*The one-sided parks are diagrammatically similar to the one-sided streets of Biskupin and Tell-el-Amarna. However, there are two important differences: 1) In the two villages the closed side is completely closed. Each block is isolated from infringement from or relation with the street in back. Connection with the public domain is explicitly through the front door. And 2) The streets are narrow enough to permit effective contact between the public network and the private domain. The expansions - the public squares - are limited in each case to the area around the village entrance.

some instructive near-misses

It would be nice to say simply: “Build cities as they were always built - city blocks, streets, parks, and all; accomodate the automobile but do not try any fancy tricks which upset traditional structural and territorial habits.” The design professions, however, have forgotten what such a simple directive might mean.

The ideas we have been exploring are fairly simple, but in putting them to use we must overcome same habits which have become ingrained. I have selected three projects which reflect some of the ideas we have been exploring. Each project, though, fails to re create the traditional urban framework. I hope by discussing their shortcomings to point out some possible mis-interpretations of this essay.

project: Bakke draget, Fredensborg, Denmark
architect: Jørn Utzon

This project, though small, and basically a “dormitory community”, does almost everything right. The units face the roads with nearly blank walls (punctured only by doors and kitchen windows) forming a distinctly protective-presentational boundary between the public network and the individual units. The almost exaggerated protectiveness of this boundary is in keeping with the short distances between public and private activities. The public network is the realm of both pedestrian and automobile, and it works very well as a focus of movement and interaction. The fields behind are separated from the private terraces of the units by low walls. It was expected that these fields would be used as pastureland, passed through only occasionally by residents: a true pastoral backdrop.

The residents were expected to confine normal movement and play to the public network. The units were developed with protective faces to the anticipated active realm and the private domain left exposed to the anticipated pasture. Unfortunately no farmer was interested in grazing cattle on the open land and the fields are used us a playground by local children. The expected pastoral backdrop has become public network - and the project is comparable to other post-Radburn townhouse projects.

Bakke draget illustrates a point made earlier: When the surrounding population increases, the impenetrability of the pastoral backdrop must be more pronounced if it is to retain its identity as land of the countryside. When the pastoral backdrop exists in the city it does so as an impenetrable buffer rather than as an open landscape. One cannot install pastoral backdrop in an area simply by letting the grass grow long.

A simple change to the existing pattern could clarify its territorial signals: Develop the pastureland as a dense wood. This would provide protection to the exposed private realm. A public park could be retained as long as it were insulated from the courtyards by the wood. This solution retains the original pattern of public network - private domain - pastoral backdrop, and the modification simply makes the pastoral backdrop appropriately impenetrable.

Buttressing the pastoral backdrop in this way is fine as an occasional solution. However, it poses some problems if pastoral backdrop becomes a prevalent border for the public network: As discussed in the preceding section, the normal street with its border of protective-presentational boundary is a unifying element in the city. It is unifying from side-to-side as the locus of social interaction, and unifying along its length as the mode of urban growth and movement.

The pastoral backdrop, on the other hand, is a divisive, isolating element. One passes by its edge and is kept away from what is on the other side:

Bodies of water and the agricultural hinterland of the single street village do not cause any problems by being divisive because they divide city from non-city. However, within the city the pastoral backdrop separates parts of the city from each other.

Pastoral backdrop can protect private domain from public network. But if streets are lined with pastoral backdrop they lose significance as urban routes and are simply by passes. A repetitive strategy of buffering private domain from public network is a sure way to make a city of enclaves.

Pastoral backdrop as an intermediate element between private open space and public network should be limited - as it is in the traditional city - to inevitable natural elements such as rivers or ravines, or to incidental areas of dense planting.

Bakke draget was built on the assumption that it was on the outskirts of the city. The surrounding countryside is still much as it was at the project’s inception. But the presumed pastoral backdrop has proven to be public network. It is a place for children to play, not a place far harvesting nor grazing, and a through road is not very far away. This transformation of pastoral land to public network is almost inevitable at the outskirts of modern cities - and in most cases takes the more drastic form of new urban growth. The pastoral backdrop, on the other hand, is a divisive, isolating element. One passes by its edge and is kept away from what is on the other side:

Rather than trying to keep the pattern of the single street village, with its agricultural hinterland stretching out behind, contemporary planning should recognize the possibe future transformation of agricultural pastoral backdrop to public network. We should build in a way which will survive and welcome urban expansion - rather than fight an up-hill battle to save the single street village. The simplest way to do this is to confront surrounding land with protective-presentational boundary - not with private open space. If the surrounding bucolic landscape can be saved so much the better. Whether it is saved or not, this pattern allows continuation of traditional urban patterns of significant integrated routes and territorially secure city blocks.

An alteration to Bakke draget which would make it more sympathetic with traditional patterns - and with the fact that cities grow - would be to sub-divide the pockets of pastoral backdrop into backyards and to build a row of housing which faced out to the park:

project: “The Californian”, Tustin, California
architect: Bachen, Arrigoni and Ross, San Francisco

This project was modelled on Chermayeff and Alexander’s book Community and Privacy. A major concern in its development was the maintenance of privacy between public and private domain. The most noticeable result of that concern is the high walls which separate units from the paths. The public and private domain remain isolated from each other by a continuous - one might say relentless – protective-presentational boundary: pattern which is apparently one of city blocks and public network.

Yet something seems to be missing.

Plan of ground-floor unit, The Californian

In the pre-twentieth century city there is a certain efficiency in the use of the public network. Some exceptional buildings - palaces, churches, and large institutions - might sit as isolated elements constituting entire city blocks. These exceptions generally have an entry on one side and face the rest of the surrounding public network with impenetrable walls. But most buildings make efficient and regular use of the public network; at a given density as little public network as possible is used to provide access, light, and ventilation to the surrounding buildings. The public network is valuable stuff, and the surrounding protective-presentational boundary is characterized by regular provision of access and penetration. The regularity and relative efficiency of use of the public network results in the normal street with its penetrable protective-presentational boundary on both sides.

One aspect of the continuous public network is that of continuous protection and presentation of the private domain. The other aspect is that of continuous interaction between public and private through the boundary, as people pass in and out, sit on stoops or porches, and look out of windows. The boundary protects the private domain; as the source of passers-by, surveillance, and general enlivenment, it also protects the public network.

“The Californian” retains the continuous boundary of city block. But penetration through the boundary is irregular. There are extensive areas of public network which have no doors nor windows in view. In these areas the protective-presentational boundary does not permit the private unit casual surveillance of, nor interaction with, the public network; the protective-presentational boundary does not protect the public network. The startling result is that the protective-presentational boundary ceases to protect even the private unit. The walls remain un-seen by surrounding units, passers-by, or loiterers. It becomes an easy matter for thieves to scale the walls take what they want and leave, only keeping a lookout for (unlikely) passers-by. And this is what has happened.

The problems with “The Californian” are, roughly, that there is too much public network for the number of units and that the isolation between public and private is too complete. These are problems of degree. Ten units built in this way and surrounded by a more traditional urban fabric would probably be fine. Occasional large elements in the city with limited penetration seem to function acceptably. But when extensive portions of the public network are bounded by blank walls we begin to lose the interaction and mutual support which normally exist between public and private territory.

The street bounded on both sides by a regularly penetrated protective-presentational boundary is direct and efficient, and certainly provides a familiar framework. Most historic exceptions to the two-sided street are limited in scale, as the palaces and institutions already mentioned, or the occasional city block with entrances on only one side which we find in some medieval villages. However, archeologists have unearthed two (to my knowledge) villages whose streets consistently had openings on only one side. The most well known of these is the Egyptian Tell-el-Amama, or Akhenaten’s village. This was a colonial settlement thought to have been created for the building of tombs and burial grounds. Various craftsmen worked in the village, each in his assigned quarter. The most popular theory to explain the unusual one-sided street is that the craftsmen worked in the streets in front of their homes. Having a blank wall opposite prevented quarrels over who had the right to use the street.

The other village with one-sided streets is Biskupin in Poland. Biskupin, an agricultural settlement of the Iron Age, was built on a small island, apparently as a defensive measure. Access to all units was through doors on the south side; the north face of all units was closed. Biskupin was a pre-historic village, and little is known about the reasons for its unusual layout. Presumably the northern walls were closed in protection against the winter winds.

One-sided streets such as those in Biskupin and Tell-el-Amarna are un-economic in that twice as much street is required for access to a given number of units. The streets have somewhat less activity, with half the usual amount of in-ing end out-ing along their lengths. The difference between these two villages and “The Californian” is that in the villages streets have access and penetration along their entire lengths, while in “The Californian” there are considerable portions of the public network with no access nor penetration on either side.

Biskupin and Tell-el-Amarna were made up of city blocks: of sets of private domains surrounded by a protective-presentational boundary. And they fit our model of traditional urban structuring. Though only half the usual number of inhabitants used any given portion of street, the streets were probably the scene of constant work and play - of regular casual territorial surveillance. In the twentieth century few societies exhibit the intensive use of the street which characterized pre-industrial societies. And in addition to arguments for familiarity, directness, and efficiency, the two-sided street has in its favour that it takes the greatest advantage of what action there is.

project: Adlikon Village, commune of Regensdorf, Switzerland
architects: Steiger Partners A.G., W.M. Forderer
landscape architects: Atelier Stern and Partner

Adlikon is obviously an example of towers-in-the-park, and less like the traditional city than either of the two preceding projects. However, the relationship between the buildings and their surrounding open space exhibits a sophistication lacking in most agglomerations of towers-in-the-park: a sophistication which brings the project surprisingly close to traditional urban models.

An axonometric drawing showing the typical structure of Adlikon.

Adlikon consists of a number of slab-apartment buildings varying in height from three to eight storeys. The slabs are forty to fifty meters apart, built on north-south axes which descend a gentle slope to the south. Pedestrian and automobile follow separate routes (at least in theory). The main pedestrian and automobile routes run transverse to the slabs - sometimes running directly under them and sometimes skirting them. Each building has entries spaced along its eastern side and private courtyards and balconies along the west. The spaces between the buildings are developed parks, with entries to buildings on one side and overlooking courtyards and balconies on the other. In order to relieve the repetitive pattern and provide some sense of location each open space is developed with a different landscape theme.

The public network is bounded on only one side by the protective-presentational boundary of entries, while the other side is bounded by private open space. The effect is that the public network is curiously lop-sided.

The buildings which front on to each side of the park are of equal mass; o n the east are courtyards and balconies and on the west are a more closed facade and ground-level entries. Courtyards and balconies are protected from the paths and overlook the park more in the manner of street-side balconies than of exposed private open space. The buildings on the east seem to have a stronger relationship with the public network they overlook than with the public network that gives them entry.*

It would be possible to alternate ground-floor units so that every-other unit was entered from the east while its immediate neighbour was entered from the west:


This would make the public network more normal in that both sides would be bounded by regularly spaced entries to units. However, each unit still faces two different “streets” - a problem which could be solved by building units back-to-back:

Either of these strategies would create a more conventional relationship between buildings and public network: city blocks surrounded by a protective-presentational boundary and overlook bounded in turn by public network.

This is a much more integrated relationship between building and its surroundings than we find in most tower-in-the-park projects. Yet it is still towers-in-the-park. The structural relationship is traditional, but the public domain seems to suffer from overkill: the amount of park per capita is phenomenal; the number of transverse routes guarantees that none is the scene of regular movement; and destinations within the development are limited to variations on a playground. The public network, though structurally legible, is exploded just as in the typical tower-in-the-park project. And the interaction and control between facade and street, which we see in the traditional city, cease to exist.

Being a dormitory enclave is probably Adlikon’s biggest problem. The scale of the public network in effect creates a necessity for important surrounding destinations - a necessity which simply cannot be met in an environment of by-pass roads, pastures, and other dormitory enclaves. If a single bay of Adlikon (with the proposed change in facades) were surrounded by conventional city blocks it would probably function quite well. Or if Adlikon were in the centre of a large city and its parks were expansions in a system of boulevards, the result might seem repetitive - a bit like the medium-rise portion of the Ville Radieuse - but it would probably prove territorially comprehensible and comfortable.