*It is interesting that all of these architects when faced with single building projects in an existing fabric had no trouble respecting the pattern of block-boundary-public network. It is only when they dealt with whole new cities or with extensive sections of existing cities that they ignored this structure.

Diagrammatic illustration of developments in site layout from das neue Frankfurt (note ‘IV’)

A portion of Hellerhof Siedlung, Frankfurt

the twentieth century

It should be clear that in traditional urban structure the public domain is not the space left over between city blocks, but an integrated, meaningful public network. Whenever a contemporary planner, architect, or landscape architect designs a new environment he is careful to label all of its parts and he usually develops construction details for them. Because he has named every place and detailed its construction he assumes that the resulting spaces are meaningful. However, the integrated public network is not created simply by giving names to all of the spaces between the buildings. In order to clarify this point it may be helpful to consider several characteristics of the public network in the traditional city:

  1. Most important is that the public network exists as routes and expansions in those routes. The routes connect the parts of the city; the expansions - the parks and the squares - are part of the system of routes (and destinations in themselves).
  2. Within an urban agglomeration the public network is continuous. The routes and expansions of the public network are bounded by the protective-presentational boundary of the surrounding private domain, or, occasionally, by pastoral backdrop. No matter what the level of the circulation hierarchy the route is bounded in this way, and this boundary provides a continuous edge as one passes from one level of route to another.
  3. The public network and its elements are intentional. This admittedly ambiguous adjective hopefully conveys the notion that the public network is not careless nor excessive. Its elements are not wasted, and the routes and expansions of the public network have some active significance in their environs. Except for very large parks such as Central Park in New York or the Bois de Bologne in Paris, very little land is actually given over to parks and squares. The major part of the public network is street and path; park and square are constrained, exceptional, intentional additions to the network.

Traditional urban structure consists of two complementary elements: the territorially secure block and the integrated public network. These two elements have a symbiotic relationship; each providing to a considerable extent definition of the other.

In the twentieth century traditional urban structure has been attacked with surprising success considering the normal inertia of past habits. The undermining of tradition has, curiously enough, come from two sides: one attacking the territorial model of the city block and the other dis-integrating the public network. Because of their mutual dependence, successful attack from either side would be, to say the least, debilitating. Destruction from both sides was so successful by the mid-20th centry that in North America and Wetern Europe between the 1940s and 1970s few new housing or commercial developments contained either city blocks or an integrated, meaningful public network.

Since the 1970s there has been a re-recognition of the success and importance of pre-20th century urban form. The “New Urbanists”, Leon Krier - and even Prince Charles - have railed against the planning and the architectural of the early-to-mid 20th century. And many projects have been developed which demonstrate an admirable return to pre-20th century planning patterns.

However, there are two “howevers”.
The most important of these concerns is that the “loss of memory” of earlier traditions was and continues to be widespread. “Advanced” planners are working regularly to re-instill - and re-install - clear traditional urban developments not completed by these knowledgeable few. Planners, civil engineers, developers - and, yes, architects - were thoroughly trained in the now-traditional urban pattern of the late 19th and early 20th centuries. These tradition breaking models have become their own traditions, and this new tradition is, in much of the world, the dominant planning and development pattern, expectation and habit.

The second concern is a bit more hidden, but equally problematic: Many of those who have fought to re-claim city form have done so using the symbolic parts of the old tradition - yet they have failed to provide a consistently integrated - territorially coherent - framework.

This problem has arisen from two specific histories:

  • One is that the lessons of the first half of the twentieth century – those “bad” lessons – have stuck, and their conflict with traditional urban structure has not been clear. Much neo-traditional urban development contains vestiges of the precepts which destroyed so much of traditional urban building during the first half of the 20th century.
  • The other history is that much of neo-traditional planning is completed in imitation of the old ways of doing things, but without a clear comprehension - or - “lens” – of the essentially territorial naissance or raison d’être of that tradition. And without that lens new-traditional urban development are frequently designed with subtle, but real, violations of the territorial map.

The following discussion outlines the key changes to the traditional urban framework which arose from the theoretical and built work of the late 19th and early 20th centuries. As noted above, these attacks came from two sides: dis-integration of private and public places, and breaking of the territorial coherence of the city block. All of these changes grew from the best intensions. The language of those good intentions persists, and has made rejection of the resultant pattern especially difficult.


Dis-integration of the public network has been going on since the 1920s. The pattern has been one of buildings placed in great “parks” with roads and other services cutting through the parks as required. Le Corbusier’s Ville Radieuse is probably the best known model, and certainly the most influential. However, Le Corbusier is not solely responsible. Most of the great architect-humanists of the century have been involved: Gropius, Frank Lloyd Wright, Sert, Aalto, Woods, and the Smithsons (and Mies, the self-proclaimed non-humanist - or at least non-sociologist).*

Part of the Ville Radieuse was to be continuous medium-rise apartment buildings. The buildings were themselves protected “city blocks”, and because of their continuous pattern:

formed a public network of streets and open spaces. The pattern perhaps lacked intricacy, joy, or surprise, and was so generalized that it reflected few of the qualities of the public networks of the past. Yet it did at least reflect the historic pattern of the protected private domain existing with and forming a public network of routes and expansions.

This portion of the Ville Radieuse, however, was not the one which inspired the profession. It was the tower-in-the-park which became the model for so much of the residential, commercial, and corporate building which has been built since.

This model of the tower-in-the-park was not the day-dream of some ivory-tower intellectual but a serious attempt to provide better environment. That it has been followed so widely is the best proof of its connection with the real issues it attacked and the desires it fulfilled. (Issues such as light and ventilation, recreational space, access and storage for the automobile - as well as those of mobility, isolation, and anonymity.) It attacked these issues, however, at the expense of the urban values and habits which had grown over the past millennia. In terms of urban structure the loss is best described as the disappearance of the street as a significant, formed element in the city - and the corollary disappearance of expansions in the street as significant elements of relief.

The tower-in-the-park is a set of private domains. These isolated buildings are usually apartments, though it is not uncommon to find them occupied by commercial and corporate activities. The buildings are protected from their surroundings by solid walls, windows, and entry-ways. Facades vary in uniqueness and decoration, but can generally be described as protective-presentational boundaries separating the private domain from the surrounding (quasi-) public domain. Seen in this way, we can almost think of such buildings as “city blocks” sitting in a non-existent public network. Of course the public network does exist as roads, green space, and often extensive parking lots. But it is an exploded network with no constriction: no signification of route and eddy. Because the network is exploded, it lacks, too, normal urban signals of gradations of possession, which rely on more active juxtaposition between the protective-presentational boundary and the public network.

Le Corbusier’s Unité d’Habitation, Marseilles

The spread of this pattern has been phenomenal, especially when one realizes that it has been here only since the Second World War with any regularity. The tower-in-the-park is not limited to high-rise towers per se: low-, medium- and high-rise buildings isolated in their surrounding green or asphalt have been built all over the world. Every continent is filled with examples - examples of “slum clearance” in Chicago; public housing in New York, Moscow, London, Hong Kong, Stockholm; luxury housing in Sydney, Paris, New York, Johannesburg; even entire cities such as Brasilia. The list could go on and on.

apartments in Näsbydal and Grindtorp subdivisions, Sweden


Low-density single-family housing on individual lots is common in North America, and it is increasingly widespread in other parts of the world. As the word suburbia implies, there is a lack of urbanity or “city-ness” in these environments. The low density, reliance an the automobile, the tendency for inhabitants to commute to work, and the homogeneity - or general lack of normal urban “anarchy” - all seem less than urbane. Suburban environments, however, often reflect the pattern of city block, protective-presentational boundary (of expansive front lawn, decorative facade, and picture window), and public network - the pattern we call urban. Because it reflects this pattern, low-density single-family housing usually presents a fairly clear territorial pattern. The signals indicating one’s territorial position are not confusing, and one can pass through these environments with a sense of ease.

One may, however, also pass through with a sense of boredom at the endless and rather un-formed quality of these environments. Though they keep the basic structure of the city block and public network, these environments are similar to the tower-in-the-park in that they fail to create firmly defined public network. This lack of firm definition results from the low density of building in the environment - and perhaps, too, from the low density of perceived variety, activity, and humanity.

In low-density suburbia and in tower-in-the-park environments buildings exist as points in the landscape rather than as continuous elements. And each, for slightly different reasons, fails to create a strong public network. Though often attacked as anti-urban, low density suburbia is less anti-pathetic with historic models, at least in territorial signalling and structuring. But both patterns have become so widespread that we now see extensive urban areas where the public network has ceased to exist as an important part of the city.

the demise of the city block

The tower-in-the-park keeps a de facto city block as a territorially defined unit, but fails to create a significant public network. Attacking tradition from the other side, Clarence Stein and Henry Wright built the public network carefully - but in so doing broke down the structure of the city block. Radburn, though perhaps less well-known than the Ville Radieuse, has had immense impact on the way cities, especially low-density and medium¬-density residential areas, are built.

The guiding principle for Radburn, and for most of Stein and Wright’s developments, was that the pedestrian and the automobile should be separated. The pattern which emerged has fingers of vehicular access coming from the surrounding roads and a complementary pattern of fingers of pedestrian access coming from a major park:

A single row of housing units fits in the space between the fingers:

The “motor way” was envisioned as a service lane. The common fate of sites developed on this pattern is that the lane provides the major access to the units - not only for adult residents, but for children, guests, and the postman. The lane also provides the major focus for interaction and play. In other words, the lane is the street. We can see this even more clearly in Stein and Wright’s design for Burnham Place, where the lane was given the gracious scale of a suburban cul-de-sac, the foot paths leading to the park were envisioned as well-used promenades and the facades of the units were built facing these park-like areas. The foot paths, however, lead to an intentionally out-of-the-way park (labelled a “private park for the residents of Radburn” on design drawings). Because they are not part of any normal, inevitable movement pattern and because the action is on the cul-de-sac, the foot paths tend to be used very little. When they are used it is for an occasional bucolic stroll - not for the comings and goings of day-to-day life in the city. They remain symbolically public network yet retain little of the action of the normal public network.

The result is that the footpaths remain a sort of down-graded public network; while the lane becomes a sort of up-graded private domain: a service lane becomes a street. The houses are surrounded by public network. They face the footpaths with the large setbacks, front porches, and more elaborate facades of the protective-presentational boundary. Yet they face the cul-de-sac, which is the real street, with garages, secondary entrances, and “backyards”. The coherent city block with its constant intermediary protective-presentational boundary between private and public domain has ceased to exist.

Since Stein and Wright’s original essays many housing developments have been built on the Radburn model. A major change has occured in many “post-Radburn” developments as a result of pressure to better-integrate coming-and-going by car: In these post-Radburn projects facades generally face the automobile access and “backyards” face the network of parks and footpaths.

This change, though better recognizing the nature of street and entry, does not change the underlying structure of Radburn. Individual housing units are still surrounded by the public network yet present a protective-presentational boundary to only one side. Backyard patios (and large patio doors) are exposed to footpaths, playing fields, schools, and swimming pools.

Stein and Wright were designing very much in the tradition of Ebenezer Howard’s Garden City movement, and it is perhaps unfair to single them out as the sole culprits in the destruction of the city block. The English architect Raymond Unwin (of Parker and Unwin) wrote Town Planning in Practice in 1909. In this book he proposed an arrangement of rectangular buildings on a curved street. A tennis court and small park sit behind the backyards of the units - the first example of “Radburn-type” structure twenty years before Radburn.

It is interesting that nothing in Ebenezer Howard’s writings proposes that in the ideal garden city traditional relationships between public and private be changed. Nor did the garden cities which were actually built vary from traditional patterns.

Part of Parker and Unwin’s 1902 project at Letchworth:

(Note allotment gardens in the block centres: garden-as-pastoral backdrop; “green”-as-park.)

Ministry of Munitions housing, West Hall Phase I:

During the same period that Stein and Wright were developing their preliminary diagrams for Radburn, Ernst May was building a number of new projects in Frankfurt. These projects were relatively high density developments and though they almost followed traditional patterns, occasional backyards facing neighbouring front yards began to appear.

Ernst May and Raymond Unwin were making proposals which stood in opposition to past practices of urban building - and which pre-dated Radburn. I nonetheless point to Radburn as the effective beginning and example of the “inside-out” city block. I have done this for two reasons: First, because with Radburn Stein and Wright emphasized their change in structural relationships, while May and Unwin only happened upon it in a peripheral way (and Unwin even seems to have rejected it). And second, because Radburn has stood as such a powerful example in recent planning and architectural history.

For the tower-in-the-park the public domain is the space left over between the buildings. In the Radburn plan there is much more careful creation of the public network. Footpath, lane, and park are formed and stand in a specific relation to each other - and in this way they can be said to be “integrated”. However, the traditional public network is not integrated and meaningful simply because it is knit into the fabric.

As mentioned earlier, a characteristic of the integrated public network is its continuity. The public network is bounded by a protective-presentational boundary, and this boundary provides a continuous edge as one passes between different levels of the public network, and between different parts of the city. In the Radburn plan a considerable portion of the public network is not bounded by facade and formal entry. And we can thus say that the public network is discontinuous.

Before Stein and Wright’s work examples of discontinuity in the public network were rare. It seems that the discontinuous public network is our chief inheritance from their work. And it affects both the private and the public domain. The private domain no longer has a protective-presentational boundary consistently protecting it from the public. This lends a certain dis-ease, or dis-comfort, to the private domain.

There is no doubt that such environments can be lived in comfortably, but they demand new ways of dealing with the environment. Private outdoor space is no longer a place to do as one pleases, but a part of one’s presentation to the public. (Even interior spaces can be forced into public presentation when glass patio doors open on to unprotected rear yards.) The challenge to the private domain is fairly obvious. What is less obvious, and probably more damaging to the urban psyche, is the challenge presented to the public network.

Elements of the public network traditionally gain significance not only as routes from here to there, but as bounded, integral parts of the city, with residential and commercial activities feeding on to them along their entire length. Since the arrival of the discontinuous public network, we see many streets bounded by service yards, backyards and patios. The street is allowed only part of its significance - its significance as a route. The significance of continuous bounding activity and territory disappears, and the public network is allowed to be only a by-pass. This may not be disorienting to the driver who actually uses the street as a by-pass. But as anyone who has dared to be a pedestrian in such an environment can testify, the result is unsettling; (especially when one is confronted with the backyards of a large housing estate on one side of the street and the facades of an older development on the other side).

When streets are simply by-passes pieces of the city sit in isolation from each other. The resultant urban form is one of enclaves with specifically defined functions. This form destroys the most exciting quality of the traditional urban fabric - the quality of continuous, constantly varying activity. A city of enclaves ignores the fact that the traditional city is always in flux, that activities and images change gradually, and that they do so along a continuous network of urban routes.

I've presented the notion of “anarchic” demographic flux. This flux demands continuity in the public network; that is, a continuous relationship between the public network and surrounding protective-presentational boundaries. The interdependence between demographic flux (or migration) and continuity in the public network may still be confusing, and a short discussion of the discontinuity in post-Radburn environments may help to clarify the point.

As one moves through the traditional city, one is constantly presented with the protective-presentational boundary of surrounding private domain. This is the case whether one moves on the paths of the Casbah in Morocco, on the boulevards of Paris, on the greens of Central Park, or on a North American single family residential street. The relationship remains constant as one moves from path-to-path, boulevard-to-boulevard, park-to-park, or from one North American residential street to another. And it remains constant as one moves from one LEVEL of route to another. (path-to-boulevard-to-park-to-path-etc.)

(Each circle represents a city block. Each block is divided into pie-shaped “slots” of private domain - each with its protective-presentational boundary facing the public network.)

The territorial clarity afforded by this pattern is, of course, the primary message of this essay. And, as mentioned earlier, it is a sense of territorial clarity, and therefore comfort, which permits some heterogeneity in use between one slot and the next. In order for activities to insert themselves as immediate neighbours of already existing activities there must be immediately adjacent slots for them to move into, and this is exactly what happens in the traditional city. Any gaps in the continuous presentation of “slot faces” to the public network results in a gap in the path of demographic migration. The greater the gap in this path, the greater is the impediment to casual “anarchic” long-term evolution of urban activities, and therfore to long-term urban viability.

The by-pass routes of post-Radburn enclaves are, by definition, gaps in the continuous presentation of “slot faces”. Radburn-type projects present a “double whammy” of discontinuity in the public network: 1) The by-pass routes which connect enclaves have no (or at best few) slots of protective-presentational boundary-defined private domain - and therefore cannot possibly be a matrix for transition and gradual evolution and flux. And 2) within the enclaves themselves the public network is discontinuous, passing equally by slots of private domain defined with their respective protective-presentational boundaries, and by exposed, unprotected private domain.

Almost all contemporary urban building, even that at relatively high density, is very close to this diagram - a fact which is often masked by the creation of “integrated” “nice tight urban spaces” (which lack any real urban continuity).

It may be argued that to descry the city of enclaves simply because it disallows traditional flux is wishful romanticism. But the city of enclaves is an urban form which does not permit the gradual development and spread of new, unforseen activities. This presents a real loss of economic potential. And, more pertinent to this essay: Its discontinuous public network presents territorial patterns which run counter to those of traditional urban forms - and counter to traditional urban habits of mind:

The result is similar to the tower-in-the-park. Enclaves of buildings sit in the landscape as isolated elements. The road system is incidental to them, acting only as a connective element and not as a significant formed street. When the city is made up of towers-in-the-park and these isolated enclaves it lacks any sense of continuity: of gradation and transition between its elements. The normal signals which one uses to guage his or her position in the environment seem to evaporate. This is especially true for the pedestrian as he or she passes through a kind of no-man’s-land between enclaves. But it is even more confusing territorially when one enters the enclave itself, where outsiders and residents lack clearly comprehensible rights of passage or of occupation.

épernay General Development Plan, Houston, Texas.

Wohnsiedlung Meerzicht-Sudwest in Zoetermeer, Holland

Post-Radburn projects are not limited to the suburbs. Even when housing is placed “in historic context” the predominant structural model is Radburn.

This project was hailed in the 1970s by the architectural press as a re-cognition of the historic way of city building. But it emulates only the visual intricacy and density of the past. Structurally it has nothing to do with pre-twentieth century urban building.

Friar’s Quay, Norwich, England, architects: Feilden and Mawson

In Paris the Ville Radieuse is still the model for how cities should be built. Paris itself has suffered incursions, but the best place to see how French urbanists think is in the suburbs and new towns. Corbusier’s stark repetitive towers have been replaced by attempts at flambuoyance and individuality and moat “parks” are now over-scaled plazas. But traditional signalling of where one stands in the city is no longer available.

Older Paris, which is a marvellous example of how we can build cities, has been ignored in favour of “modernism”.

The historic city:

Replaced by the modern city:


And Emile Aillaud’s towers in La Défense (coloured with pictures of skies and clouds):

The Ville Radieuse model, with its isolated building blocks and un-formed public domain, and the Radburn model, with its discontinuous protective-presentational boundary and ubiquitous public network, have merged and become a massive tradition themselves. In fact it seems that the tradition generated from these two patterns is more powerful, at least in the design professions, than the traditions of all previous urban building. With the merging of the two patterns we find in a single development buildings as isolated elements surrounded by meandering open space, partial isolation of pedestrian and vehicular routes, enclaves of buildings surrounded by by-pass roads, and variably exposed facades and patios.