conclusions based on the information presented. As Lynch notes, "cosmic" cities are truly magica l 1 4 in nature, designed to transform their inhabitants rather than to be transformed by them. More specifically, the "cosmic" city is intended to shape the values of those who live in it in accordance with the particular "laws of nature" held by its builders, varied as these have been, from the square universe of the Middle Kingdom and the mandalas of the Buddhists to Washington's tripartite separation of republican power into the Presidency, the Senate and the House of Representatives. Thus, a Buddhist "cosmic" city was intended to inculcate Buddhist values 14 I use the term "magical" in the sense of R. G. Collingwood (1938, 65-69). 17 amongst its inhabitants, just as the radial architecture of Versailles was designed to produce the conviction that Louis XIV was the French State personified 1 5 . Equally importantly, the values of tradition and constancy (or more critically, conservatism and stasis) also inform the cosmic city, because the "laws" informing "cosmic" cities are almost always held to be both true and eternally valid. The order expressed in such "cosmic" cities is "crystalline", "stable and hierarchical", denying change unless it occurred within "some rhythmical, ordered, completely unchanging cycle" (Lynch 1981,81) . 1.2.2. The mechanical city Lynch's second "normative theory" of city ordering is the "mechanical city". Whi le the "cosmic" city models the order of the unchanging universe, the mechanical city mimics the order found in machines Understandably, the modern wor ld in particular has been greatly influenced by the "mechanical city" metaphor, although Lynch notes some ancient examples of this type as w e l l 1 6 (Lynch 1981, 82-83). Rather than seeing the city as a unified whole, planners who subscribe to the "mechanical" metaphor look upon the city as a set of individual parts fitted together. Lynch notes that these urban components are conceived of as being discrete, individually insignificant, and are often assumed to be standardized or homogeneous in nature. The mechanical city is not a static creation. In contrast to the eternal 'cosmic' city, the individual parts of the 'mechanical' metropolis are fully expected to wear out and require replacement, just as a properly maintained 15 Interestingly, Kostof links Baroque forms in the Age of Discovery and afterwards to the new helio-centric understanding of the Universe (Kostof 1991, 215). 16 The machine was already well established in Classic civilizations — indeed, Lewis Mumford defines the advent of Ancient civilizations by their development of machine technology — including that of the organizational "megamachine" (Mumford 1963). 18 machine requires maintenance: "it can be taken apart, put together, reversed, its pieces replaced, and it w i l l run again." (Lynch 1981, 81). Planners view the city as ' running' more or less efficiently as a result of the arrangement of its parts. Such a city "grows by addition" (Lynch 1981, 79) in the same way that a factory might grow with the addition of a new production line. Crucially, planners who subscribe to the 'mechanical city' metaphor believe it to be as knowable and predictable as a machine is to its engineer: The whole machine can change, although it does so in some perfectly predictable way, as by moving steadily along some predetermined track. The stability is inherent in all the parts, and not in the whole" (Lynch 1981, 79). Lynch defers from pointing out the underlying values informing the order of the "mechanical city", stating only that the 'mechanical city' lacks the "wider meaning" which a "cosmic city" possesses. Nevertheless, it is clear from Lynch's characterization of such cities their order does indeed embody a particular value — that of production. More specifically, the mechanically ordered city is intended to maximize the efficiency with which the physical laws ordering the universe can be exploited for purposes of production. Lewis Mumford , who rejected the "mechanical" city produced by what he termed "utilitarian planning", left no doubt as to the values expressed in this form of city order: [C]ommercial-utilitarian planning [was] based on maximizing pecuniary returns from sale and rent. ... But the larger number of new towns or town extensions during the last three centuries have conformed to its ideal of providing the municipal facilities for opening land to the largest possible population to be housed with a minimal allowance of parks, playground spaces, or meeting spaces, and maximal opportunities for private investment (Mumford 1974, 988). 19 "Mechanical cities" don't necessarily have to produce money. Soon after the October Revolution in the Soviet Union, city plans were drawn up for industrial centres in which residential, transportation and leisure functions were laid out in assembly-line fashion, so that the worker-residents of each efficient high-rise neighborhood would be located adjacent to their positions in the Newtonian city-long steel mi l l (Khan-Magomedov 1983, 325, 336-338). Likewise, the "mechanical" order of Auschwitz is immediately recognizable, a settlement deliberately designed for the maximum production of deaths (Van Pelt 1994, 93-156). Despite the differences of these admittedly extreme (and decidedly unpleasant) examples, each exudes the mechanical city values of efficiency, of dynamism, of 'progress ' 1 7 and of 'change'. These are not the only values inherent in mechanically-ordered cities. As we noted above, the order of the "mechanical" city is not intended to force citizens into conformity with the laws of the universe as the "cosmic" city patently is. Nevertheless, the "mechanical city" bordered so that its residents w i l l be able to exploit the laws of the universe with maximum efficiency. Moreover, it is crucial to note that all-those who build cities according to the "mechanical" normative theory share a common understanding of 'efficiency', and calculate 'efficiency' using a common set of equations and understandings. Unlike the builders of "cosmic cities" who between themselves have believed in a tremendous variety of different "laws of nature", al l of the myriad believers in the ideal of the "mechanical city" equated scientific knowledge with objective truth. In particular, "mechanical" city builders identify 'Newtonian physics' or Newtonian systems with the laws of nature. It was this body of theory which gave 17 In the case of Auschwitz, only when seen from the Nazis' abhorrent anti-Semitic perspective. 20 rise to the modern notion of the mechanical universe, and laid the intellectual foundations for the technical developments of the Industrial Revolution. Because this particular understanding of laws of nature effectively determined how humanity went about exploiting them for productive purposes, the characteristics of Newtonian systems (and the Newtonian order such systems create), themselves became fundamental, underlying values; metavalues expressed in the order of the "mechanical city". M u c h of the rest of this thesis follows from this exclusive l inking of Newtonian systems with scientific order. Later on in this thesis we shall return to this topic in order to examine in greater detail how the particular characteristics, or axioms, of Newtonian dynamics have influenced the everyday practice of city planning. 1.2.3. The organic city Lynch's third category of city order is that of the "organic city" 1 8 . Lynch explains that this way of looking at and planning for the city arose "as one expression of the nineteenth-century reaction to the stress of industrialization, gigantic new cities, and the unprecedented leaps in technology" (Lynch 1981, 88). At first, it seems obvious that planners who regard cities as being analogous to l iving organisms would create an order quite distinct from the order produced by those thinking of cities as machines. According to Lynch, mechanistic analogies and reductionist modes of analysis are explicitly rejected by those planners who perceive the city as being "biological" in nature. The whole of the city cannot be understood from the individual analyses of its parts, for they make sense only when working together, integrated as a working system. In 18 I shall follow Lynch's own lead in equating Lewis Mumford's own ideas with orthodox "organic" normative theory, for Mumford is the most famous organicist, the figure most commonly associated with the movement, and its most prolific apostle. Nevertheless, as Hill notes in his 1992 review, the spectrum of "organic" theory is so broad as to be positively self-contradictory (Hill 1992). 21 the "organic city" it is not possible to disassociate the different tissues of the city without also ki l l ing and dismembering it. Nor can the city be understood in the same way that a mechanic can glean knowledge from a stalled engine, for the city itself is never in stasis: The whole organism is dynamic, but it is a homeostatic dynamism: internal adjustments tend to return the organism to some balanced state whenever it has been disturbed by any outside force. So it is self-regulating. It is also self-organizing. It repairs itself, produces new individuals, and goes through a cycle of birth, growth, maturity, and death (Lynch 1981, 89). The shape of the organic city can also be distinguished from that of the mechanical c i ty 1 9 : Certain physical forms are matched to these [organic city] ideas: radial patterns; bounded units; greenbelts; focused centres; romantic, antigeometrical layouts; irregularly curving, "organic" shapes; "natural" materials (that means either traditional materials, or ones close to their unprocessed state); moderately low-density housing; visible proximity to earth, plants and animals; plentiful open space. The tree is the admired model, rather than the machine (Lynch 1981, 94). Unfortunately, the clarity of Lynch's distinction between the "mechanical" and "organic" forms of city order is muddied as soon as we take a look at the list of cities which he believes best exhibit "organic" order. Unlike the other two forms of city order, Lynch can only refer to a few settlements which exemplify the "organic" form: Tapiola in Finland, Bedford Park and Hampstead Garden in England, and the small developments of Radburn and Chatham Village in the United States (Lynch 1981, 90 ) 2 0 . A number of commonalties are immediately apparent. A l l of them have been constructed within the past century. Contrary to the expectations one might have regarding organic processes, all of the examples cited are instant towns — pre-conceived, and built according to a master plan. Most significant of al l , however, is the fact that these towns seem to be rather 19 And from the "cosmic city, for that matter. 20 I myself was born in a model "organic" town which Lynch does not mention: isolated Kitimat, B.C., nestled within a cloudbound fjord on the far northern seaboard of the Canadian Pacific coast. My family left when I was only eighteen months old, so I have no memories of life there. 2 2 unremarkable, at least from the public's point of view. Unlike cosmic Versailles, or mechanical Manhattan, Lynch does not provide a single example of an "organic" city noted by anyone other than planning theorists 2 1 . Lynch himself provides an explanation for this otherwise puzzling phenomenon when he notes that many of the ideas of the "organic city" have been given " l ip service" wi thin "most modern new towns throughout the wor ld" (Lynch 1981, 90): Its basic ideas are implicit in most public discussions of city form, and have even influenced such nominally antithetical examples as [Le Corbusier's] Chandigarh and [the "cosmic" city] Brasilia (Lynch 1981, 90). In fact, a second look at Lynch's checklist of "organic" urban forms reveals that it could be equally well applied to most typical post-war suburban developments. Apparently, "organic cities" like Tapiola and Radburn go uncelebrated precisely because they look more or less like everything else that has been constructed by planners over the past century 2 2 . And yet no less an authority on "organic" form than Lewis Mumford (who did more than anyone else to popularize the concept), identifies suburbia as the product of "technocratic planning", and its mechanical pursuit of order through the r igid separation of one part of life from another — sleeping from working, material production from cultural reproduction (Mumford 1974, 993). How can we make of this confusion of Lynch's "mechanical" and the "organic" normative theories in the shaping of our real-life cities of the modern world? Perhaps the answer lies in the fact that the "mechanical" and the "organic" are not nearly as distinct as one would at first believe. O n the eve of the millennium, words like "homeostasis" and "dynamic" are conceptually linked with immune systems, rainforests and coral reefs, weather patterns and the Earth as seen from space. "Organic shapes" conjure up images of 21 And these communities' present and former residents. After all, I have always been aware of Kitimat. 22 In their defense, organicists have long complained that developers have adopted the surface patterns, rather than the deeper logic, of "organic planning" methods. 23 jellyfish and swordferns, proteins and D N A . However, none of these late twentieth-century associations was available to those who actually formulated the ideal of the "organically" ordered city. Instead, the biological understanding which informs the 'organic' city is decidedly antiquated, dating back to "the rise of biology in the eighteenth and nineteenth centuries" (Lynch 1981, 88), when the field was transformed from simple taxonomy into an analytical science. Ironically enough, the first century of biology was notable for its thoroughgoing determination to reduce biological phenomena to mechanical simplicity. The early biologists regarded hearts as pumps, nerves as wires, and animals as nothing more than intricate automatons, a mind-set which only intensified with the accumulation of knowledge. Galvani's early experiments with electricity and frog's legs demonstrated the mechanical nature of the biological, just as the artificial synthesis of urea in the nineteenth century led scientists to reevaluate the distinction between biological and mundane chemical processes. Charles Darwin, in his arguments against & Supreme Designer would term the eye "a l iving optical instrument", and confessed that "it is scarcely possible to avoid comparing the eye with a telescope" (Darwin in Jennings, 327). Even those nineteenth-century planners who regarded the city as a biological mechanism were more or less bound to do so in markedly mechanical terms. Carl Reclam's 1869 manifesto, for instance, blithely blurs the line between the biological and the mechanical 2 3 in the following passage, which argued that scientific method would ensure civic public health by determining: the volume of good air necessary for sick people, schoolchildren, and prisoners; the space necessary for graves and the correct period for rotating them as determined by the type of soil; the weight and combination of food to produce a given sum of calories and to maintain body weight while at rest or while working; the quantity of certain hazardous substances in the air, 23 The point can be taken too far: even in the heyday of nineteenth-century scientific reductionism, few would have disagreed with the idea that there is a great difference between a watch and the person who wears one. 24 water, ground and the limits of their effects in space and time; the quantified effect of ventilation and heating equipment; the area of window glass necessary for sufficient heating; the correct relationship between building height and street width, between number of residents, built-up area, and green vegetation; and much else — the most exact possible determination of these things has been sought in nearly al l civilized lands (Reclam in Ladd, 44). Nor was this conflation of the mechanical and the biological restricted to those of the nineteenth century, for it was Victorian science which went on to shape our current-day planning theory. Lynch himself notes that "giants" like Ebeneezer Howard, Patrick Geddes and Lewis Mumford both "created the organic theory of settlement in the nineteenth century and carried out its development in the twentieth" (Lynch 1981, 90). Indeed, the only genuinely twentieth century insights 2 4 to be brought into Lynch's conception of the "organic city" from the biological sciences were the concepts of cybernetics and homeostasis (Lynch 1981, 89 ) 2 5 . Even these planning concepts, however, are now over half a century old, being originally formulated during the Second W o r l d W a r . 2 6 Consequently, there is a remarkable continuity between mid-nineteenth century views such as Carl Reclam's, and Lewis Mumford's own decidedly 'mechanical' description of the "organic city", written well into the 1930s: ...the highwayless Town is based on the notion of effective zoning of functions through initial public design, rather than by bl ind legal ordinances. It is a town in which the various functional parts of the structure are isolated topographically as urban zones, appropriately designed for their specific use: with no attempt to provide a uniform plan of the same general pattern for the industrial, the commercial, the domestic, and the civic parts (Mumford 1938, 490 ) 2 7 . 24 As I will discuss below, this is not to say that other biology-based concepts have not become influential within modern planning. The insights of ecology, and particularly the concepts of ecosystems and sustainability are having tremendous influence within the field. These concepts, however, are included within the notion of "organic" order as defined by Lynch (1981). 25 We shall refer to the writings of cyberneticist Stafford Beer later on, in Chapter 4 (Beer 1974; 1981). 26 Even Lynch tacitly admits that while "[the organicists'l writings and their projects are still the classic basis of training in physical planning", they have "[begun] to seem slightly old-fashioned" (Lynch 1981, 90). 27 The divergent views of Lewis Mumford'and Jane Jacobs regarding order and planning are referred to throughout this thesis, and it is worth noting that the personal relationship between these two near-comporaries was appallingly vindictive. Jacobs did not bother to conceal her scorn for Mumford throughout The Death and Life of Great American Cities (1961). In return, Mumford gave as good as 25 In the closing passage of The Culture, of Cities, Mumford evokes a machine-built home, topped with a single sprig from a fir tree, an image wonderfully emblematic of a melded organic-mechanical world-view (Mumford 1938, 493). 1.3. Scientific Order W h y do we find this unexpected congruence between the "mechanical ' and the 'organic' city? As we have noted above, both the "mechanical" and the "organic" city are intended to be built in accordance with the laws of nature. And from the time Sir Isaac Newton's Principia was first published in the late seventeenth century until well into our own century, the "laws of nature" were understood to be both synonymous with and limited to those underlying Newtonian systems. This, then is the fundamental reason for the similarities of the mechanical order and organic order: they are both products of the same scientific conception of an ordered universe. Newton believed that al l of workings of the universe could be explained and described by quantitative means: Newton was convinced that order in nature presents itself to observation as a set of necessary relations capable of exact mathematical description. ... [Thus, n]ot only does the abstract world of mathematical relations define [say] the ideal case of motion, but the extent to which motion in the physical wor ld fails to realize the ideal is itself subject to mathematical treatment (Westfall, 78-79). Newton believed that "mathematical order is [the] distinguishing feature" of the laws of nature; indeed, the scientific discoveries within the Principia were nothing other than a demonstration of a thorough-going mathematical order in nature. It is this quantifiable, he got in the extraordinarily nasty "review" of the work, entitled "Home Recipes for Urban Cancer", which has been reprinted in The Lewis Mumford Reader (Mumford 1986). 26 mathematical order that the builders of the mechanical city and the designers of the organic city both sought in their endeavors. Furthermore, although he was hardly the first to argue it, Newton believed that the laws of nature were universal; that all things in the universe, heavenly and mundane, mechanical and organic are subject to exactly the same laws of nature (Westfall, 77) . Thus, neither trees nor electric water pumps are exempt from gravity; both can only make water flow uphil l through the exploitation of vacuums and atmospheric pressure. After three centuries, neither of these hypotheses has yet been invalidated — as far as we know, the entire universe is subject to the same laws of physics, and these laws consist of mathematical functions. What we have learned, however, is that the particular characteristics of the physical processes Newton investigated are not omnipotent: while they may operate from one end of the universe to the other, they do not apply in all situations. However, up until very recently, even scientists confused the demonstrated universality of Newtonian laws with the unfounded belief that they were the only laws in the universe. This confusion is understandable. Knowledge of Newtonian systems enabled the development of the steam engine, the electric motor, the internal combustion engine, the modern sewer, the electric light, the telegraph and the radio. It gave humans mastery over their world, and over nature itself. Looking about themselves, in the midst of the new industrial wor ld humankind was creating for itself, it is hardly surprising that people assumed all natural phenomena were subject to the laws of nature which Newton had deciphered. But it just isn't so. In the past thirty years, science has recognized a second type of natural order in addition to Newton's natural order- a previously unsuspected volume of the laws of nature. For the first time since Newton's investigations in the seventeenth 2 7 century, we have begun to perceive a wholly new type of order in the world. It is the invisible order informing the rising of smoke and the swirl of hurricanes, the orbit of a tumbling asteroid, the skip of a beating heart, and the stutter of a nervous stock market. It is the visible order we see when we look at the Earth as a globe, when we note the repeated reducing structures in a head of broccoli or a fern, when we note the balance of an ecosystem 2 8. And perhaps, it is the order of a French medieval town or a Greek fishing village, where al l the buildings are similar yet not the same, where the streets are crooked yet admirably placed, and where a lack of thorough planning has not prevented the creation of beauty 2 9 . This new type of natural order has been variously described as "non-periodic deterministic flow" or "chaos". In this paper we shall refer to it as "Lorenzian order", the product of "Lorenzian systems". This thesis argues that whether we as planners subscribe to the "mechanical" or the "organic" city as an ideal, we are still creating the same kind of order in our cities — the Newtonian order that was all we knew of in the late nineteenth century, when the latest of our fundamental planning methods was developed. It is time that planning theory considered the insights coming out of twentieth-century science — in particular, an understanding of Lorenzian systems and their characteristics. Moreover, this thesis w i l l attempt to show that in addition to our current "synoptic" method of planning, derived from the axioms of Newtonian systems, it is additionally possible to derive a second method of planning — here called "dynamic planning — informed by the characteristics of the newly recognized Lorenzian systems. By incorporating the Lorenzian paradigm of natural order into planning theory, we can broaden our perspective on our present "synoptic" 28 James Gleick explores all of the above examples in his 1987 book. 29 As we shall see later, critics like Mumford and Kostof also admire such examples of urban form — but claim them as products of synoptic planning (Mumford 1986, 115;. Kostof 1991, 10). 2 8 planning practices, and develop a new method of planning which might help to resolve the crisis of faith which has beset the profession in recent decades. Clearly, we have a number of tasks ahead of us. We must examine our current planning method of "synoptic" planning and determine how it is derived from the characteristics of Newtonian order. We w i l l also have to derive the aims and assumptions of our hypothetical method of "dynamic" planning from the characteristics of Lorenzian systems. These tasks wi l l be the subjects of the fourth and fifth chapters respectively. Likewise, if we are to make links between city planning methods and systems of natural processes, we w i l l have to be acquainted with city planning theory and practice. Accordingly, we shall examine city planning throughout its history as a professional discipline in the third chapter. Finally, of course, it is crucial that we have an adequate understanding of the characteristics of both Newtonian and Lorenzian systems before we continue further: this w i l l be the focus of the next chapter. 2 9 CHAPTER TWO: NEWTONIAN AND LORENZIAN SYSTEMS 2.1. The Orders of the Universe Science has recognized Newtonian systems since the time of Galileo, and has been able to describe and model them since the pioneering mathematical work of Sir Isaac Newton in the 1670s. As we have seen our opening chapter, the understanding of Newtonian systems in the ensuing three centuries had effects upon human civilization far beyond the field of science, creating the necessary conditions for the Industrial Revolution, and underpinning modern notions of rationality, efficiency and order itself. By contrast, the first recognition of Lorenzian systems as a second fundamental source of order in the universe came only in 1963, with Edward Lorenz's landmark article on "deterministic nonperiodic flow" (Lorenz 1963). After less than 40 years, the basic characteristics of this class of ordering processes are now fairly wel l delineated, although the nature of Lorenzian systems has yet to be fully explored. If neither "Lorenzian systems" nor "non-periodic deterministic flow" sounds terribly familiar to most people, it is because such systems are most often called by the short, sexy, but terribly confusing name of "chaos" 3 0 . Using the word "chaos" to define "processes... wh ich appear to proceed according to chance even though their behaviour is in fact determined by precise laws" (Lorenz 1993, 4), is inherently problematic considering its multi-millennial history as a term denoting formlessness, disorder and violent anarchy. These associations make it particularly inappropriate for city planners who wish to explore such processes: the term "chaotic planning" comes across as a positive threat to most 30 According to Lorenz, the ascendancy of the terra 'chaos' was "virtually assured" when Gleick employed it as the title for his best-selling book on the subject (1993, 20-21). 30 people, given the deep connection between planning and ordering discussed above 3 1 . Although it has less pizzazz, we shall use the term "Lorenzian systems", after Edward Lorenz, who first recognized and investigated this type of deterministic (i.e. non-random) systems. As we did in the previous chapter, we shall continue to refer to the long-understood processes of Newtonian physics as "Newtonian systems". In like fashion, the order which results from Lorenzian systems wi l l be termed "Lorenzian order". Order produced by the operation of Newtonian systems w i l l be called "Newtonian order". In the section below we shall briefly investigate the properties of both Newtonian and Lorenzian systems, summarizing these in a table. Where possible, I shall use the definitions of Lorenz himself, as stated in his 1993 work The Essence of Chaos. M u c h of the section below has been adapted from the lengthy and extremely lucid example of a Lorenzian system described in Lorenz's work, which looks at the movement of a loose snowboard on a ski-slope mogul field 3 2 . We shall construct a parallel example of our own in order to illustrate Newtonian systems, imagining the same snowboards on a perfectly featureless inclined run. In addition, I w i l l also draw on James Gleick's popular 1987 work Chaos: making anew science. 2.2. Newtonian Systems Let us imagine a snowy inclined plane of uniform slope. We may think of it as a trackless version of the popular ice chute which is erected behind the Chateau Frontenac in Quebec City each winter. For the purposes of our explanation, we shall need only the 31 To add to the confusion, David Engwicht has employed the apparent oxymoron "Planned Chaos" both in a satirical sense, to lampoon current planning practices, and quite differently, as a description of a future ideal based on Jacobs and Sennett's ideas (Engwicht 1995). 32 Of course, this is nothing other than a Canadian revision of Galileo's pioneering experiments with inclined planes. 31 straight steep ramp, and not the long flat run beyond it 3 3 . The glissade wi l l have a smooth surface rather than the slotted one evident in the picture in Figure 3, and the slope w i l l be wide enough that the snowboards wi l l not come into contact with the sides of the glissade. Figure 3: Idealized and Actual Glissades Image of inclined plane: Nicholas Heap 1997 Image of Quebec City glissade: National Geographic January 1958, 71 No one wi l l actually ride the snowboards — instead, we wi l l be content to stand at the top of the slope and note their respective trajectories. As a result, the snowboards w i l l be running loose, controlled only by gravity, the angle of the slope, surface friction and air resistance (the latter two factors shall be grouped together as "friction" below). Finally, we 33 Those of you who didn't experience this glissade as children should produce kids of your own, take them to Quebec City and go on the ride with them on the pretext of safety, because it is a lot of fun. 32 w i l l ensure that our experiments take place under "laboratory conditions": the course w i l l be perfectly smooth, and friction w i l l remain constant throughout the glissade. In addition, we shall prevent any random external influences like winds from acting on the boards or the slopes. We have included friction within our experimental set-up because al l real-wor ld systems are dissipative — they lose energy to the surrounding environment 3 4 . Scientists contrast these realistic dissipative systems with purely theoretical "non-dissipative", or "Hamiltonian" systems, which retain their energy perfectly. Keeping in mind, then, that our Quebec City experimental system is dissipative like other real-world systems, albeit characterized by unnaturally constant conditions and a complete lack of random interference, we shall conduct a few simple experiments. For our first experiment, we shall simply place our two snowboards next to each other at the top of the slope so that they begin to slide down the glissade. Assuming that we have placed the boards carefully so that they start from a full stop, the boards w i l l head down the slope together, parallel to one another, so that they are the same distance apart at the bottom of the h i l l as they were at the top. If both boards start down the slope at the same moment, we would expect them to arrive at the bottom at the same time. If we repeated the experiment under the same laboratory conditions, exactly the same course of events would occur again. For our second experiment on the Quebec City glissade we shall have one board begin its journey three seconds before the other. Here, the first board would reach the bottom of the slope three seconds before the second board. 34 It is difficult to imagine the setting for a Hamiltonian experiment. We would be able to eliminate air resistance on the glissade only if we could get rid of the atmosphere itself — a glissade on The Moon would be adequate for this purpose. But even on The Moon, we would have to prevent any contact between the board and the slope in order to prevent friction.. Frictionless inclined planes exist only in the realm of theory, and such a system is far removed from any Earthly application, let alone city planning. 33 The speed of the snowboards in these two experiments is worthy of note. If this were a Hamiltonian system, we would expect the snowboards to continuously gain speed, at constant acceleration, as long as they headed down the slope. However, because we are experimenting with a dissipative system, friction removes some of the gravitational energy gained by the snowboards as they travel downhil l . Because friction varies in direct proportion to speed, it is not very noticeable at the top of the h i l l , where the snowboards are moving slowly. Consequently, both snowboards accelerate almost as quickly as they would in a Hamiltonian system. However, as the speed of the snowboards increases, friction carries off more of the gravitational energy, slowing the acceleration of the snowboards. If our glissade is long enough, the energy lost to friction w i l l increase unti l it equals the gravitational energy gained by the snowboards. At this point, the acceleration of the boards is zero and their speed is constant. In scientific terms, the two snowboards have reached a stable equilibrium speed. If we push one of these snowboards momentarily so that its velocity exceeds this equilibrium speed, the level of friction w i l l also increase, and the board w i l l return to the equilibrium level again. In similar fashion, if we slow one of the boards down momentarily, its friction level w i l l decrease, causing the speed of the board to increase until it again reaches the constant equilibrium speed. In our third experiment, we shall push the second snowboard off at an angle to the first, headed directly downslope. If this were a frictionless Hamiltonian system, the nudged board would diverge at a constant rate from the other (and to the extent that it was given a forward push as well , it would end up leading the other board, increasing its lead at a constant rate). However, in a frictional dissipative system it is easy to see that friction w i l l gradually reduce the speed at which the board travels laterally across the slope 3 5 . 35 Keeping in mind that we have only given the board a single initial push, rather than continuous propulsion. 34 Eventually, this lateral velocity w i l l decrease to zero, and even the nudged snowboard w i l l travel directly downslope. Having reached this state of equilibrium, the behaviour of the snowboards w i l l remain constant, unless they are acted upon by an outside force. If we again nudge this snowboard laterally in mid-experiment, friction w i l l gradually erode this new input of lateral force so that the snowboard is again travelling straight downslope. Thus, it is clear that the snowboard-and-glissade system has an equilibrium velocity, or "speed-and-direction". Simple as these three experiments are, they suffice to illustrate the basic axioms which inform Newtonian systems, which we can list as follows in Table 1 below: Table 1: Characteristics of Newtonian Systems Deterministic Linear or parabolic Size of effect proportional to perturbation Constant or periodic equilibrium Complexity of phenomena proportional to complexity of system Graphical representation: geometric Accurate long-term prediction and control possible W e w i l l now proceed to examine each of these characteristics i n detail. 2.2.1. Deterministic Newtonian systems — more specifically, linear dynamical systems and non-linear dynamical systems characterized by parabolic functions — are non-random by definition, for scientists employ the word "dynamical" as a synonym for "deterministic" (Lorenz 1993, 208). As we have already seen in Chapter 1, it is precisely this elimination of "chance", by explaining even experimental error through deterministic processes, that characterizes Newtonian physics (Westfall 1963, 79). 35 2.2.2. Linear or parabolic When we plot the results of our three experiments on graphs [Figure 4] we produce a series of straight lines. It is this quality — the ability of such systems to be portrayed by straight lines on a graph — which gives "linear" processes their name. If we plot positions Figure 4: Results of Newtonian System Experiments: Time (t) over left-right position on slope Experiment 1 Experiment 2 Experiments snowboard 1: • snowboard 2: - - ~ \ snowboard 1: ^ snowboard 2: - -snowboard 1: snowboard 2: (t) (t) left-right position on slope (t) left-right position on slope note: snowboards graphed only when in motion. left-right position on slope of the two boards with respect to time, so that the x-axis signifies location across the slope, and the y-axis signifies time elapsed, we produce two vertical parallel lines on our graph in the first case [Figure 4, Experiment 1]. A staggered set of parallel straight lines is the graph of the second run [Figure 4, Experiment 2], with the length of the offset equal to the time elapsed between start of the first and second boards 3 6 . 3G Providing that we only begin to graph the results once each board starts moving. 36 Not al l phenomena produced by a Newtonian system have to be graphed entirely in terms of straight lines: our graph of the third experiment [Figure 4, Experiment 3] is somewhat more interesting, as the graph for the second snowboard is composed of a parabolic curve (representing the loss of lateral velocity to friction) and a straight line segment (representing the state of equilibrium). Figure 5: Results of Newtonian System Experiment 1: Velocity (m/sec) and acceleration (m/sec2) over time Graph A: Experiment 1 Graph B: Experiment 1 velocity (m/sec) acceleration o — 0 — -time time Parabolas also appear if we chose to examine the velocity (i.e. meters travelled downslope per second), of the snowboards. Figure 5, Graph A shows the downslope velocities of the snowboards in Experiment 1 over time. Here, we end up with a positive parabolic curve (acceleration greater than friction), a negative parabolic curve (friction greater than acceleration), and a straight line segment representing the equilibrium between friction and gravitational acceleration for the particular slope of the glissade 3 7 . Figure 5, Graph B 3 7 The vertical straight-line segment connecting the higher velocity curve with the origin represents the velocity gained by the second snowboard during its initial push. 37 displays the results of Experiment 1 for the two snowboards in terms of acceleration. Here, acceleration due to gravitational attraction decreases parabolically in relation to friction until the two forces are in equilibrium at zero acceleration. 2.2.3. Size of effect proportional to perturbation Effects in Newtonian systems maintain a proportional relation to their causes. Thus, in the third experiment [Figure 1, Graph 3A], the divergence of the second snowboard from the system equilibrium would be directly proportional to the force applied. If we applied a lateral force of 10 Newton 3 8 , the board would only travel half as far laterally in a given time as it would if we had applied a lateral force of 20 Newton, and only one-tenth as far as it would have had we shoved it laterally with 100 Newton of force. This relation holds true at al l times: at any point in time the board shoved twice as hard laterally w i l l have gone twice as far laterally. Because the Quebec City glissade system is dissipative, al l snowboards in this system would eventually stop moving laterally, but one shoved twice as hard would travel twice as far. Consequently, the effects would be always be strictly proportional to the initial conditions. Lorenz singles out this property as the defining characteristic of "Newtonian systems", describing them as "system[s] in which alterations in an initial state w i l l result in proportional alterations in any subsequent state" (Lorenz 1993,209) . 2.2.4. Constant or periodic equilibrium O u r snowboard example displays another characteristic of Newtonian systems — their tendency to quickly settle into constant or repeating patterns of behaviour after wh ich no "new" behaviour is produced by the dynamical, or deterministic, system. As shown in our idealized case above, the snowboards soon settled into an constant end-38 A measure of force named in honour of the great scientist. 38 state rate of acceleration. In addition, Experiment 3 showed us that snowboards sent off at an angle to the slope — given an initial lateral velocity — reached an equil ibrium lateral velocity of 0 m/sec. After the snowboards settled into this equi l ibr ium state, only external forces could change their behaviour. Moreover, as we saw when we pushed and pulled the snowboards out of equi l ibr ium, even the introduction of new forces into the system do not affect the long-term equi l ibr ium acceleration if they are only transient in nature. The equi l ibr ium of the system w i l l change only if the system itself is changed — the pitch of the slope altered, the friction levels on the snowboard adjusted, or the force of gravity amended. Beside the ways we have already seen of graphing our Newtonian snowboards-and-glissade system, there is another technique of graphing referred to as a "phase-space" diagram which is specifically designed to show the equilibrium set of the system as a whole (Gleick 1987, 49-51; Lorenz 1993, 41-43). A phase-space diagram of our Quebec City Newtonian system is pictured below in Figure 6. In the case of our Quebec City glissade-and-snowboard system, the equi l ibr ium — also known to scientists as the "attractor" or "attracting set" (Lorenz 1993, 39-41) — is composed of a single po in t . Whi le a snowboard may be sent off at the top of the glissade wi th any combination of downward or lateral velocities, the dissipative nature of the system w i l l ensure that the snowboard eventually reaches an equi l ibr ium condition, travelling straight down the slope at a constant velocity. In truth, our glissade-and-snowboard system provides a very uninteresting example of an attracting set, as other types of Newtonian systems boast periodic, rather than constant equilibria. In such cases, the end-state of the system consists of a series of conditions, l inked together into a fixed loop much like a broken record (or less 39 Figure 6: Results of Newtonian Experiment 3: "Attractor" of snowboard-and-glissade system j 5 • — - . i ' • ' • 1 I r r ^ r r I < > I • • • . | i i i i ! 1 1 1 1 | I 1 I 1 | 1 I 1 I 3 -per second) -'EED (meters | in UJ O 1 _ i CO cn 0 0 O -2 cr o -3 --1 —1 1 1 1 1 1 1 1 1 1 1 1 1 I I 1 1 1 1 ! I L . L ' 1 ! ! t ! I , , . , 1 , . , , -2 .0 - 1 . 5 -1 .0 - . 5 . 0 .5 1.0 1.5 2. CROSS-SLOPE DISTANCE (meters) 0 Image of snowboard-and-glissade attractor adapted by author from Lorenz 1993, 54 annoyingly, a very short 8- t rack tape) 3 9 . Nevertheless, such repeating forms appear to be limited in duration ("period"), and the particular order or values expressed in these repeating series remains fixed for eternity once the equil ibr ium state has been settled into. As we have seen above, where we altered the velocities of the snowboards, a Newtonian system can be perturbed from its end-state condition. Nevertheless, the very 39 Periodic and non-periodic systems are a prime focus of study by many scientists, and are central to "Bifurcation Theory" and "Catastrophe Theory". People interested in this aspect of Chaos Theory are advised to read Gleick's (1987) and Lorenz's (1993) discussions of this topic. 40 same constant or periodic equil ibr ium w i l l reestablish itself as soon as transient or random forces acting upon the system have dissipated, unless the system itself is altered. 2.2.5. Complexity ofphenomena is proportional to complexity of system Our Quebec City glissade-and-snowboard example is a very simple, idealized model of a Newtonian system. Even if we continue to ignore "random" interferences like breezes, or scattered debris on the slope, our model could only hope to produce realistic results if we included variations in slope and surface friction, the weather, the aerodynamic profile of the board, and a number of other factors as well . Depending on the quality of our work, our model might or might not produce more realistic results, but it certainly would produce more complicated results. Rather than accelerate at a constant rate, our board might slide downhil l in more naturalistic subtle fits and starts. Instead than the same simple monotonous equilibrium, we might bring a different value into existence, or even produce a periodic equilibrium. Nevertheless, the increase in the complexity of the outputs of the system comes only through a commensurable increase in the complexity of the 1 inputs to the system: were we to describe this system in mathematical formulae, it would be many times longer than that used to describe our first model. In short, if a Newtonian system produces complex behaviour, it is because the system itself is complicated. Conversely, a simple system, like the snowboards-on-the-slope example we have been discussing, can only produce simple phenomena. This characteristic has important consequences for modelling: models of systems under investigation are often simplified because of the work involved in making them truly "realistic" is often immense. Nevertheless, even models which drastically simplify Newtonian systems are often wonderfully accurate because (as we have seen above) causes are strictly proportional to effects in Newtonian systems. There is little point in massively increasing the complexity of a model if it is done to include relatively small-scale 41 phenomena, since just as in the Newtonian system itself, the influence of such factors w i l l have little if any discernible effect on the results produced. 2.2.6. Graphic representation: geometric As we have seen in Figures 3, 4 and 5 above, there is a real l ink between the lines, parabolic arcs and points produced by Newtonian systems and the geometric forms of Euclid. Nevertheless, the equation of geometry (here used to refer specifically to Euclidean geometry) and Newtonian systems has been emphasized by our cultural history: indeed, it is difficult to distinguish the scientific from the cultural. The natural philosophers of ancient Greece, convinced of the imperfection of the phenomenological world, eschewed experimentation for pure reason. Because geometry, with its purely abstract mathematical laws for the construction of figures, could be figured out a priori, the "perfect shapes" thus produced — the circle, the square, etc., were believed to embody a higher form of order than the any forms in the mundane world . The ancients saw evidence of this geometrical perfection in the heavens. The Aristotelian paradigm posited an Earth-centred universe composed of perfectly formed spheres of aether containing the stars, the sun and moon and the four known planets (Kuhn 1966, 77-83). Dur ing the sixteenth and seventeenth century, the pioneering astronomers Copernicus and Kepler replaced the Aristotelian paradigm with the more-familiar notion of a sun-centred universe. Nevertheless, science did not make a clean break with ancient geometry. While Copernicus intellectually shifted the centre of the universe (with profound consequences) in 1543, he did not dispute the ancient conviction that the orbits of the planets were perfectly circular. More than sixty years later, when Kepler established that the orbits of the planets were actually ellipses rather than perfect circles, he could nevertheless take comfort in the fact that ellipses are also geometric figures (Kuhn 1966, 42 211 -214) 4 0 . A further eighty years on, as we have already noted, Newton would establish that al l of the universe was bound by the same laws. And as we have seen these laws were graphically expressed in terms of straight lines and parabolas: to this day, it is the production of these shapes on a graph that assure scientists that they have been able to isolate particular forces and physical processes (Bueche 1981, 6-7). The particular patterns of order which Newtonian systems naturally create — points, straight lines, and perfect parabolic curves — have led scientists from the dawn of the modern age to adopt Ancient concepts regarding geometry and order. Even as modern scientists refuted ancient mystical beliefs, they reinforced the idea that Nature was in thrall to Geometry. 2.2.7. Accurate long-term prediction and control possible As we have seen above, Newtonian systems are formula-bound and deterministic in nature, which prevents random activity occurring within the system itself. They are proportionate, so that large changes in resultant phenomena can only be produced by large changes in initial conditions. We have seen how dissipative Newtonian systems converge to stable equilibria. Finally, we have taken note that complex phenomena within Newtonian systems are only produced by a complex system. The combination of these properties, remarkable in themselves, is even more significant, because they ensure that real-world Newtonian processes can be accurately modelled, enabling us to both predict and (often) control the phenomena produced. As we have already hinted, modelling Newtonian systems is relatively simple, and remarkably robust. External disruptions cannot produce anything more than a transient 40 More accurately, they are the general class of shapes of which circles are a particular variant (Kuhn 1966, 213). Kepler's conviction that the Divine order of the universe was expressed in geometry is perhaps best revealed by his (now disregarded) theory that the orbits of the planets were ordered in terms of the five "regular solids" of perfect geometry known to the ancients (Kuhn 1966,211-2148). 43 deviation from equilibrium. Here too, effects are proportional to causes, and only a large-scale perturbation from equilibrium wi l l produce large-scale (transient) effects. Because of the minimal effect that transient or small-scale perturbations have on Newtonian systems, we can effectively ignore the effect of many "extraneous" forces upon a real-world system. In those instances where these impinging forces are too powerful to ignore (and in a Newtonian system they must he powerful to be significant), then we have the ability to factor these additional deterministic forces into our model of the system, producing more realistic results. It is only in the cases where truly lawless or random behaviour overwhelms the Newtonian system that we cannot set about productively modelling the system. These characteristics of Newtonian systems means that we can effectively anticipate their behaviour. In our own example, we w i l l be able to accurately predict exactly where our runaway snowboards w i l l be at a given time if we know the general properties of the system - the strength of gravity, the angle of the glissade slope, and the coefficient of friction - and have information about the initial starting point, and velocity of the snowboards. Given this data, we can determine exactly where the snowboards w i l l be, anywhere on the glissade, at any point in the future. In theory, this holds true for any combination of starting conditions, and given a glissade of infinite length, for any point in time in the future. Nor are we simply able to predict the behaviour of the snowboards given particular starting conditions. If we are able to set the initial conditions for the snowboards — their forward and lateral velocity, and their position along the top of the glissade — it follows that we can also control the future, by forcing them to pass any particular point at any particular time we may happen to decide upon. This can be done simply by predicting what combination of initial conditions would have to exist in order for 44 such a trajectory to result. Having calculated this, a snowboard would then be sent off under these circumstances. More than any other quality which Newtonian systems possess, it is this openness to prediction and control which has made this type of natural ordering process so attractive to humanity in the Modern Era. And it has been this desire to exercise prediction and control in the cause of scientifically ordering the city which has resulted in planners assuming a Newtonian systems world-view in their work. 2 . 3 . Lorenzian Systems Having become well acquainted with the characteristics of Newtonian order through our experiments on the glissade behind the Chateau Frontenac, we now travel to an idealized mogul-filled ski slope on Whistler Mountain in British Columbia. Once again, we shall not be bothered with the random and complicating influences of wind and weather and variable snow conditions. As before, we have a large, wide uniform slope, except that in this instance the slope is composed of a uniform series of identical hillocks and depressions ("moguls") arranged in checkerboard fashion 4 1 [See Figure 7]. As we did in Quebec, we shall place the two snowboards side by side at the top of the slope, and let them travel freely down it to the bottom. Here though, the results are markedly different 4 2. A small difference in the initial position or velocity of two snowboards on a mogul-slope w i l l become magnified disproportionately as they careen down the h i l l . This result is not as mystifying as it might first seem. Any initial difference 41 The following section has been adapted wholesale from Lorenz (1993), Chapter 2. Since he is American, Lorenz prefers the ski hills of Colorado to those of British Columbia. 42 It should be noted that the results of our snowboard experiment will be "chaotic" for some, but not all such mogul slopes, depending on the magnitude and shape of the undulations. Lorenz's model mogul field, which allows for Lorenzian processes, is composed of checkerboard "hill" and "pit" sections two meters wide and five meters long, on a tilted plane with an overall slope of .25. The vertical difference between the hills and pits is one meter. Lorenz treats the snowboard as a single particle, and makes the board's coefficient of friction inversely proportional to speed so that velocity is kept constant (Lorenz 1993, 28-30). 45 in the initial conditions of the two snowboards wi l l send them on slightly different trajectories, even through the difference may not even be perceptible to the human eye. Consequently, the two sleds w i l l traverse very slightly different areas of the first mogul they come to. Because each point on a mogul wi l l have a slightly different slope and orientation, Figure 7: Idealized and Actual Mogul Fields Image of idealized mogul field: Lorenz 1993, 27 Image of real mogul field: Lorenz 1993, 29 the subsequent trajectory of the two snowboards is altered to different degrees and in different directions. Consequently, the difference in the initial conditions becomes magnified. The differences between these new trajectories gives rise to still greater disparities, as the slope and orientation of points on the mogul traversed by the two 46 snowboards increasingly diverges with distance from one another. Before long, the paths of the two snowboards w i l l be so disparate that one w i l l traverse a mogul the other entirely misses. The paths of the two boards w i l l start to cross one another in apparently random fashion, or diverge wildly, continuing on completely unrelated courses to the bottom of the h i l l . At this point, there is no longer any correlation whatsoever between the distance between the two boards and the time elapsed since the beginning of the run. Figure 8: Sample Trajectories of Snowboards in Idealized Mogul Field Graph A: Experiment 1 Graph B: Experiment 2 I i y "CIM ' V ' l ' V ' l ' V ' l ' V v 2 " \ : 5 \ ; 6 • 0 0 % 0 0 0 10 \ : 8 «• to o o | m 0 0 0 - 2 » ) o e 12 LU 2 <o E UJ 2S 1 iLOPE DISTAN* ' 0 0 ° \sli ° 0 £ 30 a LU a. ; ) DOWNS DOWNS : 22 • // 7K • 45 ; 24 26 • 0 0 r 0 /J0 \ \ 0 50 ; A M \ 20 • ]\/f I\ 55 ; / ( V ) / I , * t , / * X/l , A , ! / / A , \ , i i \ A / i - 4 - 2 0 2 4 6 8 10 12 14 C R O S S - S L O P E D I S T A N C E (maters) 60 •a -15 -10 - S O 5 10 IS 20 C R O S S - S L O P E D I S T A N C E (motors) T h e paths o f s e v e n boa rds s t a r t ing w i t h i den t i ca l ve loc i t ies f rom po in t s spaced at 10-centimeter in t e rva l s a l o n g a west-east l ine . T h e s m a l l d i a m o n d s i n -dica te the loca t ions of the centers of the m o g u l s . T h e paths o f s e v e n boards , s ta r t ing w i t h i d e n t i c a l ve loc i t ies f r o m po in t s spaced at l - m i l l i m e t e r in te rva l s a l o n g a west-east l i n e . Sample trajectories of snowboards: Lorenz 1993, 33, 34. When we retrieve our snowboards and attempt to repeat our experiment we have another surprise. By the time the boards reach the bottom of the slope the second time, 47 their trajectories bear no resemblance whatsoever to the paths the same boards took on the first run, even through we placed the boards in almost exactly in the same starting points as before. O n the virtual mogul slope Lorenz simulated mathematically, snowboards separated by 10 centimeters had begun to trace out completely independent, random-appearing trajectories after only 11 meters [Figure6, Graph A]. Snowboards separated by a single millimeter a.t the top of the slope had diverged into completely unrelated trajectories by the 60-meter mark [Figure 8, Graph B]. Even if we stayed up on the mogul slope al l day, sending the snowboards down again and again, in conditions as close to identical as we could measure, we should find that the paths they traced down the slope never once repeated. Once again, our simple experiments has given us enough material to outline the basic properties which inform Lorenzian systems, and the ways in which these systems differ from Newtonian systems. As before, we can produce a list of these properties, although given the different nature of Lorenzian systems, we w i l l explore the following points in a somewhat different order than we did for Quebec City Newtonian system above. Table 2: Characteristics of Lorenzian Systems Deterministic Size of effect disproportional to size of perturbation (a.k.a. "sensitive dependence" or "the butterfly effect") Non-linear Equilibrium constantly changes Graphic representation: fractal Complexity of phenomena disproportional to complexity of system Accurate long-term prediction and control impossible 48 2.3.1. Deterministic It is hard not to picture a fleet of runaway snowboards, each dancing a different line down a mogul field, as being subject to random (non-deterministic) forces. Indeed, Lorenzian systems gained their popular name of "chaos", precisely because these systems "appear to proceed according to chance" (Lorenz 1993, 4). Nevertheless, this is not the case: our idealized example has excluded any element of chance. As before, the only elements in our system are the forces of gravity and friction and the (varying) pitch of the slope and the mass of the snowboard 4 3 . Indeed, the deterministic nature of this system can be proven, at least in theoretical terms, simply by conducting repeated trials under identical conditions. While a millimeter of difference in initial conditions can produce totally different results, were we able to place one snowboard after another in exactly "Cos, same initial conditions, the trajectory followed would be absolutely identical (assuming that the condition of the slope also remained identical). 2.3.2. Size of effect dis-proportional to size of perturbation The problem with this is that it is absolutely impossible — rather than simply impractical — to ensure that we can actually put the snowboard in exactly the same place twice. Formulated in the early part of the twentieth century, Heisenberg's Uncertainty Principle states that we are unable to measure distances and properties on the subatomic scale without changing the position of the thing being measured (Bueche 1981, 395-397). This is as true of objects in a Newtonian system as those in a Lorenzian system, but nanoscale differences in the former kinds of processes are irrelevant, because changes in a Newtonian system produce only proportionally-sized effects. By contrast, as we have seen in the experiments above, effects in a Lorenzian system amplify extremely quickly. 43 My description here is somewhat simplified. Readers seeking a more detailed, and more authoritative examination of the forces involved in this ideal model should refer to Lorenz (1993, 27-28). 49 This phenomenon is properly described by scientists as "sensitive dependence", but it is most commonly known as "The Butterfly Effect": in a 1972 lecture to his fellow meteorologists, Edward Lorenz noted that the disproportional amplification of differences in Lorenzian systems such as the global climate regime implied that "the flap of a butterfly's wings in Brazil [could] set off a tornado in Texas" (Lorenz 1993, 181). 2.3.3. Non-linear In our Newtonian system at Quebec City graphed in Figure 4, the distance between the two snowboards in Experiment 3 is a straight-line function of the lateral force imparted to the second snowboard. By contrast, in Lorenz's graphs of his mogul-field snowboards, the tracks of the snowboards diverge and converge on an irregular basis, repeatedly changing directions and even crossing each others' paths. In Figure 8, Experiment 1, the paths of the seven snowboards have diverged by a total of two meters at the 10-meter mark, and by eleven meters at the 30-meter point, but pass within centimeters of each other in between these two points. This rate of change can neither be plotted as a straight-line divergence, nor as a smooth parabolic curve — it is instead an irregular and disproportional funct ion 4 4 . Clearly, we are dealing with something quite different than a Newtonian system. Pioneer mathematicians in the field of Chaos Theory have explained that the rapid divergence of possibilities in Lorenzian systems implies that any changes in such systems w i l l result in an infinite number of discrete results over infinite time. Nevertheless, this does not mean that just anything can happen: instead, mathematicians tell us that the infinity of potential states possible in a Lorenzian system are contained within a finite, bounded "phase-space". Like our Newtonian system in Quebec City, our Lorenzian 44 As we have seen, "parabolic" non-linear functions can be transformed into straight-line functions by taking a second order measurement - acceleration [m/sec2 ] rather than velocity [m/sec]. 50 snowboard-and-mogul system is dissipative, and the phenomena produced by it gravitate towards an attractor, or equilibrium set. Still, our Whistler Mountain attractor is remarkably odd-looking for those used to the points and loops of Newtonian equil ibrium sets; so much so that the equilibrium sets of Lorenzian systems are commonly referred to as "strange attractors" (Lorenz 1993, 48-53). A phase-space graph of the "strange attractor" for Lorenz's own snowboard-and-mogul system is shown below, in Figure 9. Figure 9: Results of Lorenzian System Experiments: "Strange attractor" of snowboard-and-mogul-field system S | i i i i | — i . . . | i i i i | i i i i | i i—i i | i i i — — i — i — [ — r — i — n — I i 4 " -4 --5 I i i i i l i i i i I i i i i i i i i i i i i i i I i i i i I i i i i i i i i i I -2.0 -1.5 -1.0 -.5 0 .5 1.0 1.5 2.0 CROSS-SLOPE DISTANCE (meters) Snowboard-and-mogul-field strange attractor: Lorenz 1993, 54. Whi le the notion of a strange attractor containing infinity within a finite area is initially hard to conceptualize, mathematicians explain this seeming contradiction through the 51 process of "stretching and folding" (Gleick 1987, 50-52). We might imagine an infinitely stretchable band packaged inside a snug-fitting box. Every day, we stretch the band to twice its length, so that when we put it back in the box it is twice as long as before. This is possible because every time we double the length of the band, we necessarily halve its width. If we were to continue doing this forever, we would end up with an infinitely long, infinitely thin band contained within a finite area — the box. It is this same stretching and folding which enables us to make sense of the non-linearity within the Lorenzian system. Referring back to our band-in-the-box analogy, let us say that we had put two dots beside each other on the band when we began. As we stretched the band again and again, these dots would move farther and farther apart along the length of the band. However, because the band itself is folded up within the box, it stands to reason that as soon as the two dots were one full box-diagonal away from each other 'as the crow flies', they would have no choice but to move closer again. As time went on and the band grew ever longer, the dots would continually shift their relative positions within the box, sometimes moving closer together, sometimes moving farther away — but never, of course, appearing beyond the confines of the containing box. The paths of the snowboards on the mogul slope, then, are "stretched and folded" trajectories, sometimes dispersing, sometimes converging and crossing each other, all seemingly at random, but actually deterministically bound to the infinite equilibrium set of that Lorenzian system. 2.3.4. Equilibrium constantly changes We have already likened the "phase-space" of our Whistler Mountain Lorenzian system to that of an infinitely long and folded elastic ribbon fitted inside a finite box. Every point on this endless loop represents an equilibrium point between left-right position and velocity for our snowboards a particular points on the mogul field. We have already seen the "phase-space" diagram for the Newtonian Quebec City system in Table 5. It displays 52 the attractor for the glissade - a single point at the centre of the diagram, towards which the behaviour of al l the snowboards is drawn. Figure 9 above shows the complete set of equilibrium states — the attractor — for our Lorenzian system of snowboards and moguls on Whistler Mountain. Quite obviously, it is more complex than the one-point Newtonian equilibrium set; in fact, as we have discovered above, the equilibrium sets for Lorenzian systems are infinite in size and complexity. Unlike Newtonian systems, the pattern produced by a Lorenzian system w i l l never repeat, let alone become static: by definition there is not enough time in the universe for al l the equilibrium points on the infinite equilibrium set of an Lorenzian system to be expressed even once. Consequently, Lorenzian systems are truly non-periodic, for they never repeat their behaviour. Unlike Newtonian systems, which settle down to a constant state, or periodic repetitions, chaotic systems never reach an end-state, but change continuously forever, even while being deterministic. 2.3.5. Graphic representation: fractal It is hard to discern much order within the trajectories traced by the snowboards in Experiments 1 and 2 of Figure 8, especially when it is compared with the comparable straight-line or parabolic graphs of our Newtonian system: indeed, to look at Figure 8, there does not appear to be any order at all there. But this is largely a matter of perception and presentation. Lorenz shows that if we graph the velocity of snowboards taken from thousands of different runs, and plot it against their cross-slope positions in relation to any given mogul, we end up with an actual portrait of the Lorenzian system's "strange attractor", or equilibrium set, which we have discussed above. In form, the attractor resembles nothing so much as a dash of vanilla which has been partially mixed into a bowl of batter [Figure 9]. This shape, and others which share its properties, are known as "fractals". Whi le this shape is rather unusual for those of us used to looking at circles, arcs 53 and polygons, the Lorenzian system governing our snowboard-and-mogul experiments has clearly produced something far more "orderly" than a random scribble. The fact that our experimental system was able to produce such a coherent pattern is not simply happy chance: it is apparent that Lorenzian systems generate "fractal" patterns in the same fashion that Newtonian processes give rise to Euclidean geometric forms. The term "fractal" was coined in the 1970s by Benoit Mandelbrot, and is defined as "a set of points whose dimension is not a whole number" (Lorenz 1993, 208). Like many other concepts in Chaos Theory, this seemingly absurd concept is not as difficult to comprehend as might be expected. A two-dimensional object can be thought of as a flat sheet of paper having zero thickness. By contrast, a three-dimensional object might be thought of as a wooden building block. But what if we crumpled our sheet of zero-thickness paper? Even though the sheet itself lacks thickness, its new creases would force the sheet into the third dimension. Depending on the extent of the sheet's crumpledness, mathematicians would calculate its "dimensionality" to be anywhere between 2 and 3 — a fractal shape. The portrait of the our Whistler Mountain fractal in Figure 9 is thus somewhat distorted by being represented as a two-dimensional shape — in its real state, the shape would extend above and below the page. Perhaps the most familiar ' icon' in Chaos Theory is the strikingly beautiful form of the "Mandelbrot Set", the graphical representation of the solution set for a simple reiterative algebraic equation which takes the overall shape of a blobby "gingerbread man" [Figure 10]. The beauty of this shape is in the details: the closer one looks at small fragments of the figure, the more complex the shape gets. The boundary of the shape breaks up into serpentine curves, ammonite swirls and fern-like branchings. Fractals are scaleless or "self-similar", as each component of the shape is composed of similar, but not identical sub-components. Wi th in the Mandelbrot Set, one even finds the overall 'gingerbread man' 54 shape replicated throughout the figure in an infinite variety of sizes. A more familiar example of fractal self-similarity is found in some species of ferns, where the shape of the whole is reflected in each of its fronds, which is further reflected in each leaf, and each leafs leaflet 4 5. Figure 10: The Mandelbrot Set Images of Mandelbrot Set: Gleick 1987, 114 Needless to say, fractal "strange attractors" of Lorenzian systems also exhibit qualities of self-similarity and infinity. Consequently, we would expect our own fractal "equil ibrium 45 While the necessities of cellular structure force an end to this "wheels within wheels" type of arrangement in the natural environment, the recursion is infinite in mathematical fractals. 55 set" for the snowboards on the mogul field to display qualities of self-similarity if portions of it were enlarged to display greater detail. A shape as amazingly complex as this cannot be adequately described in simple geometrical terms. But there can be little question that the shape of a "strange attractor" does indeed constitute a form of order. In fact, it is difficult not to conclude that the forms of order produced by Lorenzian order are more subtle, more beautiful, and even perhaps, more ordered than that produced by Newtonian systems. 2.3.6. Complexity ofphenomena dis-proportional to complexity of system In a Newtonian system, a wider variety of resultant phenomena or possible states is the result of a more complex dynamical system. Thus, we could broaden the range of phenomena produced in our Quebec City experiments by adding new forces and factors, and having them interact with one another. O n the other hand, we have seen that our idealized and simplified Lorenzian system on Whistler Mountain is capable of producing an infinite number of possible phenomena, even though the Lorenzian system employs the same number of forces and elements as the Newtonian system. The complicatedness of our idealized system is trivial compared to most others which occur in the real world. Nevertheless, it can produce an infinite number of different behaviours. 2.3.7. Accurate long-term prediction and control impossible As we have seen above, the physical impossibility of obtaining exact measurements means that we can never be sure of obtaining the same results twice in a Lorenzian system, even when there is no opportunity for random interferences. Unfortunately, the difficulties created by "The Butterfly Effect" are not limited to those who wish to relive the past. "Sensitive dependence", or the rapid amplification of nanoscale uncertainties into macro-scale phenomena ensures that long-term prediction is impossible in a Lorenzian 56 system. As we have seen from the example of the snowboards on the mogul field, the difference of a single millimeter w i l l result in trajectories 15 meters apart after only 60 meters [Figure 8]. In the real world, predictability is hampered still further by the fact that random perturbations w i l l continuously jolt the snowboard from one location to another on the system's "strange attractor", randomly altering the deterministic portion of the snowboard's trajectory. Under such conditions, how could any forecaster be expected to make any prediction whatsoever about the position of the snowboards at the bottom of a 100-meter slope, aside from the banal statement that all the boards would get to the base of the hil l? This was precisely the question which Lorenz the weather forecaster posed to his colleagues when he starting talking about the "Butterfly Effect". The answer was simple: while short and medium-term prediction was possible, albeit with some degree of uncertainty, the dream of predicting the weather months in advance, as previously hoped, was simply impossible. At present, armed with satellite data and advanced modelling techniques, Lorenz states that meteorologists can only make predictions for the weather three days ahead with any degree of confidence (Lorenz 1993, 104-105) 4 6 . The nature of Lorenzian systems implies that this is not likely to change. If possible, the situation is even more gloomy for those who might wish to control a Lorenzian system in a manner analogous to an artillery crew, who can fire cannonballs at a particular angle at a particular angle and velocity, confident that they w i l l land on target, tens of seconds and scores of miles away. Managing to perform a feat of similar accuracy in a Lorenzian system could only be accomplished with an infinite amount of luck, or with aTiterally Divine understanding of the structure of the systems' equilibrium set, and a comparable knowledge over the future random fluctuations of the surrounding 46 City planners, on the other hand, still regularly produce twenty-year forecasts: the Livable Region Strategic Plan recently produced by the Greater Vancouver Regional District is an example we shall look at in Chapter 4 (GVRD 1995). 57 environment. Only in this way could an exact measurement of the present state of the system, and a precise understanding of its future development be gathered. Lorenz himself suggests that the best hope for prediction and control in an might be to regard as "more probable" the trajectories which resemble the phenomena which developed as a result of similar situations in the past (1993, 102-110). In recent years, some scientists have tried to avoid the dilemma by perturbing the behavior of lab-table Lorenzian systems in such a way that they exhibit controllable Newtonian characteristics (The Economist 1994, 85-86; Strogatz 1995, 444). It is possible to conceive of a different approach, however. As we have seen through Lorenz's own work with the snowboards on the mogul field, it is quite possible to outline and describe the "strange attractor" of a particular Lorenzian system, even if it remains impossible to make long-term predictions for a particular snowboard operating within the confines of that system. But we can confidently predict that the behaviour of the snowboard w i l l remain within the sphere of possibilities bounded by the finite edge of the attractor. Unless the random forces impinging on the Lorenzian system are so significant as to overpower the system (making it a "random" system, rather than a Lorenzian one), they shall only succeed in jolting the snowboard from one point on the system's attractor to another. The behaviour of the system w i l l remain bounded, and predictable in the broader sense of having limits. In and of itself, this power of prediction is very modest. It is of little use for Lorenz the meteorologist to state that in a week from now, the weather for Vancouver could range anywhere from -18° C to 33° C with anywhere between 0 to 90 mm of precipitation (Oke and Hay 1994, 30). It is far more interesting to think of the advantages that we might gain if we could control the shape of the attractor of a Lorenzian system. It is worth noting that Lorenz did just this in designing his idealized snowboard-and-mogul dynamical system. Were we able to "design" an attractor to our l iking, we still would not be able to make 58 predictions about the behavior of particular elements within the system (as Lorenz himself was unable to). What we could do, however, is ensure that the finite domain of the attractor would cover only those areas we judged to be benign or actively beneficial. Were Lorenz to design Vancouver's climate in this fashion, he could restrict the range of possible weather conditions so that it would never rain more than 25mm per day, or less than 50mm per month. The temperature could be prevented from dipping below -5° C, or soaring above 30 ° C, and Lorenz could constrain the air both from hanging still and from exceeding cyclone force. Lorenz still wouldn't be able to make an accurate one-week forecast, but he could assure us that nothing untoward would occur. 2.4. Conclusion Having reviewed both Newtonian and Lorenzian systems, point by point, we are now in a position to compare and contrast the axioms of Newtonian and Lorenzian order as a whole. Table 3: Comparison of Newtonian System Characteristics with Lorenzian System Characteristics Newtonian Systems Lorenzian Systems Deterministic Deterministic Linear or parabolic Non-linear Size of effect proportional to perturbation Size of effect dis-proportional to perturbation Constant or periodic equilibrium Equilibrium constantly changes Graphic representation: geometric Graphic representation: fractal Complexity of phenomena proportional to complexity of system Complexity of phenomena dis-proportional to complexity of system Accurate long-term prediction and control possible Accurate long-term prediction and control impossible 59 Plainly, while Lorenzian order shares a few similarities with its Newtonian counterpart, we cannot help but be struck by how strange and counter-intuitive this 'new' type of ordering process is. It becomes clear that thinking of order in terms of Lorenzian rather than Newtonian processes amounts to more than having a different viewpoint: the wor ld itself appears in a decidedly different light. Although the term has been woefully overused by academics in the liberal arts ever since T. S. Kuhn popularized the term in his Structure of Scientific Revolutions, the term "paradigm" seems appropriate here. In the chapter "Revolutions as Changes of Wor ld View", Kuhn says of scientists observing phenomena by means of a new paradigm: It is rather as it the professional community had been suddenly transported to another planet where familiar objects are seen in a different light and are joined by unfamiliar ones as wel l . . . . Of course,... there is no geographical transplantation; outside the laboratory everyday affairs usually continue as before. Nevertheless,... after a [perceptual] revolution scientists are responding to a different world (Kuhn 1970, 111). Snowboards are a recent innovation. So is the intellectual achievement of "Chaos Theory". But Lorenzian systems have been in existence as long as Newtonian systems have — since the dawn of time itself. The only thing that has changed in recent years is that we have gained the understanding to recognize these processes for what they are — not random, but deterministic, not formless, but infinitely ordered. Now that we have undergone our own "paradigm shift", now that we recognize that there are two decidedly different "scientific orders", it w i l l become obvious to us how the discipline of city planning has confused the characteristics of Newtonian systems wi th science, order, and efficiency itself. In the following chapter we w i l l trace the history of modern city planning, highlighting its' growing dependence on the mechanical-organic world-view and, and the characteristics of Newtonian order which underpin this paradigm. Moreover, we w i l l discover how this half-perception of the wor ld has thrown the discipline into crisis since the late 1950s. Later on, in the fourth chapter, we shall draw 60 an explicit l ink between basic assumptions in planning method and the axioms of Newtonian systems. Having done this, we w i l l finally be able to go about the business of creating a parallel method for city planning, based on the axioms of Lorenzian order. 61 CHAPTER THREE: FROM CRISIS MANAGEMENT TO INTELLECTUAL CRISIS 3.1. A Brief History of Modern Ci ty Planning Thus far, we have dealt with concepts of order currently held by city planners 4 7 , and scientists' current understandings of natural order. At the end of the Introduction, we concluded that city-building conducted according to the "mechanistic" and "organic" normative theories of city order were both founded upon the same Newtonian scientific concepts of Newtonian order. We went on to note the basic hallmarks of this scientific conception of order — in which even the fluctuation of error is itself deterministic and ordered in behaviour, its thoroughgoing universalism, and its faith in the Newtonian nature of natural law. Following this, we went on to explore Newtonian systems in greater detail, and contrasted its characteristics with those of Lorenzian systems. What we have not yet done is explored how the "normative theories" of city order, and the metavalue of Newtonian order have affected the practice of city p lanning 4 8 . Now that we have a basic understanding of these theoretical underpinnings, it is time to see how these ideas and understandings in planning theory appear to have affected the real wor ld . We should also remind ourselves that any classification of different world-views can maintain its own pristine order only within the realm of theory. Real people in the real wor ld have multiple motivations and variable perceptions — and while we may state that the leaders of a particular city planning movement perceived the city in terms of one 47 And others. 48 In contrast to purposeful human activity, science holds that there is complete correspondence between the scientific understanding of Newtonian and Lorenzian systems ("theory"), and the actual phenomena produced by these processes ("practice"). Rather than being "rules" followed by outside actor, the scientific world-view posits that these two types of dynamical systems are descriptions of the means by which the universe orders itself. 62 particular world-view, it is doubtful that any such movement has not been influenced by a variety of different theories. 3.1.1. The origins of modern city planning By the beginning of the nineteenth century, the Newtonian scientific paradigm of Newton had substantially replaced the Biblical and Classical world-views within Western Europe and America, and the new dynamism of industrial capitalism had become the primary force in city-making and rural development. It was this new ascendant class of scientifically-minded, capitalist businessmen who would develop some of the first examples of recognizably "modern" city planning. Perhaps the most celebrated "mechanical city" of them all is Manhattan as laid out in 1811 by the Commissioners of that city. Devoid of public open space (Central Park marked a later change in policy), Kostof convincingly argues in The City Shaped'that the 1811 gr id was a purely "mechanical" conception: unbounded or unlimited,. . . it can be extended whenever there is the promise of fast and substantial profit. In this state of affairs the gr id becomes an easy, swift way to standardize vast land operations by businessmen involved in the purchase and sale of land. Public places, parks, and any other allocations that remove land from the market are clearly seen as a waste of a profit-producing resource (Kostof 1991, 121). Nearly simultaneously, on the other side of the Atlantic, another industrial capitalist saw his own venture in city building through the very different lens of the "cosmic city". In his history of the origins of modern town planning, Leonardo Benevolo writes that at the beginning of the nineteenth century factory owner Robert Owen "realized that the 'self-made man' postulated by economists and accepted by current opinion was a mere abstraction, since for the most part environmental conditions determined individual destinies" (Benevolo 1967, 39). Owen's solution was to bui ld New Lanark, a total community within which his fortunate workers worked, ate, and slept, had their minds 63 broadened by exposure to culture, and had their children educated 4 9 (Benevolo 1967, 39-46). The cosmic approach to city building was furthered in France by the explicitly totalitarian schemes of Fourier, and their substantial application by Godin in his Familistere (constructed 1859-1870), among others (Benevolo 1967, 65). There were also a number of Utopian factory towns proposed and constructed within England, and a series of Utopian settlements founded in the United States during the nineteenth century. Despite varying designs, and the differing communistic or religious ideals of their designers, al l were designed to mold their malleable populations into perfect societies (Benevolo 1967) . O n one hand, ambitious rulers of the church or the state with sufficient resources at their disposal might embark upon a program of city planning along overtly cosmic lines, creating baroque palaces, promenades, and monuments (Mumford 1961, 375-409). O n the other hand, seeing their task in a distinctly mechanical light, members of the government or the military carried out more prosaic public works — civic defences, transportation routes, aqueducts, fountains, wells and stormwater drains — within and between major population centres (Hugo-Brunt 1972, 87; Bellan 1971, 179-217). Whi le not deemed as being as historically significant as cosmic city projects, the commissioning of turnpikes, canals and city walls were also extraordinary events, considered justified only for those settlements and territories distinguished by their political, military or commercial importance. Nevertheless, such settlements were the exception: the vast majority of towns and villages in the Western Wor ld from the sixteenth to the mid-nineteenth century were essentially devoid of official planning and purpose-built public infrastructure of either the cosmic or mechanical varieties 5 0 (Barnett 1986, 2). By our own standards, conditions in 49 Intriguingly, Robert Southey, who visited New Lanark in 1819, indicates that Owen's "cosmic city" had rather "mechanical" underpinnings — Owen "literally believes" his workers to be "human machines". Southey himself does not believe "as Owen does, that men may be cast in a mold (like the other parts of his mill), and take the impression with perfect certainty" (Jennings 1987, 157-158). 50 And in pre-industrial times, towns and villages contained the vast bulk of the human population. 64 pre-industrial cities were shocking. For lack of sewers, wastes were often dumped into narrow crooked streets 5 1. Dirty coal was commonly burnt for heat and light, creating thick black smoke which hung in the air. Water was obtained from private or common wells sunk within the town itself, often drawing water from less-than-pristine sources, and liable to contamination by the filth of the city seeping downwards. Housing was predominantly cramped and crowded, especially so within the walled towns of Continental Europe (Mumford 1938, 85; Bellan 1971, 43-52, 206-214). Whi le none of this could have been pleasant, these conditions were not disastrous, simply because the limited size of pre-industrial settlements ensured that urban conditions generally remained tolerable. If the town itself was filthy, access to fresh air, clean water and unsoiled land in the countryside could generally be had by walking for ten minutes in any direction from the centre of town. The poor generally lived on the outskirts of the traditional town, often beyond its civic defences of wal l and gate, trading insecurity in times of war and curfew restrictions in and out of the gates for less cramped and more affordable accommodations in the "banlieue" or "suburbs" to that which they could have afforded within the city (Mumford 1938, 85; 1961, 482-487; Keene 1990, 97-119). 3.1.2. Planning becomes a profession Contrary to popular belief, the Industrial Revolution neither ushered in a new, ruinous method of city building, nor dispensed with previous methods of city construction. Whi le the capitalist developers of new factory towns like Manchester and Liverpool built faster, larger and more efficiently than any of their predecessors, they did not pervert the existing techniques of construction and city development any more than the growth of King Kong 51 Where sewers did exist, they were intended to drain only stormwater. Often slow-moving, the introduction of solid wastes tended to plug them up entirely. Consequently, putting wastes into the sewers was expressly forbidden in both Paris and London (Read 1970, 4). 65 (to use an organic metaphor) could be termed a perversion of simian development. Such organisms were simply far far larger than the process had originally been intended to produce. Unfortunately, both such scaled-up creations prove incapable of functioning properly. While cinematic special effects show King Kong bounding around the skyline of Manhattan, no ape of such dimensions could have even stood up without rupturing under its own weight. In like fashion, the overgrown nineteenth-century city imploded into a state of high-density squalor, congestion and endemic disease. The sheer size of the factory and manufacturing centres of the mid-nineteenth century created tremendous problems. These were so large as to preclude commuting on foot, the only means of transportation affordable to the huge class of menial labourers employed in the industrial districts of the city (Wohl 1977, 1, 29-30, 4 1 ) 5 2 . Thus prevented from living on cheaper land in the suburbs, workers and the proletarian underclass found they could afford shelter in the city only by submitting to unprecedentedly crowded conditions within the factory districts (Hall 1988, 18-19; Mumford 1961, 431-433). Narrow roads, which had been adequate for the needs of local low-density pack-animal and pedestrian traffic in the small traditional town proved disastrous in the new cities, with their exceedingly high population densities. The sheer volume of people and goods on the move created a vicious cycle of decreasing travel speeds and increased road traffic, subjecting the streets of the industrial city to near-constant gridlock (Bellan 1971, 191; Barnett 1986, 91; Benevolo 1975, 921). The miseries of travel were only compounded by the maintenance of the old regime with regard to waste disposal. Industrial cities were very literally awash in "pea soup", the infusion created when the wastes dropped onto the streets by millions of humans and 52 Conversely, Jennings includes a anecdote about a London birdseller, written about 1850, who found it increasingly difficult to commute from the city on his birdcatching expeditions (Jennings 1985, 246-247). See also (Jennings 1985, 277-279). 66 horses steeped in the stagnant water of the sewerless cities (Jukes 1990, 19). A comparable miasma hung in the skies of the city; contemporaries would describe London as the "City of Dreadful Night", blackened by a smog of coal-smoke unimaginable even to the present-day Athenian or Angeleno (Hall 1988, 14; Mumford 1961, 469-474) 5 3 . The air and water of the city were not simply unpleasantries of urban life — they were positively pathological for those trapped within it — waves of cholera swept London in the mid-nineteenth century, and the stench of putrefying sewage threatened to close down the Houses of Parliament during the 1850s (Read 1970, 14-15). Planning as we know it today developed out of Victorian society's belated recognition that the industrial city had created concentrated masses of diseased, deprived, and desperate people, in the hearts of their greatest cities no less, who had no reason to stop u short of violent revolution (Hall 1988, 24-28). Faced with an urban structure which had patently failed to serve the needs of its residents, and which now threatened society itself, a professional class of architects, surveyors, engineers and bureaucrats emerged to re-order the city, if only to prevent the triumph of insensate social disorder in the form of revolution. Curiously, given the clarity of the distinctions between the "cosmic" and the "mechanical" and "organic" cities, the actual practice of city planning initially proved to be remarkably catholic in its practice 5 4 . 53 See also Jeffries (1986, 136-137). 54 Nothing in the history of planning blends the visions of all three normative theories together as successfully as the Parisian boulevards of Eugene Haussmann, largely constructed during the 18 50s and 1860s. Credited by his biographer as being the first professional city planner, later visionaries of the "cosmic", the "mechanical" and the "organic" city have all looked back to his accomplishments for inspiration. A "cosmic" planner, Haussmann took a confusion of isolated monuments and related them to each other, by this means turning Paris into a focused and unified whole, a fitting seat of Empire. The "mechanical" Haussmann drove new thoroughfares through the densely populated metropolis, drastically increasing the transportation efficiency within the city (although not to the exclusion of military applications). Underneath the new boulevards Haussmann installed a comprehensive system of sewers and watermains, designed to improve the public health (and productivity) of working-class Parisians. While concerned to unclog the mechanism of the central city, Haussmann also intended the boulevards to accelerate the outward growth of the city: indeed, his scheme was largely 67 Given the fears and the outlook of those who ran society in the Victorian Era, it is hardly surprising that the world-view of the "cosmic city" was most prominent amongst early planners. Social revolution and the threat of the mob was to be prevented, at least in large part, by making sure that the influence of the city's form strengthened, rather than weakened the existing social and political structure. The Utopian experiments of enlightened industrialists like Owen and Fourier were much discussed during this time, although the socialist ideals implicit in these projects had already been hived off by the 1850s, when a new generation of Utopian plans for factory towns like Victoria (1849) made use of environmental determinism to forestall the socialist millennium rather than to speed it up (Benevolo 1967, 129-133). By the 1880s and 1890s American activists were pushing a decidedly conservative version of Owen's millennial doctrine of environmental reform, arguing that the provision of tidy, orderly living quarters led to the creation of contented workers with unthreateningly middle-class values — a tidy, orderly society (Boyer 1983, 18). Personal tidiness and morality was more than a individual issue, for both the symptoms and the effects of disorder were societal in scope, threatening the future of the nation as a whole. The underwhelming appearance of cities in the United States, built for convenience and profit rather than posterity, had "generated a public acceptance of municipal slovenliness and chaos," its constant change and uncontrolled dynamism preventing any feelings of loyalty from its citizens (Boyer 1983, 42). As Christine Boyer points out in Dreaming the Rational City (19S3), these activists genuinely "felt that the character of civic education financed by aggressively extending the axial routes into the vacant civic-owned lands surrounding Paris. Finally, biographer David P. Jordan stresses Haussmann's organicism: "[Haussmann] envisaged the city as an organic whole— the body was his favorite metaphor for Paris — having new and old parts, all functioning together" (Jordan 1995, 371). This perspective, Jordan argues, was particularly apparent in his infrastructure and parks - indeed, even Lewis Mumford praised Haussmann's planning of Paris (Jordan 1995, 267-296; Mumford 1968, 110). In sum, Haussmann's work forms the unifying ideological arch at the central etoile of modern planning practice. 68 and its ceremonial form would determine in the long run whether democracy and civilization would succeed in the American city" (Boyer 1983, 43). These environmental improvers intended to impose order on American cities, which were characterized by chaos, ugliness, discord and filth. The urban challenge was to control unruly growth, tidy the chaos, and return order to the arrangement of buildings.... No visual stench escaped the general crusade; every moral and spiritual consequence, every ideal of democracy and freedom were claimed to gather support and strength from these physical reforms (Boyer 1983, 46). Civic design was also seen as a societal influence by the later City Beautiful movement, which had its heyday in the two decades spanning the turn of the present century. Planners like Daniel Burnham designed elegant Haussmann-like boulevards, expansive parks and ostentatious neo-classical civic centres for the unadorned gridwork cities of capitalist America. ...against the chaos of the city with its simultaneity of land uses, jumble of vehicles, multitudes of people, corrupt politicians, and labor unrest, there stood an ideal: the city as a perfectly disciplined spatial order" (Boyer 1983, 60). Such projects were intended to create civic pride and an ambiance of culture; for "cosmic" believers in the concept of environmental determinism, the need to instill the values of American democracy in the masses of the unassimilated poor made these objectives pressing indeed. Despite their considerable differences, the schemes of the Utopian socialists, those of the bourgeois environmental reformers and the "boosters" of the City Beautiful movement were united by their "cosmic" conviction in the power of urban design alone to transform public morality. Nowadays, however, the City Beautiful movement is remembered more for its plans and proposals than its actual accomplishments. Indeed, the heyday of the movement in the first decade of the twentieth century would prove to be the last great hurrah of the "cosmic city" planners in the West. Certainly, its greatest champions since 69 this time — Albert Speer and Stalin — have made the whole concept of overt societal modification through design seem incompatible with pluralist democratic ideology. Instead, it was the "mechanical" and the "organic" conceptions of the city — and the assumptions of Newtonian systems which underlay them — which would prove to be the predominant forces in the shaping of modern planning practice. While this value system would become dominant only in the twentieth century, P.J. Smith's 1979 examination of the influence of the scientific mentality in planning traces its roots back to the utilitarian-minded activists in the "public health" movements of England during the mid-1800s. These pioneer planners were shocked by the same conditions as the "cosmic" environmental determinists, but viewed the problem in different, and distinctly "mechanical" terms: The environmental defects of the nineteenth-century city were an offense against many sensibilities, but, in the utilitarian view of the world , they were, above all , an offense against reason. Most simply, they led to enormous and intolerable waste. ... Efficiency became the practical goal of reform, spawning a flock of catchwords, such as health, economy, convenience, and order, which persist to this day as catchwords of planning (Smith 1979, 199-200). The stress these British reformers laid on taking "a scientific approach" was understandable, given that many of them were scientists themselves. They conceived of planning as "a rational scientific activity", and "advocated the use of scientific knowledge to cure the problems of the day, to bring order and economy out of chaos and waste (Smith 1979, 200). A similar group of pioneering planners stressing "scientific methods" emerged in late nineteenth century in the United States. The conservation movement was, in large measure, an effort of scientists, engineers and other technicians to substitute planned, centrally controlled resource development for the individual, inefficient, and wasteful exploitation of the past. The social ideal of the disinterested technician, the impartial servant of society, greatly influenced the thinking of conservation leaders, who viewed the expert as the key figure in a new age of planned economic growth, the savior of a society which could no longer afford the physical and human waste of days gone by (Lubove 1962, The Progressives 70 and the Slums: tenement house reform in New York City. Pittsburgh, p.215, in Smith 1979, 201). By the 1920s this new planning outlook, based on scientific Newtonian systems had clearly become the dominant planning method. The Canadian Journal of the Town Planning Institute broadcast the new mechanical-organic credo on the masthead of every issue: Town planning may be defined as the scientific and orderly disposition of land and buildings in use and development with a view to obviating congestion and securing economic and social efficiency, health and we l l -being in urban and rural communities (Smith 1979, 214). A comparison of this statement with the assumptions held by pre-prpfessional planners reveals how the new mechanical-organic vision, based on the scientific axioms of Newtonian systems, drastically transformed attitudes regarding the need for explicit city planning. As we noted above, under vernacular planning practices, it had been understood that settlements could be left to develop on their own, except where extraordinary factors compelled specific, purposeful actions. Seen from the new point of view, such an approach was as nonsensical as firing all of the workers in a factory and expecting the machinery to run itself, or surgically removing the brain of some creature on the understanding that it could continue to function as efficiently as before. The new scientific planners assumed that order could only be artificially imposed, and that without purposeful maintenance, such imposed order would quickly decay back into chaotic disorder. 3.1.3. The ascendancy of synoptic planning For mechanical-organic planners seeking to increase order in an entropic Newtonian world , nothing could be left to chance and randomness. Fortunately, the science of the time assured them that perfect control was possible. As we have already seen, Newton's axioms dictate that all phenomena, including "random variations" themselves, are the products of predictable, deterministic process. Therefore, given sufficient knowledge of present conditions were available, and adequate control over external sources of 71 interference were granted, it stood to reason that the future state of any system could be determined. Moreover, the future could be designed with confidence: within a Newtonian universe of Newtonian processes all future actions become predictable with sufficient knowledge of their initial states, making perfect forecasting, and hence perfect decision-making possible. Nor was this relation only true for the laboratory — Newtonian systems were universal. Consequently, while the task of effective city planning would require huge quantities of information, and a level of control over civic phenomena never before attained, the goal was achievable, and its benefits were enormous. The shift in planning world-view from the "cosmic" perspective to the "mechanical-organic" vision gave rise to a tremendous expansion in the scope of activities which the planner was required to perform. In order to stave off societal collapse, it was necessary to gather al l the information, consider all the possible options, and choose that option which would deterministically force future developments into the order desired, resulting in a remarkably totalitarian planning method. In Hudson's 1979 review of planning methods, this Newtonian mode of planning would be referred to as the "synoptic planning" method: a remarkably apt name for an approach where so much emphasis is placed upon "seeing a l l " . The first expressions of "synoptic planning" method in North America, pointedly termed "the City Efficient" and "the City Functional", eclipsed the cosmic City Beautiful approach by the 1920s (Smith 1979, 201). Indeed, in a manner reminiscent of a Kuhnian "paradigm shift", the old vision of order, with its non-scientific notions of magic, stasis and moral suasion no longer even made much sense to the new planners. Boyer perfectly captures the shift of the dominant planning vision of order from the 'cosmic' to the 'mechanical-organic' when she notes that after 1920, "planners became preoccupied with the physical layout and mechanical arrangement of the American city", but regarded the 72 City Beautiful urge to "ornament" the city as nothing more than "fancy excuses for authorizing raids on the public pocketbook" (Boyer 1983, 154). From the synoptic perspective, cosmic order was not fiscally efficient, and as such hardly constituted order at al l . While the particular vision of "order" had changed, the desire of planners to "order" the city was as strong as ever. Like the environmental determinists they replaced, the new orderers needed to rearrange the city in conformance with their values of efficiency and scientific rationality. Moreover, the synoptic planners needed to go about ordering the city in a far more information-intense, comprehensive and potent manner than had ever been previously required. Boyer argues that with the eclipse of the City Beautiful by the 1920s, planners shifted from an emphasis on concrete "civic center monumentality" to creating an abstract urban order through the mechanism of zoning (Boyer 1983, 154). Smith concurs, noting that zoning was seized upon precisely because it was "seen as a "scientific" tool for improving the quality of the urban environment (Smith 1979, 202). Indeed, nothing less than a "scientific" method would suffice for those who judged order by the Newtonian concepts of Newton: Zoning, it was claimed, embodied and exemplified the erection of the right building, in the right form, in the right place. ... Height districts and bulk restrictions were also needed to secure the rational order of the American city: to guarantee that the proper amount of light and air would surround tall buildings and that traffic congestion on the streets below was held to a min imum (Boyer 1983, 155-156). The practice of synoptic planning also required improved information-gathering and decision-making capacities. These aspects of synoptic planning came into their own with the onset of the Second Wor ld War , as strategic decision-making quickly became a matter of life and death for combatant regimes. Vastly increased resources were dedicated to developing decision-making structures, and qualitative modeling and prediction techniques. The new state-sponsored focus upon information gathering and analysis, 73 cybernetics and systems management continued under the new strategic rivalry of the Cold War , laying the groundwork for the Technological Revolution and the development of computers (Nash 1989, 7-9). In retrospect, Melvi l le C. Branch's 1966 book Planning, aspects and applications appears to exemplify the high-water mark of confidence in synoptic p lann ing 5 5 . Branch is almost a parody of the post-war synoptic planner: a strategic planner working for the American military, he employs the design process for a nuclear missile command centre as a case study. Despite his unconventional predilections, Branch's rationale for planning rests on scientific Newtonian systems, and he clearly sees planning in an orthodox mechanical-organic fashion, equating planning with efficiency: ...for a given expenditure of time, energy, materials, and money by municipal government, what techniques can be devised to show more scientifically the best apportionment between ... different activities... — "best apportionment" being that allocation which maximizes the benefits for the entire urban organism in accordance with established objectives (Branch 1966, 300). Plainly, for Branch, planning is a rational, scientific activity — indeed, it is through increasing rationality that progress is to be made in the field: ...with a more analytical basis of comparative evaluation progress is possible toward the substitution of greater objectivity for acquiescence to pressure, purely subjective interest, or personal opinion derived without supportive analysis (Branch 1966, 300). Branch also lays considerable stress on forecasting in planning: Projection into the future is as essential to planning as coordination. Projection is made possible by the mathematics of probability, trends based on accumulations of data, and empirical and historical evidence. ... Since increased foresight is characteristic of the advance of knowledge, more reliable prediction is also to be expected; hence the frequent application of "ability to predict" as the most meaningful measure of substantive achievement in any field. Planning, of course, is projectional by definition, 55 Termed "comprehensive" planning by Branch. 74 and its significance w i l l grow as techniques of projection become more reliable and are extended in scope and time (Branch 1966, 301) . 5 6 With increased powers of computation and foresight, synoptic planners were now confident of their ability to draw up "comprehensive plans", attempting to determine the future order of entire regions so as to make them suitable for the predicted demands of predicted future residents. Thus, the Camelot-era technocrats who developed A Plan For the Year 2000Tor the American capital did not stop at the District of Columbia, but also ordered much of Virginia and Maryland into a space-age asterisk of new cities that would be plainly visible from Sputnik or The Moon . As ever, the synoptic architects of the plan stressed the need for creating a purposefully ordered "pattern" of development: Our future depends... on the design of the Region — on the creation of a pattern of growth that w i l l produce the best possible environment for ourselves and future generations. Good design w i l l reduce traffic congestion, protect water supplies, provide adequate space for parks and recreation, create efficient commercial centers and livable residential neighborhoods, produce a suitable setting for the nation's Capital and reduce the costs of local government. Good design w i l l give us delight in the visual quality of our urban environment. Poor design w i l l increase the cost, frustration and visual chaos that each of us experiences in working and living at close quarters with several mil l ion other people (NCPC 1961, iv). 3.2. The Crisis of Faith in Synoptic Planning As it was, the confident optimism of Branch and the federal planners already seemed out-of-date to many even when these works were published. The late 1950s and early 1960s saw the publication of a series of articles arguing that the rational, synoptic method espoused by planners was not actually employed by them in the real wor ld because of its impracticality. 56 Ironically, Branch's list of areas in which data accumulation has allowed increased levels of prediction includes "weather forecasts, ... biological population estimates,... and a few economic prognostications" (Branch 1966, 301), phenomena which have since been linked to chaotic systems and unpredictable behaviour. 75 In what proved to be a watershed paper, Charles E. Lindblom began "The Science of 'Muddl ing Through'" with a poker-faced description of the steps a synoptic planner might be imagined to go through in reaching a particular decision (Lindblom 1959). In order to make a fully-informed rational decision, noted Lindblom, synoptic planners agreed that it was crucial to obtain sufficient amounts of information as well as a full range of possible options, before choosing between these options with a well-identified set of values. Ironically, fulfilling this injunction in real life was plainly inefficient. Lindblom notes that the true synoptic planner "formulating policy with respect to inflation" would have to "list al l related values in order or importance", having determined through public consultation "how much of each value is equal to how much of each other value." The planner would then be forced to consider every alternative, from "the abolition of prices" to an entirely unregulated economy and every point in between, " in the light of whatever theoretical generalisations [the planner] could find on such hypothetical economies" (Lindblom 1959, 151). Lindblom contrasts this patently impractical procedure with what synoptic planners actually do in practice. The goal could be set "either explicitly or without conscious thought" as the "relatively simple goal of keeping prices level," deliberately "ignoring many related values and many possible important consequences of his policies". Next, "those relatively few policy alternatives that occurred to [the planner]" would be evaluated. This evaluation: would rely heavily on the record of past experience with small policy steps to predict the consequences of similar steps extended into the future. ... Because practitioners of the second approach expect to achieve their goals only partially, they would expect to repeat endlessly the sequence just described, as conditions and aspirations changed and as accuracy of prediction improved" (Lindblom 1959, 151-152). While Lindblom tries to convert the necessity of "muddling through" into the virtue of planning "incrementally", the real message of the paper seems to be that synoptic planners 76 are unable to process the amount of information that they themselves maintain is crucial to proper decision-making. Other authors soon weighed in with their doubts about the practicality of planning synoptically. Alan Altshuler's 1965 paper on "The Goals of Comprehensive Planning" undermined the synoptic planners' parallel claim to control over civic development, and exposed the extent to which practicing planners found it impossible to determine in advance what the effects of their actions would be. Although the performance of their duties requires planners to "possess causal knowledge which enables them to gauge the approximate net effect of proposed actions on the public interest" (Altshuler 1965, 186), Altshuler argued that: [planners] cannot conceive means that w i l l further the operational goals of primary interest to them without also affecting innumerable others in uncontrolled fashion. Many planners recognize this, and try not to serve their stated operational goals exclusively (Altshuler 1965, 191). Like Lindblom, Altshuler concludes that short to medium-term "incremental" planning is the only viable alternative. But this too, is suspect. Constance Perin's 1967 article " A Noiseless Secession from the Comprehensive Plan", noted that the continual revising and revisiting central to the new "incrementalist" approach meant that very little actual planning was ever completed. This lack of productivity, she suspected, was the real attraction of the method for planners who felt incapable of making wise and effective decisions for the future based on current information: The 'incompetence' being masked by endless and diffuse studies may relate closely to the fact that analytic work in city planning has yet to make its peace with the tolerable range of error appropriate to each topic it deals with: fear of being found 'wrong' in a recommendation or 'incomplete' in the range of variables studied has led to an abuse of 'open-endedness' and 'flexibility' as important to the planning 'process' (Perin 1967, 337-338). 77 Taking Richard Bolan's 1967 paper "Emerging Views of Planning", as a keystone, Christine Boyer summarized the new pessimism of the 1960s over the possibility of accurate prediction: Political scientists, economists, local politicians, and planners agreed that the urban future could never be accurately predicted, that community goals in a turbulent world remained elusive, that information would always be indeterminate, that a decentralized democratic political system made comprehensive planning from a centralized authority impossible (Boyer 1983,285) . Moreover, nothing has changed since then. A generation later, in 1979, Smith wrote: There is no acknowledgment that planners must work in an environment of high uncertainty, with a future that is essentially unknowable, or that they are acting in a pluralistic society in which conflict is inevitable, a society not bound by a single, universal conception of utility but is fragmented into a multitude of individual, equally just utilities. Even the catchwords behind which planners have shielded themselves — order, economy, efficiency — turn out not to be absolutes; they are as value-loaded as good or happiness or beauty, and no more practical as operational criteria (Smith 1979, 216). By the mid-1960s it was clear to the theorists that synoptic planning method had failed to live up to its own bil l ing as a rational, scientific activity. Worse yet, these shortcomings in process appeared to be nothing less than inevitable within "real wor ld" planning practice — indeed, rather than seek to improve existing practices, Lindblom's approach struck many as a post facto rationalization of the status quo. Unfortunately, this theoretical attack on the process of synoptic planning would soon be accompanied by a landmark 1961 polemic condemning its products. Entitled The Death and Life of Great American Cities, author Jane Jacobs left no doubt that city planning, rather than rationalizing and increasing the efficiency of urban centres, merely ordered them into lifeless, crime-ridden geometrical abstractions inimical to the fine-grained economic life of a metropolis. This was not simply a question of imperfect method or inadequate results: Jacobs declared that her book was "an attack ... on the principles and aims that have shaped modern, orthodox city planning and rebuilding"(Jacobs 1961, 4) 78 Planners have ignored the study of success and failure in real life ... and are guided instead by principles derived from the behaviour and appearance of towns, suburbs, tuberculosis sanitoria, fairs and imaginary dream cities — from anything but cities themselves (Jacobs 1961, 6). In particular, Jacobs decries the Utopian visions of Le Corbusier and Howard, champions of the mechanical and organic city respectively, who had, Jacobs claimed, jointly led planners to destroy the city through low-density sprawl and homogeneity, which she characterized as "The Great Blight of Dullness" 0acobs 1961, 17-25). Again and again Jacobs excoriates planners for their destruction of cities through the red-lining of areas whose residents are trying to improve their neighborhoods, the banal uniform zoning of areas that derive their livability precisely from their diversity, the construction of anonymous Corbusian housing projects and bandit-friendly parks, and the abject worship of the automobile. Jacobs' work was not simply an attack on synoptic planning: her assumptions were inimical to the entire Newtonian mechanical-organic world-view. In seeming contravention of modern scientific truth, and in common with the pre-modern planners of more than a century before, Jacobs was convinced that cities were self-ordering: Vital cities have marvelous innate abilities for understanding, communicating, contriving and inventing what is required to combat their difficulties.... The surplus wealth, the productivity, the close-grained juxtaposition of talents that permit society to support advances ... are themselves products of our organization into cities, and especially big and dense cities (Jacobs 1961, 448). This anti-entropic perspective was obviously non-Newtonian. Given the time, four years before Lorenz's first overlooked article, it could not have been taken as anything but anti-scientific as well . The result was remarkably discomforting for planners. Whi le many read the work, and could not help but agree with many of her cogent observations, planners found it almost impossible to integrate her insights within their own practices. Jacobs held aloft the orange, but planners found that they could not relinquish Newton's apples in their quest for marmalade. 79 By the mid-sixties, academic planners were faced with the spectre of a discipline whose practices were methodologically unsound, with little hope of improvement. The products of planning had been revealed to be ideologically bankrupt and counter-productive in practice, often exacerbating rather than resolving problems of civic "blight", transportation and equity. In the wake of these wholesale critiques of synoptic planning, it became obvious to most planners that the old discipline of synoptic planning was limited, 'old-fashioned' and 'out of date'. Nevertheless, the way forward was less than clear. 3.2.1. The synoptic diaspora The last thirty years of planning theory have witnessed a veritable explosion of new planning methods and manifestos, al l advanced in hopes of overcoming the identified shortcomings of the synoptic planning model. In 1979 Hudson organised this ideological diaspora into five tribes: • The synoptic method • The incremental method introduced by Lindblom • The transactive method championed by Friedman which stresses the "social learning" which occurs between the technically expert planners and the socially and locally expert residents • The advocacy route of having planners join ideological combat against the powerful on behalf of the marginalized, and • The radical perspective which argues the futility of attempting change within a system designed to serve the interests of the dominant class (Hudson 1979). And yet, despite this plurality of ideas, Hudson concludes that none of the subsequent reactions to "synoptic planning" have ever replaced it, or offered an alternate basis on which planning might be done. Instead: advocacy, transactive and radical planning practices have appeared on the scene as countervailing methods to ongoing processes of synoptic planning, not with the result of replacing the dominant paradigm, but of introducing a broader perspective on issues and another set of voices for articulating the public interest (Hudson 1979, 396). Almost twenty years later, Hudson's typology seems incomplete, although his conclusion 80 seems as valid as ever. More and more reactions to the crisis in synoptic planning have gathered strength, but none of them dispense with its compromised ideological foundation. Significant movements with planning theory and practice now include: • "Sustainable planning" (Rees 1994; Rees and Wackernagel 1995) and "bioregionalism" (Aberley 1987;1994), which shed doubt on the Industrial Era nostrum of unlimited growth, and seek to reconcile the human economy within the ecosystem it depends upon, albeit without relinquishing the notions of prediction and control; • "Neo-traditional design" (Duany and Plater-Zyberk 1991; Calthorpe 1993), which explicitly seeks to recreate the urban order generated by traditional, "cosmic" and even "City Efficient" world-views, although largely though synoptic means, and • "Gender" (Sandercock and Forsyth 1992), and "life-cycle planning," a neo-organic movement which seeks to lessen the imbalance between the preoccupation that synoptic city planning displays with productive workers (adult males), and its lack of attention for reproductive or non-productive elements in the city (women, children, the elderly). Whi le synoptic planning itself has not been widely practised since the start of the 1970s 5 7 , a raft of modified, moderated, extended and mitigated variants ranging from "public participation" to ecologically-minded "livable region strategic plans" still comprise the dominant mode of planning relied upon today. The results have been varied. O n one hand, the underlying goals and assumptions of planning have changed far less than one might suppose when surveying the smoke and confusion out on the fields of planning theory. Wri t ing in 1979, Alberta historical geographer P. J. Smith noted that neither the phraseology nor the intent of planning legislation — "to achieve the orderly and economical development of land" — had changed in sixty years (Smith 1979, 215): planners are still going about the business of efficiency and rationality on Newtonian precepts. The other side of the equation, however, is a tremendous exhaustion and bewilderment within the profession, as well-meaning planners continue to try to do their jobs, knowingly fore-ordained to failure either through methodological shortcomings, 57 Except in the field of transportation planning, where the top-down technocratic approach is still dominant. 81 through unintended consequences, or even through the very success of a misconceived scheme. Boyer argues that the latter half of the twentieth century has seen physical planning "[abandon] once and for all its traditional focus on the physical order of the American city." The "pursuit of an order for the American city" notes Boyer, "persisted over decades in spite of its obvious failures to implement its plans." (Boyer 1983, ix). This is debatable — as we have seen — planners shall always wish to order the city in terms of their values. What is likely, however, is that planers themselves no longer know what their values are, and have been reduced to going through the motions, ordering for its own sake. Something of this vacuous despair shows through in the 1983 Community Plan for the Vancouver-area suburb of Surrey, which reaffirms the planner's central task of ordering the city without ever giving a reason why: It appears that all successful and long-lasting settlements forms resemble patterns. ... In contrast, it appears that today's metropolitan areas, outside of the historic core, either ... do not conform to a pattern or ... are the beginnings of a pattern which we do not yet quite understand.... Since most metropolitan areas do not seem to form patterns... the question arises as to whether any future form of suburbia can be predetermined. ... [But i]t is one of the convictions of this [plan] that that future of a metropolitan area needs to be shaped and an attempt was made, however tentative, to predetermine it. ... (District of Surrey 1983, 15). Planning's crisis of faith now looks like a permanent feature of the landscape — indeed there are now two generations of planners who have known no other way. It is now regarded as being rather improper to suggest a definition of planning, or to advance a set of goals for the profession 5 8 . In a manner commensurate with Kuhn's model of ideological change within the sciences, planning has been trapped in a "state of crisis" for several generations now. None of planning's recent enthusiasms have managed to institute a 58 This makes my own definition, or indeed, any definition of the goals of "synoptic planning" quite objectionable to some planning academics. 82 wholly new paradigm of planning, set upon different theoretical foundations, for none question the fundamental premises of synoptic planning, rooted in the equation of order with Newtonian order. Unt i l this happens, and a proper answer to Lindblom, Altshuler and Jacobs' critiques is made possible, the state of crisis w i l l continue. 3.3. Planning Theory and Chaos Theory In the midst of their intellectual crisis, the scientific recognition of Lorenzian systems appears to have largely passed planners by. While chaotic systems are still the "new science" James Gleick memorably described it as, Chaos Theory has been discussed in scientific journals since the late 1970s, and has been accessible to the educated public for almost a decade, through the popular science writings of Gleick, the widely distributed fractal images of Mandelbrot and others 5 9 . The new insights of Chaos Theory have been seized upon in a wide variety of disciplines, with early applications being made in astronomy [orbital dynamics] (Stewart 1989, 243-261), physics [turbulence] (Gleick 1987, 121-138), ecology [predator-prey populations] (Gleick 1987, 78-80), medicine [heartbeats] (Gleick 1987, 281-291) and meteorology [weather forecasting] (Lorenz 1993, 77-110). In the past fifteen years, an increasing range of social science disciplines have additionally been applying the insights of Chaos Theory to issues within their own fields of study, including such planning-related disciplines as economics [long-term economic development and stock market behaviour], and geography [urban modeling] (Dendrinos 1992; Batty and Longley 1994; Wong and Fotheringhaml990) G 0 . 59 Gleick published his best-seller in 1987. References to Chaos Theory have even made it into pop-culture entertainment like Jurassic Park. 60 And, as mentioned in a previous footnote, several other discipline have been actively trying to evade the more disturbing aspects of Chaos Theory by perturbing Lorenzian systems into familiar and controllable Newtonian ones (Strogetz 1995; The Economist 1994). 83 A n d yet to date, there are only a handful of articles and books relating Chaos Theory to planning theory issues. T.J . Cartwright's 1991 article, "Planning and Chaos Theory" is far and away the most important of these papers, offering a concise explanation of chaotic systems , and a brief exploration of the implications that "chaos" has for planning. Cartwright notes that the existence of chaotic systems puts the now-familiar questions about prediction and accuracy in a new light: M a n y planners — among them David Braybrooke and Charles Lindblom (1963), Andreas Faludi (1973), and Yehezekel Dror (1983) — have argued that we need to learn to plan with incomplete information and have proposed strategies and techniques for doing so. But they have done so for largely pragmatic reasons; because there is not always time to get all the facts, because it is sometimes too expensive to do so, or because we lack the necessary means or skill . ... But what chaos theory suggests is that planning based on prediction is not merely impractical in some cases; it is logically impossible (Cartwright 1991, 45). Most planners assume that, given enough information, they can anticipate what is going to happen in a particular situation and thus can determine how best to act so as to promote, defer, deflect or divert it, as may be required. Chaos theory suggests that, on the contrary, some systems are inherently unpredictable and can never be fully understood, no matter how much effort or expense is devoted to trying (Cartwright 1991, 53). In general, Cartwright is happier at pointing out the challenges posed by chaotic systems than in outlining possible solutions. Nevertheless, he does give a few indications of how he believes the discipline must react in the face of the newly-recognized laws of nature. Cartwright points out that "gathering more information or constructing more elaborate models about chaotic systems can become pointless [or] ... counter-productive, if it creates a false sense of security about planning and what it can do" (Cartwright 1991, 53). Cartwright argues for having an "ensemble of forecasts", as in meteorology, to find out "which initial points in phase space have similar trajectories and which do not" (Cartwright 1991, 53). Rather than looking for ever more detailed information on and ever more accurate models of their systems, planners should look instead for "patterns of system behaviour" even though these "attractors" may not be predictable (Cartwright 1991, 54). 84 Cartwright also helpfully points out that accurate short-term prediction is possible in a chaotic system. Clearly, Cartwright wants planners to acknowledge and accept the newly-recognized reality of Lorenzian systems — but he does not seem to countenance giving up the basic predict-and-control basis of modern synoptic planning. Cartwright appears to be more flexible on the entire issue of "order": Another fundamental implication of chaos theory is that we must learn to rethink some of our deep-rooted beliefs in the virtues of order and predictability and the 'untidiness' of chaos and disorder. In other words, we must learn to accept the possibility that a chaotic city, for instance, may be preferable to and 'healthier' than an orderly one. It may even be that humans need chaos in order to survive — that chaos is an essential ingredient in the way we manage our lives in an infinitely complex wor ld (Cartwright 1991, 54). Finally, Cartwright clearly states the scale of the ideological shift required of those who are to plan from the perspective of Lorenzian systems. Quoting Francis M o o n , it is clear that Cartwright's own view of planning history parallels the one presented within this thesis: The view that order emerged from an underlying formless chaos and that this order is recognized only by predictable periodic patterns was the pre-dominant view of 20th century dynamics until the last decade. What is replacing this view is the concept of chaotic events resulting from orderly laws [-] not just a formless chaos, but one in which there are underlying patterns, fractal structures, governed by a new mathematical view of our 'orderly' world (Moon 1987, quoted in Cartwright 1991, 54). Chaos theory promises to have profound implications for what planners do and how they do it. It suggests that the world may be both easier and more difficult to understand than we tend to believe, that noisy and untidy cities may not be as dysfunctional as we often assume, and that the need for planning that is incremental and adaptive in nature may be more urgent than we tend to think" (Cartwright 1991, 44). Although it falls outside of the discipline, Brenda Zimmerman's 1991 dissertation in Business Management, Strategy, Chaos and Equilibrium, is clearly of relevance to planning theorists, and aspects of its approach and technique have been utilized in the current paper. Zimmerman's work does not provide the same grounding in chaotic systems that 85 Cartwright does, but does clarify that she is employing Chaos Theory as a metaphorical or explanatory tool, rather than as a causal explanation of corporate practices — she thus seems to regard an detailed exploration of chaotic systems as somewhat superfluous (Zimmerman 1991, 1-2). At the beginning of the paper Zimmerman presents a list of eight characteristics of chaotic systems which she contrasts with those in "equilibrium models" arising from Newtonian order 6 1 . These characteristics, and the world-views they embody are then used to evaluate the business practices of Federal Metals, Inc. Zimmerman concludes that viewing the case study through a Chaos Theory lens can only highlight and clarify certain aspects of the behaviour of Federal Metals, but notes that these insights are inaccessible from a conventional perspective (Zimmerman 1991, 4, 330). Zimmerman also makes the point that the new perspective is valuable not because its insights are unprecedented, but because it weaves together previously isolated concepts into a robust, interlinked whole (Zimmerman 1991,4) . The only other works of planning (or management) theory to address Chaos Theory are of minor importance. Donald Scott Slocombe's 1990 thesis, entitled Complexity, Change, and Uncertainty in Environmental Planning, seeks to develop a new method of analysis in 61 Zimmerman's list is as follows: 1) Sensitive dependence 2) Non-equilibrium states leading to new patterns of order 3) Topologically transitive structures which maintain their essence while remaining flexible: "a pattern may take shape without there being any need for fixed points either temporally or spatially" (Zimmerman 1991, 31) 4) The role of fluctuations and positive feedback 5) Bifurcation points 6) Interplay of chance and necessity 7) Non-linear and recursive systems 8) Self-organizing processes (Zimmerman 1991, 3). The large and obvious differences between Zimmerman's list and the one presented in this thesis are largely a result of Zimmerman's wider scope (point 5 deals with the "transition to chaos", point 6 addresses random behavior, and point 8 refers to related, but distinct, phenomena examined in "Complexity Theory") and more relaxed approach (points 1, 4 and 7 are somewhat redundant). 86 regional and ecosystems planning by adopting a "non-equilibrium" wor ld -v iew 6 2 . Slocombe argues that this perspective, derived from theories of non-equilibrium processes such as chaos and self-organization, is suited to understanding and working within complex, evolving sociobiophysical systems at regional scales (Slocombe 1990, abstract). Regrettably, the muddy case studies together with the lack of a clear rationale, prevent the reader from gaining much insight from Slocombe's w o r k 6 3 . Thus far, the only planner or urban critic to explicitly evaluate planning practice from the perspective of Chaos Theory is David Engwicht of Australia, in his book Reclaiming Our Cities and Towns (1993). Engwicht defines the city as a device for the maximization of social interaction, and condemns the segregated, homogenized, rationalized city as one which can only provide planned interactions, preventing the fortuitous exchanges equally necessary for the continued development of the city and its inhabitants. However, in his characterization of city form and civic interaction, Engwicht appears to rely upon the ideas Richard Sennett presents in The Uses of Disorder (1970), for Engwicht describes these urban interactions as being truly random and non-deterministic (Engwicht 1993, 16-17) 6 4 . Engwicht then conflates these truly random interactions with the seemingly random Lorenzian phenomena studied by Chaos Theory, and argues for incorporating Chaos 62 While derived from Bifurcation Theory and the "transition to chaos", rather than Lorenzian systems themselves, the work of C.S. Holling should also be noted in this contect. Holling argues for a practice of 'adaptive management' in nartural resources management which accepts the inevitability of "surprise" and "uncertainty" (Holling 1978). 63 Slocombe does not provide sufficient enough explanation of the concepts of "non-equilibrium" states, self-organization, or of "chaos" to enable the reader to draw a link between these concepts and the features of his new method. This lack of clarity about the reasoning behind the method is compounded by the seemingly poor choice of case studies. Rather than using small, well-documented case studies where unpredictable "non-equilibrium" characteristics could be highlighted, Slocombe employs two huge portions of North America - the vast Kluane/Wrangell wilderness area and the (even larger) Great Lakes basin. With information unavailable or aggregated into meaninglessness, the "uncertainty" arising from chaotic dynamic is impossible to distinguish from that of inadequate data gathering. 64 While there are certainly parallels between the thinking of Sennett and the insights of Chaos Theory, it is clear that Sennett sees cities as being disordered, rather than 'differently ordered' (Sennett 1970). 87 Theory into planning theory 6 5 (Engwicht 1993, 61-63). Unfortunately, by equating "Chaos" with Sennett's conception of "disorder" (Sennett 1970), Engwicht essentially confuses order with its antithesis, tremendously weakening the strength of his argument. Consequently, we are substantially on our own in our quest to integrate planning theory with Chaos Theory. Having explored the nature of Newtonian and Lorenzian systems in the previous chapter, and traced the development of the mechanical organic method of synoptic planning up to the present day, we are now in a position to make a detailed comparison of synoptic planning practices with the axioms of Newtonian order which underlie them. Having clarified the pattern of the dependent relation between synoptic planning and Newtonian systems, we w i l l then be able to draw up an analogous method based on the axioms informing Lorenzian systems. G5 Engwicht lectured at UBC in 1993 on the topic of "Planned Chaos" (Engwicht 1995). 88 CHAPTER FOUR: NEWTONIAN SYSTEMS AND SYNOPTIC PLANNING 4.1. Introduction It is only at this point in the thesis that we can profitably begin to l ink the specific assumptions within synoptic planning method with the characteristics of Newtonian systems. Without an understanding of the connections between order and planning, the whole point of such an exercise would seem arbitrary and groundless. Likewise, unless we realize that there is more than one type of order-producing process recognized by science, the fact that the dominant planning method is based on familiar Newtonian process would seem banal. Finally, unless we can assure ourselves that it is the Newtonian foundations of synoptic planning that have led to the crisis of faith in modern city planning, our own efforts to derive an alternate method based on the axioms of Lorenzian processes would appear to be of merely academic interest. But having got this far, we still have not examined how the characteristics of Newtonian systems which inform the mechanical-organic world-view might be hypothesized to shape the day-to-day practice of synoptic planners 6 3 . This is the hypothesis we shall develop in the current chapter. This chapter w i l l be the most difficult to substantiate. As we saw in the previous chapter, there has been no lack of planning theorists wi l l ing and able to criticize the shortcomings of the synoptic planning method, particularly within the last forty years. Nevertheless, it proved very difficult to find literature l inking the practical shortcomings of planning to the peculiarities of the world-view held by planners. This is not because the ways of planners are mysterious or hidden, but because, as we have seen, the vast majority 63 Or, more accurately, "planners of the synoptic diaspora". 89 of those crit icizing current planning practice remain within the Newtonian paradigm. As a result, despite the wealth of articles on planning theory, it proved very difficult to find any which directly questioned the fundamental assumptions of the synoptic school. Indeed, these assumptions only become apparent once we have become aware that the facts they are based upon do not necessarily hold true for al l things. One cannot properly see into something unless one steps outside. Table 4: Comparison of Newtonian System Characteristics wi th Synoptic Planning Aims and Assumptions Newtonian Systems Synoptic Planning Deterministic Planning policies determine city development Accurate long-term prediction and control possible Specific future conditions of city can be determined Newtonian order results under laboratory conditions Newtonian order results if 'random' urban phenomena are controlled Complexity of phenomena is proportional to complexity of system Complex by-laws are required to control urban complexity Size of effect is proportional to size of perturbation Large-scale processes have greatest influence over city development Constant or periodic equilibrium Equilibrium is constant Graphic representation: geometric Geometry is a sign of "order" We shall examine the actions and assumptions of planners in light of the characteristics of Newtonian systems we examined in the second chapter. Table 2, which listed these characteristics, is reprinted in the left column of Table 4 above. Each of these characteristics has given rise to a corresponding axiom within synoptic planning method. These matching assumptions of synoptic planning are printed in the right column of Table 4, paired together in the appropriate order. Having drawn the connections between the 90 characteristics of Newtonian systems and the assumptions of synoptic planning, we shall illustrate each point in turn. As a consequence, we shall illustrate the practical connections between Newtonian systems and synoptic planning not by recourse to the literature, but by referring to the assumptions and practices brought to light by the actual practice of planning. 4.2. The Aims and Assumptions of Synoptic Planning 4.2.1. Planning policies determine city development At the risk of sounding simplistic, we shall start with the most fundamental assumption of synoptic planners: the belief that they can determine the growth and development of the city. No matter how random and arbitrary the functioning and layout of a city may appear, synoptic planners assume that it is the product of deterministic processes. These processes may be myriad in number, operating in a welter of cross-purposes and confusion, but each component can be isolated, addressed, and their individual behaviours controlled, thus ordering the city as a whole. Indeed, the entire idea of synoptic planning can only be entertained if planners are understood to have the ability to cause and create change in the real world. Clearly, then, the core assumption of synoptic planning derives from the Newtonian axiom that all natural phenomena are the products of lawful deterministic processes. In a 1960 article, Britton Harris explicitly stated this oft-unspoken assumption: The necessary... condition for metropolitan planning... is nothing more or less than to unravel the laws of behaviour of [cities]. What is the nature of this mechanism, and how can it be manipulated? ...[W]e need to ask three questions: How does the metropolis work at any point in time? How does it 91 change?'and How could this change be controlled and directed? (Harris 1960,268) . [Harris ' i tal ics] 6 4 We should be quite clear about this point. Synoptic planning seeks to create Newtonian order. Moreover, it is based on the assumption that the city is deterministic in nature. In its efforts to do so, synoptic planning method has not been patterned after'das. characteristics of Newtonian systems so much as it has been equated with them. Indeed, we can state the issue more strongly than this — the synoptic planning method equates a successful plan with the creation of a Newtonian system, tailored to create a specific order. A n "official community plan", a "strategic plan", or a "comprehensive plan" is intended to be a deterministic system, a set of equations which wi l l force desired phenomena to occur. This is a natural consequence of the planners' aims, for if synoptic planning aims to produce Newtonian order, it must regard its own operation as a peculiar species of Newtonian process shaped by thinking, reasoning people. 4.2.2. Specific future conditions of city can be determined As we have seen in the previous chapter, the overall goal of synoptic planning is the ordering of the city and city functioning, equating "order" with efficiency, "rationality", and predictability. However, planners see themselves as ordering the city for its future, rather than present needs. This in turn, is based on the assumption that present trends, or experiences elsewhere can be used to predict future needs. Indeed, synoptic planning assumes that in altering the functioning of the city by means of by-laws and zoning, one can effectively determine the future state of a city for some given point of time in the future. This fundamental assumption of synoptic planning underlies much of modern planning, including zoning and official community plans. In equating a plan with a 64 To be sure, this is a rather old quote to use for illustrating the basic assumptions of contemporary planners. However, it has been a long time since planners had the self-confidence to openly proclaim these sorts of statements. Since then, statements like these have been driven underground - but they have not, I would argue, been renounced. 92 Newtonian system, synoptic planners do not presume that they can force a change; rather, they believe that they can determine a particular sequence of phenomena, or create a particular pattern of outcomes. There was little doubt of this amongst the many enthusiastic observers of Kitimat during the 1950's, the "first complete twentieth-century "new town", completely new, completely modern, in North America" (Kitimat 1954). [Kitimat] is the product of design, planning and execution, rather than accident and circumstances. ... If twentieth-century knowledge, science and engineering can build the kind of a city where people like to live, Kitimat should be it (McGuire and W i l d 1959, 5). A n assessment of the town and its master plan during the 1970's confirmed Kitimat's overall conformance to its master plan over the long term, to the extent that even the long-noted shortcomings of town life — its "absence of any vitality or excitement anywhere in the community" — was attributed to the twenty-year-old plan, and not the dynamic of the community's subsequent development (Weisman 1977, 9, 13). Thus, while synoptic planners may debate whether or not Kitimat's planners succeeded in designing "the kind of city where people like to live," they do not question whether or not planners have the ability to shape long-term development towards specific targets. To be sure, we have chosen a rather singular example to highlight this assumption in its most simplistic and radical form. Few planners nowadays believe that even the most enthusiastically supported plan wi l l actually force the specific changes it was designed to. Nevertheless, planners do not appear to have abandoned the notion that they can determine specific changes, at least in principle. A %oo& example of this may be found in the Greater Vancouver Regional District's "Livable Region Strategic Plan (LRSP)", approved in 1995 by its Board of Governors 6 5 . The 65 The aims of Kitimat's head planner, Clarence Stein, seems positively humble compared to those who developed the LRSP: "Greater Vancouver can become the first urban region in the world to combine in one place the things to which humanity aspires on a global basis..." (GVRD 1995, 2). 93 main objective of the plan is to concentrate growth in "regional town centres", so as to reduce the re-development of high-quality agricultural land into low-density residential sprawl (GVRD 1995, 2-3). The plan additionally calls for the development of "complete communities" with homes, workplaces and shopping areas located in close proximity to each other. Consequently, the centrepiece of the plan is a list of population and employment "targets" for the year 2021, with every municipality and unincorporated area within the regional district given its own individual target. The bulk of the predicted population increase is to be accommodated within the already-developed "Growth Concentration Area (GCA)", and employment growth w i l l be focused on designated "Regional Town Centres" and the Vancouver "Metropolitan Core" (GVRD 1995). Thus, under the new plan, the suburban City of Surrey (1991 population: 245,000) shall have a population of 472,000 people in 2021, with a set proportion of its new population and employment confined to that quadrant of the community falling within the G C A (GVRD 1995 ,9 ,11 ) . Clearly, the "Livable Region Strategic Plan" was conceived on the assumption that planners could actually force the necessary changes and meet these targets, for there is virtually no chance that this pattern of settlement would occur 'naturally'. As elsewhere in North America, the Greater Vancouver region is currently characterized by runaway low-density suburban development. While high-density residential development has also occurred, this lifestyle is preferred only by a minority of new residents, and has not noticeably slowed the rate of sprawl. Moreover, as we saw above, the plan is centred upon the attainment of specific goals by a specific time. Like a gunner whose knowledge of the physics of projectiles w i l l enable h im to aim the cannonballs he fires, the synoptic planners responsible for the LRSP presume they can control the trajectory of a city as it speeds into the next century. 94 4.2.3. Newtonian order results if 'random' urban phenomena are controlled Although the synoptic planners' self-assessment of their potential abilities are clearly revealed by the target-oriented nature of the Livable Region Strategic Plan, this confidence is notably absent when it comes to describing how the plan w i l l actually function in the real world. Far from being an iron-clad directive determining the future development of the region, the LRSP is described as though it were merely one influence amongst many. Rather than shaping the future of metropolitan Vancouver, we are told that the LRSP w i l l "support", "call for" and "encourage" changes in the future of the region. In sum: The purpose of the Livable Region Strategic plan is to help realize this vision through Greater Vancouver's land use and transportation development (GVRD 1995, 2). [my italics] W h y does the language of actual implementation differ so markedly from the deterministic structure of the plan itself? Because synoptic planners draw a clear distinction between the viability of plans implemented 'under laboratory conditions', and those put into practice wi thin the turmoil of the existing city systems. In the case of the Greater Vancouver Regional District, any plan passed on the regional level has to be implemented by twenty-two separate municipalities and local authorities 6 6. While all municipal'governments are legally bound to comply with the regional plan, several municipaLcouncils have rejected the growth targets forced upon their communities by the plan {^ancouver Sun 1996, A9) . Clearly, the LRSP w i l l be compromised to the extent that individual municipalities fail to follow through on their respective tasks under the plan. Nor is a lack of enthusiasm the only factor synoptic planners point to when explaining the divergence between plans made and patterns resulting. As synoptic planners see it, any plan they draw up for an established area w i l l be significantly affected, and corresponding reduced in effectiveness, 66 Not to mention a myriad of other agencies, including provincial-level transit and transportation agencies, and several federal-level Port Authorities. 95 by the sheer weight of history and the tremendous diversity of divergent and random forces which such places contain. The Draft Transportation Plan (1996), explicitly designed by the City of Vancouver to implement the LRSP, is if anything, even more cautious in describing its power to effect change. Munic ipa l and Provincial governments can do much to address transportation needs by supporting transit and removing incentives which favour the use of the car where practical alternatives exist. But in the end, personal attitudes and behaviour w i l l be major factors in determining how Vancouver residents travel around the city and the impacts this travel w i l l have on neighbourhoods. ... The success of the Plan w i l l hinge on personal commitment, and willingness to accept some extra inconvenience when travelling around the city (City of Vancouver 1996a, 17). Vancouver is a well-established city, more than a century old, the heart of a metropolitan area approaching two mil l ion people. It is, moreover, a city in which the post-crisis synoptic planners have made an explicit commitment to plan only in accordance wi th the w i l l of the people 6 7 . Clearly, synoptic planners here have little choice but to accept the fact that the determinative power of their own plans w i l l be partially, or even largely, compromised by the myriad other forces in the modern city which are politically impractical to curb or control. The fact that synoptic planners conceive of their power in this fashion does much to explain their attraction to the various strategies of new towns like Kitimat, the wholesale urban renewal schemes of the mid-twentieth century, and the new totalitarian gated residential communities prevalent in the American southwest (Guterson 1992). Each of these strategies is marked by a not-so-subtle attempt to create 'laboratory conditions' for the practice of city planning, through elimination of as many random or perturbing forces as possible. Architectural Forum made the point clearly in its special feature on Kitimat in 1954: 67 Known as "CityPlan", this innovation has been hotly debated among planners (See Seelig 1995). 96 People make cities. In the process, people use so many complicated instruments that the direct connection can get quite lost between the city as built and the city as people really want it. Rarely does the process of city building start fresh. Usually there is a city already on the spot, or another close by, that distorts the calculations. To conceive a city directly for the people in it, and to start fresh, the planner likes to dream of building on an island or in a wilderness, under conditions of pioneering... (Kitimat 1954). W h y is Kitimat, far off in the wilderness, so important to town planning in settled areas? The first reason is that the wilderness, far from complicating the basic problems, swept away the usual complications that are basically irrelevant (Kitimat 1954). Arguably, the same desire to escape the myriad untamed forces of the city is also expressed in the Garden City and New Town ideas 6 8 which prescribed the construction of brand new medium-sized cities on the peripheries of major centres, and that of suburbia in general (Sennett 1970; Jacobs 1961). But it is where synoptic planners have attempted comprehensive planning within already established cities that their need to sweep away complicating factors has taken its most dramatic form. Perhaps the most conspicuous characteristic of the mid-century urban renewal movement was the manner in which large sections of the city exhibiting "blight" were rigourously scrubbed clean of context and history before they were developed anew. In his disparate reviews of two mid-century housing projects, Lewis Mumford indicates the importance synoptic planners put on the elimination of outside influences 6 9 . In his review of the suburban Fresh Meadows project, whose low population density appealed to h im, Mumford writes: Because Fresh Meadows has been built by a big corporation, it has had the advantage of large-scale organization all the way... planning this neighborhood project literally from the ground up. Piecemeal building by small investors simply cannot achieve the economies or create the collective order and beauty that a big operation can. ... The New York Life Insurance Company has given its architects a chance to show how humane and attractive a modern community can be if the designer's imagination can be applied not to isolated buildings but to the inter-relationship of people, trees, 68 Both of which influenced the design of Kitimat (Kitimat 1954). G9 Even the title of Mumford's work From the Ground Up alludes to clearing the ground of historical taint before constructing anew (Mumford 1956). 97 greens, parks, streets, and buildings, so that they become an organic unity (Mumford 1949 in 1956, 5). Conversely, the high-density urban renewal project of Stuyvesant Town in Manhattan is condemned: [the project is] absolutely uniform in every detail, mechanically conceived and mechanically executed, with the word "control" implicit in every aspect of the design. This, I said to myself, is the architecture of the Police State... [T]he City of New York aided in the acquisition of the land and allowed the new owners to wipe out the existing network of streets... In order to make the Metropolitan's control over this feudal domain absolute and inviolable, a public school and two parochial schools in the territory were also demolished. Now no citizen of New York may set foot within Stuyvesant Town except by permission of the owners, and a private police force is on duty to exercise, if the proprietors require it, this control (Mumford 1948 in 1956, 109-110). For better or for worse, Mumford makes it plain that the conditions of both projects were determined entirely by their planners, through the elimination of interfering outside forces. 4.2.4. Complex by-la ws are required to control urban complexity While the prospect of founding new towns, or spearheading massive urban renewal projects is a favorite dream of many synoptic planners 7 0 , the fact remains that most planners work in established cities and towns. Wi th limited available public funds and widespread public (and political) antipathy to change, planners within the city are confined to what is financially possible and politically prudent. For the urban synoptic planner, the task is one of maintaining, preserving and incrementally extending order, rather than building the millennial City on the H i l l . As seen by the synoptic planner, the task is a difficult one, simply because of the number of forces the planner has to contend with. The sheer dynamism and variety of the city naturally leads synoptic planners to regard the city as a terrifically complicated Newtonian system. Seen from the perspective of Newtonian order, Kitimat's admittedly 70 And the author as well... 98 unexciting ambiance is a direct result of its well-ordered, master-planned design. Because it is a fundamentally rational town, composed of a relatively small number of interacting forces, Kitimat is simply not capable of producing diverse, unexpected, or, in a word, cosmopolitan behaviour. By contrast, even a mid-sized city like Vancouver is capable of producing phenomena like Greenpeace International 7 1 , distinct gay and lesbian neighbourhoods 7 2 , riots in the fashion district, restaurant patios during win te r 7 3 , the inspiration for "Blade Runner" 7 4 , a nude beach lobby group, the concepts of "cyberspace" 7 5 , and "Generation X 7 6 " and a 100,000-person Peace M a r c h 7 7 . At this point, a comparison with our Quebec City snowboard experiments of Chapter 2 is appropriate. We performed our experiments under 'laboratory conditions', and the trajectories of the snowboards were perfectly predictable and controllable. We can imagine drawing up a very simple set of instructions setting out how the trajectory of the snowboards might be accurately controlled. But let's imagine that we subsequently performed these experiments under 'normal conditions'. Because outside influences would now act on the snowboards, we would find that we could not maintain the same control over their trajectories unless we explicitly took these new factors into account — indeed, the dynamical system of the snowboards on the 71 Greenpeace originated in Vancouver in 1970 as the "Don't Make a Wave Committee," formed to protest nuclear testing in the Aleutian Islands (Hunter 1979, 7-8). 72 The West End and Commercial Drive respectively. These neighbourhoods are hardly ghettos: both areas are known for their cosmopolitanism rather than their cultural uniformity. 73 (City of Vancouver 1996b, 3-4). Temperatures from November through February generally range from 0 to 9 degrees Celsius (Oke and Hay 1994, 28-29). 74 The now-classic science fiction movie Blade Runner (1983) was based on author Philip K. Dick's novel Do Androids Dream of Electric Sheep?, written in Vancouver's Chinatown during the early 1970's. Dick's surroundings while he wrote the book appear to have influenced his portrayal of "Los Angeles in 2019" as a high-density city predominantly populated by Asians, and subjected to near-constant rain. 75 Term coined by resident author William Gibson (Gibson 1987, 168-191). 76 Term coined by resident author Douglas Copeland (Copeland 1991). 77 Estimated attendance in the 1994 march. Approximately 60,000 marched in 1993 and 1995 {Globe and Mail 1994, M3; Vancouver Sun 1995, A3). 9 9 glissade would be different, and far more complex than the 'laboratory conditions' system was. Under these new conditions, we might realize that dips and bumps in the glissade, and imperfections in the ice prevented the trajectory of the snowboards from being adequately controlled. We might deal with this by restricting the snowboards to that part of the glissade least affected by these imperfections 7 8. Having done so we might notice that the snowboards were still off track. We might then notice, for instance, that breezes blew the snowboards "off track". We might then formulate some measures to counter this for breezes of various strengths, coming from various directions. We might add a more directives about what to do in the case of sudden gusts of wind. And so on. As time went on, and we recognized and compensated for the effects of ever more factors, we would find that we had transformed our simple guide into a lengthy manual of instructions. The more numerous the factors interacting with our snowboards on the glissade, and the more precisely we wished to control our snowboards, the more voluminous we would expect our documentation to be. Our manual might wel l prove very effective in ensuring that we controlled the trajectory of the snowboards on the 'real-life' glissade. But it would be cumbersome. In a similar manner to ourselves on the 'real-world' Quebec City glissade, synoptic planners cannot hope to evade or successfully outlaw the cauldron of phenomena real cities produce. Instead, synoptic city planners are obliged to control and direct this apparent variety of forces one-on-one, through the use of by-laws. As in al l modern planned metropolises, the City Council of Vancouver — home to Greenpeacers, cyberpunks, nudists and mid-winter al fresco diners — has passed a remarkable number of by-laws. At present, employees in the City Clerk's Office estimate that approximately 200 78 We might also resurface the glissade, but reordering the environment entirely is akin to the monumental planning of New Towns, rather than the typical day to day planning of everyday old towns. 100 by-laws are "being enforced". The text of these by-laws collectively covers approximately 10,000 pages, and occupies a linear meter of shelf space (Wylsonl997) . Not surprisingly, enforcing and elaborating this complicated control structure has long been the main task of most municipal planning departments 7 9. In order to achieve the needed control, these by-laws typically restrict citizens to a particular course of action or mandate a particular solution to a problem. Thus, in order to prevent buildings from easily catching on fire, from endangering occupants or from easily spreading to neighbouring buildings and threatening their occupants, a fire code has been drawn up. Wi th in the Vancouver Fire By-law, one can find regulations specifying construction techniques and operational precautions to (one assumes) prevent every kind of potential fire situation that might occur within the city, and to minimize the dangers created during any type of fire. It is currently 213 pages thick. Section 2.9 of the code regulates "tents and air-supported structures"; sections 2.10, 2.15, 3.7 and 3.8 deal wi th day-care centres, construction sites, "industrial ovens for baking and drying" and bowling alleys respectively. The regulations are nothing if not comprehensive: section 5.20 even regulates "nuclear weapons" despite the fact that Vancouver is a self-declared "Nuclear-Free Zone" (City of Vancouver 1992, 35, 36, 38, 55-57, 145). But we should not be too droll about the by-law, for there is no question that it works. Vancouver's fire by-law alone has by now saved hundreds, if not thousands of lives. Moreover, this particular approach to the control of civic phenomena is not simply a matter of synoptic planners making work for themselves. Rather, seen from a Newtonian systems perspective, it is the only effective means of ordering the dynamic city. 79 Between 1886 and the end of 1996, Vancouver City Council has formally passed 7689 separate by-laws and amendments regulating civic phenomena into law (Wylson 1997). 101 Because synoptic planners view the city in terms of Newtonian systems, they have progressively expanded the number of by-laws, regulations and guidelines which developments or activities must conform to, in the hopes of developing a legal edifice equal to the variety of phenomena encountered in the c i ty 8 0 . 4.2.5. Large-scale processes have greatest influence over city development Fortunately for themselves (and ourselves), synoptic planners do not believe they need to control every single phenomenon which occurs in the city, even when some such events strike them as being disordered in the extreme. Instead, city planners focus their resources upon the 'largest' developments within the city, often creating 'task forces' or 'working groups' to guide the development of large private and public redevelopment schemes. As part of their mandate, these 'task forces' explicitly try to forecast and control the impact of these large schemes on the future development of the city. Conversely, some forms of micro-scale developments are effectively unregulated altogether. Whi le it is hardly controversial to state than planners assume that large-scale processes exert more influence over future development than small-scale ones, we should take note that this assumption too stems from the characteristics of Newtonian systems. From a Newtonian systems viewpoint, a tremendous amount of the variety and dynamism which occurs in the city is simply too minor or ephemeral to affect the overall order of the city. As we saw in Chapter 2, minor perturbations of a Newtonian system remain minor, increasing (if at all) only in a static ratio to the growth of the phenomenon itself. Indeed, when viewed from a Newtonian point of view, we could expect most of the minor happenings of city life to cancel each other out — random fluctuations with virtually no effect upon the pattern produced by the overall urban system. 80 To be sure, they have been additionally prodded along by a legal system which frowns upon vague wording, and the exercise of discretion by appointed officials (Youngl995). 102 This is a hard assumption to illustrate in a straight-forward manner, because it seems completely unremarkable that large-scale processes would be planned for. Thus, in Vancouver, it is hardly surprising that four major residential redevelopment projects surrounding the downtown area 8 1 , have consumed much of the city planning department's energies over the past decade. As we have seen from the GVRD's Regional Strategic Plan, these projects are seen as crucial to the development of a lively, compact metropolitan core. These projects are very large in scale, containing a total of 11,680 new residential units, housing an estimated 20,000 people: the "Concord Pacific" development alone is one of the largest redevelopment projects currently underway in North America (City of Vancouver 1996c, 2). Because of the size and importance of these projects, planners for the city estimate that they expended 20-person years 8 2 in developing official development plans (ODPs) for the "False Creek North" and "Coal Harbour" areas (Naylor 1997). Planners were able to obtain-large scale changes to the development of these areas. Whi le the land-holders wanted to build condominiums in entirety, the Planning Department of the City of Vancouver insisted that 20% of the units built be rental stock. Thanks to their perseverance in the face of some extremely powerful corporations, the developers have donated the land for almost 2500 "non-market" residential units to the City of Vancouver. It is intended that this housing would be built as co-op housing or subsidized rental developments by the City over the 20-year build-out period of the project (Naylor 1997) 8 3 . By contrast to large redevelopment projects, "secondary suites" are built individually, each constructed within the basement or attic of a different single-family home. Often 81 The "International Village" and "Concord Pacific" developments on the north side of False Creek, and the "Eayshore" and "Marathon" developments in Coal Harbour (Naylor 1997). 82 Planners and assistant planners only - clerical and support staff were not included in the estimate. 83 Because of the general lack of funds for housing in the public sector, the City is now considering lowering the requirements on the developers in exchange for a lesser number of constructed units (Naylor 1997). 103 betrayed only by the second mailbox such suites require, these "illegal suites" (their other common description), are easily missed by neighbours, let alone city planners. Nevertheless, it is quite remarkable how overlooked they have been by city planners. Outlawed in Vancouver soon after the Second Wor ld War by zoning regulations designed to preserve the character of single-family neighbourhoods, enforcement of the ban on secondary suites was nevertheless remarkably lax. A succession of "hardship" based exemption programs allowed individual suites once they had been brought to the attention of the City (City of Vancouver 1988b, historical perspective). M a n y other secondary suites remained completely illegal, presumably undetected by city authorities precisely because these residences have such negligible impacts upon local neighbourhoods (Lewinberg 1984, 93-94). Over the years, more and more suites were built, offsetting the tendency towards the depopulation of residential areas with the shrinking of family size, offering home-owners substantial (tax-free) income to offset rising home prices, and providing affordable rental accommodations to students and other lower-income residents (Audain and Duvall 1992, 97). Nevertheless, it was only in 1988 that the city of Vancouver began to actively plan for secondary suites, setting up a "legalisation" process and inspection unit. By this time there were an estimated 21,000 to 26,000 units in the city (City Of Vancouver 1988a, overview), collectively comprising an invisible affordable housing development twice as large than the four residential megaprojects combined, and fully ten times larger than their non-market housing components 8 4. Because of obvious shortcomings in the "legalization" process 8 5 , at least 18,000 units, fully 85% of the 1988 total remain illegal and unplanned 84 There are no readily available estimates on the total population living in secondary suites. 85 Fire and safety inspectors drew up far stricter structural and fire regulations for secondary suites than exist for single-family homes, including the installation of hard-wired smoke alarms, sprinkler systems, and minimum basement ceiling heights. As a result, the cost of "legalizing" a suite has been estimated to range from a "low" of $7000 up to $60,000 for "older homes". These figures apply to the suite alone - perversely, the regulations do not require any improvements in the remainder of the house (Vancouver 1987, 2). "Legalizing" a secondary suite has thus been rendered uneconomic, as the moderate rents secondary suites earn cannot recoup this scale of investment. 104 for 8 6 . Nevertheless, secondary suites continue to be overlooked as an important factor in the development of the city. No planners are currently overseeing secondary suite policy, and there are no plans to explicitly address the issue in the ongoing "CityPlan" process (Whitlock 1997). 4.2.6. Equilibrium is constant As the 1974 presenter of the Canadian Broadcasting Corporation's Massey Lectures, cyberneticist Stafford Beer laid out the assumptions of synoptic planning with unusual clarity (Beer 1974). In his first lecture, Beer argued that the exploding complexity of the modern world had left institutions unable to generate a correspondingly flexible control mechanism. What is most remarkable about the lecture, however, is how Beer's explanation of Newtonian systems unintentionally revealed the synoptic planning assumption that system equilibria are necessarily static. Beer paints a picture of two men sitting on top of two poles, holding an elastic between themselves, "from the middle of which hangs a tennis ball suspended by an elastic thread" (Beer 1974, 8). The two men represent "an institution" and the ball literally marks the dynamical state of the system determined by the institution's actions. As it turns out, we have already explored a very similar image within this thesis, for if we look down directly upon the pole-sitters, and note the movement of the tennis ball, we w i l l obtain a real-time phase-space diagram precisely comparable to Figure 6 in Chapter 2. Beer makes it very clear that the job of the persons holding the elastic is to generate a static equil ibrium as quickly as possible: If the men on top of the poles do their respective jobs properly, they w i l l pul l correctly on the elastic. The ball — which marks the output state of the system — w i l l bob about for a bit, and then be still. The dynamic system is doing its work, and producing stability (Beer 1974, 9). 8S Only 3151 suites have been registered with the city in the nine years since 1988. Of these, 985 (or less than 4% of the 1988 estimated total) have been fully approved, and another 2166 have been allowed to continue in operation for a limited period of time (City of Vancouver 1996d). 105 As seen from above, we have a dot-like phase-space diagram identical to that produced by the snowboards in our Newtonian system of the Quebec City glissade. By contrast, Figure 9 displays the Lorenzian system of our Whistler Mountain experiments, whose equil ibrium constantly changes, and never "settles down" to a static state. This k ind of situation, according to Beer, arises out of inefficiency, error and complexity. It is most certainly to be avoided: If the men are inefficient, and cannot make up their minds how to pul l on the elastic,... then the ball w i l l bob about for ages, and may never settle down. This system is unstable (Beer 1974, 9). Beer then asks us to picture a tennis ball dangling from the focus of forty elastics each held by a different pole sitter: I think we can bet that the relaxation time w i l l now be extremely long. ...The harder all the men try conscientiously to manipulate the system so that it settles, the more unstable it is likely to become. Just imagine the chaos. ...It isn't going to work (Beer 1974, 10). Equating movement and change with instability, and stasis with equilibrium, Beer's explanation makes it obvious that the synoptic planning understanding of equil ibrium is derived directly from the characteristics of Newtonian systems. As might be expected, the identification of static conditions with equil ibrium has profound effects on the synoptic planning method. But we should not assume from the above that synoptic planners are interested in ordering cities into the complete stasis of a depopulated ghost-town, nor even the "steady-state" condition of a town that neither gained nor lost population and employment. Instead, the goal of synoptic planners seems to be that of steady, constant economic and population g rowth 8 7 , something very much akin to the constant downslope velocity of our snowboard on the Quebec City glissade in Chapter 2 condition. The latter form of "constancy" is certainly favored by our current 87 Clearly, this is also the goal of synoptic planners' capitalist and "state capitalist" superiors. 106 economic system, and also, perhaps, by the fact that planners would have nothing to do if there was no growth or change to manage. There is a further consequence of adopting a Newtonian systems perspective on equilibrium. In Chapter 2, we noted that the velocity of the snowboards in the Newtonian system reverted to a constant rate after they were perturbed from equilibrium. In a directly analogous manner, synoptic planners assume that in the absence of external disturbances, the behaviour of a neighbourhood or a city w i l l also quickly revert to equilibrium, after which no new behaviour can be expected. Thus, the impacts of a new development are assumed to come mainly during its construction and initial occupancy phases. Indeed, planners regularly deem a building or development to be "finished" at the point when it is ready for occupancy: in the case of a housing development, the whole thing w i l l be "completed" before anyone even begins to live in the new building. There are further implications to this means of viewing equilibrium. Because synoptic planners see themselves as "managing change", and completed buildings or other developments are understood to be unchanging, it follows that synoptic planners cease to be interested in developments once they are complete. To be sure, the arguments above are rather sweeping, but they do appear to correspond with the basic assumptions of synoptic planners. While rigourous substantiation would require a study of its own, several examples of this attitude came to light during a recent study by B.C. Transit, the transit authority for the metropolitan Vancouver area, to help predict ridership levels over the coming decade 8 8 . 88 Lest these examples taken from my experiences working at B.C. Transit sound too critical, it should be noted that the planners for B.C. Transit are a dedicated, competent and hard-working group of professionals. Moreover, as an intern planning student, I was the one responsible for actually conducting the study. There is nothing in this study to take issue with from a synoptic planning perspective. The example is used solely because it helps to highlight common assumptions. 107 Planners for B.C. Transit assumed that an increase in population or employment wi thin a short distance of existing rapid transit stations would subsequently lead to increased ridership. Consequently, planning departments in Vancouver and several suburban municipalities were approached by B.C. Transit planners, who requested information about current construction and future development prospects within these transit station areas (Heap 1996). It is worth noting that by focusing on future development prospects, the study automatically omitted consideration of existing developments. The reasoning for this is clear — if the buildings had been completed, then their occupants had already decided whether or not they would use transit. If they were using transit, then they were already included within current ridership totals. Slightly restated, the study assumed that the transportation-decision-process for a building's occupants would reach a steady equil ibrium as soon as the building was occupied. Nor was this assumption restricted to the synoptic planners within B.C. Transit. In several locations, long-established neighbourhoods were located by the rapid transit stations being studied. Upon being asked about these neighbourhoods, municipal planners most commonly responded that "nothing was happening" in such areas, or that they were "dead" 8 9 . In several other instances, certain long-established neighbourhoods had been purposefully targeted for "re-development" by the planners as these areas had become "run-down" — taken literally, an area entropically decomposing from stability to complete stasis. By contrast, other areas distinguished by their unfinished state and the sums of investment being poured into them, were described as being "active" and "vibrant", 89 Because one-third to one-half of all houses in some neighbourhoods within the City of Vancouver are known to harbour illegal secondary suites, I asked municipal planners in Burnaby if populations in these 'unchanging' neighbourhoods might be increasing due to construction of new suites in existing houses. Despite the obvious reality of secondary suites, municipal planners replied that this could not be a factor because dividing up single family residences was not permitted under the by-law. The planners appeared to be in denial that the by-law mechanism could not adequately control an obviously significant urban phenomena — and thus, that planners were unable to effectively order the city (Heap 1996). 108 developments actively perturbing the overall dynamical system of Greater Vancouver, albeit in a controlled fashion (Heap 1996). 4.2.7. Geometry is a sign of order There has long been a deep ideological connection between the theory of Newtonian order, and the Classical traditions of geometry and symmetry 9 0 . Moreover, the connection between geometry and planning has been strong throughout the history of city planning, within both the "cosmic" and the "mechanical-organic" world-views. Although the interplay of geometry and city planning is a broad and fascinating field in its own r igh t 9 1 , we shall limit our own investigation to establishing that the synoptic planning method regards Euclidean geometry in an manner analogous to that of Newtonian systems theory. To be more specific, both scientists and synoptic planners associate geometry with order since phenomena w i l l appear in Euclidean geometric form only when the generating system is free of perturbing random forces. It is as likely that the proverbial pack of randomly-typing monkeys w i l l re-write "Hamlet" as that a tree might shed its autumn leaves into a perfect circle or parallelogram 9 2 . As we saw in Chapter 2, a straight line on a scientist's graph signifies that the phenomenon being observed is ordered by a single process, and has been successfully isolated from the jumble of other intervening factors which usually mask or countervail its control. In analogous fashion, the appearance of arrow-straight boulevards, rigorous gridiron streetplans, a perfectly circular plaza, or the smooth french-curve ellipses of a residential district tells us that we are in a purposefully 90 Indeed, these ancient notions of order have probably been most directly transmitted through to our time in the form of architecture and city building. It is hardly coincidence that we still refer to the ancient Ionic, Doric and Corinthian architectural styles as "orders" (NSOED 1993, 1016). 91 There have been a number of important (and beautifully illustrated) books on urban form. In addition to the two works mentioned below, Kevin Lynch's Good City Form (1981) which we consulted in Chapter 1, and Leonardo Eenevolo's The Origins of Modern Town Planning (1967) are also highly regarded. 92 This image comes to me courtesy of an old "B.C." comic strip from the 1970's, in which a tree infuriates a would-be leaf raker by shedding its leaves neatly into a vertical column. 109 ordered city, whose government has been able to control and manage the disparate forces that would otherwise threaten it with anarchy and formlessness. A geometric street/plan demonstrates to inhabitants and observers that rational government and purposeful action hold the upper hand over randomness, aimlessness, irrationality and inefficiency. As a rule, the more rigid the geometrical layout of the city, the more total we perceive the power of the orderer to be. We must be specific about this. If synoptic planners really do view geometry in the same fashion as scientists investigating Newtonian systems, it is because they see geometry as a sign of order rather than order itself. Consequently, it is not enough to note that synoptic planners often plan cities into geometric shapes (which they very commonly do). Instead, we must see whether or not synoptic planners use the presence or absence of geometry in cities as an indication of the presence or absence of effective planning in cities. We can find complementary examples of this characteristic in the works of Mumford and Kostof. As we have already seen, Lewis Mumford was a harsh critic of the metropolitan city, and tirelessly promoted the Garden City and Regional Plan concepts. To demonstrate how lacking the state of planning at the metropolitan level has been, Mumford takes to the air, and challenges us to discern the order within megalopolis: Circle over London, Berlin, New York, or Chicago in an airplane, or view the cities schematically by means of an urban map and block-plan. What is the shape of the city and how does it define itself? As the eye stretches toward the hazy periphery one can pick out no definite shape... one beholds rather a shapeless mass... [S]hapelessness [is] an inevitable by-product of its physical immensity (Mumford 1961, 543). Spiro Kostof s own agenda in the picture-laden The City Shaped (1991) is quite different than Mumford ' s . Rather than draw attention to the lack of planning in modern cities, Kostof wants to "make accessible... the universal experience of making [read: planning] cities. [The book] is a discussion of some patterns and elements of urban form as 110 seen in a historical perspective" (Kostof 1991, 9). As a result, Kostof is at pains to portray the modern city as the intentionally ordered heir to the planned cities of the past. Consequently, the aerial photos which Kostof includes of Megalopolis show only the relentless cross-hatch of Manhattan, the parabolic shimmy of a suburban street lined with hundreds of identical houses, and the boxy outline of Hong Kong's skyscrapers (Kostof 1991, 81 , 94, 304). For our purposes, the merits of their respective arguments are less important than the means they choose to illustrate them. To convince us of a lack of planning, Mumford has us step back until the geometries of the modern city are lost. To demonstrate how the modern city continues to be a planned creation, Kostof hovers much closer, where geometry dominates the view. As we mentioned above, Euclidean geometry is regarded only as a sign of order and planning, albeit a very reliable one. This distinction between Euclidean geometry per se and order is highlighted when Mumford and Kostof discuss the Renaissance city of Siena. Both writers are anxious to convince their readers that cities like Siena were the ordered result of what amounts to a synoptic planning process. Nevertheless, even with its informal and irregular streetplans, each author is still reduced to justifying his claims with reference to the "geometry" of Siena's layout. While both of these synoptic theorists explicitly distinguish between geometry and order in theory, their identification of planning wi th geometry is so strong as to make the distinction difficult for themselves in practice. In his examination of the form of the medieval city, Mumford notes that: Those who dismiss [irregular] plans as unworthy of the name of plan confuse mere formalism and regularity with purposefulness, and irregularity with intellectual confusion or technical incompetence (Mumford 1986, 115). I l l While he himself can detect the order in a non-geometric town like Siena, Mumford does not seem to be so sure of his readers. He thus reminds us that such towns have "a complex ... design hardly less unified than a performed geometric pattern" (Mumford 1986, 115). Figure 11: Aerial view of Siena, Italy n^ H^ ffTrr^r'^iMMriiittiffnfT*^ ^ nmrOTtmrffiir^ rT^ i^g ^ * f ^ WrfT ~ 1 M i f f ) *" .* Aerial view of Siena, Italy: Kostof 1991, 2. Elsewhere in his discussion of medieval cities, Mumford bases his argument for the existence of "medieval town planning" squarely upon the geometrical form medieval towns often took, including the "rectangular platting" which continued to characterize those towns founded by the Romans, and the "strict checkerboard plan" into which "many towns were designed in advance" (Mumford 1986, 113-114). Geometry remains the reassuring sign of order. 112 One of the highlights of The City Shaped is Kostof s drawings depicting how ordered Roman grids incrementally and rationally evolved into the disparate irregular streetplans of the Italian Renaissance cities and the medinas of the Magreb (Kostof 1991, 49). It is therefore rather intriguing that Kostof misinterprets Mumford "who ought to have known better", as arguing that Siena was unplanned (Kostof 1991, 10). But [writes Kostof] I was not at al l surprised to discover, as I looked into the historical circumstances of Siena's origin and growth, that she was coerced to take that shape, that her urban form was one of the most highly regimented designs of medieval urbanism (Kostof 1991, 10). As proof, Kostof presents a single by-law from 1346 stating that: ...any edifices that are to be made anew anywhere along the public thoroughfares proceed in line with the existent buildings and one building not stand out beyond another, but they shall be disposed and arranged equally so as to be of greatest beauty for the city (quoted in Kostof 1991, 70). While it is difficult to see how a single by-law proves medieval Siena to be "highly regimented" (at least by comparison with post-modern Vancouver), it does demonstrate to Kostof's own satisfaction that the town was intentionally planned in a comprehensive, forward-looking and orderly manner. The edict, after al l , prescribes the creation of straight lines and smooth curves along the streetscapes of the city. Kostof takes note of the purposeful aesthetic of Siena's orderly curves at several points elsewhere in the book — not least on the cover and frontispiece (Kostof 1991, 2, 60-61, 70-71). While Mumford and Kostof can make the distinction between Euclidean geometry and order itself, neither trusts anyone else to able to do so. Geometry is not only a sign of order for synoptic planners — it seems to be virtually indistinguishable from it in practice. 113 4.3. C o n c l u s i o n Our review of the basic assumptions and goals of synoptic planning has shown how directly, and how thoroughly, they are dependent upon the characteristics science ascribes to Newtonian systems. It is these particular axioms of order that account for the similarities between the "mechanical" and the "organic" world-views in planning practice which we noted back in Chapter 1. Moreover, now that we have "uncovered" these fundamental assumptions of synoptic planning practice, we are able to see why the long-standing crisis of city planning remains unresolved. As we saw in Chapter 3, the profession's crisis of faith stems from the violation of some of these fundamental assumptions within some areas of real-world planning practice. Unfortunately, as we also saw, the reaction of planning theorists has only been to question single assumptions in isolation (if at all), rather than addressing them in their entirety. This is hardly a promising avenue of progress if, as we have seen, these assumptions are part of a coherent world-view, exactly corresponding to the scientific understanding of natural process. Nevertheless, stating that "synoptic planning is not universally workable" is completely different from saying that "synoptic planning is invalid". It is the former point which is being argued, notthe latter 9 3 . As we have seen, the method of synoptic planning appears to faithfully and accurately mirror scientific understanding of Newtonian systems. We cannot therefore hope to solve our planning problems simply by bringing synoptic planning "into line" with the axioms of Newtonian systems. Nor can we hope to resolve our problems in city planning by adjusting the assumptions of synoptic planning, universe be damned. Any fudging of these assumptions cannot be anything other than profoundly irrational, automatically invalidating synoptic planning's proud claim to scientific validity. 93 This is a point we shall return to in the Conclusion. 114 If there were only one type of scientifically-understood natural ordering process, the profession of city planning would have no option but to continue lumbering along in a state of perpetual intellectual crisis. But, as we know from Chapter 2, there is more than one dynamic in the workings of the universe. We may find it is possible to fashion a second method for city planning based on Lorenzian systems. 115 C H A P T E R FIVE: LORENZIAN SYSTEMS A N D DYNAMIC PLANNING 5 .1. Introduction In the previous chapter, we saw how the assumptions of synoptic planning are derived from the characteristics of Newtonian systems. If there is a second type of natural determinative process, then the possibility exists that a different planning method might be possible, based on its different axioms. In the following pages, we shall see how a logical and feasible method for planning can be derived from Lorenzian systems, even though the aims and assumptions of such a method may be at odds with those of synoptic planning. We w i l l derive this hypothetical planning method in a manner analogous to the way we proceeded in Chapter 4. Taking the list of Lorenzian system characteristics which we developed in Chapter 2 one by one, we shall proceed to "translate" each into an axiom for an alternative Lorenzian method of planning. We shall call this proposed method "dynamic planning' , because change and movement lies at the heart of this invented wor ld -v i ew 9 4 . Intriguingly, while our technique is novel, none of the axioms we develop in this manner prove to be unprecedented in and of themselves. It is possible to illustrate al l of the derived assumptions of our proposed planning method with quotes and proposals from planners and urban critics, synoptic and otherwise. We shall cite several anomalously 'dynamic' perceptions from the synoptic planners and critics Lewis Mumford and Clarence Stein, whom we have already met in the course of this thesis. At specific points, we w i l l refer to actual projects which have been undertaken in Vancouver and elsewhere. Nevertheless, we shall draw the bulk of our examples from two sources. As we shall see, 94 I am here avoiding David Engwicht's own unfortunate label of "chaotic planning", which sounds like an oxymoron even to those familiar with the new science (Engwicht 1995). 116 the arguments and proposals of Jane Jacobs' writings are particularly congruent with the aims and assumptions of dynamic planning, and w i l l be employed to illustrate most of the characteristics we shall derive for dynamic planning. We shall also make full use of the work of Christopher Alexander 9 5 , the renowned architect and planning critic who has headed the Centre for Environmental Structure at Berkeley since 1967 (Grabow 1983, 52-53), as his "new theory of urban design" (Alexander 1987) is remarkably compatible with the method of dynamic planning, derived from the axioms of Lorenzian order 9 6 . 5.1.1. Christopher Alexander Because of the extensive use we shall make of it in the remainder of this chapter, a brief introduction to the development of Alexander's "new theory of urban design" is appropriate before we move on to examine the aims and assumptions of dynamic planning. Although Alexander has written a series of interdependent books presenting his ideas on architecture and city planning, we shall focus on A New Theory of Urban Design (1987), his latest work, and the one which deals most explicitly with city planning issues 9 7 . In this work, Alexander writes that he and long-time intellectual companion Ingrid King discovered it was possible to produce "wholeness (with its fifteen or so geometric properties)" and on the basis of this set out to "imagine a process of urban growth, or 95 It should be noted that Alexander prefers to work in partnership with others, and has yet to publish a book by himself alone, although it is obvious that Alexander himself is very much the "first among equals" in these collaborations. Thus in addition to himself, Ingrid King and two students, Hajo Neis and Artemis Anninou are credited as co-authors of A New Theory of Urban Design. Since the numbers of these co-authors are extensive, and are typically associated with Alexander for only a single work, I will refer only to "Alexander" when citing any of his works. However, this policy downplays the role of Alexander's long-time collaborators Ingrid King and Sara Ishikawa, both credited on a number of his major works. 96 Like much else in this chapter, this is not an original insight. In May 1996, just before presenting my initial findings at the "Order and Chaos" conference in Corner Brook, Newfoundland, Sue Zimmermann informed me that presenters at other conferences had previously made the same connection. Unfortunately, neither she nor I have been able to document who these presenters were. 97 A full understanding of Alexander's approach to planning and city building can only be gained by studying his three most important works together — A Pattern language (1977), The Timeless Way of Building (1979) and A New Theory of Urban Design (1987). In addition, study of The Oregon Experiment (1975), the first book in this series, is also valuable, as many of the ideas presented in A New Theory of Urban Design are also present here in embryonic form. 117 urban design, that would create wholeness in the city"(Alexander 1987, 4-5). The pair were able to develop a planning structure designed to accomplish these ends composed of only seven basic rules 9 8 . Alexander and King tested out their rule structure in practice by means of an experiment: We then took a part of the San Francisco waterfront (about thirty acres intended for development in the near future) and simulated an imaginary process which makes use of the seven rules, to govern al l development over a five-year period (Alexander 1987, 5). To be sure, the experiment was conducted under 'laboratory conditions'. A model of the area was then built, and a class of eighteen architecture students was instructed to play the part of investors and builders. The "five year" development period was compressed into a single school term. Properties were developed one building at a time. Alexander concluded that the 'experiment' was a success: Wi th in the simulation, it is possible to see a new part of the city growing under the influence of our seven rules... Although it lacks many important details, and although many practical matters remain to be worked out, nevertheless, in broad outline it does work. It creates wholeness — or some approximation of it — in a way which is entirely different from the way that urban planning and design work today. A n d it does also seem to have the potential'for creating wholeness far more deeply than was possible in our simple experiment. We believe that it presents the beginning of a new theory for the three-dimensional formation of cities (Alexander 1987, 6). While Alexander employs the language of art and aesthetics to promote his "New Theory", it can also be explained using the language of science, for it is strikingly compatible wi th the planning assumptions which we have derived from Lorenzian systems 98 In their short form, the seven rules are as follows; 1) Piecemeal growth 2) The growth of larger wholes 3) Visions 4) The basic rule of positive urban space 6) Construction 7) Formation of centres (Alexander 1987, 30). 118 — indeed, we may say that it forms a fully-articulated and experimentally tested example of a particular dynamic planning method. 5.1.2. Dynamic Planning and its Precursors In large part, the illustrations and examples of dynamic planning aims and assumptions we shall employ in this chapter have come from people who seem to have had no access to Chaos Theory. Indeed, many of them anticipated the development of Chaos Theory itself; Lewis Mumford wrote about Siena before the Second Wor ld War , and Clarence Stein mused in the early 1950's. Jane Jacobs, whose outlook is particularly congruent with Lorenzian systems and dynamic planning, published The Life and Death of Great American Cities (1961) several years before Lorenz published the first (long-overlooked) paper examining Lorenzian systems (Lorenz 1963). The matter is less clear with Christopher Alexander, who developed his ideas contemporaneously with that of Chaos Theory" . But the fact that planners and critics have, between them, largely anticipated the assumptions and actions of a method based on Lorenzian systems should not be so surprising. The planet Neptune has existed since the origin of the solar system, and its gravity tugged on the Earth long before its discovery in the nineteenth century. Similarly, Lorenzian systems have ever behaved as they do now, and their activities and properties 99 Alexander's estrangement from modern science (as evidenced in Grabow 1983, x-xi) provokes the speculation that he has independently developed his Lorenzian ideas — a tremendous intellectual achievement if true. Alexander has never referred to "Chaos Theory" in his published work, and indeed, makes heavy use of the word "chaos" in its traditional sense, to disparage formlessness and disorder. However, the matter is unclear. A comment by Alexander recorded in his biography — "A few years ago mathematicians became aware... that even if you were to take three or four rules, you could already generate orders of complexity much greater than any mathematically describable geometry..." — implies that Alexander had some knowledge of Lorenzian systems by 1983 (Grabow 1983, 49). Moreover, Alexander's first university degree was in mathematics, one of the first fields in which the new science was explored. Most importantly, Alexander's New Theory of Urban Design was published relatively recently, in 1987. This work is substantially closer to our hypothetical structure of dynamic planning than works he published in the 1970's, and was published at a time when Lorenzian systems were becoming familiar to the educated public at large —James Gleick's bestseller Chaos: Making a New Science (1987) was published the same year. It seems most likely that Alexander was exposed to some of the salient ideas of Lorenzian systems in an informal or indirect way, and that these concepts provided him with a framework for his own compatible set of ideas and observations on architecture and urban form. 119 have always been available for observation. "For want of a nail , the kingdom was lost" is not a space-age nursery rhyme, yet its subject is that of Lorenzian sensitive dependence (Gleick 1987, 23). Observant people have long noticed that which the rest of us require an explanatory theory to see. What, then, do we gain by tying these various old observations and prescriptions to a new theory? As we have seen earlier on in our explorations, we live in a wor ld of science and rationality. It is one thing to say that our powers of intuition and observation convince us to do a certain thing. It is quite another matter to state that scientific theory dictates that we do a certain thing — even if it does amount to the same thing. 5.2. The Aims and Assumptions of Dynamic Planning The characteristics of Lorenzian systems, and the aims and assumptions of the hypothetical dynamic planning method derived from them are presented in Table 5 above. Each assumption is explored in detail within the following sections of the thesis. Table 5: Comparison of Lorenzian System Characteristics wi th Dynamic Planning Aims and Assumptions Lorenzian Systems Dynamic Planning Deterministic Planning policies shape city development Size of effect dis-proportional to size of perturbation -"sensitive dependence" All scales of processes have influence over city development Accurate long-term prediction and control impossible Future conditions of city can be bounded Non-linear Lorenzian order naturally produced by urban dynamical system Complexity of phenomena Dis-proportional to complexity of system Simple by-law structure appropriate to stimulate urban complexity Equilibrium constantly changes Equilibrium is constantly shifting Graphic representation: fractal Fractals are a sign of order; randomness need not indicate disorder 120 5.2.1. Planning policies shape city's development Our first assumption hardly requires illustration. There is relatively little difference between synoptic planning and dynamic planning when it comes to the respective methods' most fundamental assumption. Indeed, the reasoning is the same for both methods. As we have already seen in Chapter 2, the Lorenzian systems we are examining are dynamical, or deterministic, just as Newtonian systems are. Moreover, as we have argued previously, there is little point in planning anything if one is impotent to affect its actual development. If we are to have a method of planning founded on the characteristics of Lorenzian systems, in the same manner that synoptic planning is based on Newtonian systems, it seems logical to posit that dynamic planners like their synoptic colleagues would believe that their actions help to determine the development of the city they are planning for. 5.2.2. All scales of processes have influence over city development Despite the above, it soon becomes clear that dynamic planning w i l l be a qualitatively different discipline to synoptic planning. No system of dynamic planning, based upon the axioms of Lorenzian order in the city, could be based on the aim of predicting the future state of the city, let alone determining a specific future condition. As we have seen in Chapter 2, Lorenzian order is inherently unpredictable, even under laboratory conditions. Because of Heisenberg's Uncertainty Principle — that we can never know the position of a phenomena with perfect precision — and because the smallest differences in a Lorenzian system rapidly amplify, immeasurably small variances in initial states soon cause large-scale differences in the state of a phenomena (Gleick 1987, 15-18). As a result, no one worldng from the assumptions common to Lorenzian systems would be able to see much point in trying to control the long-term future development of the city. In the view of a dynamic planner, even well-funded and comprehensive actions of the city 121 planning department to effect a particular goal by a specific time within the long-term development of the city could be overwhelmed by the rapid amplification of other phenomena within the city. Such things might be tiny and easily overlooked at the time of the policy's enactment, but could develop into potent forces in the city's development a mere decade or two later. Moreover, unintended consequences of the policy itself, although minor in the policy's initial stages, might end up creating a substantially different long-term effect upon the city than that which was originally intended. Almost forty years ago, Jane Jacobs proved herself to be very much aware of such a possibility in her polemical masterpiece, The Death and Life of Great American Cities (1961). Jacobs argues throughout the work that it is small-scale processes (discounted or disliked by the synoptic planners she uncharitably demonizes), which determine the long-term behaviour of the city. Only someone deeply convinced of the profound importance of the small and seemingly insignificant processes of the city would have devoted the first three chapters 1 0 0 and 60 pages of her polemic to a discussion regarding "the uses of s idewalks" 1 0 1 (Jacobs 1961, 29-88)T Likewise, Jacobs argues that the popularity of a particular street in her neighborhood was sparked by the opening of a single theatre and night-club enterprise. These premises proved popular. They brought more people into the street during evening hours and week ends, to supplement the day people passing through, and thus helped stimulate the growth of convenience and special shops. These, in their own right, began to bring still more people, day and evening. ... 0acobs 1961, 244). Jacob's recognition that the small has a disproportionate effect upon the large extends to matters of investment and planned change. In the case of the North End of Boston, cut off from bank loans and mortgages, the only resources for repair and redevelopment came out 100 Not including the Introduction. 101 Of course, Jacobs ably convinces us both that what occurs on sidewalk is a crucially important factor in the development of the city, and that the synoptic planning nostrums of her own day were antithetical to healthy sidewalks. 122 of the residents' own personal savings, and their talents in construction and repair (Jacobs 1961, 11). However, over the long term, very considerable and positive development occurred within the community even from these seemingly insignificant sources for improvement: Twenty years ago, when I happened to see the North End, its buildings — town houses of different kinds and sizes converted to flats, and four- and five-storey tenements built to house the flood of immigrants first from Ireland, then from Eastern Europe and finally from Sicily — were badly overcrowded, and the general effect was of a district taking a terrible physical beating and certainly desperately poor. When I saw the North End again in 1959,1 was amazed at the change. Dozens and dozens of buildings had been rehabilitated. ... M a n y of the small, converted houses now only had one of two families in them instead of the old crowded three or four. Some of the families in the tenements ... had uncrowded themselves by throwing two older apartments together, and had equipped these with bathrooms, new kitchens and the like (Jacobs 1961, 9). The non-proportional success of this low-level investment is contrasted with the wrenching effects that "catastrophic money" produced within East Harlem. The aid poured in for rehousing people — some three hundred mil l ion dollars' worth. The more that poured in , the worse became the turmoils and troubles of East Harlem... More than 1,300 businesses which had the misfortune to occupy sites marked for housing were wiped away, and an estimated four-fifths of their proprietors ruined. More than 500 non-commercial store-front establishments were also wiped away (Jacobs 1961, 307). While this deluge of money certainly produced more dramatic immediate effects than the trickle in the North End, there appears to have been little long-term effect at a l l , at least in terms of improvement. Before redevelopment, East Harlem was a poor, crime-ridden, undesirable area. It remained so 300 mill ion dollars later 0acobs 1961, 307). Here too, the effects of investment are non-proportional. Christopher Alexander takes a decidedly sunnier approach to the "Butterfly Effect", and takes particular pains to accommodate "sensitive dependency" in city development. Indeed, the maxim that a l l scales of development have an influence on overall 123 development of a city is explicitly included as the first of his basic seven rules governing city development: ...In the ideal version, the rule has a logarithmic character, which requires that the total amount of construction in small, medium and large projects, is kept equal. In this ideal version, for every $3 mil l ion spent, $1 mil l ion w i l l be spent on large projects (one project, say), $1 mil l ion w i l l be spent on medium-sized projects (ten projects, say), and $1 mil l ion w i l l be spent on small projects (a hundred projects, say). However, the circumstances of our experiment would have made it impossible to follow this extreme rule, and we replaced it wi th a more modest one, namely: There are equal numbers of large, medium and small projects (Alex. 1987,32-33) . [Alexander's italics] Nor is this simply a matter of maintaining a human scale in built urban form. Alexander truly believes that each new increment of building, no matter how small, w i l l help determine what type of development should follow it. Because this is a decidedly Lorenzian assumption, and foreign to most of his students playing the part of land developers, Alexander found he had to explain himself in the course of the 'experiment': So far, almost al l your proposals, even when they are based on a genuine inner vision in your mind, are still essentially solitary. What I mean by that, is that they exist more or less independently of their surroundings. ... not one of you has realized yet, that your proposal should be enormously sensitive to the exact moment in sequence when it comes, and that a certain proposal might make sense as P n , if it comes after P n - 1 , but as soon as even one other proposal comes in between, even in a place fairly far away from that location, then a properly executed project at the place where Pn was w i l l have to be enormously different from P n . It is even possible that the whole idea of what you proposed as Pn might no longer be relevant at all — because as a result of P n~ 1, the gestalt of the whole has shifted so enormously (Alexander 1987, 60-61). A dynamic planner cannot take comfort in the fact that they might be able to increase the quality of their predictions by increasing their control over the phenomena in the city. The disproportional nature of cause and effect within Lorenzian systems means that any variation in initial conditions, or any measurement error in any element of the dynamical system w i l l , within an unpredictable (but finite) time period, create results as disparate as 124 that system can encompass. Prediction and control are simply not valid aims within the practice of dynamic planning. 5.2.3. Future conditions of city can be bounded As we have seen previously in Chapters 3 and 4, it is exactly the task of predicting and controlling the city's future which synoptic planners regard as their primary task. Can one do away with prediction and control without doing away with planning entirely? Fortunately for our make-believe discipline, the answer is yes. As we shall see below, the concept of the "strange attractor" in Lorenzian systems allows the dynamic planner to give up the task of forcing particular changes, while retaining the mission of managing change. As we noted above, Lorenzian order is every bit as deterministic as Newtonian order. In our exploration of Lorenzian systems in Chapter 3, we also discovered that while there is a literal infinity of possible phenomena that can be generated by a single system, al l of these possibilities are ordered and bounded. Lorenzian systems, random though they may appear to the casual observer, cannot just produce any result at al l — they are tightly constrained to produce only those which fall within the systems' "attractor", or equil ibrium set. We also know that "strange attractors" are contained within a given, finite area. Finally, we know that a Lorenzian system is robust; while a tiny random fluctuation may well give rise to markedly altered results over the course of time, these phenomena will still fall on the attractor of the system. As a result, it seems obvious that the professional dynamic planner would seek to shape the civic system's attractor, rather than the behaviour it produced at any given time. Although they could not reasonably hope to direct the city's development towards a specific end, dynamic planners could reasonably hope to shape the city's "attractor" by means of by-laws and other instruments. If done successfully, al l of the infinite possible future states of the city — that is, all the states on the city "attractor" — would be beneficial 125 to the city, in keeping with the values and aims of its citizenry. While even the most proficient dynamic planners would not wish to hazard a long-term prediction about the future course of their city, they might be quite confident in stating that al l would certainly turn out for the best. A moment's thought on the subject w i l l make it clear how different, and how unsettling this assumption could be to planners holding orthodox synoptic assumptions. At the beginning of this thesis we took a look at some meanings of the word "order" wh ich simply stressed arrangement, but they accompany other definitions with more normative connotations. After al l , "order" is not just a description: it is also form of action: • Put in order; arrange methodically or suitably; spec, marshal, draw up in order of battle, arch. M E . . . . • Set or keep in order or proper condition; adjust or carry on according to rule; regulate, govern, manage, settle. ... • Take a certain course with (a person or thing); treat, deal with, manage, esp. in a specified way. • Cause to submit to lawful authority; chastise, punish (NSOED, 2017). These definitions are common to all of us who speak English. But these nuances are particularly comforting to synoptic planners, for every time the word "order" is uttered, the marriage of organization to that of control is confirmed: the very vocabulary we use makes it hard for us to perceive of an order without also imagining that someone made it so. But under dynamic planning, the creation of new order in the city would be something that planners would only shepherd - not actively create. Lorenzian order is the order of parenting a city, not building it. Dynamic planners believe in the self-disciplined city, not the metropolis that must constantly be kept in line. Dynamic planning is the act of setting limits and letting go. A reluctance to disentangle the notions of control from those of organization may well be a fundamental reason why planning, of al l the professions, has been so slow in applying the insights of Chaos Theory. 126 It is both surprising and gratifying then, that this concept of planning — perhaps the most dissonant in terms of the synoptic method of prediction and control — has been anticipated by planners and theorists in the past. In his article on medieval towns which we explored in the last chapter, it seems that Lewis Mumford could countenance, at least an obsolete practice, a mode of planning in which the specific end-state is unknown, though contained within a certain sphere of possibilities: In organic planning, one thing leads to another... it does not begin with a preconceived goal: it moves from need to need, from opportunity to opportunity, in a series of adaptations that become increasingly coherent and purposeful.... Each medieval town grew out of a unique situation, presented a unique constellation of forces, and produced, in its plan, a unique solution. ... [But] for al l their variety, they embody a universal pattern ... The consensus is so complete as to the purposes of town life that the variations in detail only confirm the pattern. That consensus makes it look, when one views a hundred medieval plans in succession, as if there were in fact a conscious theory that guided this town planning (Mumford 1986, 115). Jane Jacobs' vision of what planning should aspire to also appears to embody notions similar to that of dynamic planning, stressing the encouragement of a wide sphere of positive possibilities, rather than the enforcement of a single optimal future: In our American cities, we need all kinds of diversity, intricately mingled in mutual support. ...Most city diversity is the creation of incredibly numbers of different people and different private organizations, with vastly differing ideas and purposes, planning and contriving outside the formal framework of public action. The main responsibility of city planning and design should be to develop — insofar as public policy and action can do so — cities that are congenial places for this great range of unofficial plans, ideas and opportunities to flourish, along with the flourishing of the public enterprises (Jacobs 1961,242) . Ironically, considering Kitimat's place in the short list of master-planned communities, it harder to find a better statement of our dynamic planning ideal than that found in Architectural Forum's 1954 article on this synoptic showpiece 1 0 2 : 102 No author is credited for the article, although the paternal tone taken towards the project suggests it might be chief planner Clarence Stein himself. 127 ... Too rigorous a plan, too fine a finish, may forestall development, dampen citizen initiative. ... A less-tailored town may result, but one wi th greater vitality. Kitimat planning ... holds that the planner is an earthly strategist and not a heaven-sent seer. Accordingly he does not presume to reduce his data to absolutes which no one can guarantee. If a plan is tailored only to absolute figures which are later undershot or overshot, the plan is vulnerable. He must find rational minima and maxima between which his plan is designed to work at any point. His research and that of his specialists should be to establish the statistical envelope within which a plan w i l l be secure. This is common sense. Once recognized, it saves arduous pin-pointing and frees the mind (Kitimat 1954). Finally, Christopher Alexander also stresses the need for both unpredictability and overall guidelines within his "new theory". Alexander believes that the development of the city must be unpredictable if it is to develop properly: When it starts coming into being, it is not yet clear how it w i l l continue, or where it w i l l end, because only the interaction of the growth, and the whole's own laws, can suggest its continuation and its end (Alexander 1987, 14). In The Timeless Way of Building, Alexander compares the unpredictable development of cities with the growth of biological organisms: ...it is never certain just exactly where a given pattern w i l l appear. Nor is it certain just what form any pattern w i l l take, in any one particular place. We do know ahead of time, what general form it has.... In this sense, it is like the natural order of an oak tree. The final shape of any one particular oak tree is unpredictable.... We know in general that it w i l l have the overall form of an oak, because its growth is guided by the pattern language for an oak tree (its genetic code). But it is unpredictable, in detail, because each small step is shaped by the interaction of this language with external forces and conditions... And a town which is whole, like an oak tree, must be unpredictable too. The fine details cannot be known ahead of time. We may know, from the language which is shared, what kind of a town it w i l l be. But it is impossible to predict its detailed plan: and it is not possible to make it grow according to some plan. It must be unpredictable, so that the individual acts of building can be free to fit themselves to all the local forces which they meet (Alex. 1979, 508 -509 ) 1 0 3 . 103 This is, of course, an explicitly biological metaphor. Nevertheless, it is very difficult to picture a synoptic organicist like Lewis Mumford agreeing with Alexander about the positive need for unpredictability, or with Alexander's implicit faith in the self-ordering capacity of cities. These are not differences on small points of detail; they are indicative of two very different ways of seeing — even though they are looking at the same thing (cities and trees). 128 Because of this inherent unpredictability, Alexander makes it clear wi thin A New Theorythat he has no interest in drawing up a synoptic plan which specifies exactly what w i l l happen where, and when. Nevertheless, Alexander has no intention of letting his experiment develop anarchically, for he has a number of specific end qualities which he wishes the resultant urban form to embody. As we have already seen, Alexander demands that equal numbers of small, medium and large projects be built within his experimental district of San Francisco. Several other strict constraints are also included as part of his 'piecemeal growth' rule. No building increment can be more than $5,000,000 or 100,000 square feet (Alexander 1987, 32). Moreover, Alexander also prescribes a pre-determined mix of uses within the area. While the experimental district is not "zoned" in the sense that particular land uses are limited to particular areas, or distanced from others, a percentage quota for each building type in square feet is prescribed. The schedule is reprinted below: Alexander notes that the "ideal distribution would vary from community to community, according to [their] wishes" (Alexander 1987, 34), and noted that running totals of construction for each building type were kept. At each moment in time, actual running totals are either above or below the level specified by the ideal distribution. New projects which tend to move the actual distribution towards the ideal distribution, are encouraged. New projects which tend to move the actual distribution away from the ideal one are discouraged (Alexander 1987, 34-35). Alexander is simultaneously concerned with unpredictability and setting boundary constraints for development. Alexander seeks to ensure certain goals, like a particular mix Housing Parking Offices Community Functions Manufacturing Shops and Restaurants Hotels 26% 19% 16% 15% 12% 9% 5% (Alexander 1987, 34) 129 of uses in the study area, but does not seek to predict or determine precisely how these goals w i l l be met — indeed, within a city ordered to his l iking, he does not believe prediction would even be possible. This non-restrictive shaping of development described within A New Theory is clearly akin to our own conclusion that dynamic planners would seek to shape the "attractor" of the city through adjustments to the by-laws and regulations of the city. 5.2.4. Lorenzian order naturally produced by urban dynamical system Our next point is implicit in the previous two assumptions. If dynamic planners believe that accurate prediction of the city's trajectory over the long-term is impossible, and yet also believe they can shape the attractor of the urban dynamical system, it follows that dynamic planners must assume the urban dynamical system to be Lorenzian (rather than Newtonian, or random) in its behaviour. Consequently, dynamic planners would expect the order produced by the system to be Lorenzian in nature. This too is not a novel idea. Articles by Couclelis (1987) and Cartwright (1991), have already introduced the concepts of Lorenzian processes to practitioners in geography and planning respectively, and have highlighted the capacity of simple Lorenzian functions to give rise to infinitely complex behaviour. Both articles note that the process of city development may be inherently "chaotic" — complex rather than complicated. Couclelis in particular stresses the very limited success that increasingly complicated Newtonian models have had in replicating city patterning and process, and implies that the behaviour of cities may be determined by a relatively simple Lorenzian process instead (Couclelis 1987, 105-108). As we have already seen in Chapter 3, Cartwright also implies that cities may constitute Lorenzian systems (Cartwright 1991, 53). Neither article, however, actually proposes a new model for the city, and limit themselves to the discussion of simple population models for purposes of clarity. 130 Intriguingly, while the modellers above were aware of the new science, the perception of the city as a non-Newtonian system 1 0 4 predates Chaos Theory itself. In The Death and Life of Great American Cities (1961), a book published several years before Lorenzian systems were recognized scientifically, Jane Jacobs effectively argued that the city was something akin to what we would now term a Lorenzian system: The history of modern thought about cities is unfortunately very different from the history of modern thought about the life sciences. The theorists of conventional modern city planning have consistently mistaken cities as problems of simplicity and disorganized complexity, and have tried to analyze and treat them thus. No doubt this imitation of the physical sciences was hardly conscious. It was probably derived, as the assumptions behind most thinking are, from the general floating fund of intellectual spores around at the time. ...(Jacobs 1961, 4 3 5 ) 1 0 5 . Cities happen to be problems in organized complexity, like the life sciences. They present "situations in which a half-dozen or even several dozen quantities are all varying simultaneously and in subtly interconnected ways." Cities, again like the life sciences, do not exhibit one problem in organized complexity, which if understood explains all . They can be analyzed into many such problems or segments, which, as in the case of the life sciences, are also related with one another. The variables are many, but they are not helter-skelter; they are "interrelated into an organic whole (Jacobs 1961,433) . Once again, we find that Christopher Alexander's ideas are, at the very least, compatible with the dynamic planning method. Since Alexander never refers to Chaos Theory we can hardly argue that Alexander equates city processes and Lorenzian systems. O n the other hand, Alexander rejects the idea of the city as a Newtonian system in more explicit terms than even Jane Jacobs. In the foreword to Grabow's sympathetic 1983 biography Christopher Alexander: The search for a new paradigm in architecture, Alexander states: Our present cosmological picture is largely the one which was first sketched out by Descartes. This picture has been immensely influential — perhaps responsible — for the great success of science in the last three hundred years.... Nonetheless, I believe this view is so partial, so badly flawed, that one might have to say, that in one respect at least, it is altogether wrong. ... I have now become convinced that it is possible to construct a 104 That is, a deterministic, ordering system, but not a Newtonian system. 105 In this quote she has also casually anticipated the first two-thirds of this thesis. 131 modified view of nature, which does pay proper attention to the issue of order.... I am also convinced that a proper theory of architecture, a way of understanding the production of buildings as the production of deep order, can only be created, within this shifted world view (Grabow 1983, x-x i ) . Alexander specifically links this 'new order' to cities within A New Theory of Urban Design, and at much greater length, in The Timeless Way of Building. Significantly, Alexander describes the natural urban order-creating process in terms of a dynamical process — something that any student of "Chaos Theory" w i l l find familiar: When we look at the most beautiful towns and cities of the past, we are always impressed by a feeling that they are somehow organic. This feeling of "organicness", is not a vague feeling of relationship with biological forms ... [or] an analogy. It is instead, an accurate vision of a specific structural quality which these old towns had ... and have. Namely: Each of these old towns grew as a whole, under its own laws of wholeness ... and we can feel this wholeness, no only at the largest scale, but in every detail.... (Alexander 1987, 2-3). Whi le the vocabulary employed by Couclelis, Cartwright, and this thesis (all acquainted with Chaos Theory) differs from that of Jane Jacobs and Christopher Alexander, it is clear that al l of them perceive the city to be a natural generator of complex order, rather than the root cause of confusion and random disorder. As such, their work anticipates a key assumption underlying dynamic planning. 5.2.5. Simple by-law structure appropriate to stimulate urban complexity Dynamic planners, aware of the diversity which can result from a single function, would take a fundamentally different approach to by-laws and regulations, although it is admittedly an approach at some odds with current legal trends. As noted above, the aim of dynamic planners is to maintain (and to adjust as necessary), by-laws which collectively circumscribe a civic "attractor" containing an infinite number of desirable futures for the city. A l l developments or activities which took place within this bounded realm of permissible activities would ensure that the needs of city's residents were met, and that 132 ongoing development would benefit the city and its inhabitants. This same attractor would exclude (and thus forbid) by by-law or regulation, a myriad number of activities held to be against the future interests of the city. Because simple systems of Lorenzian processes routinely produce a limitless number of phenomena with given bounds, maintenance of a civic "attractor" not nearly as daunting a matter as it might appear to a synoptic planner. The best single example of this assumption may be provided by the beautiful fractal shape generated by the Mandelbrot Set, which we have already discussed in Chapter 2. As we saw, the shape is infinitely complex, and w i l l never be fully explored. Yet this infinitely complex whole can be perfectly described in extremely simple terms: the Mandelbrot Set is the graphical presentation of all numbers which do not exhibit Newtonian positive or negative feedback behaviour for the function: X 2 + x = x' Simply by adjusting or maintaining this simple one-line function, we can exercise complete control over the infinite variety the function produces (Gleick 1987, 223). Likewise, instead of the mass of restrictive regulations currently employed in synoptic planning, aimed at controlling phenomena within the city, dynamic planners would instead favor a vastly fewer number of what might be termed "objective-oriented" by-laws, aimed at controlling the processes that gave rise to these phenomena. What would these by-laws look like? Remarkably, the National Research Counci l of Canada (NRC) is currently working on a wholesale revision of their fire code along "objective-oriented" lines. Arguing that the current regime of micro-management stifles innovation and prevents the development of more effective and more economical solutions to fire safety in buildings, the NRC intends to replace it wi th "performance criteria". Under the proposed new system, walls could be constructed in any fashion as long as they met certain standards — they could not burn through for a set period of time, for instance (Ellis 133 1997). While these criteria are still far more detailed than the theoretical "single-page" by-law we looked at above, it does clearly represent a radical reduction in the complicatedness of the by-law control mechanism for fire control. In the previous chapter, we took a brief look at the current, synoptic fire code for the City of Vancouver, currently 213 pages long (City of Vancouver, 1992). By contrast, we might imagine a fire control by-law written by a dynamic planner to be contained on a single sheet of paper. It might read thus: No building w i l l be constructed unless the Board of Planning can be convinced that its design w i l l prevent the building from a) catching fire easily, b) endangering its occupants by exposure to fire, c) spreading a fire to neighboring buildings and d) endangering passers-by, or occupants of other buildings. We may be sure that dynamic planners would accept any building that was built according to the current fire code of the synoptic planners, because this set of standards was designed to meet these objectives. But even if the board of planners was as cautious as can be imagined, it is still possible to conceive that a virtually limitless number of other valid solutions to the fire risk might also be approved. Nor would the planners be forced to generate these new solutions themselves. If a developer wished to employ new building materials, create an innovative building form so as to passively reduce fire risk, or exploit a novel method for extinguishing fires, she would be responsible for obtaining and presenting credible evidence that such innovations were effective. Thus, one by-law could potentially allow any number of resultant phenomena, all of which satisfied dynamic planning's central aim — the maintenance and enhancement of desirable phenomena within the city. We can turn to Alexander for another example of these "objective-oriented" by-laws. The rule system that King and himself devise for their experimental section of San Francisco 134 is far simpler than the complex neighborhood that is built in accordance with this framework. Alexander's rule structure consists of only seven rules, each of which consist of between none to 11 and 25 sub-rules (these latter sub-rules prescribe how to construct buildings and how to design large buildings). If printed in the manner of most by-laws, Alexander's rule system would amount to three or four pages. To some extent, we might well expect Alexander's rule system to be simple, for he is conducting his self-described "experiment" under laboratory conditions. The rules need only to control the creation of architectural models and their placement on a plywood replica of the San Francisco waterfront. Governing an uninhabited scale model, Alexander saw no need to include a fire code at al l , objective-oriented or otherwise. But we cannot explain away the simplicity of Alexander's rule-system simply by citing the simplicity of his experiment. As we have seen, Alexander is interested in producing complex urban form and unpredictable development, not Newtonian order. More importantly, Alexander has had a long history of devising simple rule systems specifically for the creation of "infinitely" complex order. Alexander's 1974 work A Pattern Language contains 253 patterns with which to design everything from a bedroom to a nation-state and is set out in a logical order, with the individual patterns roughly arranged from largest to smal les t 1 0 6 . Viewed from a synoptic point of view, 253 rules would hardly suffice for a decent village. Designing an entire country according to these dictates would therefore seem to entail an Orwell ian homogeneity and insensitivity to surroundings. However, 10G Those who are setting out to design a structure using the "Pattern Language" are advised to work through the book in the same order, establishing the large spaces before working to the smaller details. This would be a "Newtonian" type of top-down arrangement, were it not for Alexander's insistence that the process of entire design be done without resorting to plans any more accurate than a back of the envelope sketch. The reason for this is simple — if one were to draft an accurate plan, the shape and orientation chosen for the large spaces would automatically constrain or entirely prevent the placement of the smaller patterns which were to compose it. By leaving the house plan as amorphous as possible for as long as possible, Alexander enables the requirements of the smallest patterns to affect the placement and nature of the largest, in a fashion akin to that of the "sensitive dependence" exhibited by Lorenzian systems. 135 Alexander maintains that because even the most rigorously defined "pattern" w i l l be affected by the different needs of different people, by its placement amongst the other patterns, and by the nature of the surrounding environment, any "pattern" w i l l nevertheless allow people to "use... [it] a mil l ion times over without ever doing it the same way twice" (Alexander 1977, x): The individual processes are standardized, and very simple. But the actual parts which are produced are infinitely various — they are infinitely different manifestations of the patterns which the processes define (Alexander 1979,463) . Synoptic planners look at complex cities, and discern a complicated process underneath which can only be controlled on a phenomena-by-phenomena basis. Synoptic planners — and those who have anticipated them — see the same complexity as the product of relatively simple functions, which can be simply managed, by means of a small number of carefully-thought-out "objective-oriented" by-laws. 5.2.6. Equilibrium is constantly shifting We should not expect dynamic planners to be goal-oriented in terms of percentages or target dates. Given that Lorenzian processes themselves are characterized by infinite, continuous change, it seems logical that dynamic planners would not assume that there was any 'finished' state in buildings or in the city as a whole. Nor does it seem that a dynamic planning perspective would be compatible with goal-setting at a l l , in terms of effecting a certain amount of change in a certain future by a certain date. Instead, the goal of the dynamic planners would be to ensure that al l legally-permitted futures of the city were healthy and desirable for the city and its occupants. However, it stands to reason that a dynamic planner would not even assume this could ever be fully achieved. As the city continuously and unpredictably developed, even within the beneficial bounds set for it by the planners, the vision of what was suitable or beneficial for the city would likewise 136 evolve. Instead of accessing projects upon 'completion', dynamic planning would by nature be characterized by continual review. Following on from this, we might additionally expect dynamic planners to have an implicit appreciation of the uses and value of developments at al l points in their life cycle, from being shiny, new, high-rent and fashionable to being old, obsolete and run-down. The dynamic planner, armed with the conviction that the city itself is in continual adjustment from its present state towards an indeterminate future condition, would seem to be far more attuned to the modification of existing buildings and the redevelopment of extant neighborhoods. Here too, the assumptions we have derived for dynamic planning have been anticipated by Jane Jacobs, who clearly sees the city in terms of continuous change 1 0 7 , of active growth and decline (Jacobs 1961, 220): This book has discussed cities, and their components almost entirely in the form of processes, because the subject matter dictates this. For cities, processes are of the essence (Jacobs 1961, 440). This assumption of constant change is especially pronounced in Jacobs' discussion of what she calls "slumming" and "unslumming". Rather than seeing 'prosperity' and 'poverty' as the static end-states of 'development' and 'blight' respectively, Jacobs sees even the immured slum as the expression of an active, ongoing process: The constant departures [of those who can from the slum] leave, of course, more than housing vacancies to be filled. They leave a community i n a perpetually embryonic stage, or perpetually regressing to helpless infancy. The age of buildings is no index to the age of a community, which is formed by a community of people. In this sense a perpetual slum is always going backward instead of forward... (Jacobs 1961, 277). 107 Rather unkindly, Jacobs interprets the synoptic planners' static conceptions of city development in terms of conspiracy: It was the very fluidity of the new nineteenth-century industrial and metropolitan society, with its profound shiftings of power, people and money, that agitated [Ebenezer] Howard so deeply — and his more dedicated followers (like the American Decentrists and Regional Flanners) after him. Howard wanted to freeze power, people, and the uses and increments of money into an easily manageable and static pattern. ... The restoration of a static society, ruled — in everything that mattered — by a new aristocracy of altruistic planning experts... (Jacobs 1961, 289-290). 137 The processes that occur in unslumming depend on the fact that a metropolitan economy, if it is working well , is constantly transforming many poor people into middle-class people, many illiterates into skilled (or even educated) people, many greenhorns into competent citizens 0acobs 1961, 288). Considering Jacobs' explicit assumption of constant change and shifting equilibria in the state of the city, it is hardly surprising that she is also acutely conscious of the changes and processes catalyzed by 'old ' neighbourhoods and 'finished buildings'. Far from seeing old buildings as 'obsolete' or simply as potential sites for future 'development', Jacobs argues that structures which have paid off their construction costs are crucial "incubators" for dynamism within the city. Only within such low-rent spaces can the small-scale experimental, marginal and speculative operations which w i l l drive the long-term future of the city be developed. The relevant chapter is simply entitled "The need for aged buildings" 0acobs 1961, 187-199). Although the structure of his experiment in A New Theory of Urban Design is somewhat misleading, Alexander consistently denies that there is an end state in construction and urban form. In The Timeless Way of Building, Alexander argues that no one can have a perfect conception of a building they are imagining in their head. Completion of the initial construction phase, in his eyes, merely allows the builder to see the features of the building which can be improved. "Completion" is therefore only the beginning of an endless process of alterations, renovations, and repair. This cannot come to an end, simply because the renovations in themselves w i l l create new, unanticipated possibilities and problems. Even at the level of a single building then, Alexander argues that development and change is both unpredictable and eternal (Alexander 1979, 479-480) This is also true, on a much deeper level, of the " pattern language" which constrains the creation of these buildings. The pattern language itself is dynamic, continually 138 changing and evolving, as new and better patterns are adopted, and older, less effective patterns are discarded. This change in the pattern language w i l l in itself create new conflicts, and hence, a need for new, unanticipated patterns to resolve these conflicts. Alexander sums up his argument with the statement "there is no final equil ibrium" (Alexander 1979, 346-347). Despite the above, the experiment in A New Theory of Urban Design is not a perfect parallel of our own dynamic planning method. Taking place on a large plywood model over a single term of university classes, the experiment ended when virtually all of the open space on the model had been developed into buildings, roads and open space. Moreover, none of the 90-odd individual projects focused upon modernizing or re-using one of the handful of structures which already existed on the site. As a result, Alexander's "new theory" does not seem to address the building 'life-cycle' issues of repairs, expansion and remodelling that we singled out as a feature of our own method. In large part, the end-state assumptions in A New Theory... may simply be a consequence of the fact that the experiment was conducted as part of a university course, which itself 'finished' after a single term. Similarly, the lack of attention to repair and renovation may be attributable to the nature of the architecture students' course curriculum. Nevertheless, the end-state 'problem', if it is one to Alexander's way of thinking, is never mentioned, even i n a short self-critique which Alexander offers at the end of the work. As such, we must conclude that Alexander has adopted the "end-state" assumption of the synoptic planners, despite emphasizing the ever-evolving nature of cities earlier in his career. Synoptic planners see the active vibrant city continually jolted by the impact of new forces and the imposition of new and more complicated systems. In contrast, dynamic planners and their precursors perceive the city travelling — relatively smoothly — along 139 the endless path of its infinite attractor, continuously moving, yet continuously in equil ibrium with the forces acting upon it. 5.2.7. Fractals are a sign of order; randomness need not indicate disorder If dynamic planners were to regard urban processes as the workings of a Lorenzian system, it stands to reason that they would also view the phenomena produced by the city in a substantially different manner from their synoptic colleagues. Rather than associating geometry with order, we would expect dynamic planners to associate fractal shapes, and even "random" patterns (such as those traced by the snowboards descending the Whistler Mountain mogul-field) with order. Moreover, a stress upon heterogeneity naturally follows from the dynamic planner's acceptance of constant change, diversity of solutions, and self-similarity within built form. Rather than sameness, a striking melange would be the signal feature of a well-ordered city to a dynamic planner. Associating 'random' patterns with order seems counter-intuitive from the standpoint of synoptic planning. We recall that Caws' dictum from Chapter 1 specified that "order" is: [an] arrangement with respect to which it would matter if it were otherwise (Caws 1968, 108). How can this be reconciled with 'random' patterns, whose specific arrangement is of no consequence? The answer is simple enough — the dynamic planner would perceive (at least some of) the 'random' patterns generated by the city not to actually be random, but rather, be the random-appearing products of a deterministic urban system. Despite the modern vocabulary, this too is not a new perception. In the late nineteenth century Camillo Sitte argued that the irregular streetscapes of medieval and Renaissance-era towns were not 'random' or disorderly, but were the purposeful and ordered creations of architects and city builders: 140 ...city planning should not be merely a technical matter, but should in the truest and most elevated sense be an artistic enterprise. Such it was in Antiquity, in the Middle Ages, in the Renaissance; indeed, wherever the arts were fostered. It is only in our mathematical century that the process of enlarging and laying out cities has become an almost purely technical concern (Sitte 1965,4) . Sitte maintained that these expressions of "aesthetic" order were in fact superior to the explicitly geometric arrangements favored by the new synoptic planners of his time: [In the Medieval era] ... quite incomprehensible choices of location [for statuary] were made, and yet one must grant that a fine sensibility guided the choices since... everything always harmonized beautifully. Thus we are presented with a mystery - the mystery of the innate, instinctive aesthetic sense that worked such obvious wonders for the old masters without resort to narrow aesthetic dogma or stuffy rules. We on the other hand, come along afterward, scurrying about with our T-square and compass, presuming to solve with clumsy geometry those fine points that are matters of pure sensitivity (Sitte 1989, 21). In even more explicit fashion, Jane Jacobs also anticipated this assumption of dynamic planning when she argued that civic order was not to be found in geometry, but rather in the diversity which her contemporaries took to be random behavior: We are constantly being told simple-minded lies about order in cities, talked down to in effect, assured that duplication represents order. It is the easiest thing in the world to seize hold of a few forms, give them a regulated regularity, and try to palm this off in the name of order. However, simple, regimented regularity and significant systems of functional order are seldom coincident in this world. To see complex systems of functional order as order, and not as chaos, takes understanding. The leaves dropping from the trees in the autumn, the interior of an airplane engine, the entrails of a dissected rabbit, the city desk of a newspaper, al l appear to be chaos if they are seen without actual comprehension. Once they are understood as systems of order, they actually look different (Jacobs 1961, 375-376). Intricate minglings of different uses in cities are not a form of chaos [i.e. disorder]. O n the contrary, they represent a complex and highly developed form of order 0acobs 1961, 222). Intriguingly, considerable work has been done within the fields of urban geography and economics over the past fifteen years, l inking "random" and fractal patterns produced by Lorenzian processes to known urban processes and actual city form. Although many of 141 these models require some random inputs to generate random-looking behaviour they also incorporate Lorenzian processes as w e l l 1 0 8 . In their 1984 work The Social Logic of Space, Hillier and Hansen attempt to divine the processes which give rise to built form. The authors develop a very simple rule system corresponding to the built form of the villages in the Vauclose region of southern France. When replicated randomly, the simple rule structure gives rise to streetforms analogous to those in the villages (Hillier and Hansen 1984, 52-81). A more complex model of this sort, intended to replicate urban form on a metropolitan scale, was presented in White and Engelen's 1993 paper "Cellular Automata and Fractal Urban Fo rm" 1 0 9 . The two modellers' simulated cities are composed of vacant, residential, commercial and industrial pixels, with each land-use type having differing affinities for the other varieties. Each run of the model started with a village of six pixels, and grew at a pre-set rate. A particular level of randomness or 'chance' was also set for each run. White and Engelen obtain "cities" of pixels, self-separated into industrial districts, commercial cores (often including a prominent "central business district"), and intervening residential areas. The resulting pattern of urban growth is remarkably fractal in form, which they liken to the land-use pattern of existing North American c i t ies 1 1 0 . Very different cities can be 'grown' from the 108 Michael Batty of University of Wales, together with other investigators, has published a number of papers and recently written a textbook in which he grows fractal "cities" on computer through the random accretion of "built up" pixels onto an initially designated "urban core" pixel (Batty 1991; Batty and Longley 1994; Batty, Longley and Fotheringham 1989). The resultant cities do indeed look like low-resolution plans of cities, although as White and Engelen point out, Batty et al.'s methods "[do] not seem to correspond very closely to any actual urban growth process"(White and Engelen 1993, 1177). Wong and Fotheringham's 1990 paper on "Urban Systems as Examples of Bounded Chaos", likewise uses the random accretion of "rural" populations to a gridwork of "cities" to demonstrate the compatibility of fractal processes with the rank-size parameter which plays a central role in theories of urban growth (Wong and Fotheringham 1990). 109 With cellular automata, the state of any "active" cell in an array is dependent upon the states of the neighbouring cells. The phenomena produced by the actions of cellular automata are thus inherently spatial, well suited to geographical applications. 110 Nevertheless, the coarse resolution and relative simplicity of the model ensures that White and Engelen's silicon cities lack transportation networks, public space, or any means of addressing densification or physical geography, all of which have long been assumed to be important determinants of urban form. It should also be noted that White and Engelen's model assumes that zoning is a natural phenomenon rather than an artificially imposed one (White and Engelen 1993). 142 same village because of the stochastic disturbance provided in the program, a result likened to the "sensitive dependence' exhibited by Lorenzian systems (White and Engelen 1993). These modelling efforts and others do not prove that the city is in actual fact a Lorenzian system 1 1 1 . They do seem to indicate, however, that some academics (though not planners) have started to view the "random" patterns of the city as being ordered, determined at least in part by Lorenzian processes. Once again, after we have made allowances for Alexander's decidedly non-scientific terminology, the congruency of his "new theory" with that of dynamic planning is striking. Rather than talk about 'ordering' cities, Alexander states that "every increment of construction must be made in such a way as to heal the city," and goes on to argue that this must be done by creating "wholes" or "centres" in the city (Alexander 1987, 22). Alexander describes his second basic rule, termed "the growth of larger wholes" in the following manner: Every building must help to form at least one larger whole in the city, which is both larger and more significant than itself (Alexander 1987, 38-39). No single building is to create such a "whole", and these centres are to arise "spontaneously"; centres w i l l not be planned in advance, but w i l l be encouraged once the potential of a given location becomes apparent. Ideally, a new building w i l l help to complete an identified whole, help to delineate another whole and begin the creation of a third centre I 1 2 (Alexander 1987, 43-44). In similar fashion, Alexander's seventh and final rule, entitled "formation of centres", where he defines a "centre" as follows: 111 In 1982, before she was aware of Lorenzian systems Couclelis stresses the shortcomings of extrapolating concepts and processes in the physical sciences to "analogous" applications in the social sciences. If one works by analogy, instead of from first principles, then any "fit" between the model and observed reality may only be an artifact in the model's construction (Couclelis 1982). Still, the modellers' success in replicating the fractal structure and dynamic nature of modern cities with simple rule structures, and their apparent ability to reproduce sensitive dependence in their modelling is certainly promising 112 Nor is this wholeness to be applied to buildings alone. The fourth rule: "Positive urban space," ensures that the shape of the outside public spaces formed by the creation of buildings is as important as the shape of the buildings themselves. "We may express this rule 143 a. It is whole in itself, in an obvious, relaxed way, with its own symmetries. b. Its main parts are also whole, and have their own symmetries. c. The space or buildings next to it, in so far as they are themselves whole, have their own symmetry. d. The whole is always part of some still larger whole, which is itself a centre, possessing certain symmetries (Alexander 1987, 94). Whi le these "wholes" are ordered, they are not geometric in the Euclidean sense. Instead, they are fractal: so much so that when Alexander details the characteristics of "centres", he appears to be describing the self-similar Mandelbrot Set with its infinite numbers of gingerbread men within gingerbread men. As we have seen in previous chapters — and in the world about us — synoptic planners equate straight lines and Euclidean geometry with order, for random activity cannot produce such shapes. Because they discern order behind seemingly random events, dynamic planners cannot recognize order through the presence or absence of geography. Instead, the dynamic planner recognizes the sign of order in fractal shapes, capable of infinite variety, yet united in self-similarity, where each part of the whole resembles the whole. It does not take a planner to notice the difference between a city like Brasilia and one like Brazzaville, the former designed to resemble a bird when seen from the stratosphere, the latter looking like nothing so much as the finely-veined leaf of a tree when viewed under a microscope. Nor would most people have difficulty in spotting a fraternity of city form amongst Brazzaville, the medieval towns of Western Europe, the O ld City of Jerusalem, the favelas of Sao Paulo - maybe even the fishing villages of Newfoundland and the Greek Cyclades. The order of the synoptic planner is displayed in the boulevards oi Paris: the order of the dynamic planner is apparent throughout the quartiers in between. simply as follows: 'buildings surround space', NOT 'space surrounds buildings' (Alexander 1987, 44). By specifically referring to public space as an entity in itself, Alexander also intends that these objects be subject to "wholeness". 144 5.3. Conclusion We have gone through the list of characteristics of Lorenzian systems, working from it to derive aims and assumptions for a new method of dynamic planning. In doing so, we were consciously paralleling the process by which the aims and assumptions of synoptic planning were derived from the characteristics of Newtonian systems. As we have seen, the method of dynamic planning has proved to be qualitatively different from that of synoptic planning. A comparison of the two planning methods point by point confirms that many of the aims and assumptions of dynamic planning are actually at odds with those of synoptic planning. Table 6: Comparison of Synoptic Planning Aims and Assumptions with Dynamic Planning Aims and Assumptions Synoptic Planning Dynamic Planning Planning policies determine city development Planning policies shape city development Specific future conditions of the city can be determined Future conditions of city can be bounded Newtonian order results if 'random' urban phenomena are controlled Lorenzian order naturally produced by urban dynamical system Complex by-laws are required to control urban complexity Simple by-law structure appropriate to stimulate urban complexity Large-scale processes have greatest influence over city development All scales of processes have influence over city development Equilibrium is constant Equilibrium is constantly shifting Geometry is a sign of order Fractals are a sign of order; randomness need not indicate disorder 145 And yet, we have also seen that each individual tenet of our invented method, derived from first principles, has actually been anticipated by planners and planning theorists. In some cases, initiatives resembling the individual components of dynamic planning-method have even been carried out in practice. We can, therefore, take some assurance that "dynamic planning" is not simply a sterile exercise in the manipulation of theory, but may actually correspond to articulated needs and proven methods. This reliance upon the words and works of previous planners and urban critics to bolster a radically new conception of planning may seem rather conservative. Whi le the assumptions informing the new method are often very different from those of synoptic planning, it seems that the actual methods prescribed by dynamic planning prove to be not so distant from the familiar practices of "muddling through", "incrementalism", and "establishing a vision for the future". This may even be disappointing for some. After al l the implicit criticism of synoptic planning methods and talk of paradigm shifts, we have nevertheless ended up developing a relatively modest alternative. We seem to have ended up at a tea party rather than a revolution. A n d yet, there can be no question that the practitioners of dynamic planning w i l l see their task, and see order itself in a completely different manner from that of the synoptic Visionary' or 'muddler'. Dynamic planners can justify their actions and outlook from first principles, as a scientific way of promoting order in the city. By contrast, synoptic planners who proceed on the basis of stop-gap adjustments developed in the face of crisis are unable to justify their practices except on the grounds of practicality. Thus, the worth of Lorenzian systems, and the method of "dynamic planning" one can derive from Lorenzian system characteristics, is that it provides a scientifically valid, internally consistent framework that brings together approaches existing elsewhere in piecemeal or scientifically suspect form. 146 CHAPTER SIX: ORDERS A N D PLANNING 6.1. Looking Backwards From its origins through to the present-day, the discipline of city planning has always defined its mission as 'ordering the city'. Planners have differed as to what ends the city should be ordered, but the benefit of order itself has rarely been questioned. Wi th in the last century or more, there has been even greater agreement. If we are to order the city, it should ordered scientifically. This thesis has argued that the aims and assumptions of our present practice of synoptic planning is based on the characteristics of Newtonian systems, the only type of natural ordering process known to science at the birth of professional city planning. Moreover, this thesis argues that the profession's 'crisis of faith' — which has beset planners since the practical and theoretical shortcomings of the synoptic method were made clear in the 1950's and 1960's — cannot be overcome until the assumptions of the Newtonian paradigm are explicitly acknowledged. While the synoptic method of planning (and its related diaspora of revisionist practices) is based on the characteristics of one form of natural ordering processes, late-twentieth-century science now recognizes a second, distinctly different, type of natural ordering process. Both create equally forms of "order" but they differ from each other as much as cabbages do from kingfishers. Known to scientists as "chaos", this second variety of ordering processes have been referred to here as Lorenzian systems. We now know that both Newtonian and Lorenzian systems are involved in ordering the natural world. We have also determined that the planning method we presently use -based on the characteristics of Newtonian systems alone - is not effective or desirable for al l 147 planning purposes. As a result, the question becomes obvious: should we not investigate what a planning methodology based on promoting this non-linear order might be like? Work ing from the first principles informing recently-recognized Lorenzian systems, we have derived a new, "dynamic" method of planning. Dynamic planning is intended to create urban order, just as our current method of synoptic planning is. Moreover, if Alexander's work with his remarkably similar "new theory" is anything to go by, this method of dynamic planning appears to be at least a potentially workable methodology. M u c h of this thesis has argued that a cogent and logical new method of "dynamic planning" can be developed, based upon the characteristics of Lorenzian systems. Is dynamic planning itself, then, the reason for all this discussion and argument that we have put ourselves through? Not really. As we have seen in Chapter 5, the assumptions and techniques of dynamic planning have been available for decades in the form of Jane Jacobs' critiques and Christopher Alexander's prescriptions. Instead, this thesis has attempted to present a new perspective for thinking about planning. Rather than simply show how we might go about ordering the city, this thesis asked why we order the city the way we do. The real aim of this paper has been to suggest that the scientific paradigms of Newtonian and Lorenzian systems have an important, even crucial importance to planning theory. Indeed, this thesis forcefully argues that many of the problems currently facing the theory and practice of planning have resulted from an outdated exclusive reliance on the Newtonian paradigm, and an ignorance about the characteristics of Lorenzian natural ordering processes. This thesis posits that if we as planners seek to create scientifically-valid order, then we should be as effective when going about planning in the manner described by Jane Jacobs and Christopher Alexander as we are when acting in accordance with synoptic aims and assumptions. 148 As planners today, our vision of order is not wrong, but it is incomplete, as though we were viewing the city with a single eye. We are half-sighted: we lack the needed depth of perception to distinguish civic order in all its forms. Our blind spots prevent us from being alerted to the crucial details of urban process. For the previous two generations, we have been reduced to tapping our way forward incrementally, or recklessly rushing forward, trusting our expert-made maps and guidebooks to see us through. But on the whole, in order to prevent disaster, we have simply stood still. This paper argues that it is time that we as city planners opened our eyes to the "new science" of Chaos Theory which offers such promise for the future of the profession, as wel l as the old science of Newtonian systems, which have shaped so much of our past and present. Once we acknowledge that our actions and goals as planners are not self-evident, but the outcome of how we interpret the world about us, then the irresolvable simply becomes a question of tactics, and the impassable just a matter of the proper approach. A definitive proof of these assertions can not be presented within the scope of a master's thesis, if at al l . The arguments of this thesis remain very debatable propositions, and w i l l require a great deal of supporting information before they are widely accepted. Nevertheless, these are topics about which there have been remarkably few remarks. This thesis was not intended to resolve a debate, but to provoke one. 6.2. The Orders of the Ci ty In championing the cause of dynamic planning, this thesis has emphasized many of the shortcomings in synoptic planning. Nevertheless, it is most unlikely that any variant of dynamic planning based on Lorenzian systems would prove itself to be a complete solution to the challenge of ordering the city, any more than synoptic planning on its own has 149 proved to be. The universe is not purely Lorenzian, any more than it is purely Newton ian 1 1 3 . Kuhn himself reminds us that that new paradigms consistently fail to handle certain issues as wel l as the old paradigms they replaced were able to (Kuhn 1970, 153-157). As sketchy as our outline of dynamic planning is, the relevance of Kuhn's statement is apparent. If, for example, Alexander is correct, and his dynamic "new theory" is the modern equivalent of the pre-modern "timeless way of building", then we must remind ourselves why we don't order the city by this method today. As we saw in Chapter 3, the traditional mode of urban planning utterly failed to provide the crucial basic infrastructure that residents required. Safe water, clean power, sewerage, solid waste management and efficient medium and long-distance transport were simply not provided for in pre-industrial cities. In this regard, it is worth noting that Alexander, our adopted prime exponent of "dynamic" planning, never addresses the toilet, the sewer, the power line, or the water pipe. The issue of personal sanitation, for instance, seems to be mentioned only in passing within a discussion of the need for social bathing areas (Alexander 1977, #148) 1 1 4 . It appears that if planners were to order the city solely in terms of dynamic planning, at least as Alexander has envisaged it in his "new theory", we could well be thrown back to the age of metropolitan cesspit deathtraps. Clearly no one is about to do this, not even Alexander himself, for it is obvious that he assumes, together with his readers, that synoptic planners w i l l continue to provide the utility hookups, the boulevards and sewerage that they have done so well for the past century. In turn, Alexander hopes the synoptic 113 Moreover, we risk hubris to state even now that we have complete knowledge of the ways in which Nature orders itself - there may yet be more than two orders to the universe. 114 While a transportation planner will find the treatment cursory, streets, arterial roads, cars, buses and even parking lots are acknowledged in A Pattern Language (Alexander 1977,#11, 16, 17,20,22,23,49-57, 92, 97, 103, 113). 150 planners w i l l stand aside when it comes to ordering the neighbourhood, the local street, and the shape and setback of a single house. 6.2.1. A tale of two orders We have talked planning theory and Chaos Theory for more than one hundred and fifty pages now. We have even gone to the trouble of deriving a entire hypothetical method of planning, an illustration of the new perspectives available when we acknowledge the roots of planning theory in scientific paradigms of natural ordering systems. But our task is finished now — we are at the end of the thesis. Surely this is the time for a brief vacation! We are city planners, so let us travel to a place of buildings and streets. Moreover, let us go to a place where people shall understand this strange language we have been speaking of late - our dialect of "Newtonian" and "Lorenzian systems", of "synoptic" and "dynamic planning". Let us go to the city of S i r ap 1 1 5 . We arrive by plane, and just as Mumford did when he circled above Megalopolis, we peer out of the little portholes, anxious to see if we can discern any order in the city below. We are surprised perhaps, and certainly relived, to see that yes indeed, with the whole city below us spread out like a map, there is a definite structure and purpose to the way in which the wide boulevards traverse the urban area. They are spaced at generous but regular intervals, and are wide enough to carry the produce and the people of the city efficiently from one part to the other. It seems to be a most rational arrangement. Before we know it, we have passed through the terminal, for the airport too is designed to efficiently move us from plane to downtown train. One transfer later, we stand outside the neighbourhood transit station on one of the same boulevards we saw from the air, wondering where our lodgings might be. We have booked into a small place - a pension -115 As we shall see, the inhabitants have named their city for their habit of putting in the boulevards first, and only then constructing irregular arrondisments between them. It is quite the reverse of Parisian practice. 151 located in the tight nest of streets that lie behind the station. We are glad for the map at the station, and the streetsigns at every turning, or else we would be sure to be lostl The streets curve and twist as we head up the h i l l , and the streets are only wide enough for two cars to pass: the delivery trucks beep us as they sidle along. We haven't gone much more than two hundred metres before we come to a little square with a tiny restaurant where some kids are noisily playing street hockey. Our four-storey pension is located on the smaller, convex north side of the 'square', butted up against an apartment on one side, and a bakery and computer repair shop on the other. There is a small rectangular blockhouse off to one side of the square which intrigues us. One of the hockey players tells us that this is the power, water and sewer block for the immediate area, servicing the surrounding hundred meters or so. She says it was the first building in the neighbourhood, back when just the boulevards had been laid out, and people hadn't yet begun to build here. The people on her block al l got together to install a common conduit to the blockhouse during the construction of their homes. At the pension, our hosts tell us that the city hasn't grown very much in the past few years, but if and when growth picks up again, the planners w i l l extend the boulevards to the west, and construct and service some more blockhouses. The hills to the east, and the seashore to the north are being preserved in perpetuity as park and wilderness. We have rooms on the top floor, which were built, the owners say, so as to catch the last of the afternoon Apr i l sun. The neighbouring apartment has a deck on the second floor instead, in order to make room for the spreading limbs of the large tree out front. Someone across the square is putting up a second floor on her house, which w i l l shade the dietitian's office across the way this summer, so it looks like the doctor w i l l enclose the old deck and bui ld a new one on the roof instead. Someone's always making renovations. 152 Last year our hosts were extended a development loan by the city's Chamber of Commerce, and they opened up a second and third room to tourists. Because the little restaurant now has three tables virtually guaranteed at dinner, they've been able to hire a decent chef. Now the locals eat there more often too. As we look to the lovely rose-tinted sunset above the roofs of town, our hosts talk of opening a little coffee room on the ground floor. First, however, they'll first have to demonstrate to the city's satisfaction how they'll prevent the added noise from disturbing the neighbours at night. The man of the house said the planner made herself quite clear: " M e n of disorder need expect no indulgence from m e " 1 1 6 . 6.3. A Final Note It is time that we as planners opened our eyes to the 'new science' of Chaos Theory as well as the familiar wisdom of Newtonian physics. These are both invaluable tools in our task of ordering the city. But by the same token, we must make sure we see that which is really in front of us. In The Structure of Scientific Revolutions, there is a discussion about a cleverly executed drawing which can be interpreted either as a duck or a rabbit. Kuhn notes that having gone through the "gestalt switch" from one mode of vision to another - seeing the bunny, then the duck, and then the bunny again - the viewer is able to perceive the drawing in a third way as well . The observer can see the drawing for what it is; a collection of lines printed on a sheet of paper (Kuhn 1970, 114). We have already considered that in a 116 The planner is quoting Georges-Eugene Haussmann, who ordered a decidedly chaotic Paris with his network of linear boulevards. To be honest, Haussmann didn't have city planning in mind when he spoke the words "les hommes de desordre ne doivent attendre aucune menagement de moi." Instead, Haussmann was intent on keeping his position as Prefect of the Gironde in the wake of Louis-Napoleon's coup-d'etat in 1851(Malet 1973, 93;Jordan 1995 136-139). Nevertheless, this statement is an succinct summary of his, and our, outlook as city planners. 153 universe of Newtonian and Lorenzian order, it is unlikely that we shall be able to effectively plan our cities in terms of one order alone. But if we as planners develop our craft so that we must constantly alternate our view of the city, sometimes in terms of Newtonian systems, sometimes from perspective of Lorenzian systems, we may eventually come to see the city itself, not as something ordered or disordered, as an example of a Newtonian or a Lorenzian system, but as what it actually is: a city. 154 BIBLIOGRAPHY Aberley, Douglas Carroll . 1987. A bioregional approach to community economic development. Vancouver, B.C.: Social Planning & Review Council . , ed. 1994. Futures by design: The practice of ecological planning, ed. Douglas Carrol l Aberley. Gabriola, B.C.: New Society Publishers Alexander, Christopher. 1975. The Oregon experiment. New York: Oxford University Press. . 1977. A pattern language: Towns, buildings, construction. New York: Oxford University Press. . 1979. The timeless way of building. New York: Oxford University Press. . 1987. A new theory of urban design. New York: Oxford University Press. Altshuler, Alan. 1965. The Goals of Comprehensive Planning. Journal of the American Institute of Planners (August): 186-197. Audain, Michael and Elaine Duvall. 1992. New directions in affordability: Report of the Provincial Commission on Housing Options.. Victoria, BC: Crown Publications. Barnett, Jonathan. 1986. The elusive city: Five centuries of design, ambition and miscalculation. New York: Harper & Row. Batty, Michae l . 1991. Generating urban forms from diffusive growth. Environment and Planning A 23:511-544. Batty, Michae l and Paul Longley. 1994. Fractal cities. London: Academic Press Ltd. Batty, M . , P. Longley and S. Fotheringham. 1989. Urban growth and form: Scaling, fractal geometry, and diffusion-limited aggregation. Environment and Planning A 21:1447-1472. Beer, Stafford. 1974. Designing freedom. Massey Lectures, thirteenth series. Toronto: CBC Publications. . 1981. Brain of the Firm. 2nd ed. Chichester: John Wiley & Sons. Bellan, Ruben C. 1971. The evolving city. Vancouver: Copp Clark. Benevolo, Leonardo. 1967. The origins of modern town planning. Trans. Judith Landry. London: Routledge and Kegan Paul. Boyer, M . Christine. 1983. Dreaming the rational city: The myth of american city planning. Cambridge, M A : The MIT Press. Branch, Melvi l le C. 1966. Planning: Aspects and applications. New York: John Wiley & Sons. Bueche, Frederick J. 1981. Understanding the world of physics. New York: M c G r a w - H i l l Book Company. Calthorpe, Peter. 1993. The next american metropolis: Ecology, community and the American Dream. New York: Princeton Architectural Press. Cartwright, T. J. 1991. Planning and Chaos Theory. Journal of the American Planning Association 57( l ) :44-56. Caws, Peter. 1968. Order and Value in the Sciences. \n The concept of order. Ed. Paul G. Kuntz. Seattle: University of Washington Press. Pp. 106-108. City of Vancouver. 1987. Proposed RS-1S revenue suite enforcement program and RS-1 illegal suite phase-out. Report to Council. December 17. . 1988a. Secondary suites in RS-1 areas: Enforcement and family suites: Report to Council . August. Photocopy. 155 City of Vancouver. 1988b. Secondary suites in RS-1 areas: 1986 and 1988 city wide policy. Report to Council . September. Photocopy. . 1992. Vancou ver fire by-la w: A by~la w to maintain acceptable standards for fire safety in buildings and occupancies within the City of Vancouver. Vancouver: Vancouver Fire Department. . 1996a. Draft transportation plan. Part one: the transportation strategy. Vancouver: City of Vancouver Transportation Planning Team. . 1996b. Urban structure policy report (Sidewalk cafe review). Report and recommendations of General Manager of Engineering Services. September 23. Photocopy. . 1996c. Planning for a downtown in transition. Vancouver: City of Vancouver Planning Department. Spring. . 1996d. Secondary suite statistics to December 31, 1996: Development permits. Vancouver: City of Vancouver Permits and Licensing. Photocopy. Collingwood, R. G. 1938. The principles of art. London: Oxford University Press. Coupland, Douglas. 1991. Generation X: Tales for an accelerated culture. New York: St. Martin's Press. Dear, Michae l and Allen J. Scott, eds. 1981. Urbanization and urban planning in capitalist society. London: Methuen. Dendrinos, Dimitrios. 1992. The dynamics of cities: Ecological determinism, dualism and chaos. Routledge: New York. District of Surrey. 1983. Surrey official community plan. Surrey, B.C.: District of Surrey Department of Planning. Duany, Andres and Elizabeth Plater-Zyberk. 1991. Towns and town-making principles. New York: Rizzoli. The Economist. 1994. Balancing broomsticks, June 25. Ellis, Michae l . 1997. Interview by author. Vancouver, February 2. M r . Ellis is a graduate student in the Advanced Studies in Architecture program, UBC School of Architecture. M r . Ellis is researching researching Canadian initiatives in implementing "performance standard" bylaws. Engwicht, David. 1993. Reclaiming our cities and towns: Better living with less traffic. Gabriola, B.C.: New Society Publishers. . 1995. Lecture: "Planned Chaos". Vancouver, November 3. Gibson, Wi l l i am. 1987. Burning Chrome. New York: Ace Books. Glieck, James. Chaos: Making a new science. 1987. New York: Viking. Globe and Mail. 1994. Demonstrators Jam Vancouver Streets, Apr i l 30 :M3. Grabow, Stephen. 1983. Christopher Alexander: The search for a new paradigm in architecture. Stocksfield: Oriel Press. Guterson, David. 1992. No place like home: O n the manicured streets of a master-planned community. Harper's Magazine (November) :5 5-64. [GVRD] Greater Vancouver Regional District. 1995. Livable region strategic plan. GVRD: Burnaby, B.C. Hal l , Peter. 1988. Cities of Tomorrow. New York: Basil Blackwell. Harris, Britton. 1960. Plan or Projection. Journal of the American Institute of Planners 26(4):265-272. Heap, Nicholas Ian. 1996. Author's personal work experience. Surrey, B.C. September-December. The author was a part-time student intern working with the Strategic Planning Department, B.C. Transit. H i l l , David R. 1992. America's disorganized organicists. Journal of Planning Literature 7(1):3-21. 156 Hill ier , Bi l l and Julienne Hanson. 1984. The social logic of space. New York: Cambridge University Press. Holl ing, C. S. 1978. Adaptive environmental assessment and management. Toronto: Wiley. Hudson, Barclay M . 1979. Comparison of current planning theories. APA Journal (October):387~405. Hunter, Robert. 1979. Warriors of the rainbow: A chronicle of the Greenpeace movement. New York: Holt, Rinehart and Winston. Jacobs, Jane. 1961. The death and life of great American cities. New York: Vintage Books. Jennings, Humphrey. 1985. Pandaemonium: The coming of the machine as seen by contemporary observers. Eds. Mary-Lou Jennings and Charles Madge. London: Picador. Jordan, David P. 1995. Transforming Paris: The life and labors of Baron Haussmann. New York: Free Press. Jukes, Peter. 1990. A shout in the street: An excursion into the modern city. New York: Farrar Strauss Giroux. Keene, D . J . 1990. Suburban Growth. The English medieval town: A reader in English urban history, 1200-1540. Eds. Richard Holt and Gervase Rosser. London: Longman. Khan-Magomedov, S. O. 1983. Pioneers of Soviet architecture: The search for new solutions in the 1920's and 1930's. New York: Rizzoli. Kitimat: A New City. 1954. The Architectural Forum. Reprint of articles from July, August and October 1954. Kostof, Spiro. 1991. The city shaped: Urban patterns and meanings through history. Boston: Little, Brown and Company. Kuhn, Thomas S. 1957. The Copernican revolution: Planetary astronomy in the development of Western thought. Cambridge: Harvard University Press. . 1970. The structure of scientific revolutions. 2nd edition, enlarged. Chicago: University of Chicago Press. Ladd, Brian. 1990. Urban planning and civic order in Germany, 1860-1914. Cambridge: Harvard University Press. Lewinberg, Frank R. 1984. In your neighbourhood. Toronto: [Ontario] Minis t ry of Munic ipa l Affairs. Lindblom, Charles E. 1959. The science of 'muddling through'. Public Administration Review (Spring):151-169. Lorenz, Edward N . 1963. Deterministic nonperiodic flow. Journal of Atmospheric Science. 20:130-141. . 1993. The essence of chaos. University of Washington Press: Seattle. Lynch, Kevin. 1981. Good city form. Cambridge: MIT Press. Malet, Henri . 1973. Le Baron Haussmann et la renovation de Paris. Paris: Les Editions Municipales. McGui re , B. J. and Roland W i l d . 1959. Kitimat - Tomorrow's city today. Ottawa: Canadian Geographical Journal. Reprint of article from November 1959 issue. Mumford , Lewis. 1938. The culture of cities. Harcourt, New York: Brace and Company, Inc. . 1956. From the ground up: Observations on contemporary architecture, housing, highway building, and civic design. New York: Harcourt, Brace and Company. . 1961. The city in history: Its origins, its transformations, its prospects. New York: Harcourt, Brace, Jovanovich. . 1963. Technics and civilization. New York : Harcourt, Brace & Wor ld . . 1968. The urban prospect. New York: Harcourt, Brace and Wor ld , Inc. 157 Mumford , Lewis. 1974. Theories and Ideals of Planning. In Encyclopedia ofurban planning. Ed. Arnold Whittick. New York: M c G r a w - H i l l . Pp. 985-996. . 1986. The Lewis Mumford reader. Ed. Donald L. Mi l l e r . New York: Pantheon Books. Nash, Michael . 1989. Computers, automation, and cybernetics at the Hagley Museum and Library. Wilmington, DL: Hagley Museum and Library. Naylor, Michael . 1997. Interview by author. Vancouver. Apr i l 7. M r . Naylor is a Planning Assistant in the Planning Department, City of Vancouver. [NCPC] National Capital Planning Commission. 1961. A policies plan for the year 2000. Washington: National Capital Regional Planning Council . [NSOED] 1993. The new shorter Oxford english dictionary on historical principles. Ed. Lesley Brown. Oxford: Clarendon Press. Oke, T im and John Hay. 1994. The climate of Vancouver. B.C. Geographical Series #50. Vancouver: University of British Columbia Department of Geography. Perin, Constance. 1967. A Noiseless secession from the comprehensive plan. Journal of the American Institute of Planners (September):336- 346. Rees, W i l l i a m E. 1994. Sustainability, growth and employment: Toward an ecologically stable, economically secure, and socially satisfying future. Vancouver: University of British Columbia School of Community and Regional Planning. Rees, W i l l i a m E. and Mathis Wackernagel. 1995. Our ecological footprint: Reducing human impact on the earth. Gabriola, B.C.: New Society Publishers. Read, Brian. 1970. Healthy cities: A study of urban hygiene. Glasgow: Blackie. Sandercock, Leonie and Ann Forsyth. 1992. A gender agenda: New directions for planning theory. Journal of the American Planning Association 58(1):49~59. Seelig, Michae l . 1995. Citizen participation as a political cop-out. Globe and Mail, February 27. Sennett, Richard. 1970. The uses of disorder: Personal identity and city life. New York: Knopf. Sitte, Camillo. [1889] 1965. City planning according to artistic principles. Trans. George R. Collins and Christiane Crasemann Collins. New York: Random House. Slocombe, Donald Scott. 1990. Complexity, change, and uncertainty in environmental planning: From the Great Lakes to the Kluane/Wrangells. Ph.D. diss., University of Waterloo [Ontario]. Smith, P. J. 1979. The principle of utility and the origins of planning legislation in Alberta, 1912-1975. In The Usable Urban Past. Ed. Alan Artibise. Toronto: Macmi l lan . Pp.196-225. Stewart, Ian. 1989. Does God play dice?: The mathematics of chaos. New York: B. Blackwell. Strogatz, Stephen H . 1995. Ordering chaos with disorder. Nature (November 30) 378(30):444. Vancouver Sun. 1994. Richmond, GVRD take step to resolving growth issue. January 27:A9 . 1995. Wet weather hampers peace march. Apr i l 29:A3. Van Pelt, Robert-Jan. 1994. A site in search of a mission. Anatomy of the Auschwitz death camp. Eds. Yisrael Gutman and Michael Berenbaum. Bloomington: Indiana University Press. Pp.93-156. Weber, M a x . 1966. The theory of social and economic organization. New York: Free Press. Westfall, Richard S. 1968. Newton and Order. In The concept of order. Ed. Paul G. Kuntz. Seattle: University of Washington Press. Pp.77-88. 158 White, R. and G. Engelen. 1993. Cellular automata and fractal urban form: A cellular modelling approach to the evolution of urban land-use patterns. Environment and Planning A 25:1175-1199. Wiesman, B. 1977. The plan that shaped the town: Kitimat, British Columbia: Implications for resource frontier towns. Vancouver: UBC School of Community and Regional Planning. W o h l , Anthony S. 1977. The eternal slum: Housing and social policy in Victorian London. Montreal: McGil l -Queen 's University Press. Wong, D. W . S. and A. S. Fotheringham. 1990. Urban systems as examples of bounded chaos: Exploring the relationship between fractal dimension, rank-size, and rural-to-urban migration. Geografiska Annaler 72B:2-3, 89-99. Whitlock, Rob. 1997. Interview with author. Vancouver, Apr i l 7. M r . Whit lock is a Re-zoning Planner for the Planning Department, City of Vancouver. Wylson, Rae. 1997. Interview with author. Vancouver, January 21. M s . Wylson works in the Office of the City Clerk, City of Vancouver. Young, Ray. 1995. Lectures: The Legal Context of Planning. Vancouver, September-December 1995. M r . Young is a sessional instructor at the UBC School of Community and Regional Planning, and works as a lawyer in Vancouver. Zimmerman, Brenda. 1991. Strategy, chaos and equilibrium: A case study of Federal Metals, Inc. Ph.D. diss., York University (Ontario). 159 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087744/manifest

Comment

Related Items