There are three possible mis-interpretations of this essay which I have done my best to discourage. However, I will take advantage of this final opportunity to re-state what should be clear by now: 1. The reader may feel that communal or “semi-private” open space in the centre of the conventional city block is acceptable within the structural precepts we have explored; it is not, and the main point of this essay is precisely that it is not acceptable. 2. The reader may feel, because so much reference has been made to past cities, that the stylistic mannerisms of the past are the only true way to build cities: this essay is a discussion of the structural relationship between parts of the city, and that structural relationship can be respected with any number of styles, scales, and densities. And 3. the reader may feel that we have been exploring only low- and medium- density primarily-residentia1 development; this is not the case. All urban building before this century respected these structural patterns, and there is no reason to limit the significance of these patterns to housing.

The range of styles, densities, and functions to which our model applies should be clear. Miesien steel and glass is as appropriate to conventional urban patterns as is Greek village vernacular. And the Beaux Arts facades of 19th century Paris are no more nor less correct than the facades of any evolving architetural ‘ism’. Single family houses, medium-density row-houses, high density apartments; corner stores, markets, retail outlets, department stores; bureaucratic, religious, educational, and philanthropic institutions: All are part of the continuing evolution of the city, and all have traditionally sat in the urban fabric in the territorially and structurally coherent pattern we have explored.

Environmental style and building use make no appreciable impact on the inherent structure of the traditional city. The first issue, though, that of communal “semi-private” open space, may still cause some confusion:

It may be tempting to make city blocks which are surrounded by a protective-presentational boundary and which actively define their surrounding public network - yet which have a “communal” private open space in their centres.

In general such blocks would satisfy the traditional structure we have been exploring. Sunnyside Gardens (see p.181-182) would fit such a model if the public were simply prohibited entry to the common green at the centre of the block:

The original Sunnyside Gardens (left) and an hypothetical "Sunnside-communal"

This, however, is not what happened.

The courtyard of the Seville apartment building (see p.28) provides an interesting comparison with such a closed-communal-private open space. The Seville apartment is diagrammatically similar, but it has three characteristics which “Sunnyside-communal” would not have:

It has a strong boundary between individual units and the communal courtyard. In some buildings this boundary is almost an extension of the protective-presentational boundary, with entry to individual units giving on to balcony-corridors - which in turn overlook the courtyard.

In other buildings the units give directly on to the courtyard with windows and overlooking balconies. In either case the private unit is clearly separate and insulated from the communal courtyard; the strong boundary permits a degree of territorial clarity between individual and communal property. The strength of this boundary stands in marked contrast with “Sunnyside-communal”, where private courtyards (outdoor rooms of the private units) are exposed directly to the communal open space of the courtyard.

The Seville apartment has a much higher density of users. The higher density lends a certain anonymity to the courtyard’s users. The changing identity of the individual user makes little difference to the courtyard’s significance. The traditional functional and cultural definition of the courtyard is strong enough to withstand a changing and relatively anonymous population. “Sunnyside-communal”, on the other hand, would not have an anonymous population; it would have to have an identifiable group of right-thinking people who would agree to change existing patterns and to maintain a set of rather ambiguous etiquettes. Over an extended period this original group would change, and the original coherent-anti-traditional spirit and etiquette would have little likelihood of surviving the normal fluctuations of an urban population.

A third characteristic of the Seville courtyard is that it can be reached directly from the entry of the building; it has the character of an “entry court” more than the character of a “backyard court”. The significance of this distinction is that the communal court is most legible when it is an intermediate zone between the public network and the private unit.

Two types of “entry court”; in the first the courtyard is an extension of the entry hall - though it is not passed through as one enters individual units; in the second the courtyard is actually part of the entry hall, and passed through on the way to the individual unit:

Communal open space is not a necessary element in the urban fabric, but when it exists it should mediate between the privacy of the individual unit and the hustle-bustle of the public network. It should be an extension of the protective-presentational boundary rather than an intrusion on exposed private domain.

communal space as part of the portal.


The destruction of traditional patterns has been widespread since the 1920s. And the clarity of what has been lost seems to evade even those with humanist or nostalgic points of view. However, there has continued to be urban building which respects the structural model of past cities. Most small-scale in-fill projects impose no shift on their pre-existing framework. Except for the shopping centre, most commercial buildings confront streets and walkways with their presentational fronts, and keep garbage, storage, and bulky services away from the public network. Office and residential buildings seem most consistently to have lost their traditional relationship to the city - but there are also numerous examples of housing estates and office buildings which sit in a traditional relationship with their environs.

In the past three decades there has been renewed respect for historic urban form and greater distrust of the precepts of the modern school. Most rejection of the Radiant Garden City Beautiful has failed to retrieve the binding structure of past cities, and has relied on visual intricacy and nostalgic style. But there have been some developments which indicate a growing awareness of the importance of integrated public network and territorially unambiguous environment. As we move toward a more open use of historic reference in urban design, it will be important to respect more then the issues of “townscape”. One of the problems associated with a re-discovery of past urban form is what seems to be the position of the “contextualists” and proponents of “integration”: a belief that simply by putting buildings closer together and by carefully controlling the physical form of public open space it will be possible to retrieve past urban form. The significance of traditional urban patterns runs deeper than rhythm, scale, and enclosure in public places and deeper than a ratio between land covered by building and land allotted to public interaction.


The subtitle of the essay, “The Territorial Structure of Cities,” may have suggested something quite different to the prospective reader than the actual contents of this essay. The more sociologically inclined might well have expected a dissertation on the “Field Theory” approach to individual and group interdependencies. Those with a more psychological point of view may have expected an analysis of the perceptual (or experiential or existential) models with which the individual interprets his position in society and in the environment. And urban activists or social workers may have expected an examination of economic pressures, role models, and expectations and their effects on different classes in the city. Instead of these many fruitful avenues of inquiry we have dealt almost entirely with the way the physical pieces of the city fit together. (And even the concentration on physical issues has been severely restricted; there is a whole range of unexamined questions regarding townscape, overall urban form, and the effect and direction of large-scale long-term planning.)

The message of this essay is simply: That there is a well-tried way of building cities which provides an inherent territorial clarity. That a clear almost-unconscious understanding of one’s territorial position makes life in the city easier. And that a clear territorial framework is - almost ironically - the ideal framework for the anarchic evolution of cities.