Introduction: the problem

Traditional Urban Design 'theory' is anarchistic and insubstantial. This is a situation, which has been ongoing for the best part of 50 years, offering unprecedented opportunity for debate and resurrection. Urban Design is a discipline where almost without exception its major proponents have failed to engage with any substantial theory in the cognate disciplines of economics, social and political science, psychology, geography, or the humanities. We can push this idea even further and say that it has not even embraced what today would be recognized as significant sub-disciplines, such as urban geography, urban economics, urban sociology, or cultural studies, the latter only recently emerging as a major force in design theory. Urban design practice is perilously close to a social technology without the grounding in social theory that would allow critical self-reflection to flourish. Problematically, this effectively locates urban design as being several realms removed from any substantial theory at all. At its weakest it could be seen as merely an extension of the architectural imagination or the physical consequence of state planning policies. Both of these are somewhat nihilistic attitudes that fail to accept, as I hope to demonstrate, that urban design can stand on its own as a legitimate theatre of activity with an acceptable theoretical signature. Nonetheless, the enduring dependency on architecture and planning remains, largely due to the historical relation of urban design to these two professions. The first position one could take on this situation, as indeed many practitioners and academics do, is to accept that: Its power derives from the fact that, irrefutably, it is a deeply embedded social practice that societies have valued from time immemorial, and therein lies its value. As such it does not have to justify its existence through reference to a discrete set of home grown theory (Cuthbert, 2003, p. 10)

While at one level this is true, the question is not how the practice of urban design is to be understood, but how one is to determine its object; either real or theoretical, on which basis the discipline can establish its credibility. The next task is to generate some explanatory framework rooted in substantial theory that explains this object. In order to do this we must begin by defining what we mean by 'Urban Design', its relationship within a hierarchy of practices, from architecture through urban design into urban and regional planning, and the social function of each within a larger and more substantial social context. While the stress here is on a theory of urban design rather than theories for urban design, the reason is simply because the former has been ignored and the latter has suffered in consequence.

In pursuing this line of reasoning, significant propositions are suggested below which will probably expand rather than confine debate, and due to their nature will unlikely be answered. For example, is urban design an art or a science? If it is an art, this removes the discipline entirely from the realm of scientific enquiry. On the other hand, if it is a science, to which branch should it belong, and how does it fit within the philosophy, logic and rationality of scientific endeavour? Such questions have no simple answers, and the mystery that surrounds them will remain. My approach is therefore to offer some propositions about the current state of urban design knowledge, emerging from our best efforts over the last 50 years, with the sole purpose of increasing its integrity and legitimacy. I see this first and foremost requiring a commitment by those who teach and practice the subject, to recognize that it needs a foundation which is presently lacking. At the moment, urban design is largely fragmented in its practices, theories and methodologies. Following Alfred North Whitehead's dictum that it is more important that a proposition be interesting than that it be true (although a true proposition is more apt to be interesting than a false one) – I suggest five propositions below that form the substrate of this paper. I maintain that these are central both to the legitimation of urban design and to its relevance as a social practice.

Proposition 1:

Urban design is self-referential and is neither informed by, nor committed to, any external authority in intellectual terms.

As a result, the discipline is denied any autonomy. To quote Paul Feyerabend: How can we discover the kind of world we presuppose when proceeding as we do? The answer is clear: we cannot discover it from the inside. We need an external standard of criticism, we need a set of alternative assumptions, or as these assumptions will be quite general, constituting an entire alternative world, we need a dream world in order to discover the features of the real world we think we inhabit......the first step in our criticism of familiar concepts and procedures, the first step in our criticism of 'facts' must be an attempt to break the circle (Feyerabend, 1975, p. 22)

As both theory and praxis, urban design is caught in such a circle, and this paper constitutes an attempt to find an appropriate way out. As it stands, urban design has no external standard of criticism that complements or addresses its internal fractures and inconsistencies. It is a real scientific anomaly, a region of knowledge with an inside but no outside. Whatever legitimation it does have is stitched together by imagination in academic life and regulation in practice. This seems to me to be insufficient. Following from this,

Proposition 2:

Urban design must reorient itself to social science as its wellspring, specifically urban sociology, geography, and economics.

The simple rationale for this position is that the organization of cities reflects the organization of society. Social space and form is our fundamental object of concern. Abstract exercises in spatial form must weld to the demands of social process.

In relating urban design to the social sciences, it is clear that the precepts of scientific enquiry apply, and I am aware of the difficulties involved in this position. Later we will see how the attempt to create a specifically urban sociology during the 1980s was fraught with conflict, the normal situation when any serious scientific or philosophical enquiry is undertaken. At least the concept was debated for some 10 years, whereas in urban design the term remains meaningless since most urban designers have been insulated from social theory by their education. Ask most urban design practitioners what urban means and one falls immediately into a morass of subjectivity and confusion. Borrowing from urban sociology, specifically Manuel Castells, we can suggest:

Proposition 3:

To be scientific, a discipline must have either a real or a theoretical object of enquiry.

While this seems to be quite obvious, it is one problem seldom if ever debated within Schools of Architecture or Planning where most urban design programmes are located. Part of this is due to the fact that the intellectual environment within the social sciences and humanities does not exist in professional built environment faculties to anything like the same degree. Questions related to the third proposition above become irrelevant to urban design practice to the extent that ideologies are embedded in production, with whatever inherent logic or consequences they contain. In such a context this third proposition becomes an unnecessary indulgence. This definitely leads to a happier life. Nonetheless, the question of a real or theoretical object remains, and can be suggested as follows:

Proposition 4:

The theoretical object of urban design is civil society, and its real object is the public realm.

If indeed this proposition constitutes the very essence of the discipline, then any urban design education should revolve around these two principles. It should begin by answering such questions as – What are the fundamental principles governing civil society? How did the concept of civil society arise? What types of civil society can be identified, what particular social forms does civil society engender and how is it constructed? What are the specific relationships that allow civil society to produce spatial forms, specifically those of the public realm? How does civil society project meanings into space through urban form? How is globalization affecting the organization and design of the public realm and social life? – and other such questions. Finally, it is clear that if the above ideas are acceptable, then logically there are some fairly substantial implications for education and practice alike. It demands for example a complete reassessment of urban design traditions as historically relevant, having emerged from the material circumstances of the time, yet potentially irrelevant to the globalization of capital in the digital world of the third millennium. Once again I want to be clear that this does not mean, as it might sound, that we must throw out everything we have been taught and start again, or that we are faced with an either/or proposition – that the notion of a theory of urban design stands in opposition to a theory for urban design. Quite the opposite is implied. I see this not as an 'either or' proposition, but 'both and'. Emerging from this:

Proposition 5:

Our understanding of the production of design outcomes must change from a modernist, Beaux Arts obsession with form, the eureka principle, and the cult of master/disciple to one where the organic production of urban forms and spaces are homologous with the production of society.

Devolving from this, several points need to be clarified in relation to the above propositions and observations, since they have certain wide-ranging implications. Nor am I suggesting that there are no social or economic consequences to specific types of spatial design practice. Of course, it would be nice to re-orient some research in urban design to discovering exactly what these are. Re-orienting our analytic lens to spatial political economy would not only allow a significant assessment of projects before they are ordained, it would also permit a more user-centred approach via an in-depth appreciation of culture, as well as appropriate methods of predicting or assessing outcomes of large-scale urban redevelopment projects. Second, we must accept that the use of the term theory has a huge archaeology of levels from the astronomy of black holes to making the best cup of coffee. This is also true within urban design. I have suggested below some of the more powerful tools available to urban designers via theories for urban design, and the actors central to their use, for example, Hillier's analysis of crime levels in complex layouts, in deprived areas which exclude surveillance, Alexander's use of patterns as templates, etc (it is also instructive however that Hillier and Hanson (1984) refer to their 'space syntax' as a A Configurational Theory of Architecture not Urban Design).

While much urban design theory has considerable integrity, in other cases, it is clear that claims to 'theory' are merely descriptions of common urban features or processes. They are axiomatic and have no universal application, qualities which I feel pervades most of, for example, Kevin Lynch's work. Clearly there is a lot of theorizing that remains to be done at this level, and the very best of it makes some effort to engage with more substantial concepts in psychology, economics, and social theory. The important consideration here is that science demands a generally agreed rule system that rises above, for example, Alexander's pattern language yet relates to it. Both are necessary, both should exist, but both need significant integration if we are to succeed in specific and coordinated urban design knowledge.

In order to do this I will adopt the following process. To begin, I try to put to rest the endless problem of 'defining' urban design. This covers the entire range of so-called 'definitions' that have been ongoing for the best part of a century with no resolution. The suggestion is made that the task has been fruitless to date for several good reasons that are then explored in some depth. Such exploration leads directly to the inescapable question of appropriate theory, without which any definition is insubstantial and largely content free. It also gives a clue as to why prevailing definitions remain largely axiomatic and deprived of analytical rigour and interest. Discussion then ranges over pre-existing ideas of theory in urban design. I have selected the four most compelling efforts in two areas. First, that of attempts at synthesis, and second, individual positions which have in some manner attempted a unified theory of urban design. More recently, several new theoretical positions have emerged which challenge knowledge to date with new ideas and hypotheses. Following Karl Popper's dictum that science advances through disproofs or falsification, the creation of new theory demands that the best of the old falls under the microscope.

Taken together the status quo in urban design may be seriously challenged by a generational position called political economy. This discipline emerged in the work of Adam Ferguson, Adam Smith and David Hume, part of an intellectual ferment originating in Scotland some 300 years ago, which at least one author has sourced as 'the wellsprings of the modern mind' (Herman, 2002, p. v11). Some of this history is traced and connected to the idea of a specifically spatial political economy, a discipline which up until now has been largely confined to aspatial phenomena, as was most social science until relatively recently. Taking this concept from aspatial phenomena into a sociology of space has been a daunting and worthwhile task, despite immense intellectual conflict. I then argue that in order to legitimize urban design as a discipline, we must escape from the shackles of personality cults and ideologies into a shared theoretical base, with all its warts and flaws. In conclusion, I suggest that if urban design theory is to progress, it must be by substantial interaction between spatial political economy and urban design, since space cannot be separated from its social production in specific urban forms.

Urban design – definitions

As in all disciplines, the most difficult problems arise over the simple question of definition – what is being defined? – an institution, a process, a theoretical construct, or an object of investigation? At the most elementary level, we have to consider the very fact of naming the discipline, and there are three related phrases that have had varying levels of use in signifying the design of cities, namely Civic Design, Urban Design and City Design. Civic Design was the terminology used in regard to modernist architecture, where large urban projects were carried out for the state and were therefore 'civic' in a very direct sense. The term was predominantly used in Britain, prior to Liverpool University establishing a foundation course in 'Civic Design' in 1909. This constituted the beginning of urban design as an academic discipline, and the stimulus for all other programmes in the UK. By 1970 the concept had become outdated, since its close connotation to municipal government and functions such as 'Civic Centre', were seen to be too restrictive. Concurrently, Harvard University in the United States was the first to establish a programme in Urban Design in 1960, and the term was then imported into the UK. Hence the more general term Urban Design came into use. In some senses this neutralized the discipline by removing its inherent political significance – the new term did not have any necessary connotation to citizenship, the modern state and urban administration in general that were implied in the term 'civic'. In a major sense, the emergent discipline was mystified by apparently removing it from the social realm. One of the most applauded theorists in urban design, Kevin Lynch, had difficulty with both concepts. He preferred to use the term 'city design' which to him was a more encompassing term (Bannerjee and Southworth, 1990). He had this to say about city design: City or environmental design deals with the spatial and temporal pattern of human activity and its physical setting, and considers both its economic-social and psychological effects (of which latter the sensuous aspect is one part). The concepts and techniques for manipulating this complex pattern, are as yet half formed. The ambiguity of our graphic notation system, its lack of inclusiveness, is one symptom of this inadequacy. The goals for which this pattern is manipulated are not clearly stated, and their relation to pattern is imperfectly known. Our vocabulary of city form is impoverished: the need for innovative ideas is correspondingly strong. Yet it is clear that city form is a critical aspect of the human environment, and design it we must (Bannerjee and Southworth, 1990, p. 483)

While this statement was written in 1968, Lynch remained true to this basic ideology throughout his entire life, and frequently conflated the term urban design to city design. Lynch had a significant influence over several generations of architects, planners and others involved in the design of cities, and his concept of urban design was one synonymous with project design or the design aspect of urban planning. There are, of course, a whole series of flaws with this kind of analysis, predominantly the question of defining any significant process merely in terms of its scale, using concepts that now have little meaning such as city, and not having a concrete definition of what urban signifies, let alone when it is tied to the equally problematic concept of design. In addition, few people today would accord to Lynch's use of the term city design for other reasons. Lynch also correlated city design with environmental design, where the more commonly used phrase is now sustainable development, leading to additional confusion (see Cuthbert, 2006, Chapter 7).

Nonetheless, the term urban has had a much more significant intellectual history than the general descriptor city which has little residual value in communicating any meaningful functions, processes, forms or symbolic content. In contrast, the term urban, apart from the fact that it originates in the Latin word urbs meaning city, has had significant accrued meaning since Lewis Wirth first wrote his legendary paper Urbanism As A Way of Life in 1938. Interrogating the term 'Urban' also formed the basis for one of the most meaningful investigations into urban structure, that of Manuel Castell's now iconic book The Urban Question, first published in French in 1972. After its English debut in 1977, it set a debate in motion for the next 10 years over the idea of a conceptually valid urban sociology, one that still resonates today, although much of the territory has now been captured by urban geography. So I will continue to deploy the term urban since it remains a more relevant and conceptually challenging term than either civic or city when applied to design, one whose meaning will hopefully become clearer over the remainder of this monograph.

Therefore from its somewhat confused identity, the discipline of urban design is usually defined without reference to any meaningful theory – and by analogy in its relation to physical scale, to administrative matters, in regard to urban functions or to a skill-based process in support of urban planning. Definitions such as the following underwrite most of the literature in urban design, and one searches in vain for anything of real substance: 'Urban design is the art of three dimensional city design at a scale greater than that of a single building'; 'Urban Design links planning, architecture and landscape architecture together, to the extent that it fills whatever gaps may exist among them'; 'Urban design is that part of city planning that deals with aesthetics, and which determines the order and form of the city'; 'Urban design is the design of the general spatial arrangement of the activities and objects over an extended area, where the client is multiple, the program is indeterminate, control is partial and there is no state of completion'; 'Urban design is primarily concerned with the quality of the urban public realm, both social and physical, and the making of places that people enjoy and respect'; 'Urban design is the art of making or shaping townscapes', and so on (Cuthbert, 2003, p. 12)

In Popper's terms, these sorts of statements are structured for extremely low levels of refutability, the mark of insignificant theory. The literature on urban design is swamped by axioms such as these, which are obvious even to the uninitiated. And so the search for an appropriate definition as the foundation for urban design theory has been ongoing over the last 50 years, a task doomed to failure in the absence of a substantial theoretical foundation. In spite of endless attempts at defining the discipline over the intervening period, progress towards developing a satisfactory hypothesis, a set of guiding constructs or principles, or a reasoned manifesto of ideological practices has been absolutely glacial. Virtually all definitions begin and end in dogma, and the so-called 'crisis in urban design', like the endless 'crisis in urban planning' continues, and nobody asks the obvious question as to why the crisis is seemingly continuous and endless, and what its source might be. So urban design theory continues to be impelled by a dearth of critical and dialectical thinking, an emballage of anarchistic practices, an obsession with skill-based learning, and a continuing belief in physical determinism.

Here two papers stand out simply because of their titles; David Gosling's 1984 paper Definitions of Urban Design, and Alistair Rowley's paper of the same name exactly 10 years later (an altogether superior work can be found in Punter, 1996). In his paper, Gosling has adopted a wholly architectural perspective, as if only architects had any right to define the discipline. While it may seem unfair to criticize this paper, now 20 years old, it remains significant precisely because it embodies one of the most powerful and enduring ideologies still dominating the field of urban design. While beautifully written and executed, the work is nonetheless a self referential, ideologically biased and atheoretical rendition of the genre, alienating every major theorist concerned with urban development, structure, and form, to the bleechers. Similarly, potential 'models' of urban design, for example, defined through civil society or the public realm, as a socio-spatial matrix, as symbolic representation, etc, are wrapped and made accessible only in and through the work of architects and their critics.

So Gosling's paper presented a picture of urban design that could only be defined and understood in reference to the entire payroll of modern and postmodern architecture, beginning, as is the norm with Le Corbusier, and continuing through the work of Oscar Niemeyer, Rob and Leon Krier, Robert Venturi, Lucien Kroll, Bernard Tschumi, Emilio Ambasz, Miguel Angel Roca, Aldo Rossi, Matthias Ungers and a host of others. The paper closes with 'a new definition of urban design' which is not stated, but is inferred again with reference to architectural projects such as Cergy-Pontoise (Lucien Kroll), the Byker Wall in Newcastle (Ralph Erskine) Carlos Nelson Dos Santos (Bras de Pina Favella, Rio de Janiero), all of it wrapped into having something to do with fair and equitable access to well designed state-sponsored housing. While most of the projects discussed are undeniably brilliant in formal terms, their collective outcomes are not evaluated, nor does any meaningful definition of urban design emerge from any association.

Similar criticisms can be applied to Alistair Rowley's paper, which exhausts the superficial to such an extent that only substantial theory remains to be addressed elsewhere. This is in itself a singular service, because in the 20 pages of the paper, Rowley condenses virtually every kind of definition of mainstream Urban Design since Spriergen's seminal work in 1965, and there is nothing left to say. Most definitions are axiomatic to the point of humour. Even on the first page, and 20 years after the huge debates on the term urban raged within urban sociology, involving some of the best social theorists of our time, we are still presented with a definition of 'urban' as something (we know not what) in contrast to 'rural' development, a relationship demolished in the postscript to The Urban Question by Manuel Castells in 1977. Quoting Ruth Knack, we are informed that 'Trying to define urban design is like playing a frustrating version of the old parlour game, twenty questions' (Rowley, 1994, p. 181). In the section 'The substance of urban design' we are offered two insipid definitions that illustrate how notions of the subject have developed since the 1950s: 'The purpose of town design is to see that (the urban) composition not only functions properly, but is pleasing in appearance' and in contrast, 'Urban design is essentially about place-making where places are not just a specific place, but all the activities and events that make it possible' (Gibberd, 1953; Buchanan, 1988 in Rowley, 1994, p. 182)

Definitions of urban design in Rowley's paper are also seen to be dependent on the repetition of endless taxonomies of various kinds that exhaust the stock of available adjectives to describe urban form such as '87 planning considerations grouped under 14 second tier headings which in turn were grouped under six broadly based first tier considerations.......Cook has written about the four qualities that urban design, as process, seeks to achieve – visual, functional, environmental and urban experience' and the essence of cities – 'complexity, surprise, diversity, and activity and..........Kevin Lynch's five performance dimensions, habitability, sense, fit, access and control, Bentley's seven qualities – permeability, variety, legibility, robustness, visual appropriateness, richness and personalisation......the Prince of Wales ten principles, the place, hierarchy, scale, harmony, enclosure materials, decoration, art, signs, lights and community' (Rowley, 1994, p. 185)

In Definitions of Urban Design, Rowley concludes with 10 more characteristics (why not have 20), by which point it should be apparent to the intelligent lay reader that the discipline is in serious trouble. Once again another taxonomy replaces any critical thinking on the subject. The last of these notes that urban design education demands literacy in the social sciences, law, economics, public policy and business administration, none of which are deployed in the paper. The problem with all of these attempts to define urban design is that they are content-free, depthless and incapable of moving us forward, except perhaps into another set of the so-called basic values, functional qualities, descriptive properties, performance dimensions or other qualitative groupings, usually claimed to have universal significance which of course they don't. As a collectivity, the result is akin to running on the surface of a sphere, at some point and on a random basis you have to arrive back where you started. It is not that these observations are untrue or uninteresting, simply that they are trivially correct, that is, so devoid of content that it is almost impossible to devise any empirical test which would prove them false.

In this sense they are also immune to Popper's concept of refutability, which would accord them some kind of theoretical justification (see Popper, 1959, 1974). As Thomas Kuhn has suggested, this does not even consist in its proof but of its probability in the light of evidence that is available to us. Hence definitions such as those above contain no propositions of any real content. Not only are they wholly isolated from significant theory, they also block the possibility that any will ever emerge by perpetuating this kind of thinking endlessly into the future, a process which has now been going on at least since 1945, and is still alive and well in the annals of the urban design literature (Schurch, 1999). One quote from Donald Appleyard however does stand out, that boundary definition in the case of urban design is a negative activity. It is indeed more enriching 'to identify, clarify and debate the central beliefs and activities of a field than to hide behind a simplistic mask' (in Rowley, 1994, p. 181). What these are have not yet been stated with any real conviction. Given that the least meaningful definition of urban design used by professionals is that it occupies the space between architecture and planning, it is worth pausing for a moment to consider John Punter's insightful paper of 1996, Urban Design Theory in Planning Practice: The British Perspective which traces its origins from the Garden City Movement and Raymond Unwin's classic text Town Planning in Practice: an Introduction to the Art of Designing Cities and Suburbs (1909).

Before considering the implications of the above situation which are clearly substantial, I will briefly amplify on the reasons why I feel that urban design is in such a moribund position. Rowley's explanation of urban design vis-à-vis the British planning system concentrates on two traditions. First, that of the picturesque, which had close ties to landscape painting. This in turn was closely related to the practice of English landscape gardening and such artists as Uvedale Price, George Repton, and Capability Brown. Second, to the tradition that evolved from it called the Townscape Tradition, one that concentrated on the idea of serial vision as the organizing principle (Cullen, 1961, 1967 and Smith, 1974, 1976).

The definition that emerges from both of these is purely visual and experiential. There is a nostalgic fixation on appearances where the perfect model would seem to be the idealized English village located in a beautiful landscaped garden (of somewhat limited application in a globalized world). This vision dominated planner's definitions of urban design at least until 1973 when the Essex Design Guide synthesized prevailing attitudes into the first comprehensive attempt to embed urban design in the UK into planning practice. As with architectural definitions of urban design, the ideas that evolved from Raymond Unwin and Gordon Cullen also remain alive today. The principles of the British picturesque, visual, townscape tradition are enshrined in the highly publicized design interventions of His Royal Highness, the Prince of Wales, supported by various architects such as Francis Tibbalds, Quinlan Terry and Rob Krier (Hanson and Younes, 2001). This neo-traditional movement is now sweeping the globe in the form of the New Urbanism, a movement that seeks to construct the future based on nostalgic visions of the past.

Coming right up to the last 10 years, the fact that all of these definitions remain alive is well represented in the UK. In December 1999 the Department of the Environment, Transport and the Regions (DETR) commissioned two reports on urban design by Arup Economics and the University of Reading, respectively. These were combined in a final report in 2001. The completed document estimated that there were some 180,000 individuals in Britain who may be involved in some way in urban design activity. The fact that a large engineering practice was chosen to run the project says a lot in itself. They include among the professions, architects, planners, surveyors, civil engineers, and landscape architects, who together with the Urban Design Group and Civic Trust, have formed the Urban Design Alliance. The combined membership of these five professions is 216,000 persons with about 5000 applicants commencing first degree courses at the universities and colleges each year, potentially leading to a professional qualification in these disciplines (DETR, 2001, p. 1)

The final paper encapsulates all of the above descriptors and because of the arcane problems surrounding any reasonable definition, begins by saying that 'It is easier to say what urban design is not, – architecture, civil or highway engineering, landscape architecture, surveying, town planning, – than what it is: urban design is both more and less than any one of these long established professional activities'. This is a bit like saying that 'an elephant is definitely not a giraffe, a zebra, or a camel. It's much heavier than all of them but cannot run as fast'. Moving on into even deeper waters, many other definitions are given in the paper, for example, 'Urban Design is an interdisciplinary process and activity', 'Urban design should be taken to mean the relationship between different buildings; the relationship between buildings streets, squares, parks, waterways and other spaces which make up the public domain; the nature and quality of the public realm itself; the relationship of one part of a village, town or city, with other parts; and the patterns of movement and activity thereby established'; 'There are different styles or types of urban design. The variations between each type reflect a range of factors including the role of the urban designer in the urban design process; the objectives of those employing urban designers; the situation and issues that will be addressed; and the criteria used to measure success'. The paper concludes by paying a tribute to Jonathan Barnett's 'memorable description' of urban design as 'designing cities without designing buildings' (DETR, 2001, p. 1).

Forget any meaningful theory, in every case the above attempts to define urban design are seriously divorced from any theory at all. Variously, they constitute radically empiricist, functional, technocratic, historicist, or practice and skill-based definitions. Many are deprived of any meaning whatsoever. Most are tautological or axiomatic, where no learning is possible. What we can assume from this, given that many highly intelligent minds have had a go over decades, is that seeking a definition exemplified by the many attempts given above is not only counterproductive but impossible, in the absence of a theoretical framework that would make them 'real'. Like any other social process, urban design does not remain a static phenomenon, and the need to hit a moving target is also omnipresent. However, this is not all bad, and it is quite clear from the fairly extensive rubric above, that one can indeed get a glimpse of how urban design fits into the overall processes of urban development while simultaneously being deprived of any explanation at all as to how or why it occurs as it does.

With few exceptions, traditional definitions constitute descriptions of perceptible surface structures in administrative, professional, aesthetic or other arrangements, a bit like trying to define gravitation in terms of apples falling to the ground, by what colour the apples are, what type of apple, how they compare to apples falling off other trees etc, in the absence of a supporting hypothesis. Overall, the processes of science, critical thinking, and focused research are all held in suspense. Even the basic principle of formulating a significant hypothesis about the discipline, one to be tested as the foundation for theory, is ritually avoided. One may search in vain for any author who begins an article 'my hypothesis as to a satisfactory theory of urban design would be the following'. But as in all theory, we are not looking for some immutable or unchanging truth, just a satisficing summary of the object in question that can be debated and tested, so that another horizon in the development of knowledge can be established (Simon, 1969).

So the problem remains. In the absence of any rigorous theoretical framework that links urban design activity to the historical process, to social development and to other professions, the same basic positions and approaches outlined above will be recycled ad finitum. Conversely, explanation using the method described, demonstrating some theoretical coherence might begin by aligning urban design theory to the material production of the built environment as an ongoing historical continuum. On this basis the need to continually restate definitions would disappear, and the endless process of grinding out yet another taxonomy could terminate. But such an approach demands that we consider at least three basic ideas.

First, we should abandon any attempt to define urban design in any of the forms outlined above, and conflate the term urban design to the production and reproduction of urban form. In general, this is how I will use the term in this paper. The central reason for such a distinction is to make it impossible to consider design independent to other processes (as exemplified in Gosling, and Rowley, above). Overall, the arguably superficial nature of urban design theory stems from this one fact, that is, the separation of form from content.

Second, we need to set aside the professionalization of knowledge/process for the simple reason that professions are by definition, territorial. They capture regions of intellectual capital supported by legislation, membership, and the arcane languages of practice (Cuthbert, 2006, Chapter 10). Any emergent theory or even any satisfactory explanation of the production of built form would have to incorporate a critique of professions qua ideological structures. Critical self-reflection is not about to emerge from this source since professions are knowledge monopolies linked to big capital. Hence, developing a substantial theory of urban design, might be experienced by the architectural and planning professions as a form of self-immolation.

Third, to a large degree the same is true of the academy and tertiary education in general. The ideological nature of tertiary education is on the one hand heavily penetrated by professional intervention via accreditation processes. On the other hand, neo-corporate interests intrude in the general definition, production and ownership of intellectual property. In this sense, the boundaries within tertiary education are no better than those within the professions themselves. Indeed, it can be easily argued that the professions play an active part in maintaining territorial imperatives in tertiary education via accreditation (legitimation) processes.

For the moment the thorny problem of what constitutes science and what does not, and whether urban design knowledge can be defined scientifically will be set to one side. What is required is some significant framework or structures of knowledge that satisfactorily account for the production of specific spatial arrangements, which result in identifiable urban forms and systems of representation. Since these will ultimately be proven false in the necessary development of knowledge, science requires that such a formulation should be highly resistant to empirical refutation. Fortunately, much of the required infrastructure has already been addressed in other areas of knowledge as we shall see, but with few exceptions it remains invisible to urban design administrators, educators and professionals.

To paraphrase the words of Scott and Roweis in their seminal article Urban Planning in Theory and Practice – A Reappraisal (1977), 'we should attempt to discover the sociohistorical meaning, rationality and limits of urban design, rejecting definitions of the type proposed above. Such definitions are counter-productive in that they all configure urban design theory as essentially normative. Using an analogy to urban planning, what is implied is that to date, urban design has presented itself as that reality, not so much one that is false, but one which is trivially correct or otherwise structured for a low level of refutability. In contradiction, it is not an independent and autonomous urban design theory that produces the various facts of actual urban design; it is rather the realities of contemporary urbanization that give rise to urban design as a necessary social activity, and hence its explanation as a social fact'. The only person to my knowledge that has actually tried to do this is Manuels Castells, and his propositions below bear comparison with the plethora of attempts to define urban design given above.

In relation to spatial political economy, Castells frames the fundamental question 'on the basis of the fundamental concepts of historical materialism, how can we grasp the specificity of the forms of social space?' (Castells, 1977, p. 235). In concert with this question he also offers by far the most encompassing and theoretically rigorous definition of urban design to date, one which informed both The Design of Cities and The Form of Cities (Cuthbert, 2003, 2006). In contrast to every other attempt documented earlier in this paper, he avoids the pitfall of any definition that is not situated within a hierarchy of socio-spatial concepts that allow context, relativity, and process to be included: We define urban meaning as the structural performance assigned as a goal to cities in general (and to a particular city in the inter-urban division of labour) by the conflictive process between historical actors in a given society. We define urban functions as the articulated system of organizational means aimed at performing the goals assigned to each city by its historically defined urban meaning. We therefore define urban form as the symbolic expression of urban meaning, and of the historical superimposition of urban meanings (and their forms), always determined by a conflictive process between historical actors. We call urban social change the redefinition of urban meaning. We call urban planning the negotiated adaptation of urban functions to a shared urban meaning. We call Urban Design the symbolic attempt to express an accepted urban meaning in certain urban forms (Manuel Castells 1983, pp. 303–304)

Rather than resort to definitions of urban design such as those previously discussed above, where the various qualities, properties, dimensions, etc of cities are used to delineate urban design as praxis, Castells' great contribution was to define it theoretically as an embedded part of other urban functions and processes (notwithstanding the fact that what exactly constituted 'urban' remained unresolved, and the fact that the concept of 'urban meaning' as a workable principle remained somewhat opaque). Importantly, Castells also assigns the term 'meaning' (not economy as one might expect) – as the ultimate measure of the performance of cities, and associates such meaning with the outcome or representation of conflict. While Castells can be criticized for not making explicit what he meant by urban meaning, that is, turning it into a formula, the term is made in the context of his own massive output of scholarly works. I take the term meaning here at its broadest compass to signify the actual material expression of the history of capitalist development, writ large in the built form of cities using the medium of urban design, or more succinctly, the accrued history of its symbolic capital. This would include first and foremost in Castell's terms, the idea of class struggle, collective memory, and the expression of social distinctions. Inherent in this would be included war – oppression, liberation and reconstruction, representations in the realms of science, art and philosophy and other forms of semiosis. Physically we are dealing with the entire panoply of urban form, individual architectural elements, monuments, street sculpture, including spaces and places as well as their naming and associations. In this regard Allen Scott's classic paper of 2001 'Capitalism, cities and the production of symbolic forms' offers additional insights. Prime also is the concept of the urban landscape, the classic work in this last category being Dennis Cosgrove's Social Formation and Symbolic Landscape (1998). One may hazard that the urban designer's central activity is in the production and consumption of symbolic capital through the medium of fixed capital investment, reproduction and exchange (Harvey, 1989).

Concepts of symbolic capital and symbolic space and form, are closely related ideas (Bourdieu, 1977; Schusterman, 1999). Bourdieu advanced the idea of symbolic capital, which he describes more accurately as the symbolic effects of capital at great length in Outline of a Theory of Practice and the term remains in widespread use today. A crude way to look at symbolic capital would be say that it is the value of the added symbolic expression and associations of a product, in contrast to its material cost. In many cases this may outweigh its cash value which can become irrelevant over time. For example, the symbolic capital of the great pyramid of Cheops in Egypt is priceless but its material value is zero. Bourdieu also introduced the concept of taste, exhibited through the ownership of symbolic capital, Harvey points out that symbolic capital is both fetishistic and ideological since it conceals 'the real basis of economic distinctions'. Furthermore, Conjoining the idea of symbolic capital with the search to market Krier's symbolic richness has much to tell us, therefore, about such phenomena as gentrification, the production of community (real, imagined or simply packaged for sale by producers), the rehabilitation of urban landscapes and the recuperation of history (again real imagined or simply reproduced as pastiche). It also helps us to comprehend the present fascination with embellishment, ornamentation and decoration as so many codes and symbols of social distinction (Harvey, 1989, pp. 80–82)

So concepts of urban meaning, symbolic capital, ideology, social reproduction and the actual design of the built environment are all integrated and lie at the heart of urban design practice. By relating function, form and meaning, Castells provides us with a complex of relationships where we can clearly see the interaction of associated elements in the urban process, rather than as the fixed properties of physical form. Whereas most other definitions of urban design discussed above are static observations that go nowhere or are based on happy certainties that cannot be refuted, the complexity of Castells' insight automatically leads to other hypotheses or speculations rooted in substantial theory – 'What is urban meaning and how is it derived?', 'In what manner does the redefinition of urban meaning result in changes to the built environment?'; 'How does the symbolic expression of urban form relate to class and other societal conflicts?' Through which processes does such urban meaning materialize, using what kinds of content? Castells encompassing hypothesis therefore stands out as a singularly insightful attempt to connect the process of designing cities to the overall process of the production of space within capitalism, or to paraphrase Scott and Roweis once again: We cannot assume that Urban Design emerges, acquires its observable qualities, and evolves, according to forces that reside solely within itself. Urban Design is not invented in a vacuum, but is structurally produced out of the basic contradiction between capitalist social and property relations (and their specifically urban manifestations) and the concomitant necessity for collective action (Castells, 1977, p. 1011)

In The Urban Question, Castells had already offered a rare analysis of urban spatial forms as products of basic economic processes – production, consumption, exchange and administration, and the reflection of ideological structures in symbolic configurations, elements and places. Castells' typology of urban space is extremely well structured and is solidly grounded in an extensive theoretical foundation. There is no doubt that it represents a qualitative advance on its predecessors by several levels of magnitude. This does not imply that it is ideal or cannot be improved, or that all of the epistemological problems have been solved. Nonetheless, it was the first serious attempt to incorporate urban design and urban planning within a unified and coherent definition, linked to a significant body of social theory based on the morphing of political economy into the dimension of urban space and form.

Before moving these ideas forward, we will now take a greater in-depth look at some of the more prominent mainstream 'theories' in urban design in order to maintain a progressive historical perspective on the discipline. A more substantial approach to theory will then be engaged, following Scott and Roweis lead into recent developments in spatial political economy.

Urban design – Theory?

It is not my intention here to write a normative history of urban design (one that is yet to be assembled), but to illustrate some of the more influential and prototypical discourses that traditional theory still clings to for support. All of these texts are classics in their own right, and constitute significant markers in the journey towards an improved understanding of urban design. Historically, each represented a significant attempt to correct what was seen to be a dominant problem at the time they were written. Much of what is contained between their covers will remain valid for years to come, for the simple reason that even the simple principles embodied in, for example, Gordon Cullen's Townscape (1961) remain widely ignored 40 years after the book was written. As we approached the end of the second millennium however, three things became very clear.

The first was that the ideologies represented in the collective corpus of work traditionally associated with modernist urban design had lost much of their explanatory power. Jon Lang's classic text Urban Design, The American Experience (1994), marked in a very real sense, the last significant breath of the modernist position.

Second, since that time, that is roughly over the last 10 years, a new era in urban design theory has surfaced, although this remains to be articulated in any empowering manner. Nan Ellin's book Postmodern Urbanism (1996) and Ross King's Emancipating Space (1996) represent two memorable texts written in the intervening period, the latter the only one that has a dialectical relationship to theory.

Third, the upsurge in things urban in disciplines that had previously been wholly disconnected to the design of cities began to produce a significant corpus of work. Urban sociology, economics and geography, cultural studies, art history, landscape architecture and other disciplines from anthropology to philosophy were all involved. Overall Urban Sociology and Human Geography have been the two key players since the early 1980s. Leslie Sklair's recent articles The Transnationalist Capitalist Class and Contemporary Architecture in Globalising Cities and Iconic Architecture and Capitalist Globalisation (Sklair, 2005, 2006) demonstrate the extent to which significant explanations of urban design in the information age have been abstracted out of the hands of urban designers into urban economics and urban sociology. While the design process is not discussed in any of this work, what constituted the real content of the urban (design) process was being articulated in all kinds of ways and from many different sources. This progression has resulted in the undeniable observation that more significant theoretical paradigms about the shape and form of urban space were originating from outside the discipline of urban design than from the inside, and not before time either. It also offers a partial explanation as to why so few key texts on urban design have emerged over the last 20 years.

In Designing Cities (Cuthbert, 2003), I therefore made a clear distinction between what I consider to be normative theory in urban design over the 30 years period from 1960 to around 1990 or so, which I will summarize below, and other more significant theory of urban design and urban form that addresses urban spatial theory, has been elaborated in significant detail in The Form of Cities (Cuthbert, 2006). In order to contextualize the knowledge represented in mainstream urban design theory, I also suggested an elementary taxonomy of 40 scholars whose work had significant influence over mainstream urban design which I have slightly modified and upgraded to 40 on the basis of feedback from colleagues over the last few years (see Table 1).

Table 1 - Forty classic texts in urban design.

Full table

In addressing mainstream theory, we must briefly look at the functional relationships between architecture, urban design and urban planning that configure much traditional thinking about the significance of each. To facilitate this comparison, a systems view of the three related disciplines, couched in terms of Herbert Simon's irreducible elements of systems referred to in his book The Sciences of the Artificial (1969) is given in Table 2. Like all attempts to create a simple taxonomy or table of relationships one has to resort to some fairly pragmatic statements. Nonetheless, clarifying the purpose of each in relationship to its social function allows a transparency of meaning denied to definitions that conflate each to the other. On this basis it is fairly clear that some significant distinctions between the three disciplines can be made even at this fairly rudimentary level of analysis.

Table 2 - A systems view of professional boundaries.

Full table

Within this arrangement, Architecture is limited to the design of individual buildings, which are conceived primarily in terms of the design parameters of artificially controlled environments. Importantly, the term 'artificial' used in this context does not connote false, but man-made. Despite the endless claims by architects that their buildings integrate with nature and interact with their environment, and whatever aesthetic is adopted, buildings, for all practical purposes are closed boxes. The essential function of architectural components is private and defensive, predominantly from the weather and from other people; hence buildings generally operate as closed systems with human, electronic or physical means of surveillance used to mediate external relations. Urban design on the other hand is represented as an open system that uses individual architectural elements and ambient space as its basic vocabulary. Whereas architecture is predominantly concerned with social closure and protection, urban design is by its very nature, focused on social interaction and communication in the public realm. Urban planning may then be conceived as something fundamentally different again, as the agent of the state in controlling the production of land for the purposes of capital accumulation and social reproduction, in allocating sites for the collective consumption of social goods such as hospitals, schools, religious buildings, and in providing space for the production, circulation and eventual consumption of commodities.

While Architecture represents the locus of exchange values locked into fixed capital assets, Urban Design qua the public realm appears as use values representing the space of civil society. Urban Planning is unique in the sense that it exists only in the degree to which it is represented as an ideological structure focused on two things, social control and profit. It relies solely on the state for its existence and is legitimated through a complex system of planning law that consolidates capitalist social relations in regard to space. As symbolic art forms and containers of history, none of this infers that architecture and urban design do not rise above these functional criteria, nor that urban planning cannot focus on the greater social good. Clearly it has played a significant role in improving the general well being of labour while simultaneously guaranteeing superprofits from land development in return for minimal regulation. There are however significant philosophical, technical and functional differences among and between them, and neither architecture nor urban planning can make facile assumptions about their own ability to expedite urban design strategies as praxis. Within this general paradigm, we can make three kinds of distinction about urban design theory for the purposes of the ensuing discussion.

First, there have been several courageous attempts to synthesize the entire field of urban design: the most notable epitomized in the following four examples:

• Rowe and Koetter's Collage City (1979)
• Gosling and Maitland's Concepts of Urban Design (1984)
• Roger Trancik's Finding Lost Space-Theories of Urban Design (1986)
• Geoffrey Broadbent's Emerging Concepts in Urban Space Design (1990).

Second, there are claims to primacy, by which I mean some claim to a theory of urban design by individuals

• Kevin Lynch's A Theory of Good City Form, (1981)
• Rob Krier's Urban Space (1979)
• Bill Hillier and Julienne Hanson's The Social Logic of Space (1984)
• Christopher Alexander's A New Theory of Urban Design (1987).

Third, there has been a significant movement in approaches to urban design over the last 10 years. Overall there have been two major trajectories, both practice led. The first is the overall trend towards sustainability, largely considered the arena of architecture until relatively recently. As sustainability is by now generic to most professional environmental disciplines, its implications are not directly related to any specific theory of urban form, its principles being incorporated as praxis into engineering, architecture, planning, building, landscape and related disciplines. Second, the movement called The New Urbanism is a fast maturing ideology with wide-ranging implications with specific theoretical implications for urban design. As a movement, it may be argued that the assumptions and practices of the New Urbanism have the potential to overshadow most of the paradigms under consideration (Clarke, 2006).

Attempts at synthesis

The four major attempts at synthesis tried in various ways to generate typologies, even if the idea of collage does not readily fit the category. Taken together they offer an interesting summary of theory in urban design since the Second World War. Each book is beautifully produced with a multitude of illustrations, collectively encompassing most of the significant major urban projects up until 1990. It is also telling however, that without exception, the authors are architects. Each is rampantly modernist in his approach. Unsurprisingly architectural ideologies dominate. Because of this there is endless referencing of the rationalist and contextualist schools of thought (as well as neo-rationalist and neo-contextualist) – and a continual recycling of the work of major architects such as Walter Gropius, Le Corbusier, Frank Lloyd Wright, Aldo Rossi, Rob Krier and others. In two cases (Rowe and Koetter, 1979 and Broadbent, 1990) there is at least the acknowledgement of some theory that exists outside the aesthetic and formal aspects of architectural design.

Collage City by Rowe and Koetter (1979) is an incredibly erudite piece of prose that derives much of its creative impulse from associations with literature, poetry, painting, and philosophy, as well as architecture. While Collage City is on the one hand, totally committed to architecture as an original reference point, there is also an ingrained pessimism and a deeply held sense of failure about its accomplishments, an observation which can nonetheless be welcomed as a positive contribution. This paradox of celebration and failure that exists throughout the work is clearly expressed at the outset: The city of modern architecture (it may also be called the modern city) has not yet been built. In spite of all the goodwill and the good intentions of its protagonists, it has remained either a project or an abortion; and, more and more, there no longer appears to be any convincing reason to suppose that matters will ever be otherwise. For the constellation of attitudes and emotions which are gathered together under the general notion of modern architecture and which then overflow, in one form or another, into the inseparably related field of planning, begin – in the end – to seem altogether too contradictory, too confused and feebly unsophisticated to allow for any but the most minor productive results (Rowe and Koetter, 1979)

Collage City is fundamentally a critique of urbanism defined as the relationship between architecture and planning. But the sense of failure that permeates the text is both unnecessary and misplaced. The reason for this is manifest – that architects are not responsible for the failure of the modern city (if indeed it is failing) and any perceived failure emerges from the modernist arrogance that places architects at the centre of the world. Clearly architects are not responsible for the production of urban space, which is essentially an economic and political process. Even their overall role in manufacturing built form is minimal, given that architects are only involved in some 15% of all construction processes. The idea that cities are composed of fragments as in the French concept of bricolage has exalted claims – 'together these fragments create a conceptual framework for our experience of the city as a form of phenomenological cubist collage, in our individual and collective consciousness. Each fragment was a lived world, the product of history, which could be inhabited. The differences between fragments and their inhabitants made for the diversity and vibrancy of city life' (Shane, 2000). Whether one could ever examine a phenomenological cubist collage sufficiently to render even a bit of it operational is open to question. In addition, Collage City presents no manifestos, theories, methods or taxonomies that would actually assist in building some encompassing vision of architecture or urban design, nor is there any material presented that might support the concepts. Its central thesis that the building of cities occurs in fragments that assemble into a collage of parts, each with its own identity, flies in the face of most urban geography as to how cities grow and change. Also the book has no outcome, even the title Collage City and what to do with it remains a mystery. Ultimately the work is both a eulogy on modern architecture and a sermon on its demise, conflating in the process, architecture with urban design.

Given the free-floating nature of Collage City, Gosling and Maitland's Concepts of Urban Design (1984) is much more highly structured. In contrast to Gosling's article discussed above, the book has much greater intellectual scope and depth. Nonetheless, it maintains the same basic relationship to modern architecture as the source of all inspiration for urban design, and in the very first chapter there is an immediate reference to urban design being a question of physical scale and of the imperialist assumptions of architecture beyond that of the individual building We have suggested that urban design is concerned with the physical form of the public realm over a limited physical area of the city, and that it therefore lies between two well established design scales of architecture, which is concerned with the physical form of the private realm of the individual building, and town and regional planning, which is concerned with the organisation of the public realm in its wider context (Gosling and Maitland, 1984, p. 9)

The book proposes a tripartite division of theory in chapter two into:

• Natural Models
• Utopian models
• Models from the Arts and Sciences.

By natural models the authors basically mean feudal or pre-feudal settlement forms that have grown organically on the basis of location, geography, defence, religion, ownership, natural disasters, wars, etc, either in toto or in various combinations. Utopian models translate into several different forms, but in essence refer to places that can exist only in the imagination: So also Campanella, Bacon, Fourier, Le Corbusier, Wright and Howard all devised their utopian models during periods of either enforced or voluntary obscurity and isolation .........we could add the profound philosophical objections of Karl Popper to utopianism, namely that it is historicist in nature, unscientific, oppressive, unable to learn from its mistakes, and based upon a number of doubtful propositions, such as 'the colossal assumption that we need not question the fundamental benevolence of the planning Utopian engineer' (Gosling and Maitland, 1984, p. 32)

Many of the utopias discussed are also included in an entire book on the subject by C.A. Doxiadis (1966), who felt compelled to add a few more topias (places) of his own (dystopia, eftopia, entopia).

The third category of models from the arts and sciences is really a catch all for everything left over from the other two categories. The authors suggest that two predominant processes are at work, namely analogy and translation. Hence we find a disparate collection of influences from architecture, mathematics, semiotics, psychology, anthropology, ecology and other disciplines. These are rapidly set aside in favour of a retreat to the Bauhaus at Dessau – (Siegfried Giedion, Paul Klee, Gyorgy Kepes, Laslo Moholy-Nagy and others) – as the initiator of the whole trend towards serial vision. This was to become one of the most dominant forces in picturesque urban design exemplified in the work of, for example, Philip Thiel, Donald Appleyard, Kevin Lynch, and Gordon Cullen. Later in chapter four, Gosling and Maitland give eight potential directions for the future (which is now with us twenty years later) as follows:

• Urban Design as Political Statement
• Urban Design as Technique
• Urban Design as Mediation
• Urban Design as Private Display
• Urban Design as Public Presence
• Urban Design as Theatre
• Urban design as the Guardianship of Urban Standards
• Urban Design as Collage.

Despite the popularity of Concepts in Urban Design, based upon a singular contribution to the field, the analytical position sketched out in the appendix is somewhat more enlightening (Javier Cenicacelaya Marijuan, 1978). Marijuan suggests a series of approaches to urban design consolidated into three broad groups:

• Dependent on a particular politico-cultural system
• Marxist
• Utopian
• Capitalist

• Related to a variable politico-cultural system
• Mathematical, economic
• Descriptive, functionalist
• Morphological
• Historical

• Not dependent on any politico-cultural system
• Perceptual

While Concepts of Urban Design constituted an heroic attempt to consolidate the diversity of urban design practices, it remained seriously confused as to how the design of urban form fitted into any larger paradigm in economics, politics, social science, or even a reasonably coherent social theory of architecture. Many of the examples given in the eight potential directions for the future are themselves utopian, but many of the projects illustrated, both fantasy and reality, are nonetheless brilliantly conceived and ingenious in their synthesis of complex physical design problems. The central hiatus is that no matter how ingenious the examples, they do not add up to a satisfactory explanation of how urban form comes about. Despite the avalanche of creativity displayed in the chosen examples, the work as a whole remains deeply empiricist and historicist in its message. This is most clearly demonstrated in the relationship between the proposed Theories of Urban Design in chapter one, and the Directions for the Future in chapter four, where the designated theories bear no relation to where urban design may be heading. Somehow the future became totally unplugged from development as a continuum, and the seeming detachment of the production of urban form from the reproduction of society remained intact.

Roger Trancik's Finding Lost Space (1986) comes two years after Concepts of Urban Design (1984). He begins with the five causes that he says have contributed to the lost space of our cities, spaces that have basically disappeared from the drawing board because of urban dereliction in some form or another. He denotes these as our failure to deal with the automobile; the Modern movement in design; zoning and urban renewal practices of urban planning; the denial of responsibility for the public realm, and the problem of abandoned land in and adjacent to central business districts in American cities. While this is no doubt true, the arbitrary collection of problems, no matter how accurate, does not constitute theory in any meaningful sense. Despite this, Trancik's book makes an invaluable contribution to the process of documenting in some coherent manner, the physical manifestation of space in the 20th century and the designers concern with what he refers to as Three Theories of Urban Spatial Design. He denotes the three theories as:

• Figure Ground Theory
  • Grid
  • Radial concentric
  • Angular
  • Axial
  • Curvilinear
  • Organic
• Linkage Theory
  • Group form
  • Megaform
  • Compositional form
• Place Theory

Figure ground theory is basically a two-dimensional concept allied to the idea of Gestahlt where black and white, positive and negative, solid and void or in Taoist terminology, Yin and Yang, are formed by opposites that define the other. The Gestalt concept has had many applications, and is probably best known in relation to the psychotherapeutic method developed by Fritz Perls and his wife Laura in the 1940s, and the Gestalt theoretical psychology of Hans-Juergen Walter. In urban design the figure–ground concept constitutes the fabric of the city where a harmonious ground plan may be arrived at through a balancing of spatial relationships and contexts in order to generate requisite variety within a larger whole. The classic application of this idea (and also the epitome of contextualist principles) was Gian Battista Nolli's 1748 map of Rome. Rowe and Koetter resort to the use of this idea throughout Collage City (see, eg, pp.74, 75, 82, 131, 168–171), and a special issue of the Architectural Design was devoted to Nolli's plan in 1979.

So architects have used the concept of Gestahlt for centuries as a method of seeing their designs more clearly. In essence, what it does is reverse the impact of the visual image which usually delineates buildings over spaces, akin to producing a negative image of this page, whereby the letters appear as white shapes, but in fact are defined by the black background. In the figure–ground relationship, the spaces tend to predominate, thus heightening the impact of the public realm. Trancik defines this situation well in regard to Nolli where he says 'In Nolli's map the outdoor civic space is a positive void and is more figural than the solids that define it. Space is conceived as a positive entity in an integrated relationship with surrounding solids. This is the opposite of the modern concept of space where the buildings are figural, freestanding objects, and space is an uncontained void. In Nolli the void is figural' (Trancik, 1986, p. 98). He then enunciates six typological patterns of solids and voids (grid, angular, curvilinear, radial concentric, axial and organic as a method whereby some more analytical rigour can be obtained. Much of this is a restatement of Camillo Sitte's writing of 1889 (see also Collins and Collins, 1986).

The second element in his equation is what he terms linkage theory or: the organization of lines that connect the parts of the city and the design of a spatial datum from these lines that relate buildings to spaces. The concept of datum in spatial design is analogous to the staff in music, upon which notes are composed in an infinite number of ways. The musical staff is a constant datum, providing the composer with continuous lines of reference. In urban spatial design, the determinant lines of force on a site provide a similar kind of datum from which a design is created (Trancik, 1986, p. 106)

The subgroups identified here are what are termed Groupform – arising from an incremental and historical accumulation of urban fabric, usually organically structured on the basis of roads, pathways or open space armatures. Megaform is the artificial version of groupform, where hierarchic linear frameworks are consciously designed and imposed on the landscape. The classic version of this design idiom is Candilis, Josic and Woods famous project for Toulouse-le-Mirail in France, or Kenzo Tange's unrealized project for Tokyo Bay. Compositional Form is derived from functionalist planning methods such as those advanced by Ludwig Hilberseimer, Le Corbusier and other modernist architect-planners, whereby the actual two dimensional graphic organization of space is implied rather than overt, in other words it plays a secondary role to the architects' chosen building blocks.

The third category is what Trancik calls Place Theory (undefined, not to be confused with Walter Christaller) where he proposes a theory of place derived from its cultural or geographic context; 'For designers to create truly unique contextual places, they must more than superficially explore the local history, the feelings and needs of the populace, the traditions of craftsmanship and indigenous materials and the political and economic realities of the community' (Trancik, 1986, p. 114). There are no subgroups in this section and meanings must be inferred in reference to an eclectic grouping of various theorists and practicing architects, for example, the work of Alison and Peter Smithson, Herman Herzberger, Kevin Lynch, Leon Krier, Hans Hollein, Lucien Kroll, and Donald Appleyard. The book concludes with four urban case studies that serve as laboratories for the spatial design theories enunciated above (Boston, Washington D.C., Gothenburg, Sweden, and Byker in Newcastle, England).

Overall Trancik's book is interesting and useful with a clear exposition of the ideas through which he configures his approach to urban design. However, it sustains the normal practice of recycling the concepts and practices of Modernism, all of which embraced the philosophy of physical determinism while paying lip service now and then to 'social and other factors'. Despite Trancik's promotion of three theories of Urban Design in figure ground, linkage, and space theory, each fails to rise above a fairly rudimentary attempt to justify urban design theory solely in terms of its spatial arrangements. Nor do they constitute theory in any meaningful sense, being largely detached from any larger picture of society, explanations as to how such forms come about, or indeed to any substantial theory within the related disciplines of architecture or urban planning. Figure–ground is not so much a theory as an interesting graphic volte-face which allows a more accessible analysis of the public realm in terms of its spatial organization. There are no assumptions of any significance that can be made from figure–ground relationships apart from throwing geometry, form, and structure into higher relief, none of which means anything without some supporting concepts.

Questions also need to be asked about how particular forms of space are more functionally suited to specific uses, what psychological or physiological assumptions can be formed on the basis of specific two-dimensional shapes on a plan, or how social meanings might be transmitted and recognized on the basis of Gestalt relationships. The same is true of linkage theory and hierarchic structures, which demand similar levels of explanation as to efficiency in use and flexibility in organization. Taken together, the first two theories, figure ground and linkage, are useful devices used by designers that have some interesting historical and contemporary referents. But in no sense do they constitute 'theory' in any meaningful way. The last category place theory, is almost wholly detached from the first two, and tentatively suggests that 'cultural and human characteristics' might have some relevance, without actually defining what these might be. Modernist architects apart, a few sentences are devoted to Phenomenology (Martin Heidegger and Christian Norberg-Schulz), followed by references to Ian McHarg, Kevin Lynch, Donald Appleyard and Gordon Cullen. While the first two theories stand as good expositions of design techniques, there is no coherence to 'place theory', despite the fact that quite a lot could have been deduced even at that time, from the work of Amos Rapoport (1969, 1977, 1981), Jane Jacobs (1961), Oscar Newman (1971, 1973, 1976, 1980), Herbert Gans (1962), E.T. Hall (1959, 1969, 1976) and others.

Geoffrey Broadbent's book Emerging Concepts in Urban Space Design (1990) follows the general pattern established above, allocating primacy to architectural rather than social or other concepts. So the text is a fairly straightforward rendition of ideas, which had already been well expressed, for example in the three previous syntheses discussed above, and in significantly greater detail by individual scholars who have specialized knowledge of the selected material. In a very real sense, this was probably the last book on urban design that could perform this task, for two main reasons. First, because Broadbent's book is the best of its kind and does not leave much left to be said within its chosen paradigm. Second, because the entire deterministic construct of modernist urbanism has finally been exhausted of its ideological and aesthetic content. The great planning disasters of modernist architecture recounted in P. Hall (1982) and Dunleavy (1981) which came to a catastrophic and symbolic finale in the dynamiting of the Pruitt-Igoe housing estate in St Louis Missouri, have slowly given rise to a new consciousness. In essence, the architectural imagination has had to realize its significant limitations, which were impacted by new knowledge emerging from other disciplines such as environmental psychology, urban geography and urban sociology. Beyond the design of individual buildings, it started to dawn on many architects that a wholly different kind of knowledge is required to understand the modern city and the production of urban form, and that the vocabulary of architecture was seriously lacking in this respect.

Urban Space Design is therefore symbolic of the end of an era, and it is unlikely that another book of this kind would require to be written. It reflects the functionalist philosophy of modernism, ritually emptying the production of urban form of its economic, social, political and symbolic content. This is also reflected in the curious fact that 18 years after the emergence of postmodernism, there is no mention of it in the index to Broadbent's book, although the distinction is made between rationalists and empiricists, and neo-rationalists and neo-empiricists. Despite this criticism, there was a kernel of awareness that something important was being ignored. Part one is a fairly straightforward enunciation of the historical evolution of urban form emphasizing the 20th century city. Part three is almost completely dedicated to an exhaustive presentation of architectural projects up until the time the book was written. While both parts recycle the same historical and contemporary examples, at the conclusion of part two, which carries over eight pages into chapter three, there is a brief and belated recognition that designers might just require some additional theoretical assistance.

Under 'Urban Realities' Broadbent discusses the work of a few scholars not directly connected to modernist orthodoxy – Jane Jacobs, Christopher Alexander, Nicholas Taylor, March and Trace, Peter Cowan, Oscar Newman and Alice Coleman. This collection is a pot-pourri of the scholarship of the time, a mixture of architects and non-architects that were concerned variously with social cohesion, system theory, mathematical modelling, environmental psychology, social housing and other factors. Their work was not connected in any meaningful way to mainstream architectural design, but overall it reflected a concern with the production of humane environments on the basis of at least some scientific knowledge grounded in substantial theory. Similarly, section three begins with a discussion of ideologies. Broadbent offers an extremely brief summary of the work of Marx and Engels, Charles Baudelaire, and George Simmel, whom he considers were instrumental to the intellectual foundation of La Tendenza, probably one of the most theoretically informed of all movements in modern architecture. Its name derived from its tendency towards an emerging neo-rationalism, later manifested most powerfully in the work of Aldo Rossi, although the intellectual powerhouse behind the movement was Manfredo Tafuri, a historian who became Chairman of the Institute of History at the School of Architecture in Venice in 1968. Broadbent is dismissive of what he calls 'Lamborghini Marxists' and 'bourgeois Marxism', due to a demonstrable ignorance of the development of Marxist thought since 1887.

As this is not the place to debate his position in this respect, reference can be made to a more sympathetic critique of La Tendenza by Massimo Scolari (in Hays, 2000, p. 124). From La Tendenza, Broadbent then progresses to a detailed study of Aldo Rossi and Carlo Aymonimo whose work is representative of that school of thought. Apart from these brief excursions into paradigms not directly related to architecture, Urban Space Design remains firmly rooted to an architectural definition of urban space and form, one that is largely impervious to any external influences. Its basic philosophy is encapsulated in an insightful quotation from Massimo Scolari about modern architecture in general and the debates held within Tendenza in particular What clearly emerges from this is first of all, an overall critical vision. One realizes that in the university architecture departments – because of, on the one hand, the objective marginality of their institutional and economic role and the subsequent lack of development of an explicit demand for research aimed in this direction, and on the other, the cultural backwardness of professional arrangements within the discipline – no comprehensively organized and systematic work of research has ever been developed that might fit into the whole as a way of advancing the dispositions of the discipline (Scolari in Hays, 2000, p. 135)

This statement could equally apply to urban design today. Overall, what emerges from an assessment of the four major attempts at synthesizing urban design theory up until 1990, either wholly or partially, is a stunning insistence that architecture as art, technology or science represent the only potential pathways. While each author attempts to broaden his position in regard to the problem with occasional reference to a favourite philosopher, to associated disciplines such as landscape architecture or urban planning, or to some enlightened critics such as Jane Jacobs or Alice Coleman, there is an unshakeable commitment to modern architecture and its adopted ideologies. It is also notable that every work was written after postmodernism had been officially christened, and that there is a singular refusal to recognize that epochal changes in the world economy were affecting society and space in ways wholly unrecognized by any synthesis. In addition, research in other areas such as proxemics, environmental psychology, architectural anthropology, and social theory, is nowhere mentioned. Here we would have to include many authors who had a singular focus on the organization and design of urban space such as Sommer (1969), Hall (1969, 1976), Proshansky et al (1970), Perrin (1970), Rapoport (1981), Jameson (1988).

Not only this, but the revolution that was taking place in urban social theory throughout the 1980s that we will discuss later, goes wholly unrecognized. The problem here however, is the danger of throwing the baby out with the bathwater, and unwittingly consigning modernist architectural urban theory to the scrap heap in its entirety, insofar as it provides any substantial explanation of the urban dimension. This is not my intention. The aesthetics of architecture have an unparalleled place in all societies, and there is nothing so uplifting to the human spirit as direct contact with an architectural masterpiece such as King's College Chapel, Cambridge, the chapel at Ronchamp or Frank Gehry's Guggenheim museum in Bilbao. Indeed, there are many astounding insights in the four texts just discussed, and each has been selected because at the time it was the best available attempt at presenting some coherent synthesis of the urban prospect in terms of its formal arrangements.

In the last instance however, not only have these ideas failed to produce a synthesis, there remains a general reluctance to accept that modernist ideology was on many fronts, a wholesale disaster zone. Not only did it generate an astounding array of bad buildings and planning disasters, it exhibited a serious incapacity to accommodate the explosion of highly relevant theoretical debates and discoveries in related fields. Within the same period that the above attempts at synthesis were being set out, four individuals were to stake claims to unified theories of urban design in various areas. Their work in many ways still represents the core of mainstream urban design theory for many academics and practitioners over the last half century, and it is to this contribution we now turn.

Unified theories

Through the analysis of the work of many scholars and practitioners, the above texts all attempt some synthesis of urban design theory with almost exclusive reference to architecture. In addition, there have also been four quite clear attempts by individuals to propose theories of their own – unified theories. The four theories that I refer to are Kevin Lynch's A Theory of Good City Form, (1981), Rob Krier's Urban Space (1979), Christopher Alexander's A New Theory of Urban Design (1987), and Bill Hillier and Julienne Hanson's Social Logic of Space (1984). Each is fascinating and unique, proposing altogether different theoretical paradigms for understanding urban design. In addition and without exception, each text that embodies the basic argument is both preceded and succeeded by other important contributions in the form of books, articles and/or empirical research of some kind. We now turn to the first of these, a person who has an enormous following in the professions and whose work is frequently taken as being synonymous with the field of urban design.

Kevin Lynch

Kevin Lynch probably shares the laurels with Christopher Alexander as the most cited scholar in the field of urban design. Lynch died in 1984 and his life work has been edited by Tribid Bannerjee and Michael Southworth (1990) into a massive collection containing his writings and projects called City Sense and City Design. The collection is arranged in seven thematic groupings dealing with city design – form, experience, analysis, theory, etc and covers the 30 years of his professional life from 1954. On the cover credits Philip Langdon states that 'Lynch has made the greatest personal impact of any planning teacher in the past thirty years; he established the ways in which urban designers examine a city's physical form, and he stimulated two decades of productive research in a variety of disciplines' (my italics). Ignoring the fact that Langdon conflates planning and urban design, the editors, both former students of Lynch (who in turn was a student of the great American architect Frank Lloyd Wright) – state in the introduction that Lynch 'was the leading environmental design theorist of our time'. Lynch wrote seven books and some 25 articles over the course of his life, and indeed there is no doubt that he had a profound influence over several generations of environmentalists, architects, planners, urban designers and landscape architects.

Notwithstanding his last book A Theory of Good City Form (1981) it is arguable that his first two books The Image of the City (1960) and Site Planning (1971) had by far the greatest impact of all his writing. Despite this significant influence, his work is extremely difficult to categorize. There appears to be two main reasons for this. Firstly, he chose to discuss a huge range of topics, making any overall categorization difficult, covering everything from horticulture to nuclear war. Secondly, most of his writing is eclectic and self-referential, with homage paid to few other individuals. Even after 30 years of writing, the bibliography to A Theory of Good City Form, lists authors using their first names, rendering the entire book somewhat inaccessible. His eclecticism is no doubt an outcome of his education. After leaving high school he went to study architecture at Yale, which he gave up after his sophomore year to work with the world famous Frank Lloyd Wright at Taliesin. After 18 months he became disenchanted with Wright, then studied engineering at Renssellaer Polytechnic Institute for a short time, and soon switched to studying biology at the same institute, prior to being drafted into the army in 1944. By this time Lynch had studied architecture, engineering and biology over a 4-year period without completing a degree. These interests however continued to influence his writing. For example, the biological analogy is evident in his two 1958 papers Environmental Adaptability and A Theory of Urban Form (with Lloyd Rodwin), two of the best papers he ever wrote. Even his experience in the military is reflected in an article published in the year of his death entitled Coming Home: The Urban Environment after Nuclear War, where he places himself in the existential position of a survivor of a nuclear holocaust (Bannerjee and Southwarth, 1990). Lynch eventually graduated with a Bachelor of City Planning from MIT in 1947.

In referencing City Sense and City Design it is extremely difficult to classify his work due to the fact that in the 850 pages of the book there is no chronological statement of what he wrote, or the projects he was involved with. There is no Lynch bibliography that allows the reader to track the changes in his interests or indeed the development of his intellect. Similarly, and probably more telling, is the fact that one can range hundreds of pages through the book without finding a single bibliographic reference or serious discussion of any substantial theory or theorist. This creates the impression that much of the writing is intuitive and anecdotal, pieced together from a variety of interests and experiences, but largely deprived of any significant theoretical foundation. Little homage is paid to anyone. This basic weltanschauung permeates his opus magnum regarding urban design, namely A Theory of Good City Form, published in 1981. In enunciating his own approach, Lynch offers what he calls 'three normative theories' which he says have structured city form since time immemorial, namely the cosmic, the mechanical and the organic: The form of any permanent settlement should be a magical model of the universe and of the gods. It is a means of linking 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, p. 73)

Lynch cites India (Madurai) and China (Beijing) as the two best-developed branches of 'cosmic theory' (curiously, Beijing had become singularly non-cosmic and atheist with a large museum at its core). Lynch describes his mechanical model (a favourite of many architects from Saint' Elia and Le Corbusier to Archigram and beyond) as follows: Thinking of the city as a practical machine, on the other hand is an entirely different conception. A machine also has permanent parts, but those parts move and move each other. The whole machine can change, although it does so in some clearly predictable way, as by moving along some predetermined track. The stability is inherent in the parts, but not in the whole....The machine model is not simply the application of a grid layout, but rather a characteristic view about parts and wholes and their function (Lynch, 1981, pp. 81–83)

Despite this qualification, most of the examples given have gridiron frameworks (Greek Colonial towns, Roman military camps, towns created under The Law of the Indies, New York City, etc).

Finally, Lynch offers the biological model, where human life, consciousness and action, is conflated to that of the plant and animal kingdoms. This had already been accomplished in The Chicago School of human ecology, beginning with Robert Park's seminal statements in 1916, thereafter developed by Louis Wirth and Roderick Mackenzie. But by 1950, Park's ecological approach to the city had almost completely expired (Saunders, 1986). The problem was that Lynch was writing in 1981: The third normative model is much more recent, even if it is already two centuries old. This is the notion that a city may be thought of as an organism, a notion that came with the rise of biology in the eighteenth and nineteenth centuries. It was one expression of the nineteenth century reaction to the stress of industrialization, gigantic new cities, and the unprecedented leaps in technology. The force of this current persists, as evidenced in the spreading political influence of the idea of ecology, or in the academic struggles over subsuming human culture into the new field of sociobiology (Lynch, 1981, p. 89)

He cites few examples of this model, but refers to Tapiola (Finland), Hampstead Garden Suburb, and Greenbelt, Maryland as prototypical (all the product of artifice and as far from the organic as one can get).

A catalogue of models of settlement form in Appendix D reinforces these three basic normative theories. All three theories suffer to a greater or lesser degree from a combination of essentialist thinking with modelling. Any form of modelling is a negotiation between the model and reality (in Sayer, 1976), the relevance of urban modelling is systematically demolished). The flaw throughout is that rather than starting with how cities actually grow and function, some dominant feature of cities is adopted, suppressing all other features to this single concept. The idea that any contemporary settlement should correspond to a 'magical model of the universe' is quite absurd and does not warrant discussion. Similarly the concept of the city as a molar machine is also ridiculous. The mechanical model has more currency but still collapses round the analogy that cities are machines and can therefore be treated functionally, objectively and in accordance with simple and predictable laws. People do not enter into the equation. Definitive urban processes such as urban politics are simply ignored. The organic model probably arose from Frank Lloyd Wright's influence over Lynch when he was a student, Wright being the father of 'Organic Architecture'. Here Lynch falls into the standard trap of many architects, planners and urban designers in two ways. First, in adopting the idea that a city is merely a somewhat larger building, and second, in assuming that buildings and cities can be compared by analogy in some significant manner to the natural world, as in the architecture of Wright, the writing of Lewis Mumford, and the research of The Chicago School of Human Ecology. Somehow Lynch missed out on the most overwhelming model of all, that cities are the stage where the class struggles of capitalism are played out.

Origins apart, Lynch's 'theory' is founded at two levels. First, there are the characteristics on which the performance dimensions of any city should be established – 'The characteristics should be as general as possible, while retaining their explicit connection to particular features of form' (Lynch, 1981, p. 113). Second, he states the six basic elements of his system as follows:

1. Vitality
2. Sense
3. Fit
4. Access
5. Control
6. Efficiency and Justice.

Each of these dimensions has some secondary breakdown or subsidiary classification, for example of 'Control' into five subgroups – the right of presence, the right of use and action, the right of appropriation, the right of modification and the right of disposition. Each of these is explained in detail both relationally and how they would be applied in practice. There is probably no single grouping of elements in urban design that has had a greater impact on the consciousness of urban designers than these six.

Lynch's theory of urban form then unfolds on the basis of an extended explanation of what he means by each of these terms and how they are deployed in analysis. For example, he says of 'Fit' that 'Simple quantitative adequacy is the elementary aspect of fit. Is there enough housing of standard quality? A sufficiency of playgrounds? Room enough for the factories that will be built?' (Lynch, 1981, p. 153). The limitations of the theory are anticipated by Lynch himself when he notes the lack of a complementary theory on how cities came to be and how they function, and perhaps also in his comment that 'It is a peculiar fact that much of the literature on the theory of city form is outstanding for its stupefying dullness'. It is also quite possible that the dullness that Lynch refers to, just might result from a theory of good city form being normatively generated in the absence of some profound links to 'how cities came to be and how they function' (Lynch, 1981, p. 342).

Rob Krier: Urban space

Rob and Leon Krier, both brothers and architects alike, are two of the most popular of the European Avant-Garde in contemporary architectural theory. While they share a similar intellectual focus, Rob Krier is older and more utopian than his brother Leon. Also his written work is more prolific and politically left of centre (Krier, 1978; Krier and Culot, 1979). But here we are concerned with a single work by Rob Krier first published in the original German under the title of Stadtraum in 1975, followed by the English version in 1979. While Krier makes no claim to a new theory of urban design, his book was quite revolutionary when it appeared, offering urban designers a complete interpretation of urban form in actual graphic examples. The book's subtitle Typological and Morphological Elements of the Concept of Urban Space indicates his contained approach to urban design theory, and this is laudable. Krier's hypothesis is that all forms of urban space are already in existence and therefore no new forms can be created. Over the passage of historical time, every physical relationship between form and space has been worked out, possibly thousands of times over in some cases. The assumption we can glean from this statement is that we must pay our respects to history for providing us with the essential vocabularies we need to construct meaningful urban spaces for the foreseeable future. Krier states his intention simply and directly 'It is not my intention here to generate a new definition (of urban space), but rather to bring its original meaning back into currency' (Krier, 1979, p. 2).

Stadtraum constitutes the most complete statement of urban space typology that we have to date, and it simultaneously offers a theory of urban form that is more both limited and unified than that of Lynch and Alexander, but without the claims to prominence. Despite Krier's politics, his approach is highly functional, trying to clarify the concept of urban space independent to aesthetic criteria, and therefore arguably from all applied meaning. One might question why this seems a reasonable thing to do, since depriving anything of meaning would appear to render it catatonic – incapable of expressing anything at all. Krier makes his case as follows: If we wish to clarify the concept of urban space without imposing aesthetic criteria, we are compelled to designate all types of space between buildings in towns and other localities as urban space....Every aesthetic analysis runs the risk of foundering on subjective questions of taste. As I have been able to observe from numerous discussions on this topic, visual and sensory habits, which vary from one individual to the next, are augmented by a vast number of socio-political and cultural attitudes, which are taken to represent aesthetic truths....However my observations indicate that they are almost always identified with the social structure prevailing at the time in question (Krier, 1979, pp. 2–3).

Krier's argument is that we must ab initio seek a social foundation for urban form, for the simple reason that while historically derived urban forms may be defined and used rationally as a vocabulary, their meanings are wholly contextual. The implication behind this is that meanings are not given to buildings and spaces by designers, but are semanticized by populations. Since meanings are historically and socially reproduced, one would have to think in significant depth as to the correctness of borrowing any urban form along with its historically designated meanings. This situation was brought home to me quite vividly in 1984 on a visit to Shanghai while walking with a Chinese friend on the Bundt. On the other side of the road I saw an exquisite building designed in the Classical Greek style, and asked what the banner meant that was hanging outside it – the banner read 'The Headquarters of the Communist Party of the People's Republic of China'. After a long analysis of the symbolism of the façade and the functions that were going on behind it, I decided rather reluctantly that they were not altogether inappropriate.

But in separating form from meaning in Stadtraum, Rob Krier emphasizes that the real task of the urban designer, having been handed the vocabulary of form, is to select the most appropriate urban forms to fit any specific site, without agonizing over symbolism. Despite his fundamentalist approach, it is difficult to be so absolute about the designer's role. After all, the designer is a social product as well, and embodies a constellation of meanings that drive their actions. Krier is clearly ambiguous about this when he says – 'It is certainly worth trying to establish why certain kinds of urban space were created in the 17th century which we now identify with that period. And it would be even more interesting to examine the real reasons why 20th century town planning has been impoverished and reduced to the lowest common denominator' (Krier, 1979, p. 10, see also Shane, 1976).

As an urban design theorist, Krier therefore denies the simple classification of rational/contextual by simultaneously occupying both camps at once. On the one hand he argues for a historical and contextual urbanism, criticizing Le Corbusier's 'confused historical sense', yet at the same time he adopts a wholly functionalist stance to his analysis of urban form. The advantage that Krier has over both Lynch and Alexander is that he is a practicing architect of significant international reputation and has managed to apply his own theoretical concepts in major projects across Europe (Amiens, Luxembourg, Berlin, Amsterdam, The Hague, etc). Curiously, and on the one hand, while his theory of urban form is much more limited than that of Lynch or Alexander, on the other, it is founded on a much deeper understanding of how society actually functions using many of the concepts of political economy as intellectual sources. To this extent his ideas, like those of Rem Koolhaas, are seriously driven by concepts which lie outside his immediate focus of attention, and represent one of the most sophisticated attempts to combine the social with the physical.

Bill Hillier and Julienne Hanson

Bill Hillier, former Professor of Architecture at the Bartlett College, London University, is probably the least known and cited of the four individuals theorists under discussion, for the simple reason that his work is extremely difficult to understand, much of it buried in mathematical concepts and formulae. On the other hand, it constitutes one of the most erudite and scholarly expositions of urban spatial theory. In his first major book Space Syntax (with A. Leaman, 1976), a basic framework of ideas is sketched out. The concepts are further elaborated with Julienne Hanson in The Social Logic of Space (1984), which concludes with notes towards a general theory of social space. These ideas are also condensed in a fine article called The Architecture of the Urban Object (1989), and are pursued in significantly greater depth in Space is the Machine – A Configurational Theory of Architecture (1996). It is also notable that in the massive 800 page Architecture Theory Since 1968 (Hays, 2000), Hillier's work goes unmentioned, indicating that at least one editor does not consider his work part of the last thirty five years of theory in architecture, despite his three books and numerous articles on the subject. This is unsurprising, since the word social is anathema to most architects. Hillier's work is in many respects remote from specific disciplinary boundaries, and is not easily confined to academically or professionally defined territories.

However, as Hillier and Hansen state in the preface to The Social Logic of Space, 'The book is thus a statement of a new theory, and a sketch of new methods of spatial analysis' pointing out that further case studies are necessary which would include 'the social logic of settlements, the social logic of housing, and the social logic of complex buildings' (Hillier and Hanson, 1984, p. xi). In greater detail, they suggest that what might appear to be associated fields of research such as Alexander's Pattern Language (above) actually have limited relevance to their notion of fundamental syntactic generators. Alexander's notion of a pattern is too bound to the contingent properties of the configurations to be useful for us; while at a more abstract level, his preoccupation with hierarchic forms of spatial arrangements (surprising in view of his earlier attack on hierarchical thinking in a City is not a tree (1965) – would hinder the formation of non-hierarchical, abstract notions of spatial relations, which in our view, are essential to giving a proper account of spatial organization (Hillier and Hanson, 1984, p. xi)

Curiously, Hillier does not mention Alexander's analysis of an Indian village in the last chapter of A Synthesis of Form, which appears to bear a singular resemblance to Hillier's analysis of the Tallensi people of Northern Ghana and the Ndembu of Northern Zambia (Hillier and Hanson, 1984, Chapter 8). In spite of the denial, there is a definite link between Hillier and Alexander, and also to the idea of urban patterns as fractals (Batty and Longley, 1994; Batty and Xie, 1996), and to the Theory of the Urban Web (Salingaros, 1998).

Hillier's concerns begin and end with architecture, which he views as the only discipline that has the capacity to take account of the underlying morphological structures in the relationship between society and space. Hillier and Hanson maintain that the reason why the socio-spatial integration remains problematic is because of the missing link between the domain of society and the purely physical nature of urban space, between the social element in the physical and the physical element in the social. The objectives of the syntax model are noted as follows:

• To find the irreducible objects and relations or 'elementary structures' of the system of interest – in this case, human spatial organization in all of its variability.
• To represent these elementary structures in some kind of notion or ideography, in order to escape from the difficulty of always having to use cumbersome verbal constructs for sets of ideas which are used repeatedly.
• To show how elementary structures are related to each other to make a coherent system.
• To show how they may be combined together to form more complex structures (Hillier and Hanson, 1984, p. 52).

The proposed theory is purely descriptive (as opposed to analytical, methodological, predictive, etc), and builds a relational conceptual model of pattern types, thereafter.